EPA-670/2-74-022
July 1974
COMPUTER MANAGEMENT OF A. COMBINED SEWER SYSTEM
By
Curtis P. Leiser
Municipality of Metropolitan Seattle
Seattle, Washington 98119
Project No. 11022 ELK
Program Element No. 1BB034
Project Officer
James C. Willman
U.S. Environmental Protection Agency
Region X
Seattle, Washington 98101
For
Storm and Combined Sewer Section (Edison, N.J.)
Advanced Waste Treatment Research Laboratory
Cincinnati, Ohio 45268
NATIONAL ENVIRONMENTAL RESEARCH CENTER
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO. 45268
For sale by the Superintendent of Document's, U*S. Government
Printing Office, Washington, D.C. 20402
-------
REVIEW NOTICE
The National Environmental Research Center-
Cincinnati has reviewed this report and approved
its publication. Approval does not signify that
the contents necessarily reflect the views and
policies of the U.S. Environmental Protection
Agency, nor does mention of trade names or
commercial products constitute endorsement or
recommendation for use.
XI
-------
FOREWORD
Man and his environment must be protected from the
adverse effects of pesticides, radiation, noise and other
forms of pollution, and the unwise management of solid waste.
Efforts to protect the environment require a focus that
recognizes the interplay between the components of our
physical environment—air, water, and land. The National
Environmental Research Centers provide this multidisciplinary
focus through programs engaged in
• studies on the effects of environmental contami-
nants on man and the biosphere, and
• a search for ways to prevent contamination and
to recycle valuable resources.
One source of water contamination has been sewer over-
flows occurring during rainfalls. This report describes a
computer-controlled "total systems management" complex
designed to reduce combined sewer overflows by providing
the controls that permit storing the overflow in sewers.
A. W. Breidenbach, Ph.D.
Director
National Environmental
Research Center, Cincinnati
111
-------
ABSTRACT
At the conclusion of a ten-year construction program which
affected much of Seattle's combined sewer system, a computer-
controlled "total systems management" complex was proposed, funded
and constructed. CATAD takes advantage of storage in the sewers
to limit overflows, and selects overflow points based on water
quality data.
Since the control system began operating in 1971, receiving
water quality, especially dissolved oxygen and coliform levels, has
shown significant improvement; overflow volume has decreased by
50-60 percent during supervisory control, and in excess of 90 per-
cent during three months of limited automatic control. Eight pollu-
tion loading parameters were measured and found to be 68 percent
less than before advanced control techniques.
Capital costs totaled $2.6 million for the control system at
the 36 remote stations, or $5.3 million, including construction of
15 gate-driven regulator stations. Annual maintenance and opera-
tion costs total $270,000. The system is worth roughly $40-$245
million in equivalent sewer separation.
Work continues on a fully automatic optimizing model to add
predicative capability to program decisions so the system could
maintain an 80 percent overflow reduction.
This report was submitted in fulfillment of Project Number
11022 ELK, by the Municipality of Metropolitan
Seattle, under the sponsorship of the Environmental Protection
Agency. Work was completed as of August 1973.
iv
-------
CONTENTS
Sections
I
V*
II
III
IV
V
VI
VII
VIII
IX
X
XI
XII
XIII
XIV
XV
Abstract
List of Figures
List of Tables
Acknowledgments
Conclusions
Recommendations
Introduction to the System
Data Acquisition Requirements
Project Construction
Operating Procedures
Program Development
Implementation-Operation Problems
Water Quality Studies
Discussion of Results
Costs and Benefits
References
List of Inventions and Publications
Glossary
Appendices
Page
ii
iv
viii
x
1
4
6
24
39
92
115
203
262
277
350
360
365
366
374
v
-------
LIST OF FIGURES
Figure No.
1
2
3
4
5
Title
Page
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
West Duwamish Collection System 11
Initial Project Costs and Planning 14
Revised Demonstration Grant
Planning and Progress 15
CATAD Project Coordination Network 17
Dynamic Regulator Types 19
Interior-Exterior Photos of
Denny Way Regulator Station 20
Interior-Exterior Views of s
Lake City Tunnel Regulator Station .... .^ 21
Metro System - Primary Facilities ' 25
Flow Computations at Pumping Stations ..... 27
Flow Computations at Regulator Stations . , . . 28
Rain Gaging Sites 36
Block Diagram - CATAD Hardware 4U
CATAD Central Control Facility 51
Command Module Push Buttons 53
Alarm and Status Module Indicator Lamps .... 54
Selector Module ..... 55
Segment Selector 56
Map Display 57
Segmented Lamp Codes . 58
CATAD Location Map 62
Telemetry Message Word Format 64
West Point Satellite Terminal 67
Renton Satellite Terminal 69
Motor Driven Gate Details 74
Outfall Gate Controller 84
Automatic Monitor Stations 86
Blockhouse for Monitor 87
Monitor Pump and Float 88
CATAD River Monitor Log 91
Station Hydraulic Display 93
Alarm Display 94
Rainfall Display 95
CATAD Central Floor Plan . 102
Daily Maintenance Work Program 103
Environmental Protection System 105
VI
-------
Figure No.
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
List of Figures (Continued)
Title
Page
Local Control
Supervisory Control !!•!!!!!! 109
Automatic Control '.'.'.'.I'. 109
Block Diagram - CATAD Software !.!!!! i 1 ne
CATAD Executive Structure . ...*!!![[ 127
Program Hangup Detection Flowchart ..... 133
Primary Interval Computations !! 142
Secondary Interval Computation Module . . i i 143
Tertiary Interval Computation Modules . . * i 145
West Point Console Display ] igi
Hourly Log 166
Hourly Storm Log ...... 168
Sluice Gate Flow Regimes [ 171
Storage Volume Model *...! 173
Typical Pump Control Programs ] \ 175
Storm Control with Rule Curves ....... 188
Installation of Palmer-Bowles Flume ..... 193
Flume Rating Curve 193
Equipment Problems [ 197
Plan of Fremont Inverted Siphon & Vicinity [ . 200
Profile of Modified Facilities,
Fremont Inverted Siphon & Vicinity .... 201
Lake City Tunnel System Schematic ...... '. 220
Green Lake Trunk Sewer Diversion
Structure Profile 222
Lake City Tunnel System Profile [ 223
Well Level Control for Variable
Speed Pumping 227
Pump -Speed Analog Circuit 230
Calibration Equipment in Use ......... 232
Dexter Avenue Regulator Station ...„,.. 237
Regulator Gate Control Modes -
Dexter Avenue Regulator 238
Metro Telemetry Unit Simulator ....... 240
Contractor's Pump Station Simulator 241
Flow Limitations Affect Control Strategy . , 244
Isohyetal Map of Seattle 246
Water Quality Monitoring Stations Associated
with the CATAD Demonstration Grant ..... 263
Programmer Disk 265
VII
-------
List of Figures (Continued)
Figure No.
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
Title
Storm Sewer Separation Sites 269
Freeway Drainage Studies 270
Urban Storm Drainage Areas 273
Map of Duwamish Monitoring Stations 276
Green-Duwamish River Dissolved Oxygen Range . 280
Duwamish River Flow 281
Green-Duwamish River Temperature Range at
16th Avenue South, Surface 282
Green-Duwamish River Temperature Range at
16th Avenue South, Bottom 283
Green-Duwamish River Minimum Dissolved Oxygen 284
Green-Duwamish River Bottom Chemical
Oxygen Demand ...< .....286
Elliott Bay Coliform Levels and Standards . . 289
Duwamish River Coliform Counts 292
Median Coliform County, Lake Washington
Ship Canal 294
City of Seattle Rainfall Frequency,
Intensity, Duration 296
Average Monthly Precipitation - City of
Seattle ..... 298
Summer Rainfall Distribution 301
Storm Drainage Loading 309
Norfolk Regulator - Long Term Overflow Quality 310
Overflow Characteristics of Large Storage
Stations 313
Overflow Characteristics of Small Storage
Stations 314
Connecticut Street - Overflow Quality Variation 315
Chelan - Overflow Quality Variation 316
Hanford - Overflow Quality Variation .... 317
Denny Lake Union - Overflow Quality Variation 318
Michigan - Overflow Quality Variation .... 319
Regulator Drainage Basins 321
District Overflow Profiles 322
Nitrate - Comparative Peak Loading 326
Ammonia - Comparative Peak Loading 327
Phosphate - Comparative Peak Loading 328
Vlll
-------
List of Figures (Continued)
Figure No.
101
102
103
104
105
106
107
108
Title
Chemical Oxygen Demand - Comparative
Peak Loading
Solids - Comparative Peak Loading
Overflows During All Control Phases Combined
Overflows for Each Control Mode
Regression Lines for Each Control Mode . . .
Loading Trends During Local Control
Loading Trends During Advanced Control
System Performance Summary .
329
330
333
335
336
346
347
349
IX
-------
LIST OF TABLES
Table No. Title Page
1 Original Grant Cost Estimate , 12
2 Predicted Control System Efficiencies ...... 13
3 Storage Availability in Collection System .... 31
4 Computer System Equipment 41
5 Central Computer Main Storage Allocation 42
6 Interrupt Assignment 44
7 Input/Output Channel Assignments 47
8 Remote Station Status Inputs 63
9 West Point Terminal Equipment '. . . . 68
10 Renton Terminal Equipment 69
11 Program Clock Functions 124
12 Contents of Physical Data Files 153
13 Cathode Ray Tube Displays 156
14 Console Command Checks 159
15 Classification of Stations for Automatic Control . 180
16 Force Main Calibration Stations 228
17 Pump Speed Calibration Equipment 231
18 Pump Speed Calibration Data 233
19 Sequential Sampler Timing 266
20 Duwamish River Automatic Monitor Analysis of
Variance on Dissolved Oxygen Data 285
21 Adult Salmon Returns 287
22 Trawl Catches of English Sole (In Green-
Duwamish River) 288
23 Median Coliform Counts - Elliott Bay Shore
Stations 288
24 Median Coliform Counts - Duwamish Estuary .... 290
25 Maximum Goliform Counts in Duwamish River,
East and West Waterway 291
26 1971-1972 Elliott Bay Median Coliforms/100 ml -
One Meter Depth 293
27 Total Annual Storm Data 297
28 Storm Comparisons 300
29 Storm Pattern Variations 303
30 Stormwater Runoff Quality - South Bellevue
Interchange 304
x
-------
List of Tables (Continued)
Table No. Title Page
31 Loading Factors - South Bellevue Area, ...... 305
32 Loading (Ibs/storm) 308
33 Regulator History and Characteristics 312
34 Variation of Overflow Quality with
Rainfall Intensity 324
35 Total Rain in Inches for Estimated Peak
Loading All Stations 321
36 Estimated Overflows and Reductions for a Storm . . 334
37 Program Analysis of Performance at Different
Storm Ranges 333
38 System Performance for Summer Periods . . i . . . 340
39 System Performance for Different Storm Patterns . 341
40 System Performance for Different Storm
Characteristics 342
41 Reduction in Loading Parameters 343
42 Comparison of Loading Peaks 347
43 Metro Combined Sewer Area Improvements 350
44 Eligible Project Costs 351
45 Total System Costs 352
46 Hardware Cost Details . 353
47 Operating-Maintenance Costs 354
48 Average Overflow Volumes for Control Modes .... 354
49 Overflow Solutions and Comparisons 356
XI
-------
ACKNOWLEDGMENTS
Appreciation is extended to the following persons and their organi-
zations for their assistance and cooperation during the course of
this study:
PROJECT GUIDANCE
Municipality of Metropolitan Seattle
Charles V. Gibbs, Executive Director
Theodore W. Mallory, Director, Technical Services
Office of Research and Development - U.S. Environmental Protection Agency
William A. Rosenkranz, Director
Municipal Pollution Control Division
James C. Willmann, Project Officer, Research and Develop-
ment Office, Pacific Northwest Region
ENGINEERING DESIGN AND DRAFTING
Metropolitan Engineers, Consultants for the Municipality
Stuart M. Alexander, Executive Engineer
James F. Lynch, Project Engineer
Paul C. Leach, Chief Electrical Engineer
Chris Chantrill, Senior Engineer
Ron Brooks, Senior Engineer
City of Seattle, Department of Engineering
Harvey W. Duff, Senior Supervising Engineer, Head of
Sewerage and Drainage Section
Municipality of Metropolitan Seattle
John Wolff, Electrical Engineer
George Fields, Engineering Associate
WATER QUALITY AND PROGRAMMING
Municipality of Metropolitan Seattle
Glen D. Farris, Superintendent Water Quality and
Industrial Wastes
Ralph Domenowske, Process Research Engineer
Bruce Burrows, Water Quality Technician
James E. Stapleton, Programmer
REPORT PRODUCTION
Theresa Murphy, Editor, Metro
Girmy Bradberry, Secretary, Metro
Roy Montgomery, Senior Engineering Associate, Metro
MAINTENANCE AND OPERATION
Thomas G. Rice, Superintendent, West Point Div., Metro
Gary D. Isaac, Superintendent, Renton Division, Metro
Paul J. Bride & William E. Nitz, CATAD Console Operators
COMMERCIAL FIRMS
Philco-Ford Corporation, Western Development Laboratories,
Palo Alto, California
Xerox Data Systems, El Segundo, California
Northwest Digital Systems, Seattle, Washington
xii
-------
SECTION I
CONCLUSIONS
5.
A 13,120 acre combined sewer area within the city of Seattle
has become an example of a successful application of advanced
computer control techniques to reduce combined sewer overflows.
Fifteen regulator stations and one major pumping station were
controlled to reduce overflow volumes by 50 to 60 percent
under supervisory control, and in excess of 90 percent under
automatic program control and during unusually dry weather.
Loading analysis, reveals that 80 to 90 percent of the peak
loading has been reduced, and the peak loading has been shifted
to a higher rainfall rate which occurs less frequently. Total
loading in pounds has been decreased an average of 58 percent
for ammonia; up to 76 percent for COD. The average loading
reduction for eight measured pollutant factors was 68 percent.
A total of 514 storms and 762 overflow events were analyzed.
Rainfall studies show that the city averaged about 150 measur-
able storms per year at an average of 0.23 inches per storm.
The weather studies also indicated that in the final year of
automatic control, rainfall volume was 50 percent below normal;
however, the number of storms remained approximately the same.
A few intense storms during summer periods severely influenced
the performance data. Further studies are necessary to deter-
mine whether system performance was actually as ineffective
during summer periods as the limited data indicated. Addi-
tional data may contradict the information accumulated during
this study, which indicates that the system was most effective
during winter storms. This may be due to the inability of
local control systems to recognize storm or collection system
conditions outside of the stations' immediate area.
During this period of regulator local control, the tendency
was for infrequent but large volume overflows. Supervisory
and advanced storm control strategies allow for regular monitor-
analysis-command cycles. It was found that frequent, short
duration, small volume overflows shaved storm flow peaks and
resulted in an impressive reduction in total system overflow
volumes.
Rainfall intensity has a considerable effect on overflows.
Considering the average rainfall rate of a storm, the total
system reduced overflow volumes by 73.6 percent in supervisory
control, 97.2 percent in automatic control, and 85.8 percent
under combined advanced control modes.
-------
7. The many variables which influence the overflow quantity
generated by a system during a given storm make standard sta-
tistical correlations very difficult. Tabular analysis of
overflow events for different rainfall ranges was quite success-
ful as was a moving average program which effectively indicated
loading trends.
8. In analyzing a specific sample station, a characteristic time
lag, build up and wash-out was observed for all measured
pollutants (often termed the first flush effect). However,
the peak concentrations for some pollutants such as ammonia^
evidenced no obvious relationship with the more typical solids
analysis pattern during a storm.
9. Each station tended to show a "fingerprint" effect for sequen-
tial overflow data. This fingerprint was generally unique for
each station and usually repeated itself for different storm
types. The data indicated that the first flush of materials
is often diverted to the interceptor in a combined system
rather than overflowing to the receiving water.
10. Overflow priorities were based primarily upon volume reduction.
Station by station priority varied considerably depending on
which pollution factor was the basis for establishing priority.
11. During the course of the study, the Duwamish River receiving
water has improved dissolved oxygen content by 1 to 2 milli-
grams per liter.
12. During the supervisory control period, coliform counts in the
Elliott Bay area have dropped significantly at all near-shore
stations, but have remained relatively constant at offshore
stations. Coliform counts in the Duwamish River have fallen
more than 50 percent in the estuary area where overflows
generally occur. >
13. The success of the computerized control management system is
due to:
(a) surveillance and early attention to potential
mechanical and control problems,
(b) early action and preparation for storm events,
(c) logical use of all available system storage to
minimize combined sewer overflows.
14. System cost was $2.6 million for the computer controls and
station control equipment. Additional regulator station con-
struction would increase costs to $5.3 million.
-------
15. Capital costs averaged about $200 per acre; maintenance and
operation costs averaged $5 per acre or $.88 per thousand
gallons of overflow captured and diverted to the treatment
plant.
16. Interfacing was the most difficult hardware problem
encountered. About two years of sporadic work were required
to pinpoint and solve electrical noise interference problems.
17. Equivalent costs for complete separation to equal the CATAD
system volume reduction is between 40 million and 240 million
dollars. Separation does eliminate overflows compared to
CATAD1s 60 - 90 reduction, however, separation also diverts
many nutrients and other pollutants into receiving waters
effectively lowering the performance of separation to a
level with the CATAD system. (Editor's ttote: This comment
is also made in Table 49).
18. Intangible benefits include the newly available routine
data processing capability, a flexible system readily
adaptable to new requirements or expanded facilities,
possible improved plant operation and capacity by controlling
dry weather flows, improvements to receiving water, improved
public relations, and the possibility of delaying or pre-
venting the large cost of sewer separation.
19. Data transmission checking and protection features are nec-
essary. Local control capability and 24-hour monitoring
must assure that breakdowns in any portion of the system are
covered in some manner.
20. The storage rule curve control technique is feasible and has
been very effective on this control system.
21. It is often difficult to isolate system problems to either
hardware or software. A systematic problem solving tech-
nique using the computer as a diagnostic tool effectively
isolated such problems.
22. Excessively rapid equipment reaction to level changes risks
creating a water hammer or other potentially dangerous and
expensive problem within the collection system.
-------
SECTION II
RECOMMENDATIONS
1. Development of a computerized control system necessitates
passing through various levels of control. A flexible, in-
formative console is essential during the developmental phase
to provide system information and to allow a human operator
to perform control functions and enlarge the control data
base.
2. To minimize hardware/software problems, a single contractor
should be selected for complex computer control systems wher-
ever possible.
3. Studies should be initiated to predict overflow water quality,
and to minimize pollutant loading factors.
4. The first flush capture in combined sewer systems versus storm
drains should be examined. Sequential samplers located at the
site of a trunk diversion and overflow point should be triggered
automatically and remotely by a computer system if possible.
5. New control systems should make maximum use of digital sensors
to provide data for direct transmission to the central computer.
Modern noise filtering techniques should be applied to minimize
transmission and/or command difficulties.
6. When installing a CATAD-like control system, early data collec-
tion and instrument calibrations should be specified to elim-
inate one major source of errors.
7. Urban drainage studies should be continued to determine how
loading changes from separated and urban areas affect regula-
tor performance.
8. The performance characteristics of Michigan and Brandon Regu-
lators where the sewer profile has been modified to drain
stagnant sewage back to the interceptor after an overflow
should be studied.
9. Summer storm response to the advanced control system should be
studied further to determine whether the unimpressive perfor-
mance is to be expected or is a result of unusual storm acti-
vity during the automatic control period.
10. More definitive methods of classifying storms should be made
available to simplify analysis of regulator and system per-
formance to various storm characteristics.
-------
11. Improvement resulting from the optimizing simulation model
being developed by Metro Seattle should be monitored. It's
effectiveness as a design tool in determining runoff coeffi-
cients and selecting optimum storage volumes should be deter-
mined.
12. The effect and performance of sewer separation of North
Interceptor and West Point Plant flows should be evaluated
and compared with centralized control influence on the
Elliott Bay interceptor.
13. Standardized data collection, display hardware and software
would greatly reduce costs in developing similar computer
control systems.
14. Investigate the possibility of an intertie between the
collection system control model and the Duwamish River
Estuary model in determining overflow priorities.
15. Standardized sampling equipment should be improved to avoid
data gaps which are a major deterent to thorough analysis
of all storm and control situations.
16. Software development should follow a logical sequence to
minimize delays:
a. System design
b. Systems programming
c. Applications programming
17. A thorough information and training program should be
initiated as early as possible. The training program should
be run con-currently with the development, and should
include all documentation and standardized operator's manuals,
18. Strong and continued leadership and staff support are
required to successfully complete a major research and
development project such as the CATAD system.
19. Federal and/or State guidance in the form of overflow
pollutant guidelines or limitations would be beneficial
in determining overflow priorities in automated control
systems where alternative overflow decisions are consistent
and can be easily programmed.
-------
SECTION III
INTRODUCTION TO THE SYSTEM
SYSTEMS IN GENERAL
Computer proliferation/ significant improvements in output
capacity versus investment/ and the success of computer techni-
ques in complex and varied applications around the country have
brought the term "systems" into common use. In recent years/
urban analysts and experts have advocated management information
systems and the use of systems analysis and planning-programming-
budgeting in city governments. Analysis of whole systems rather
than their individual components is becoming increasingly popu-
lar in many fields including pollution control (1). Systems
methodology, also referred to as management science/ operations
research/ systems analysis/ and systems engineering is an analy-
tical approach for large/ complex/ multi-variable problems. The
technique provides systematic definition and evaluation of
alternative solutions to these complex problems for use by de-
cision makers. Even a superficial review of the literature
reveals that pollution control and other environmental quality
problems are complex enough to benefit by the systems approach
(2). Insights thus gained may lead to much better problem solu-
tions. Large/ complex, multi-dimensional problems are not new,
but they could not be clearly pictured even a decade ago. Rules
of thumb and handbook criteria were used to design and operate
water pollution control facilities. In the past decade, techno-
logy has changed those standards. New mathematical, economic
and computational tools demonstrate the fallibility of earlier
design criteria. Today's methodologies and computer technology
analyze many variable and relevant factors to determine engineer-
ing design and operating specifications. This then, is the
systems approach.
Hierarchical management levels of all organization struc-
tures require four basic levels of information for operating an
organization as identified by Tolle (3).
1. Policy or strategic information — is required at top
management levels for defining priorities.
2. Planning and forecast information — is required or
initiated by middle management to identify prospective
avenues for future undertakings.
3. Performance or control information — is required by
direct supervisory personnel to determine how well
objectives are being met.
-------
4. Operations, production/ transactions or working informa-
tion — is required for instruction of or initiated by
workers directly involved with the organization's tasks.
The tremendous information volumes and levels of refinement
have led to the use of the computer based management information
system (MIS) in many industries. Evolution of these intelli-
gence systems can be compared to development of human intelli-
gence. The basic decision process includes four identifiable
stages of activity:
1. Observation
2. Inference
3. Evaluation
4. Choice.
Management information systems generally evolve from relatively
simple data banks where observations are summarized so that a
user can apply inference and evaluation to make a decision. A
second stage data bank system involves the computer in making
inferences so that a manager can propose "what if" questions and
receive an array of responses for his evaluation and decision.
A third level is often a large scale optimization model system
which encompasses action recommendation. Procedures are pro-
grammed to evaluate alternatives against assigned go.als. The
manager either implements a decision based on the recommendation
or he rejects the alternative and requests further analysis for
decision. In the final stage of maturity, the entire decision
process is automated within the information system. Observation,
the ability to choose to initiate action, commit resources and
monitor results are programmed into operating procedures. The
complexity of the decision process for any given application
limits the development of a fourth level (fully automated) system
to very well defined environments (4).
Federal, State and local governments and other decision
making bodies have shown increasing interest in modern systems
techniques. The systems approach, however, has just begun to
solve management problems in water pollution control. Although
is has often demonstrated technical success, the results have
frequently been rejected by the political process (2). Metro-
politan engineering-economics systems analysis of all aspects
of urban water resources has been found feasible (5), and com-
prehensive simulation models are being designed in many areas
around the country. These systems methodologies produce simu-
.lation models which are static or dynamic mathematical repre-
sentations of some process. Gordon (6) presents an excellent
description of the model classifications now employed in systems
studies. Simulation model sizes may vary from simply depicting
-------
the time-varying position of a stretched and released spring
to the largest and most complicated dynamic environmental plan-
ning models in the country today (7). The water balance model,
possibly a portion of the total environmental management model,
is pictured by McPherson (2). The staggering amount of data to
be collected, stored and evaluated in producing and running a
model of this size dictates a computer based systems approach
for either a one-time solution or for real-time process control.
"SYSTEMS CONCEPT" IN COMBINED SEWER REGULATION
Past and present applications of systems information and
control techniques provide a wide base of experience which can
be concentrated to apply systems concepts in combined sewer re-
gulation and control. Combined sewer systems were designed and
constructed on the principle that overflows at frequent locations
could effectively provide prompt and local relief of collector
sewers and interceptor lines. This concept resulted in a multi-
plicity of overflows and regulators, but also reduced the size
of interceptors in various sections of the total sewer collec-
tion system. This "each unit of a sewer system for itself"
concept does not integrate the various facilities into a master
management plan. Defects of single unit control include:
1. Multiplicity of overflow points
2. Lack of priority sequence of locations to minimize
environmental harm and hazards
3. Failure to achieve the full hydraulic capacities of
each section of the total sewer system
4. Inability to utilize the whole sewer system in rela-
tion to the patterns of storm and runoff in various
areas of a community; particularly large sewer areas
with variable topography (8).
By including the full function of the combined sewer system
control program, the manager can use the "systems concept" to
achieve maximum benefits of regulator control. "Total systems
management" of combined sewer systems uses not only the trans-
porting capacity of sewer systems, but their retention capacity
as well. Frequently, past improvement to a sewer facility in
one location caused greater problems elsewhere in the system,
but this fact was ignored due to the difficulty in analyzing and
achieving total systems management. The concept envisions man-
agement or control of all elements and facilities ofvthe sewer
system to:
-------
1. Maximize utilization of interceptor sewer capacity to
carry combined sewage to the waste water treatment
plant
2. Maximize utilization of all in-system storage
3. Give priority in the interceptor sewer to flows with a
higher pollutional load whose overflow would result in
adverse conditions upon receiving waters
4. Integrate use of combined_sewer overflow storage or
treatment facilities.
Communication is the key to a systems approach. Modern
communication methods provide the basis for integrating the in-
dividual units into a master system. Avoid oversimplification
of the systems concept in combined sewer system management and
overflow control. Systems control dictates dynamic regulators
so that facilities can efficiently route, limit, divert, trans-
fer and store waste waters within the sewer system according to
a predetermined plan. Measurement and status determination are
accomplished by sensors which utilize electrical signals to
represent such conditions as flow, head, pressure differences,
gate positions and receiving water levels. Data collection,
storage, correlation and display are accomplished by supervisory
personnel or by computer programs they prepare. Decisions -are
executed by field crews or by remote control automation facili-
ties which receive instructions and carry them out. These in-
structions are verified by feedback communication from field
crews or automated communication channels.
Anticipating more stringent Federal controls on sewer over-
flows, numerous cities have begun or have implemented "total
systems management" projects. In-depth descriptions of these
control systems are available (8, 9, 10). Cleveland and Minne-
apolis/St. Paul have some type of supervisory control under con-
sideration or now in operation, and an automatic control model
is completed. Such models are difficult to lift off the draft-
ing board and implement in actual hardware. Later sections of
this report describe the country's most advanced fully automated
feedback control system. In Seattle's Computer Augmented Treat-
ment and Disposal System (CATAD), the human operator is" an
observer who is involved in the control loops only in the event
of systems failures or unusual unprogrammable environmental con-
ditions which require his specialized technical knowledge.
-------
EVOLUTION OP SEATTLE DEMONSTRATION GRANT
Federal demonstration grants stimulated recent developments
in computerized total systems management projects in combined
sewer technology. Demonstration grants have a large impact on
cities considering development of such control systems. Grigg
(11) describes many positive and negative incentives available
to and used by our government. Without the impetus and encour-
agement of Federal funding arrangements, Seattle and other cities
might not have had the interest or funding to embark upon such
research and development projects. Federal and State demonstra-
tion grant funding levels when this project began were consider-
ably less than funds made available by the 1972, amendments to
the Water Quality Act. Significantly, the money available in
1967 has provided six years of technological advancement in
pollution control that might otherwise not be available for
cities presently considering such systems.
When the Federal Water Pollution Control Administration
(FWPCA) was formed in 1965, Seattle was already four years into
a construction program to eliminate pollution of local receiving
waters and to reduce overflows from a heavily overloaded com-
bined sewer system. Design considerations and construction
activities are discussed in the first major report on the CATAD
system, "Maximizing Storage in Combined Sewer Systems" (12).
Computer system control of regulator storage within inter-
ceptor sewer lines was proposed by Metro early in 1966 in a re-
quest to the FWPCA for construction grant funding for an inter-
ceptor, regulator and pump system in the West Duwamish area of
Seattle (Figure 1).
The agency denied Metro's request for construction funding
for the regulator portion of that system, but expressed consid-
erable interest in this type of construction. Later in 1966,
public information about demonstration grants was widely distri-
buted in the Federal Register (13). In October, 1966, a mutual
resolution of the Metropolitan Council authorized application
for a demonstration grant, and an application was submitted to
the administrator of FWPCA. An FWPCA official visited Seattle
and indicated interest in the project which had begun five years
before with construction of an interceptor system with storage
capacity. This official reviewed interceptor construction and
three dynamic regulator structures with mechanically driven
sluice gate mechanisms which had been operating for over three
years. In November, 1966, FWPCA forwarded copies of demonstra-
tion grant application and guideline materials, and expressed
interest in the program.
10
-------
FUTURE I
EXTENSION j
Siphon to
P.S. and
Treatment
DUWAMISH
LEGEND
OUTFALL — -.
INTERCEPTOR
TRUNK SR.
STORM OR.
FIGURE 1
WEST DUWAMISH COLLECTION SYSTEM
Metro officials selected a technical consultant. A hydro-
electric process control computer system in the Wenatchee, Wash-
ington, area was visited. Hydroelectric upstream water storage
concepts from this and other control systems were applied to
Metro's CATAD system•control computer. Late in 1966 the exact
details of the computer augmented treatment and disposal project
were revised and resubmitted. In an official communication on
December 29, 1966, the FWPCA commissioner extended the $1 million
limitation for demonstration grant projects to the CATAD system.
The grant offer was for $1,400,000 or 53.5 percent of the esti-
mated eligible project costs.
To demonstrate changes in concept and costs since the ori-
ginal 1966 conception, some details of the demonstration grant
request will be explained. Table 1 shows the breakdown of con-
struction and services foreseen in the original estimate. Later
additions to this breakdown outlined regulator construction costs
of $1.5 million and water quality study expenses of $30,000.
Primary features of the demonstration grant, agreed to by both
parties, included:
11
-------
Table 1. ORIGINAL GRANT COST ESTIMATES
Total
$
CONSTRUCTION
65-11 (W. Duwamish Reg.'s)
Later Contracts-Construction*
Computer and Monitors
SUB-TOTAL
TECHNICAL SERVICES
W. Q. Reports
Engineering Plans & Specs @ 6%
Supervision @ 3%
Post-Construction Studies
.SUB-TOTAL
O & M (1 year)
Legal and Fiscal 1%
Administration 1%
Contingency 10%
Tax 4.2%
Site
SUB-TOTAL
GRAND TOTAL
589,000
1,550,000
600,000
200,000
164,000
82,000
133,600
30,300
27,000
27,000
274,000
115,000
100,000
Eligible
1,550,000
500/000
$2,739,000 $2,050,000
123,000
61,500
133,600
$ 579,600 $ 318,100
20,500
20,500
205,000
$ 573,300 $ 246,000
$3,891,900 $2,614,100
*Later Contracts Detailed:
est. Hanford
Lander
Connecticut
King
Denny
$ 450,000
400,000
250,000
150,000
300,000
$1,550,000
Compared to Actual:
Harbor
CheIan
Eighth
W. Michigan
$ 90,000
75,000
65,000
60,000
$290,000
12
-------
1. Five automatic regulator stations
2. One computer control system
3. One central supervisory station
4. Water quality monitoring equipment
5. Controls and wiring for an automated CATAD storage and
regulation system.
In one of the original submittals to the Federal Water
Pollution Control Administration, the Metro staff predicted
the sort of results expected from the CATAD control system.
Table 2 summarizes this data and shows that the performance
anticipated an optimistic 90 percent reduction in critical
summertime, overflows and a 77 percent reduction during the
winter months, for an overall annual reduction of 80 percent
overflows. In a later section, this report compares the pre-
dicted performance with actual results under different operat-
ing modes.
Table 2. PREDICTED CONTROL SYSTEM EFFICIENCIES,
Rainfall Range
Inches/Day
0.1 - 0.24
0.25 - 0.49
More than 0.5
On Site
Regulation Only
Winter
ABC
30 45 45
21 0 45
14 0 45
Summer
ABC
10 53 53
7 0 53
2 0 53
Annual
ABC
40 47 47
28 0 47
17 0 47
Computer
Controlled Regulation
Winter
ABC
30 45 45
21 32 77
15 0 77
Summer
ABC
10 53 53
7 37 90
2 0 90
Annual
ABC
40 47 47
28 33 80
17 0 80
A — Number of Days
B — Incremental Percentage of Overflows Contained
C — Cumulative Percentage of Overflows Contained
Source: Table 3-9 "Metropolitan Seattle Sewerage and Drainage Survey" (27)
Note: Percentages based on statistical average days of rainfall in excess
of 0.1 in/day which was: Summer - 19; Winter - 66; Annual - 88.
13
-------
The original plan was for a three-year project beginning
in January of 1967 with design, construction, evaluation and
report breakdowns as shown in Figure 2. For reasons explained
JAN
3,614,000 -
(02,000,000-
§
o
o
cc
a.
UJ
—
o
1.0OO.OOO-
500,000-
PHASE
OPERATIONaEVALUATION
PHASE II
CONSTRUCTION
40 50 60 70
PERCENT COMPLETION
FIGURE 2
INITIAL PROJECT COSTS AND PLANNING
in later sections of this report, progress on the demonstration
grant lagged behind schedule. In February of 1971, the demon-
stration grant was extended until August, 1973. The newer sche-
dule (see Figure 3) called for an interim report describing pro-
gress to that date. This report, "Maximizing Storage in Combined
Sewer Systems," was completed in 1972 and published late that
year (12). These schedule extensions disrupted the 1967 budget
plan because of additional maintenance, operation, administra-
tive, engineering and programming expenses. Cost details are
covered in detail in Section XI of this report.
14
-------
25
50
75
100
PAMTDAPT FVDITMntTI IDC
FACILITIES
CONSTRUCTION
CATAD 8 INTERFACE
INSTALLATION
SAMPLER
INSTALLATION
RECEIVING WATER
MONITORING
OVERFLOW
SAMPLING
RPI ATPH CTiinicc
DEVELOP CONTROL
PROGRAM
REPORT OUTLINES PRE
PARED a APPROVED
LOCAL CONTROL
SYSTEM
EVALUATE CONSTRUC-
TION/LOCAL CONTROL
WRITE-PUBLISH
INTERIM REPORT
SUPERVISORY
CONTROL
COMPUTER CONTROLS
QY^TFM
EVALUATE COMPUTER
WRITE-PUBLISH
FINAL REPORT
1
1
1
I
1
|
1
I
3
1
|
•
1
1
|
1
1
|
\
=|
J
JAJOJA J O J AJ OJAJ OJ AJOJAJ OJAJ OJ
1967 " 1968 1969 1970 1971 1972 1973
FIGURE 3
REVISED DEMONSTRATION GRANT
PLANNING AND PROGRESS CHART
15
-------
Metro discovered the considerable staff effort required to
manage and complete a demonstration grant of this magnitude and
complexity. The first study requirement was a "Post Construc-
tion Studies Evaluation" report (14). This report was prepared
by staff people normally assigned to other jobs. The prepara-
tion effort plus the additional work required to respond to
questions of the federal awarding agency demonstrated the need
to assign permanent responsibility for this project. A staff
engineer was assigned coordination responsibility in September
of 1967, and a regular reporting procedure on CATAD progress
and construction form submittals was initiated. Construction
plans were assembled and coordination of report production
(Figure 4) was initiated. Regular monthly reports were sub-
mitted with special reports as conditions dictated. All reports
associated with the production of this demonstration grant are
referenced in this report or in the first project report (12).
16
-------
to
z
o
a s
8
d
-_i
Ul
X
o o
u ui
Ul <£
CD
Z
5*1
si
_j 5
-1 o
o
CO
Ul
1-
H
CO
||
•""••"•""
1
1
» 1
0 * '
111
Q.OJ |
too: 1
1'
JC.V/ 1
INATOR
rnu
COORD
"V
_^
7
Ul
o
z t-
Z Ul
o u.
u to
Ul
o
II
co
o
Q.
o
85*
Ul O
Ul O
h-
<
0
CC
£
^^ .
|1|
C3 o OT
S = a
CO
£ cc
PROGR/
OPER
£
SIONS
ACTOR
REEMEN"
CM 8 O
Z
s
to
0
CO
5
CM
zii
— o ^~
Ul Q. <
Q CO CC
Z 0
CC iij
Ul J
h- Q.
m o
ui ^ i^
0 S
uj g
01 >
o
D
O
H
W
g;
O
H
EH
g
o
o
§
O4
P
H
-------
CONSTRUCTION PROGRAM
The original grant request appropriated approximately 70
percent of the budget to construction with the remainder applied
to the supervision, technical studies and services associated
with the grant requirements. A brief overview of the construc-
tion work will simplify later explanations of control concepts.
Section V provides a detailed explanation of the system hardware.
Project construction work can be divided into three catagories:
1. Entirely new regulator structures built to develop
storage within the system — designed to allow remote
control from the computer
2. Control modifications to existing remote stations that
overlaid CATAD remote control on the existing pneumatic
control systems
3. The central computer and supervisory control station
itself — the heart of the system.
Regulators
The CATAD concept maximizes utilization of in-system stor-
age to reduce or eliminate overflows. A detailed description of
the typical Metro dynamic regulator was presented in the initial
report on this system (12). Several different regulator struc-
ture types were considered in programming as well as construc-
tion. Figure 5 shows regulator types from a typical diversion
regulator to a storage-only regulator placed directly in inter-
ceptor lines. A list of regulators with hydraulic and cost
data is presented in Appendix A. The first new regulator con-
structed was the Denny Way Regulator (Figure 5F). It is unique
in that two separate trunk lines are controlled at this station.
One trunk line has a large storage volume and is low in eleva-
tion, while the second trunk line has essentially no storage
and a high elevation. If the two trunk lines were physically
connected, storage in the larger Lake Union trunk would be lost.
Separate controls at this location avoid this loss. The com-
puter treats these two regulators as separate stations even
though they are inside the same physical structure. Figure 6
shows interior and exterior photographs of this facility, which
cost approximately $500,000.
18
-------
A. COMMON DIVERSION
C. STORAGE ONLY
(TMM MAM
HANFMD I
U.S.
E. DIVERSON ONLY
LEOENO
NOIIMM. FLOW
WTllllilTTINT (OVIRFLOW)
•LUICI (ATI
•UH.OIM* OUTLINC
B. TWO-STRUCTURE DIVERSION
mULATOM
D. STORAGE w/DOWNSTREAM OVERFLOW
F. DUAL REGULATOR
FIGURE 5
DYNAMIC REGULATOR TYPES
19
-------
FIGURE 6
INTERIOR-EXTERIOR PHOTOS OF DENNY WAY R.S.
Within six months of the grant award, studies showed that
storage locations within the collection system had been over-
looked. Considerable interceptor storage was available in the
Lake City tunnel. Metro requested that grant contingency funds
be allocated to construct the Lake City tunnel regulator within
the CATAD project. This facility is typical of in-line storage
type regulators. Figure 7 shows views of this structure which
cost approximately $150,000. The interesting, unexpected hydrau-
lic problems presented by this installation should be reviewed by
agencies contemplating this type of sluice gate regulator station.
Construction of four additional sluice gate diversion type regu-
lators was originally planned through demonstration grant funds.
Some original budgeted costs were realigned; and these four sta-
tions along Elliott Bay were removed from demonstration grant
budgets and funded from other sources.
20
-------
FIGURE 7
INTERIOR-EXTERIOR VIEWS OF LAKE CITY TUNNEL R.S.
Remote Station Control Modifications
When imposi-ng centralized computer control on existing
regulator stations and pumping stations, computer or telemetry
failures originating in either hardware ,or software must not
leave the remote station in a situation less desirable than
existing proven controls. Alexander and Gibbs (15) described
the original concept of superimposing computer control on an
independent local control system while retaining reliable methods
.for reverting to local control in the event of computer system
failures. Improvements to control hardware are described in
this report (Section V). Briefly, regulator controls were con-
verted to electrical instruments providing analog voltages to
be converted and transmitted to the central computer. At the
same time, an interface was provided to filter out expected
remote station electrical transients.
21
-------
The grant project's first construction contract was to
supply electrical controls with remote sensors and transducers
which would provide signals to the computer and would allow
ramping type return to local control in the event of failure.
Metro electricians and technicians installed this equipment
because of their familiarity with station instrumentation and
control philosophies.
The second contract provided cabinets and relay contacts
plus other interface equipment. This interface contract also
provided for a multi-conductor cable connection between the
regulator and outfall gate control stations at two locations
where these gates were 800 to 1000 feet apart. Coordination
difficulties resulted from the large number of consultants,
contractors and staff people involved in three or more simul-
taneous jobs, all trying to match up physically and electri-
cally at some point on a drawing or map.
Experience gained in installation of these regulator sta-
tion controls enabled engineers to combine all remote controls
for the computer interface at each of eighteen pumping stations.
Assigning the entire project, including supply and installation,
to a single contractor provided better project control and an
improved end product.
Computer Control Center
The last major construction was the computer control center.
Foreknowledge of the time required to provide the computer and
the programming might have segmented this project early in the
demonstration grant so that control model programming could
parallel hardware development. The computer contract called
for a process control computer with peripheral equipment in an
office building. A supervisory control console would display
information from all remote stations. The contract, required
provision at each remote station of a control, coding and trans-
mitting unit which would be the dividing line between computer
equipment and Metro station equipment control and maintenance
responsibilities. The contractor was to provide software (pro-
gramming) which would demonstrate that the hardware could per-
form all basic functions required by the detailed operating
specifications of the CATAD system (16). Other contractor pro-
vided features should also be considered by agencies contemplat-
ing such a control system. The contractor was required to train
the agency staff people assigned to operate this system; and to
provide a renewable annual maintenance function for the entire
system.
22
-------
A series of changes requested and implemented early in the
progress of this contract have added immeasurably to the bene-
fits and flexibility of the system. The prime contractor,
Philco-Ford, within four months of the notification to proceed
on the contract, suggested revision of the supervisory control,
reducing the number of cathode ray tube displays and providing
a separate console and wall map display similar to those in the
NASA Houston Control center. Control push button arrays were
suggested to permit flexibility in determining which stations
and which commands would be utilized at various times in the
life of the system. When Metro changed its timetable for
future regulator construction, it became obvious that numerous
telemetry control units would be idle for years, providing no
information to the agency, while they deteriorated through dis-
use. The designated list of active terminal types was modified
and telemetry unit allocation revised for installation in exist-
ing pumping stations as monitoring units. The map board and
alarm displays were modified to improve the board's functional
arrangement and appearance and allow a readily interpreted four-
color segmented lamp display for each station without excessive
board congestion. The map was enlarged to include future sta-
tions and extensions to the Metro system. Great flexibility
allows future map revisions and changes to the system as con-
struction is accomplished. Lesser changes to the CATAD computer
room, both internal and external, are covered in other sections
of this report.
Along with the station and control installation, the demon-
stration grant required considerable data accumulation for
analysis of the success or failure of the system. Initial data
collection equipment consisted of automatic sampling machinery
installed at existing regulator stations. Samplers were trig-
gered into operation by opening any outfall gate. Initially,
a combination of sequential and composite type samplers was pro-
vided at seven stations. Improved versions of these samplers
installed at other stations corrected deficiencies in the ori-
ginal design. A three-pen recorder was installed at the regula-
tor structures to obtain overflow volume information from the
regulator sites. The recorder plots water levels upstream and
downstream of an outfall gate; and the outfall gate position.
Standard orifice formulas allow calculation of flow through the
outfall gate. Techniques for digitizing plotted information and
converting it to flow data are described in this report based on
three years experience with the sampling and recording equipment.
-------
SECTION IV
DATA ACQUISITION REQUIREMENTS
GENERAL CONCEPT
This report intends to provide any agency considering
implementation of a "total systems management" project with
information to begin the accumulating and organizing require-
ments for data and equipment for the system. This section con-
centrates on the goals/ objectives and data requirements which
were an integral part of the development of the Metro Computer
Augmented Treatment and Disposal system. A systems study of
this magnitude not only generates information required to design
the hardware to perform the desired functions, but it also pin-
points areas where information is lacking and suggests new types
of sensors or additional manual studies which will provide the
needed information.
The following paragraphs summarize the basic objectives and
the concepts which provide guidelines for investigation of data
requirements and system design.
The Municipality of Metropolitan Seattle, generally known
as Metro, is a regional agency providing sewage collection treat-
ment and disposal services to 11 cities and 23 local districts
which provide sewer services directly to residential and business
customers. Metro's "wholesaling" activities are performed by
incorporating a large interceptor sewer system (Figure 8) which
collects sewage flow from local trunk lines and then conveys
this sewage to treatment plants. Interceptor sewer design capa-
city is based on the amount of sanitary sewage flow expected
with ultimate planned development of the areas served. Most of
Seattle is presently served by trunk or collector sewers carry-
ing combined storm water and sanitary sewage. Metro's intercep-
tion system cannot handle these combined flows. Peak sanitary
sewage flow is only about 5 percent of the total flow in the
trunk under storm flow conditions.
The flow of sewage into the interceptors from trunk sewers
carrying combined flows is controlled by regulator stations which
perform two functions:
1. Regulating the volume of flow diverted from trunk sewers
into the interceptor sewer
2. Providing overflows into the Duwamish Estuary or into
Elliott Bay for excess trunk sewer flows during storms.
24
-------
ft
} )
1
y -r- i
' 1 J
r
f
^
_' — :•; i_
. S
'•.. \
.j 'A
'"' x-x
/\
\
•\/ ./ , v
V: \ i V-
\ "-\ . ^A
; ; :
Snohomish County
^
.1_
County
\>
—\
<•
Metropolitan facilities
to be constructed
and operated by others
\. -- ~i >
^ L-
!-^1
| Auburn
.--
/'
FIGURE 8
METRO SYSTEM - PRIMARY FACILITIES
25
-------
These unattended regulator stations operate under local station
automatic pneumatic-electric type controls. Schematic diagrams
of existing controls are presented and described in Section V.
Sewage flow from the trunk sewer through the diversion con-
duit into the interceptor is controlled by an adjustable regula-
tor gate operated by automatic controls triggered by a preset
maximum water level in the interceptor. Under normal conditions
(no heavy storm water flows), the regulator gate is fully open
and all trunk sewer flow is diverted to the interceptor. When
the interceptor sewer level rises to the setpoint, the regulator
gate closes in steps to maintain the preset level in the inter-
ceptor. When the trunk sewer flow exceeds the regulated flow
into the interceptor, excess sewage is stored in the trunk until
the level in the trunk reaches a preset point. An outfall gate
then opens in steps to maintain the trunk level at the setpoint.
Outfall conduits from some of the trunk sewers into the Duwamish
Estuary are submerged by tidal variations of as much as 15 feet.
In order to prevent saltwater or river water backflow into the
trunk sewers, outfall gates at these stations -have overriding
controls which prevent outfall gates from opening when the tide
is above the trunk level (16).
Note that the' trunk setpoint is manually established at a
regulator station so that at the worst possible combination of
high tide, heavy storm flow in the trunk, and a fully laden
interceptor line, the outfall gate opens soon enough to permit
combined flow from the trunk line to escape from the sewer system
through a fully opened outfall gate without flooding or backup
conditions upstream in the trunk sewer drainage area. That set-
point remains essentially fixed through the various seasons un-
less manually altered. In the local control mode, each station
is operated independently by instruments located within the
station in response to signals from local sensing devices.
It would be difficult and time consuming to have persons
visit each of the widely separated regulator sites to adjust the
setpoints to compensate for dynamically changing water level
conditions occurring during each storm. If it were possible to
do this, setpoint adjustment could allow a higher water eleva-
tion to gain additional storage and reduce or prevent overflows.
A centralized control facility utilizing a high speed computer
makes these remote station adjustments in response to, or in
advance of, variable storm patterns. The report "Maximizing
Storage in Combined Sewer Systems" (12) describes storage within
a combined sewer system under varying storm patterns using regu-
lator gates and pump controls. Figures 9 and 10 show some nomen-
clature of these stations, the sensors which are monitored to
generate information about storage, and the control signals which
drive the machinery to achieve efficient use of that storage.
26
-------
.NFLUENT
Time_2./2_ A. LEVELS
DOWNSTREAM
INTERCEPTOR
-UPSTREAM
INTERCEPTOR
PUMP STATION
CROSS-SECTION
UPSTREAM
INTERCEPTOR
CROSS-SECTION
FIGURE 9
FLOW COMPUTATIONS AT PUMPING STATIONS
27
-------
UPSTREAM
REGULATOR
GATE
DOWNSTREAM
REGULATOR
PLAN VIEW
TRUNK
CROSS-SECTION
LEGEND
••^ COMPUTED FLOW (a Name)
UE=SJ SLUICE GATE
BUILDING OUTLINE
FIGURE 10
FLOW COMPUTATIONS AT REGULATOR STATIONS
28
-------
The primary objectives of the Computer Augmented Treatment
and Disposal system (CATAD) are:
1. To utilize the maximum storage capability of large
combined sewer trunk lines and ultimate capacity of
interceptor sewers to reduce or eliminate overflows
caused by rain storms.
2. When overflows cannot be avoided because of storm
intensity, the CATAD system selects the overflow
. points which will cause the least harmful effects
on the receiving waters, beaches and marine life.
Overflow decisions are based on water quality data
monitored and stored within the computer. This in-
formation is regularly updated to reflect latest
receiving water conditions.
3. Regulation of daily flows to treatment plants to aid
in stabilization of treatment processes and effec-
tively increase the dry weather capacity of existing
treatment plants.
4. The CATAD system may eliminate or reduce cost of
separating combined sewers which would be especially
costly and disruptive to commercial/industrial areas
and central business districts. Sewer separation
may not have to be 100 percent, and/or it may be
stretched over a longer period by optimizing use of
the combined system as separation progresses.
If these objectives do not encourage a Municipality to
embark upon "total systems management" concepts, proposed guide-
lines (17) regarding point sources such as combined sewer over-
flows or bypasses and stormwater outfalls will drastically in-
crease the volume of data which will be required. Statements
from the President's Council on Environmental Quality and EPA
(18) and the General Accounting Office (19) outline the types
and levels of control now proposed for these long ignored pollu-
tion sources. Sewer separation may compound the problem of
urban waste collection and treatment since storm runoff may soon
require some mandatory degree of treatment.
29
-------
CONTROL STATION SELECTION
The design of a system which will carry out the objectives
just described must establish whether all stations in the system
will have both monitoring and control functions, or whether
there will be some mixture of these functions within the collec-
tion system. Adding control functions to monitoring functions
increases the remote facility cost by $3/000 to $6/000, depend-
ing upon controls necessary. For Metro's CATAD system, control
instrumentation decisions were based primarily upon storage
availability determined from a preliminary investigation of the
system. This information is shown in Table 3 together with the
final storage values determined by detailed studies and operat-
ing experience. Supervisory operation pinpointed connections
into the system which had never been documented and which ser-
iously affected the safe available storage in a trunk or inter-
ceptor line. Some of the greatest volumes of storage are avail-
able within interceptor lines such as the Lake City Tunnel and
the Elliott Bay Interceptor upstream of Interbay Pumping Station.
The largest trunk storage is available at the Hanford Regulator
Station where approximately one million gallons of storage is
available above the base average dry weather flow.
Cost factors affect command station decisions. The cost
of controlling a multi-pump station with a very complex mode
control arrangement may not be justified to gain a small amount
of storage. It may be more economical to gain storage at some
other site in the system or to consider installing permanent
storage basins (20) underground or inflatable holding tanks (21,
22) for additional controlled storage. Engineering and economic
factors must be weighed in deciding which stations will be com-
mand stations.
Metro system decisions resulted in the selection of the
command stations shown in Table 3. The selection of these com-
mand stations did not preclude storage at other regulator sta-
tions in the system (not shown in Table 3). In fact, storage
is generated at these stations by direct command of the regula-
tor and outfall gates. Volume of storage is the main reason
for calling a station a storage station.
For safety precautions, the special ramping type control
system described in Section V was specified so that the large
storage volumes would not create an overflow risk in case of
computer or telemetry failures. Ramping controls gradually
release a large volume of storage. At direct control stations,
the volume of storage is considered too small to require this
ramping procedure. In the event of computer or telemetry fail-
ure, any excess stored volume drains from the trunk to the inter-
ceptor without overflow risk.
30
-------
S
W
EH
CO
>H
CO
S3
O
EH
O
W
rl
. T
M
o
o
S3
H
!x
r .
CH
H
rl
H
*^4
rH
H
<
w
CD
3
0
EH
CO
•
CO
0)
i-H
Xt
5
S CO
p<
0
•rl
•P
co
H
m
4J
o
rtj
- f
•rl
CO
0)
Q
rH
m
o
d
•H
0)
Q
i-H
jjj
o
O cncoo covoin
OCO OinrH rH«*rH
rH CM
O^J* OCMVO COCMt^-
OO OOO OCOO
^1* **J* ^^ LO CO ^3*
O O rH 0 CO rH
-P
CD
rl
•P
CO
GO) (1)
(033
tn c d
•rl d O in oo
CM rH r- o in
rH rH
CM 0 ON rH CO
CM CM VO «* VO
vo o in o o
rH CM CM CO O
rH
CO CO VO *3* 00
OOO O CM
oo rH oo in
OOO VO
C3
o
•rl
C! rH
JD rH Q)
^ O 3
-P 1-3 rl EH
j3
O >i >i >»
•rl (d (0 -P
•P IS 12 -H
0 0
i >i
d tn c d 0)
f4 p4 fH f4 ^t
O -rl 1 CO
rl 43
•P 0) -P
CO 4-> -P
-------
DATA COLLECTION REQUIREMENTS
Since the real-time computer deals with numbers/ determin-
ing the types of data to be collected/ measured, transmitted
and stored in the form of numbers and characters (represented
by digits) is extremely important in implementing a control
system like CATAD. The agency must establish physical para-
meters to be monitored/ forms of commands to be issued/ and
other data needs to get the system off the drawing boards/ into
specifications and under construction. Information require-
ments are:
1. Dynamic data includes the physical parameters monitored
during each scan of remote stations; the command infor-
mation sent to the station following the scan; and any
other decision making processes in the program for
supervisory control procedures.
2. Fixed data includes base information required to cal-
culate elevations/ flows/ calibrations of instruments
and maintenance factors; and
3. Strategy type information might initially be in tabu-
lar form to be used by a human operator in making
control decisions. It might later be an equation or
a computerized table used for automated command deci-
sions .
The types, volumes and frequency of change or update of
system information will have a great impact on selecting instru-
ments and computer equipment to perform data gathering, analy-
sis and storage.
DYNAMIC DATA NEEDS
The dynamic data requirements for the sewage collection
system influence the selection of data communications equipment;
the transmission line characteristics, and computer speed and
capacity specifications. Water levels at many locations are
probably the most important single category of information
required to calculate flows and trigger certain types of alarms.
By incorporating Manning's equations, various orifice and weir
formulae and pump efficiency curves, it is possible to calculate
flow at almost any location in the system. The computer checks
water elevations with reference to overflow weirs to generate
pre-alarms before overflows take place and/or actual alarms
after the overflow has begun. Monitoring, control and modeling
of the entire system depends upon flow information from many
32
-------
locations/ some of which cannot be obtained from water level
measurements alone. In the CATAD system, these additional on-
line flow measuring techniques are employed:
1. Flow measuring weirs are installed at various locations
2. A calibrated propeller type flow meter is monitored at
the West Point treatment plant site.
3. Force main pressure calibration methods are used at
pump stations where there is sufficient friction head
loss to calibrate force mains at various flow ranges.
Flow information and the water level data collected from
these many locations throughout the system enable the programmer
to calculate the amount of storage available in the various
trunk lines and interceptors. Once this information is avail-
able/ system storage can be monitored and controlled as described
in Section VIII.
Once the primary transmission and receiving equipment has
been installed/ the data collection system can monitor equip-
ment operating rates and status. An important pump station
parameter is the pump rotational speed in revolutions per minute.
This data is necessary in calculating flows using pump efficiency
curves. At regulator stations, the most important information
is gate travel. Since there are a variety of gates (some open-
ing upwards and others opening downwards, and each gate has a
different cross sectional appearance), it was decided to report
gate positions as a "percentage open" figure. For stations with
indirect, overriding control systems (15), it was necessary to
provide "state-of-the-art" controls incorporating pulse counter
devices. These pulse counters are monitored at all times so
that a perfect change-over from remote control to local control
is assured in any event.
The risks of encountering explosive atmospheres in com-
bined sewers either through biological generation of methane
gas or spills of explosive type fuels or cleaning agents, re-
quire an atmosphere monitoring system. Explosion hazard moni-
toring units were incorporated at various prime locations in
the system to detect potentially hazardous conditions and to
aid in the monitoring and investigation of some illegal hazard-
ous material dumps into the collection system
A list of data monitored and definitions is summarized in
Appendix B. Up to this point, the dynamic data refers to analog
signal monitoring. Stations within the system range from four
to eight active analog signals per station with an average of
33
-------
five and a maximum of twelve analog signals possible. A digital
totalizer, which is driven by the pulse generator, advances and
can be scanned to determine the number of pulses in a timed in-
terval.
The status of equipment and different alarms at the remote
station is monitored quite easily by providing relay contacts
connected to various alarm sensors and status indicators. In
all cases, a relay contact was interposed to prevent excessive
voltages from entering the solid state electronic monitoring
devices installed at each remote station. Examples of contact
status information include:
1. A float switch monitor for high water levels
X,
2. A "power on" indication
3. Air pressure sensor which indicates loss of instrument
pressure
4. A status indication that a gate is moving either up
or down.
These alarms will be discussed in detail in other chapters.
For a list of contact status reported, refer to Appendix B.
The number of status and alarm contacts scanned ranges from
10 to 48 monitored contacts per station with an average of
21 and a maximum of 64 possible.
Command procedures require additional dynamic data report-
ing to the central computer. In general, the computer command
involves closing a hardware contact in a remote station followed
by a physical movement of an instrument setpoint or gate posi-
tion. The commanding logic continuously scans these contacts
and equipment positions to assure that the command has' gone to
completion. Command output contacts range from zero to six per
station with an average of four and a maximum of twelve possible.
System programs perform a three step command check. The
contact closure is scanned to assure command initiation; secondly,
the rate of movement of the setpoint or gate is checked at vary-
ing intervals to insure movement at a proper rate; and lastly, a
test determines command action completion. The commanding pro-
cedures and various checks which comprise the dynamic data re-
quirements for this function are reported in detail in Section
VII of this report.
34
-------
Before instrument selection and design, the engineer must
know what degree of precision is desired and! how frequently
collection system information and commanding activities must be
reported. Sewage collection and treatment precision is not
quite as critical as in manufacturing processes; however, it
cannot be ignored. Metro decided on instrumentation with a
two to five percent precision and on remote station scanning
at a minimum frequency of one minute, adjustable in four steps
to ten minutes for dry weather conditions when system parameters
are more stable. To achieve desired overall system precision,
individual instruments in the system had to have precision on
the order of 0.1 to 1 percent to compensate for the error multi-
plying effect from source to output.
During 1969 a series of weather analyse's investigated those
meteorological quantities which would provide the best informa-
tion for predicting storm intensities and actual wet weather
flows in the combined sewer system. Theoretically, a lead time
in excess of two hours permitted excess storage being used for
leveling dry weather flows to the treatment plants to be quickly
released and the system drawn down as far as possible to gain
the maximum amount of storage in preparation for the predicted
rainfall. The results of the study were reported in reference
(12). The correlations indicated that some combinations of
barometer, anemometer, and wind direction could give predictive
capabilities in excess of two hours. Since there is no advan-
tage in competing with the weather bureau in predicting rainfall
events, the four hour prediction reports issued by the weather
bureau along with "chance of rain" indications are entered into
the program logic by operators in deciding what levels of storage
to maintain. Dynamic meteorological data requirements are limited
to rain gaging stations at critical sites in the Seattle area.
Tipping bucket type rain'gages are utilized. Each 0.01 inch
of rainfall triggers an electrical pulse which is counted and
stored at the remote station. The computer checks the rain gage
totalizer during each scan to observe accumulated rainfall amounts
and average rainfall rates. Six rain gages were installed init-
ially. Further studies indicated that more rain gages were
needed. Since the City of Seattle's manually operated system
contained seventeen gages/ Metro planned a minimum number
of gages, assuming that serious gaps in the rain gaging system
could be avoided by relocation or telemetering City gage data
into the Metro system. This will provide detailed coverage in
various drainage basins as required after flow and hydrograph
predictions in the mathematical model routines are tested.
Present rain gage locations are shown in Figure 11.
35
-------
SNOHOMISH_ COUNTY
"KING COUNTY
KIRKLAND " R§fflJOND
LEGEND
O METRO GAGE
• SEATTLE GAGE
FIGURE 11
RAIN GAGING SITES
36
-------
Water quality data is important, primarily from the Duwa-
mish River south of Seattle. If the river has oxygen deficit
conditions at certain locations, these sites are considered
critical areas and nearby overflow points are assigned a very
low overflow priority. The computer or supervising operator
selects the alternative available. The hourly data was to in-
clude temperature, dissolved oxygen, conductivity, pH, and
turbidity at four river locations in the combined sewer area.
River flow in cubic feet per second was considered for water
quality monitoring purposes. The combination of the water
quality monitors and river quality information from a United
States Geological Survey (USGS) gaging site some 20 miles up-
stream could provide sufficient model input data to predict
the effects of overflows on the receiving water and thus aid
in the selection or rejection of overflow sites when alterna-
tives were available. The idea was rejected because of the
complexity of modeling a tidal estuary and because USGS was
working on such a model. A cooperative effort whereby Metro
collected data and USGS worked on the model was initiated in
1964.
•
In addition to the base elevation data for remote stations,
basic information is needed about equipment speeds and limits
as well as pump efficiency curves and existing control schemes.
The computer can then compare rates, speeds and other control
data with fixed information stored on disk files for rapid access
and comparisons. Fixed data collection for the Seattle collec-
tion system and station equipment was aided by the relative
youth of the system. Except for the trunk sewers, most of the
system was less than ten years old when the information was* com-
piled. Fixed data accumulation may be difficult for municipal-
ities considering such a system if their files have not been kept
up to date with all changes since original installation.
Lastly, drainage basin size and runoff characteristics are
necessary to enable the computer to generate or select an inflow
hydrograph due to rainfall from a particular storm. For the
CATAD system, approximately forty drainage basins within areas
tributary to the Elliott Bay and Duwamish interceptors were
analyzed.
37
-------
STRATEGY DATA
A fourth major category of system data which 'effects selec-
tion of instruments and computer hardware to satisfy all require-
ments is called "Strategy Information". This data is filed for
ready access and provides either a human operator or the com-
puter modeling logic with restrictions and control strategies
which allow optimum overall system control decisions. In a super-
visory control mode where human operators must make decisions,
they must have available at a moment's notice such information
as "safe" storage volumes or water levels and limits on the
speed of movement where this type of information might be of
considerable importance. Pump changes performed too quickly
may result in a dangerous and potentially damaging pump cavita-
tion. The CATAD system provides the operator with a detailed
summary of local control instrument setpoints so that changes
from one control mode to another are based on actual dynamic
data which predict the consequences of switching from remote
to local control.
For automatic operation, a controlling logic termed a "rule
curve" as described by Alexander (24) was chosen as the basis
for control strategy. Rule curves were designed and stored for
each control location in the system and for each type of storm
situation considered. Rule curves are discussed in Section VII.
For municipalities considering some other form of control logic,
rule curve data would not be required, but some form of command
strategy information would have to be retained for each command-
ing situation.
Another factor which influences the selection of computer
equipment for the control system is the amount and duration of
data retention — which items are to be saved and how they will
be stored. Data retention media include anything from an hourly
log on paper forms to a dense magnetic tape or disk storage.
Data storage decisions can be combined with previous decisions
regarding dynamic, fixed and strategic data collection require-
ments to develop specifications for the system. The next sec-
tion describes Seattle Metro CATAD system selection and assembly
based on the data requirements decisions made in the early
phases of this demonstration grant.
38
-------
SECTION V
PROJECT CONSTRUCTION
GENERAL
Approximately 68 percent of the cost of the CATAD project
was expended for equipment or facility construction and modifi-
cation which, for the purposes of this report/ are categorized
inclusively as construction. System hardware refers to the
equipment which/ after installation/ allowed a computer based
central monitoring and control capability. Project construc-
tion included two regulator stations and control modifications
to all remote stations to complete the communication and con-
trol links which are discussed in detail in this section.
CATAD SYSTEM HARDWARE
The principal hardware items included in the CATAD System
controls are as follows:
1. Central Computer
2. Secondary Storage
3. Peripheral Equipment
4. Central Control Facility
i
. 5. Telemetry Sys,tem
6. Satellite Control Terminals
7. Water Quality Monitor Interface
A system block diagram is shown in Figure 12.
CENTRAL COMPUTER
Central Processor
CATAD System operation is directed by a general-purpose
computer provided with special hardware features for real-time
applications; including hardware interrupts, a real-time clock
and power fail safe (see Table 4).
39
-------
CM
H
D I
O
H a
O
<
H
Q
O
O
40
-------
Table 4. COMPUTER SYSTEM EQUIPMENT
Name
Computer System
Central Processing Unit (CPU)
Real-time clocks (4)
Power-fail interrupts
Memory protect feature
Parity error interrupt
External interrupts
Multiply/divide hardware
External core memory
Direct I/O system
Buffered I/O channels (12)
Watchdog timer
Rapid Access Disk (RAD)
Storage System
Magnetic Tape Units
Keyboard Printer
Buffered Line Printer
Paper Tape Input/Output System
Card Punch
Card Reader
Plotter
Manufacturer
Xerox
Xerox
Model
Number
Sigma 2
8001
Xerox
Daconics
(Hewlitt-Packard)
Xerox (Teletype)
Xerox
Xerox
Xerox
Xerox
Calcomp
7204
2914
809KKSR35)
7440
7060
7160
7122
563
The central processor provides 8 high-speed general registers,
two of which may be used as index registers, 24 input/output
registers and 16 memory protection system registers.
Main Memory
The basic storage unit is the 16-bit binary word. Main
memory is magnetic core storage with a total cycle time of 900
nanoseconds. The system includes 49,152 words of core memory;
16,384 words of internal memory and 32,768 words of external
memory. External memory may have independent data paths for
external devices. Memory may be expanded to a maximum of
65,536 words. ;
The allocation of main storage among computer functions is
shown on Table 5. About 50 percent of main storage is allocated
to simple tasks which permanently reside in core storage. The
other half is shared by tasks which reside in bulk storage and
are loaded into main storage before execution.
41
-------
Table 5. CENTRAL COMPUTER
MAIN STORAGE ALLOCATION
Function
Memory Allocation
Resident Overlay Total
Monitor System Resident
Monitor Extensions
Public Library
Common
Hardware Interrupt Handlers
Applications Programs
Background Batch Processing
& Application Programs
Unassigned
TOTAL
7,552
1,376
4,608
6,912
1,056
2,816
24,320
512
8,960
7,552
1,888
4,608
6,912
1,056
11,776
12,288 12,288
3,072 3,072
24,832 49,152
The total storage required for applications programs is
24,064 words. However, 12,288 words are used only for a real-
time system model which is executed at half-hour or one-hour
intervals during storms and at two-hour to four-hour intervals
or not at all during dry weather periods. This storage is
available much of the time for background batch processing,
including compiling and assembling new real-time tasks. Many
of the subroutines most widely used by programs coded in For-
tran are in the public library for use in both real-time and
background tasks, leaving a substantial amount of background
storage for engineering studies.
Approximately 3,000 words of core storage are unassigned
and available for additional real-time tasks pending completion
of real-time programming. Unused core storage will be assigned
to background tasks.
Instruction Set — The instruction format provides 4 bits for
the instruction operation code, resulting in a maximum of 16
unique instructions. Eleven operation codes are used for memory
reference type, and five operation codes are used to develop
instruction groups as follows: conditional branch instruction
42
-------
(8); shift instructions (8); register-to-register instructions
(15); direct control instructions (2); and input/output instruc-
tions (5). Optional hardware multiply and divide instructions
are included in the memory reference instructions. Three basic
addressing methods are available: relative, indirect and indexed/-
these may be combined. One level of indirect addressing is
available.
Priority Interrupts — A system of hardware interrupts is essen-
tial for a computer used for real-time control to permit async-
hronous communications between the computer and external devices.
A hardware interrupt signals the computer on the occurrence of
an internal or external event and causes the computer program
to transfer to a unique location in main memory after completion
of the current instruction, unless inhibited (see below). After
determining the cause of the interrupt, the computer may either
begin a new task, if its priority is high, or defer the task and
place it in a queue if its priority is lower than that of the
task currently being executed.
Interrupts may be caused by internal hardware, such as a
clock, by parity error circuitry, by memory protection circuitry,
by program control or by external devices, such as console
switches, teletypes, etc. A maximum of 146 hardware interrupts
can be installed on the CATAD System computer, of which 34 have
been furnished. A list of interrupts assignments is shown on
Table 6.
Real-Time Clocks — The central processor is provided with four
real-time clocks. Each clock operates at a different frequency:
Clock 1
Clock 2
Clock 3
Clock 4
Frequency
hz
500
2000
60
8000
Interval
Milliseconds
2
0.5
16.7
0.125
Each clock comprises a hardware counter which is incremented
by the' real-time clock at the clock frequency. When the counter
increments to zero, the hardware automatically triggers a second
counter = 0 hardware interrupt. Each counter consists of a de-•
dicated core memory location which can be set under program
control such that the counter = 0 interrupt may be triggered at
43
-------
Table 6,
INTERRUPT ASSIGNMENT
Internal Group
Power On
Power Off
Counter #4
Counter #3 (60 Hz)
Counter I2 (2 KHz)
Counter #1 (500 Hz)
Parity Error
Protection Violation
Buffered input and output
Control panel (computer)
Counter #4=0
Counter #3=0
Counter #2=0
Counter #1=0
Watchdog timer
Telemetry (future)
Remote Station Telemetry
Water Quality Monitor
External Group 1
Teletypes (Events and Logging)
Renton Satellite Terminal
West Point Satellite Terminal
Metro Executive (Applications Programs)
Level 0 Commands
Level 1 Process alarm
Console Push Buttons
Executive (Applications Programs):
Level 2 Digital Filtering
Level 3 Alarm Message
Level 4 CATAD Console
Level 5 Renton Terminal
Level 6 West Point Terminal
Level 7 Logging
Level 8 Flow Calculations
Level 9 System Model
Non-Resident Foreground
Monitor
Level
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
44
-------
any even multiple of the clock interval. A user interrupt
handler routine at the counter = 0 interrupt resets the clock
counter and initiates real-time tasks at'programmed intervals.
Memory Protection — Memory protection is required for a com-
puter system used for concurrent processing of real-time con-
trol tasks and background batch processing. Memory protection
is also required for testing and de-bugging real-time programs
on-line. Because the CATAD system controls are developmental
in nature/ facilities for on-line debugging are necessary to
permit various applications to be developed and placed on-line
without interfering with programs previously developed.
Memory protection prevents a background task or an unde-
bugged real-time task from either altering a location or trans-
ferring (branching) to a location outside of the area of core
memory allocated for execution of the task. Memory protection
may be implemented by programming, in which each instruction in
a background task is checked for validity before execution; but
such means increase processing time from 10:1 to 20:1 (or more).
Such methods are not suitable for on-line debugging where pro-
gram execution may be time-dependent. Hardware memory protec-
tion is therefore essential. Some machines protect program
location against alteration but not against invalid transfer
to these locations which can cause serious problems.
The memory protection feature on the CATAD system computer
provides protection in "pages" of 256 words of main memory.
The processor handles protected pages differently from unpro-
tected pages in three ways:
1. Control may transfer from an instruction stored in an
unprotected page to an instruction stored in a pro-
tected page only in response to an interrupt.
2. The contents of protected pages may be changed only
by an instruction stored in a protected page.
3. An input/output (including internal control) instruc-
tion may be executed only if stored, in a protected
page.
Any violation of the protection status of a protected area
will cause an interrupt to the central processor. The protec-
tion status of an area may be altered under program control of
the computer.
45
-------
Power Fail Safe — Because the CATAD system will operate un-
attended most of the time/ it must be able to resume operation
after a momentary power interruption. Although the computer
core memory is bistable (that is, it will not change state when
power is turned on or off), the computer operating registers
are volatile; and if power is cut off, they will be in a random
state when power is restored. Therefore, a power interruption
causes program error unless the status of the registers is re-
stored when the power supply is restored.
Two means are appropriate for protecting real-time systems
such as the CATAD system controls:
1. Providing a non-interruptable power supply; i.e.,
supply outside power to a battery charger which
powers a static inverter with back-up batteries.
2. Providing automatic power-off/power-on facilities.
The first method is used for power system controls where
the equipment must continue to operate through a power inter-
ruption to the computer. The second method is suitable for
systems such as CATAD where a short interruption is acceptable.
The power supplies to the computer logic contain suffici-
ent capacitance to continue computer operation for several
hundred microseconds after power is cut off. Voltage level
detectors interrupt the processor at separate priority levels
for both low-voltage (power off) and return-to-normal-voltage
(power on) conditions. These interrupts enable the monitor to
direct an orderly storing of volatile register contents and of
interrupt status as system power fails and to insure an orderly
computer restart after peripherals have come back up to speed,
thus maintaining system continuity through a power failure. An
external battery-powered real-time clock can be read during the
power-on and power-off sequences to determine the duration of
the outage. The programmer can define the length of power out-
age permitted for restoration of the system to automatic control
after power is resumed.
Input and Output Control
The computer provides two methods of input/output control:
1. A buffered input/output channel which operates under
control of the input/output processor.
2. A direct input/output channel which operates under
program control.
46
-------
Buffered Input/Output Channel — The buffered input/output chan-
nel is a high-speed channel having a maximum data transfer rate
of 4000,000 bytes per second. Data transfers between core memory
and external devices via the buffered I/O channel are under the
control of the input/output processor. Data transfers, once
initiated by the central processor, are carried out by the input/
output processor independent of program execution by the CPU.
The data transfers take place between clock cycles of the cen-
tral processor (i.e., by cycle stealing), and the speed of the
central processor is reduced accordingly. As the data transfer
rate approaches the maximum transfer rate of the channel, the
data processing speed approaches zero.
Each device communicating with the central processor through
the buffered I/O channel is connected via a sub-channel con-
troller. A maximum of 20 sub-channels may be connected to the
buffered I/O channel, of which 12 have been provided on the
CATAD systems. Sub-channel controllers may control a single
device or multiple devices; however, multiple devices may be
connected only to sub-channels 0 through 7. Sub-channel device
assignments are shown on Table 7.
Table 7. INPUT/OUTPUT CHANNEL ASSIGNMENTS
Device
Sub-channel
Number
0
1
2
3
4
5
6
7
8
9
10
11
Keyboard printer
Line printer
*Disk
Paper Tape
Card Punch
Card Reader
*Magnetic Tape (2)
Console CRT's
Spare
Spare
Plotter
Spare
The high-speed devices on the buffered I/O channel are the
fixed-head disk and the magnetic tape units, with maximum trans-
fer rates of 170 bytes per second and 38 bytes per second,
respectively. Extensive disk transfers tend to slow down
processing.
*Multi-device controller
47
-------
Direct Input/Output Channel — Data transfers on the direct
input/output channel are made -under program control. A single
write instruction transfers a 16-bit word from a hardware re-
gister to a device specified by a 16-bit address. Similarly,
a single read instruction transfers a 16-bit word from a device
specified by a 16-bit address to a register in the computer.
All inputs and outputs directly related to real-time control
functions, except console CRT displays, are made via the direct
input/output channel.
Watchdog Timer — The watchdog timer detects system malfunctions
resulting from a program hang-up or from failure of the system
to respond to a command over the direct input/output channel.
The watchdog timer is attached to the direct input/output chan-
nel of the computer and is provided with an interrupt to the
central processor to signal channel malfunction. Program hang-
ups are monitored by requiring a command to the watchdog timer
at regular intervals. Failure to issue the command results in
a contact closure, signalling a program malfunction. Program
hang-up features of the Executive are described in Section V.
In the CATAD system, the contact closure causes a relay to open,
cutting off the telemetry to remote stations and placing the
system in local control. A direct input/output channel mal-
function is detected by monitoring the acknowledge signal from
the channel. If the timer fails to receive an acknowledge sig-
nal within 64 microseconds, the timer simulates an acknowledge
signal from the channel, so that the system does not hang up,
causing an interrupt to the computer. The computer takes appro-
priate action as programmed.
SECONDARY STORAGE
Secondary CATAD system storage comprises a fixed-head disk
drive and two magnetic tape drives (see Table 4). Both disk
and tape drives are attached to the computer buffered input/
output channel through multi-device controllers.
The disk has a capacity of 1.5 million words organized into
512 tracks. Each track is subdivided into 16 sectors of 180
words each. A separate read-write head is provided for each
track, and the average latency time to the beginning of a ran-
dom sector is 17 milliseconds (one-half revolution). Data
transfer between main memory and the disk is at a rate of
170,000 bytes per second. Write-protect switches permit record-
ing heads to be disabled in groups of thirty-two tracks, thus
ensuring that the operating system, permanent data files and
CATAD system control programs cannot be destroyed inadvertently.
Under normal operating conditions, 160 tracks are write-pro-
tected.
48
-------
Magnetic tapes are driven by direct-drive reel-and-capstan
motors with tension arms and tape guides for feed and takeup.
Transport rate is 37.5 inches per second. Data is recorded in
a 9-channel format at a density of 800 bits per inch, resulting
in an effective transfer rate of 30,000 bytes per second. The
drives and the interface to the input/output controller were
not furnished by the computer manufacturer; however, the inter-
face is designed to permit the drives to be compatible with
those of the manufacturer and to operate under control of the
standard input and output routines.
PERIPHERAL EQUIPMENT
Peripheral devices for data processing input and output
include a programmer's console, card punch, card reader and a
line printer, a paper tape punch and reader, and a drum plotter.
Manufacturer model numbers are listed on Table 4 while char-
acteristics of the peripheral devices are summarized in the
following paragraphs. Each device is connected to the buffered
input/output channel through a single device controller.
Programmer's Console
The programmer's console is a Teletype KSR 35 modified by
the computer manufacturer. Maximum printing speed is 10 char-
acters per second. Most computer control commands are entered
through the operator's console. All system hardware and pro-
gram diagnostic messages are printed on the console.
Card Punch
The card punch punches in either of two modes, the binary
or the Extended Binary Coded Decimal Interchange Code (EBCDIC).
Cards are punched row by row at a rate of 300 cards per minute.
An 8-bit EBCDIC character is translated automatically to its
equivalent 12-bit card code. An automatic read after punch
checks the punched data against input data. Two output stackers
are provided to permit preselection of the stacker under program
control.
Card Reader
A table-top photoelectric-type card reader reads cards in
two modes, binary or automatic, at a rate of 400 cards per min-
ute. Automatic read mode permits switching from EBCDIC to
binary facilities and intermixing of EBCDIC and binary punched
cards. A .1,200-card hopper and 1,000-card output stacker are
provided.
49
-------
Line Printer
The drum-type printer with 132 print positions prints from
628 to 800 lines per minute, depending on the characters being
printed. The character set comprises 56 printable characters
plus a blank. The printer controller includes a full line
buffer to minimize tying up the input/output channel.
Paper Tape Input/Output
The paper tape input/output system comprises a paper tape
punch and paper tape reader. The tape is in eight-channel for-
mat with no parity bit generation or checking. Any combination
of eight punches is considered valid; therefore, all checking
must be done by program in the computer.
Punch — The punch operates at a maximum rate of 120
characters per second and is provided with takeup and
supply reel mechanisms.
Reader — The reader reads paper tape photoelectrically
at a maximum rate of 300 characters per second.
Drum Plotter
The 30-inch incremental drum plotter operates at a speed
of 300 steps per second with a 0.005-inch step size.
CENTRAL CONTROL FACILITY
The central control facility provides the means for communi-
cation between the CATAD system operator and the remote station
data acquisition and control system. The principal features of
the central control facilities shown in Figure 13 are an opera-
tor 's console, map display board and logging and events printers.
The operator can monitor data and command all operations at
remote stations from the console. In typical operation, the
CATAD system controls operate fully automatically. However,
during the normal eight-hour work day while routine maintenance
is in progress, the operator can check stations for proper opera-
tion, sensor calibration and status indication. The operator's
console has been used extensively for testing hardware and pro-
grams during system development.
50
-------
§
H
EH
H
1-3
H
§
EH
§
O
Q
•^
<
U
51
-------
Operator's Console
The CATAD operator's console is a five-bay desk-type con-
figuration. Two eight-inch cathode ray tubes are mounted in
each of four console bays. Seven of the eight tubes display
alpha-numeric information; the eighth tube is a spare which
can be activated by remounting in an active slot.
The control and display panels include indicating lamps,
illuminated push buttons, digital displays and a keyboard-type
numerical entry unit mounted in four of the five bays. The
panels are arranged in functional modules as follows:
1. Digital entry module
2. Command module
3. Alarm and status module
4. Station select modules:
West Point System
Renton System
Alki System
5. Segment select module
Operation of the illuminated push buttons and display lamps
and the digital displays and keyboards is under program control
of the computer as described in Section V. An audible alarm is
also operated under program control. The fifth bay is a communi-
cations bay which includes a telephone handset and radio-tele-
phone equipment.
Digital Entry Module — The digital entry module comprises a
four-digit projection-type digital display, a telephone-type
push button keyboard and a group of illuminated push buttons
labeled as follows:
Lamp Test (2)
Enter
Cancel Entry
Display Power
Alarm Acknowledge
Invalid Command (indication only)
.52
-------
Command Module — The command module comprises a group of 21
illuminated push buttons which are the vehicle for entry of
supervisory commands to remote stations, data entry and display
control. Command module push buttons are shown in Figure 14.
s
COMMAND PANEL
FIGURE 14
COMMAND MODULE PUSH BUTTONS
Alarm and Status Module — The alarm and status module comprises
a group of 42 indicating lamps on the operator's console. The
lamps show the status at a selected station under program con-
trol. The function of each lamp is shown in Figure 15.
53
-------
FIGURE 15
ALARM AND STATUS MODULE INDICATOR LAMPS
Station Select Modules — The three station select modules com-
prise the following numbers of illuminated push buttons:
Module
West Point Station Select
Renton Station Select
Alki Station Select
Number
48
40
8
The station select push buttons control the alarm and status
module display; sequence entry of supervisory commands to remote
stations; and perform other display control functions (see
Figure 16).
54
-------
FIGURE 16
SELECTOR MODULE
Segment Select Module — The segment select module comprises 20
two-segment indicating push buttons. The white upper half of
each push button is illuminated when the segment is being dis-
played; the red lower half indicates the presence of an alarm
condition at some station in the segment. The segment push
buttons select the particular group of stations for which data
is to be displayed on the console CRT's. The stations within
each segment are logical groupings of related remote stations;
e.g., stations along one continuous section of the interceptor/
or stations in the vicinity of a major junction point or similar
stations such as treatment plants (see Figure 17).
55
-------
FIGURE 17
SEGMENT SELECTOR
CRT Displays — CRT displays are eight-inch television monitors
with a maximum display capacity of 24 rows of 32 characters per
row for a total capacity of 768 characters. The displays are
operated under computer program control which displays data at
a group of stations determined by the segment select push button
depressed by the operator. The operator may select one of three
display formats via the command module:
1. Station hydraulic data
2. Station alarms
3. Station miscellaneous data
56
-------
The operator may also designate any station as a "Test Sta-
tion" for test and diagnostic purposes, causing a display of all
analog data at the station in the form of voltages, with no con-
version or scaling.
Map Display
The map display is a 12-foot by 7-foot graphic panel posi-
tioned behind the CATAD operator's console. Figure 18 shows a
plan of existing, interim and future Metro facilities included
within the CATAD system superimposed on an outline map of the
FIGURE 18
MAP DISPLAY
area showing major bodies of water and major political bound-
aries. Each pumping station, regulator station and sewage
treatment plant is designated by a graphic symbol. A four-
segment, four-color indicating lamp is mounted within the
symbols for existing or future CATAD stations. The indicating
lamps are operated under computer program control. The design
and interpretation of the segmented lamps is shown in Figure 19,
57
-------
•WILBURTON
FIGURE 19
SEGMENTED LAMP CODES
Logging and Events Printers
Two printers adjacent to the operator's console print perio-
dic data logs and record important operator actions and alarm
messages. The printers operate at a speed of 15 characters per
second on a 72-character line. The printers are controlled via
the computer direct input/output channel using special-purpose
interface logic located in the operator's console and interface
rack.
Each character is output in ASCII code to either printer by
a single direct output instruction. The least significant half-
word (8-bit byte) received by the interface equipment is shifted
serially to the printer after adding a start and stop bit. After
the last serial bit has been received by the printer, an inter-
rupt is transmitted to the computer to indicate that the trans-
mission is completed and that the interface is ready for the
next character.
58
-------
Input and Output Control
Data transfers between the computer and the operator's con-
sole and map display, are made principally via the computer direct
input/output channel; however, output to CRT displays is made
via the computer buffered input/output channel. Space contact
inputs and outputs provide miscellaneous command and display
functions at the central control facility. The electronic cir-
cuits for transferring data between the computer input/output
system and the console, map board and printers, are located in
logic drawers in the rear of the console and in racks located
behind the map board.
Lamps— Lamps on the console or map display are operated by
the computer, which transmits two 16-bit data words to the lamp
circuits via the direct input/output channel. Six bits of the
first word designate the binary-coded address of one of 64
groups of 16 lamps. Each bit of the second word subsequently
determines the state of a corresponding lamp in the selected
group. After the second word is received from the computer,
its data is strobed into the selected 16-bit lamp-driver re-
gister, which stores the state of the lamps for display. Thus,
capability is provided for expansion to 1,024 lamps on the con-
sole and map board.
Switches — Depressing a push button closes a switch, causing
a interrupt to the computer, which initiates an interrupt-
handler routine to determine which switch has been closed.
Lamp and switch data transfers occur as word pairs via the
input/output system. To input switch data, the interrupt-
handler program responds to a switch-generated interrupt by
reading two words of data which are generated simultaneously
with the interrupt. The first word assembled by the switch
data circuits has one bit set, representing one of 16 hard-wired
groups of 16 switches. The second word indicates which switch
in the selected group was pushed. Thus, capacity is provided
for expansion to 256 console switches.
After the closed switch has been determined, the computer
initiates a routine associated with the push button. The pro-
gram turns console lamps on or off, causes the display, or
initiates the control function associated with particular push
buttons depressed by the operator in the prescribed sequence.
CRT Displays — Alphanumeric data displayed on the console CRTs
is output from the computer through the buffered input/output
channel. The displays are controlled by a CRT device controller
and by a CATAD display electronics unit. The controller handles
communication with the computer channel while the display elec-
tronics unit stores, decodes and demultiplexes information to
the several displays.
59
-------
The CRT device controller contains a channel subcontroller
which provides address recognition and connection to the com-
puter's buffered I/O system. Information for display is re-
ceived from the computer in blocks of eight-bit bytes under the
direction of the device controller and is held in a twelve-
byte buffer. The controller provides control, status, inter-
rupts and other signals required by the system.
The display electronics unit contains equipment for all
seven display channels and is expandable to include one more
channel. Control circuitry writes received data into specified
locations in the refresh memories which are separate, full-
screen delay-line memories provided within the unit for each
display channel. All channels time-share a common symbol gen-
erator which converts each memory byte into a seven-line alpha-
numeric character for display. Seven separate video amplifiers
compose the final signals from information received from the
symbol generator and synch generator circuits and drive the
console-mounted monitor screens.
Spare Contact Inputs — Spare contact inputs have been provided
at the central facility for miscellaneous digital inputs to the
'computer via the direct input/output system. The contact inputs
are arranged in 16 groups with 16 contact inputs in each group
for a total of 256 spare contact inputs. Three of these groups
are presently used for the following purposes:
1. Input of time from battery powered external real-time
clock.
2. Input of status from environmental protection system
for monitoring fire, smoke, etc.
Spare Contact Outputs — Spare contact outputs at the central
facilities provide miscellaneous computer control functions via
the direct input/output channel. The contact outputs are arranged
in 8 groups with 16 contact outputs in each group for a total of
128 spare contact outputs. One of these groups presently con-
trols switching of the central station telemetry equipment on
the various telephone circuits.
60
-------
TELEMETRY SYSTEM
The telemetry system provides communication between the
central control facility and remote stations for data acquisi-
tion by the computer and issuance of control commands to the
remote stations (see Figure 20). All data is transmitted
serially in on-off digital bit .form. A single telemetry inter-
face unit at the central station communicates with all remote
stations.
Telemetry Control Units
The remotely located telemetry control unit (TCU) provides
interface between the CATAD System and the control equipment
and sensors at each remote station. Thirty-seven TCU's were
furnished under the CATAD System contract. Two TCU's are required
for one of the regulator stations which regulates flows from two
trunks; therefore, a total of 36 remote stations may be con-
trolled using existing equipment. Each telemetry control unit
is mounted in a 2' x 2.5' x 1.51 sealed enclosure with a locking
door. A lamp display on the door permits maintenance personnel
to read the on-off condition of each contact and the pulse rate
of each pulse input.
Original plans called for installing the equipment in the
West Point and Alki Systems. A future telemetry system was
planned to monitor and control the Renton System from the exist-
ing central computer. As a result of construction delays in
the West Point and Alki Systems, nine TCU's have been tempor-
arily installed at key stations in the Renton System to evaluate
the benefits of central control of this facility.
Data Acquisition
The data acquisition system within the telemetry control
unit at each remote station provides facilities for acquiring
three types of data:
1. Contact status inputs
2. Analog inputs
3. Pulse count inputs
Contact Status Inputs — Contact status inputs indicate operat-
ing status and high or low alarm limits. Digital inputs are
summarized on Table 8. A minimum of 40 and a maximum of 65
contact inputs are available at the TCU.
61
-------
62
-------
Table 8. REMOTE STATION STATUS INPUTS
Pumping Stations
Pumping status (6)
Power monitor (4)
Instrument air
Emergency generator
Pumping unit (6)
Wet well level (2)
Influent gate closed
Starting air monitor
Bar screen differential
Fluid power system (4)
Explosion hazard
Setpoint (2)
Remote/local (2)
Water system monitor
Flood hazard (2)
Operating mode (12)
Telemetry failure
Generator set (3)
Pump discharge valve (12)
Crossover valve (2)
Reg'uTator Stations
Regulator open
Regulator close
Regulator opening
Regulator closing
Outfall gate open
Outfall opening
Outfall closing
Outfall bypass open
Computer/local
Power monitor (3)
Instrument air (2)
Selector switch auxiliary (2)
Emergency generator (2)
Fluid power system
Explosion hazard (2)
Trunk high level (2)
Setpoint (2)
Remote/local (2)
Computer lockout
Service air
Equipment room flood
System master control
Telemetry failure
Analog Inputs — Analog inputs are received from water-level,
pump-speed and gate-position sensors. Signals from each type
of sensor are input to an analog-to-digital converter as 2V to
10V variable signals. The analog-to-digital converter provides
a 10-bit binary output at a maximum rate of 20,000 conversions
per second. (Note: The conversion rate far exceeds the spec-
ification requirements, but is standard with the supplier.)
Analog signals are multiplexed through a single A-D converter.
The multiplexer converts each signal in turn and initiates
data transmission to the central terminal. A maximum of 11
analog signals may be input to each TCU. In addition to pro-
cess analog signals, a constant reference voltage input to the
A-D converter at each station is read during each data scan to
verify proper converter operation. i
Pulse Count Inputs — Pulse inputs are received from two sources:
1. Pulse generators used in control circuits at pump and
regulator stations which operate under remote control.
63
-------
2. Tipping-bucket type rain gages used to obtain rainfall
measurements.
Pulses are collected in a nine-bit binary counter (512 counts).
Up to four pulse counters may be installed in one TCU.
Control
All remote controls for the CATAD System are on-off con-
tacts which energize or deenergize interposing relays. Each
relay has a rating of 5 amps @ 120 VAC or 28 VDC and is pro-
vided with 2 form "C" contacts. Up to 12 contact outputs may
be installed at each TCU, including 2 outputs which determine
the station's control mode.
Message Format
A very high level of security is provided by the follow-
ing procedures:
1. Each data or command message contains a parity bit.
2. Each message is transmitted twice and compared bit
by bit.
3. The second transmission is complemented (inverted).
Each data or command message contains 10 bits of informa-
tion. Three control bits are added to each message/ including
a parity bit, resulting in a 13-bit message. A single trans-
mission comprises a 31-bit word made up of the 13-bit message,
its 13-^bit complement and a 5-bit Barker code for synchronizing.
The Barker code is a 5-bit synchronizing code (11001) which is
detected by the hardware to indicate the beginning of a message
(see Figure 21).
f*
i i olo 1 i
IMurtMMt.
^/^_
6 16 ^ f 18
0
\ DATA \ \
ADDRESS/ PARITY ADDRESS/
DATA DATA
INVERTED
A.
^
• T
DATA 1
t=ARITY
FIGURE 21
TELEMETRY MESSAGE WORD FORMAT
64
-------
Message Transmission
All message transmissions to remote terminals are initiated
by the computer under program control. The computer initiates
all data scans and issues all commands to the remote station.
Commands — Each command transmission to a remote station com-
prises three 31-bit words and includes 30 bits of information.
The first word transmitted contains the station address. The
remote station echoes the message back to the central station
to acknowledge its ready-to-receive status. The second and
third words each contain ten information bits indicating de-
sired contact operate commands. Eight data bits in the third
word are unused and ignored by the remote TCU, resulting in a
total of 12 data bits used for commands.
At the remote station TCU, the Barker code is recognized,
bits 6 through 18 are stored for comparison with bits 19 through
31, and parity is checked. If the ten information bits are re-
cognized as a command addressed to the station, the 31-bit word
is recomposed and returned to the central computer. If the next
two words satisfy the redundancy and parity checks, the twelve
information bits are strobed into a register controlling twelve
relay drivers. Six relay drivers are in use at each regulator
station, with four used at each pumping station. If the message
does not satisfy the security checks, the commanded changes
are ignored.
Data Acquisition — Normal data acquisition is done by polling.
At programmed intervals, the computer addresses a remote ter-
minal, receives all data from that station and repeats the pro-
cedure in sequence for all remote stations. During control
operation, the computer may address the station at which the
control operation is in progress. During a poll, the computer
addresses the stations at about one-second intervals. A com-
plete data poll takes approximately 37 seconds.
A station scan is initiated by an address word from the
central station to the remote station and an echo back from
the remote to central. The acknowledge word from the TCU is
followed by 24 data words separated by circuitry in the TCU into
6 groups of 4 words each in the following sequence:
Group
1
2
3
4
5
6
Data
4 Contact status words
4 Pulse counter words
4 Contact status words
4 Analog data words
4 Analog data words
4 Analog data words
65
-------
One bit in each status word entering the computer is used
to indicate failure of the corresponding telemetered word to
satisfy validity checks at the receiving end. The computer can
use this bit to detect a transmission failure and reinitiate
the scan.
Modems
Data transmission uses frequency-shift non-return-to-zero
(NRZ) cioding at a transmission rate of 1,200 bits per second.
Modems are three-frequency type with a center frequency of
1/700 Hz shifting + 600 Hz from mark to space. A continuous
2300-Hz mark signal transmitted from the central terminal to
all remote stations is monitored by carrier detection modules
at the remote stations. Loss of carrier signal opens a relay
at a remote station which transfers the remote station to local
control operation.
Communications Channels
Communication between the central station and remote ter-
minals employs leased telephone circuits. Circuits are four-wire
multi-point unconditioned Type 3002 circuits and comprise two
simplex unidirectional pairs. Up to eight remote stations are
grouped on a single multipoint circuit using an eight-way Bridge.
A single modem at the central facility is switched between six
multipoint telephone circuits under computer control through '
contact outputs.
SATELLITE CONTROL TERMINALS
Satellite terminals to the CATAD System are located at the
West Point and Renton sewage treatment plants. Each terminal
includes a cathode ray tube (CRT) terminal unit and a hard-copy
printer. Operators at each plant may monitor the complete
system and may enter display requests, specific items of data
and specific commands as programmed. All CATAD System control
monitoring outside the normal eight-hour workday when the cen-
tral control facility is manned, is done at the satellite ter-
minals .
The following functions are presently performed by the
satellite terminals:
1. Display and recording of alarms.
2. Data display under operator control.
66
-------
3. Commanding (changing control mode).
4. Data entry for control and for recording.
The West Point Satellite Terminal, which was procured under
the main CATAD System control contract, operates under fixed
wired logic control. The Renton Satellite Terminal/ which was
procured subsequent to the main contract, operates under mini-
computer program control.
West Point Satellite Terminal
The West Point Satellite Terminal comprises a CRT input/
output terminal and a teletype printer (see Figure 22). The
FIGURE 22
WEST POINT SATELLITE TERMINAL
67
-------
CRT terminal display provides an ASCII character set with a
capacity of 1536 character delay line memory and a 64-charac-
ter solid state row memory. The display is refreshed at a rate
of 60 refreshes per second. All communications control logic
is built into the CRT terminal. The. terminal input keyboard
is arranged in standard typewriter format.
The printer speed is 15 characters per second with 72
print positions. The printer receives character input from the
delay line memory under control of the CRT display logic.
A list of equipment at the West Point Satellite Terminal
is shown in Table 9.
Table 9. WEST POINT TERMINAL EQUIPMENT
Item
Display Unit
Printer
Modem
Manufacturer
Philco-Ford Corp./ WDL Division
Teletype Corporation
Radio Frequency Laboratories
Model
Number
D21
RO-37
2099
Renton Satellite Terminal
The Renton Satellite Terminal comprises a minicomputer
with a communications interface, a CRT terminal and an ASR
printer (see Figure 23). Additional planned functions for
the minicomputer at the Renton Sewage Treatment Plant include;
1. Monitoring contact status
2. Monitoring analog variables
3. Direct digital control of process functions via
analog output system'and on-off controls.
68
-------
V FIGURE 23
RENTON SATELLITE TERMINAL
A list of equipment at the Renton Satellite Terminal is
shown in Table 10.
Table 10. RENTON TERMINAL EQUIPMENT
Item
Control Computer
Printer
CRT display
Modem
Manufacturer
General Electric
General Electric
General Electric
Radio Frequency Laboratories
Model Number
GE-PAC 3010-2
Terminet 300
Datanet 786
22DB
The central computer facility may be used to analyze pro-
cess operations and initiate direct digital control operations
via the satellite terminal.
Minicomputer — The operation of the Renton Satellite Terminal
is under control of a minicomputer which includes hardware fea-
tures required for real-time applications, such as hardware
interrupts, power fail safe and a real-time clock. The computer
is also provided with a real-time programmed operating system.
The basic unit of core storage is an 8-bit byte which is directly
addressable on an 8-bit byte or 16-bit half-word basis.
69
-------
Control of the central processor instruction execution is
implemented through microprpgrams written into very high-speed
read-only memory (ROM). The microprograms use ten basic micro-
coded instructions which are prpgrammed to emulate the nominal
instruction set available to the user. The microprogram accesses
processor instructions from core memory, decodes them and pro-
cesses them in the general registers and core memory locations.
Microprograms can be..used to emulate any desired instruction.
The computer supplier has in this manner implemented a very
extensive instruction set and an automatic input/output scheme.
Actual user programming is done using the emulated instruction
set/ and the programs are stored in read-write core memory.
Core memory access time is one microsecond.
The computer is provided with automatic input/output hard-
ware which is used by the real-time multiprogramming operating
system (RTMOS). Data transfers between the computer and an
external device, once initiated by the central processor, will
proceed under control of the automatic input/output system
until completion without further processor action. The auto-
matic input/output system signals the computer when a data
transfer is complete through an interrupt. Thus while input
or output is in progress with one task, the computer is free
to process other tasks.
CRT Display — The CRT Display has a capacity of 22 lines of
92 characters per. line or a total of 2024 characters. The dis-
play controller uses a delay line memory to store the characters
displayed on the terminal, which are refreshed 30 times per
second.
The terminal keyboard is not arranged in a conventional
typewriter pattern. Alphabetic keys are grouped together in
alphabetical order, numeric keys are grouped into a 10-key
numeric keyboard, and special character keys (punctuation,
symbols) are in a separate group. The shift key permits dual
use of alphabetic, numeric and special-character keys for
additional punctuation and symbols. Transmit control and
cursor keys are also included.
A special group of 15 action keys (expandable to 45)
initiates transmission of the top line of the display and
inserts a special two-character function code into the message.
Each function code inserted by an action key can be interpreted
by the computer as a specific request. These action keys are
not presently used, but are reserved for implementation of
direct digital control.
70
-------
Printer — The printer device includes a paper tape punch and
a paper tape reader. The printer operates at a rate of 30 char-
acters per second and can print up 'to 118 characters per line.
Characters" are transmitted to the printer in seven-bit ASCII
code. The paper tape punch tape punch output eight-bit binary
characters at a rate of 30 characters per second. The paper
tape reader can input 8-channel characters at 30 or 120 char-
acters per second. The printer does not print copy when the
reader is in the high-speed mode.
Data Format
All data is transmitted as 10-bit characters, including a
7-bit ASCII code, a parity bit start and stop bits.
Data Transmission
All data transmission between the central control facility
and the satellite terminals is under program control of the
central computer. The central computer polls each terminal
at regular intervals by transmitting a sequence of characters
consisting of a start byte, two terminal address bytes and a
command byte.
The command byte is interpreted by the satellite terminal
as an instruction to enter either of two receive states or a
transmit mode.
When no message is to be transmitted to the satellite ter-
minal, the command byte places the terminal in a transmit mode.
Unless the satellite terminal operator has requested message
transmission by depressing the enter key, the satellite terminal
responds to the poll by sending an end-of-transmission byte.
Upon receiving a command to enter the "erase/write" receive
mode, the terminal erases the screen, positions the cursor in
the upper left corner, and transmits a ready character to the
computer, which initiated transmission of the text. If the
command is to enter the "row address write" receive mode, the
full screen is not erased and the first byte of the text con-
tains a row index which positions the cursor at the beginning
of the specified row.
Text including non-printing characters is received in a
character string followed by a longitudinal redundancy char-^
acter (LRC). The receiving terminal checks parity on each
character which is received at the end of the character string.
If an error is detected, an error byte is sent to the central
terminal. If no error is detected, an acknowledge byte is
sent to the central terminal, which returns a byte signifying
end of transmission.
71
-------
To transmit from the satellite, terminal to the central
terminal, the operator composes a message on the screen using
the keyboard, terminating the message by. depressing the "enter"
key. If the enter key has been depressed, the satellite ter-
minal responds to the next poll by. transmitting a byte string
comprising an address byte, text between delimiters, and the
LRC character.
Information is written on the satellite terminal printer
(when the satellite terminal is in a receive mode) by trans-
mitting a "print" byte.
Central Station — Data transmission equipment for each satell-
ite terminal xs~ interfaced to the computer direct input/output
channel with identical interface logic. The interface equip-
ment receives characters serially bit by bit from the modems
and assembles them into parallel bytes for computer input.
When an eight-bit byte has been received, the interface logic
causes an interrupt to the computer, which reads the byte
under program control.
Characters to be transmitted to either satellite terminal
are output by the computer to the appropriate interface logic,
which adds control bits and shifts the data out serially to
the modems.
Satellite Terminals — Although satellite terminal hardware
differs,functional operation is similar.
Modems
Modems similar to those used by the telemetry system trans-
rait data at a rate of 1,200 bits per second.
Communication Channels
Communications between the central station and each satell-
ite terminal are over separate four-wire unconditioned Type 3002
circuits.
WATER QUALITY MONITOR INTERFACE
The Municipality's hard-wired water quality monitoring
system was installed approximately seven years prior to pro-
curement of the CATAD System. The existing water quality system
monitors up to ten variables at each of five remote monitoring
stations along the Duwamish waterway. The monitoring system
central station equipment was interfaced to the CATAD system
computer so that water quality data could be input directly to
72
-------
the computer. This data may then be stored for statistical
analysis or used in the control system as one factor in deter-
mining locations where sewer overflows: will result in the least
adverse effect.
Water quality monitor data are input to the computer via
the direct input/output channel using a specially designed inter-
face. The interface processes messages character-by-character
into the computer. Transfer is initiated by raising a ready
signal in the water quality monitor equipment. The interface
equipment receives and stores the first character in parallel
form and interrupts the computer. In response, the computer
executes a routine which reads the character into a storage
buffer. Each character in the water quality message is handled
similarly until the whole message has been entered into the
computer.
REMOTE STATION MODIFICATION
Existing Local Station Controls
All regulator and outfall gate control stations were de-
signed and constructed after 1961 and station controls were all
of similar design/ differing only in details to meet special
station requirements.
All pumping stations to which CATAD controls were applied
were also designed and constructed after 1961. Older pumping
stations in the original Seattle sewerage system were incompa-
tible with the control system due to lack of backup storage or
were in the process of being modernized. It was more economi-
cal to delay installation of the monitoring equipment than to
modify these stations on an interim basis.
Types of Stations
Remote stations designated for installation of control and
monitoring equipment were regulator stations, outfall stations,
combined regulator and outfall stations and pumping stations.
Regulator and Outfall Station Gate Modulation Power
In general, regulator station and outfall station gate
movement (or modulation) is powered, by a motor-driven gate
operator (see Figure 24). Through gear reduction, the threaded
shaft is turned and the sluice gate below moves up or down in
response to the direction of shaft rotation. Four installations
were provided with a medium pressure (750 to 1,500 psi) oil
hydraulic power system wherein the gate is connected directly
73
-------
A. PHOTO OF MOTOR AND GATE STEM
B. PHOTO OF SLUICE GATE
FIGURE 24
MOTOR DRIVEN GATE DETAILS
74
-------
to a power piston that moves in a cylinder. Oil pressure is
admitted either above or below the piston to produce gate
movement.
Regulator Station Controls
The original regulator station controls were an electric
contact controller with an adjustable fixed setting (setpoint)
which it matched against a level measurement in the interceptor.
The level was measured upstream of the regulator station to
help avoid turbulence.
Several regulator stations designed to meet special cri-
teria have overriding features or more complex control systems.
These are Dexter Avenue regulator, Hanford No. 1 regulator and
Lake City Tunnel regulator.
Outfall Station Controls
Outfall station controls are more complex than regulator
station controls because they operate with more parameters and
malfunction is more critical. The outfall gate limits the
maximum trunk sewer storage level during high flow conditions
by allowing overflow to the adjacent receiving water such as
Elliott Bay. Excessively high levels in the trunk sewer may
cause flooding of private property and excessively low maximum
trunk levels limit storage, causing frequent unnecessary over-
flows. An added complexity is the tidal influence of the waters
in the bay and estuary. Backflow of the tidal waters^ into the
system must be prevented. The low elevation of the river delta
land and resultant low elevation of trunk sewers with respect
to maximum tide levels can cause the water level outside the
gate to be higher than the normal trunk setpoint. The normal
diurnal tidal range is about 11 feet. A prolonged storm can
increase the water level several feet above normal. The dif-
ference between the record high and low tides is over 19 feet.
Pumping Station Controls
Pumping station controls were designed to operate either
constant speed pumps with on-off control or adjustable speed
systems. All but one of the pumping stations designed for the
Municipality had adjustable speed operation for most of the
pumps. One station has engine-driven pumps; one station has
engine-generator drive for the largest unit; several stations
have eddy-current couplings and several have liquid rheostat
speed control.
75
-------
Alarm Monitoring
Regulator stations/ outfall stations and pumping stations
were originally equipped with varying degrees of annunciator
supervision to aid in detecting station problems. A tone tele-
metry system installed in 1965 was designed to give simple but
effective reporting of two levels of trouble. Alarms were
classified and connected as "priority 1" and "priority 2"
status. Priority 1 alarms included power failure and in some
stations, a high wet well condition. This system, termed
"Metrotel," is still operating, but will be phased out in favor
of the broader scope CATAD data collection system. Comparisons
between Metrotel and CATAD monitoring capabilities are shown in
the Operating Manuals, Volume III of this report.
OPERATION
Regulator Stations
The level signal from the upstream bubbler in the inter-
ceptor was (and is) the measured variable of the electric con-
tact controller. The controller allowed the regulator gate to
be wide open as long as the backwater interceptor level was
below the instrument setpoint. However, if the interceptor
level measurement reached the setpoint (plus a small differ-
ential or dead-band amount) the gate would begin to close.
The gate would then either continue to close or modulate until
the interceptor level arrived at a stable setpoint value. The
gate might be in any position between fully open and fully
closed to meet the setpoint requirement.
The setpoint was adjusted by Metro to approximately meet
the expected interceptor flow requirements at the particular
station. Since the Elliott Bay interceptor is the same dia-
meter (42 inches) for several miles, the regulator farthest
upstream (Norfolk St.) had a low setpoint (percentage full)
and Brandon Street, the most northerly pre-CATAD station, had
a high setpoint adjustment. It was not practical to readjust
the setpoint for short term changes in weather conditions,
so settings were picked for "eyeball" system balance, assisted
by design calculations of system hydraulics.
Outfall Stations
The original control system utilized a single controller
with auxiliary pneumatic control instrumentation equipment.
The controller was a'Foxboro large case electric contact unit
with a pneumatically-moved setpoint. Trunk level was measured
by a bubbler system and the reaction pressure fed to the instru-
ment as the measured variable, which the instrument compared
76
-------
with the setpoint. If the measured variable (trunk level) was
be.low the setpoint, control relays would close the outfall gate
or, if it were already closed, would maintain the closed posi-
tion. If the level exceeded the 'setpoint, the gate opened
until the discharge flow lowered the. level to the setpoint.
The gate could modulate about the .setpoint, but sufficient dead
band was provided to prevent continual hunting.
The pneumatic setpoint was adjusted by a manual pressure
regulator to a value considered safe for ordinary high water
storage under any storm (flow) conditions. A pneumatic mea-
surement of the upper range of the tidal level was made for
each station affected. This signal was compared with the fixed
setpoint signal by a sensitive pneumatic selector relay. If
the tidal level approached within six inches of the manual set-
point level, the selector relay would substitute the tide level
signal for the manual setpoint. A further increase in tide
level would result in a corresponding increase in the setpoint
signal to the gate controller, thereby maintaining a six-inch
differential between the gate controller setpoint and the tide
level. Under these conditions, the gate might be open (or par-
tially opened) and only an outward flow could occur. The system
proved very reliable and effective.
Pumping Stations
Pumping station control system operation differs too
greatly from station to station to discuss in other than
general terms. Station design criteria generally required
adjustable speed drives and pumps were programmed in a lead-
follow sequence allowing the influent sewer and set well depth
level to vary in direct proportion to the flow. This allowed
the influent sewer to be used for storage while maintaining a
reasonably constant velocity in the sewer to provide adequate
scouring.
At some stations limited usable storage in the upstream
sewers, or Metro operational decisions/ have introduced con-
stant-level control systems. This is a minor controller change
for stations with adjustable speed drives.
DESIGN CONSIDERATIONS
Storage Availability
The purpose of the CATAD system is to minimize overflows
by utilizing available storage existing in the old trunk sewers.
A study was made to determine the safe storage capacities of
each major segment of the collection system. A summary of
these data is listed in Table 3. Those stations with suffici-
77
-------
ent storage were labeled "command" station so that special
remote control instruments could be installed to insure opti-
mum utilization of that storage.
Pour stations/ West Michigan Street Outfall/Regulator/
Harbor Avenue Outfall/Regulator, King Street Outfall/Regulator
and Denny Way Local Outfall/Regulator, were found to have no
usable storage. The Duwamish pump station was also designated
a "non-command" station partly because of limited storage, and
partly because maintenance considerations require good scouring
sewage velocities in an inverted siphon upstream of this station,
Interface with.Telemetry System
Instrumentation for the necessary analog and status infor-
mation for transmittal was determined by studies and tabula-
tions of both common and unique station requirements. Remote
terminal requirements were first delineated. The tabulation
and study resulted in five categories of regulator-outfall
station remote stations and five pumping station categories.
Definitions and functions of remote station equipment listed
in the specifications are:
(1) Functions of Equipment — The equipment will perform the
following basic functions:
(a) Periodically collect operating data from remote
stations and transmit contact position and feedback quantities
from the remote stations to the computer.
(b) Transmit control signals from the computer to remote
stations and transmit said data to the central computer.
All functions shall be under the control of the central
computer.
(2) Classification of Remote Terminals
(a) Purpose of Classification — The number of data
items to be collected and transmitted from a remote station/
and the number of control points at the station, will depend
upon station operating functions. For the purpose of this
work, equipment for pump stations and regulator outfall
stations has been grouped on the basis of input and output
requirements. The objective of the specifications is to
procure modular equipment so that, insofar as practicable,
equipment at stations of all classifications will be identi-
cal.
78
-------
(b) Regulator-Outfall Stations — Five remote station
classifications for regulator-outfall stations have been
assigned:
1. Type R01. Includes combined outfall regulator
stations with storage capability. Storage regulation is accomp-
lished by direct control of the regulator gate position using
raise-lower contacts in the motor control circuits to regulate
the amount of flow diverted from the trunk to the interceptor.
Indirect control of the outfall gate is effected through con-
trol of trunk water level setpoint by raise-lower contacts in
circuits controlling pulsed input to a stepping motor.
" 2. Type R02. Includes separate regulator-outfall
stations having storage capability and provided with common
telemetry equipment. The distance between the outfall and
regulator stations is not greater than 1,000 feet. The Muni-
cipality installed a multipair individually shielded cable
between the outfall and regulator stations so that a single
TCU could be installed at the outfall station. Control opera-
tions are the same as ROi remote stations.
3. Type R03. Includes combined regulator-outfall
stations having no storage capability. Remote control is
accomplished through direct control of the regulator and
outfall gate positions via raise-lower contacts in the gate
motor control circuits.
4. Type RO 4. Includes the Denny Way regulator
outfall at which two trunk lines feed into the interceptor.
One trunk line has storage capacity, the second line has no
storage capacity. The position of the diversion gate from
each trunk to the interceptor is directly controlled. The
outfall gate for the local trunk (no storage) is directly con-
trolled. The outfall gate for the Lake Union trunk (with
storage) is operated through control of the trunk water level
setpoint.
5. Type R05. Includes regulator stations having
storage capability with no outfall. Storage regulation is
effected by direct control of the regulator gate position.
(c) Pumping Stations — Five classifications for pumping
stations have been assigned:
1. Type PI. Includes pumping stations with three
electrically-driven pumps having storage capability in the
conduit upstream of the station. Storage regulation is accomp-
79
-------
lished by indirect pump control through a simulated water level
signal to existing control equipment in place of the local con-
trol signal from a wet well level bubbler device.
2. Type P2. Includes pumping stations with three
electrically-driven pumps and having no storage capability in
the conduit upstream of the station. The equipment is used
only to monitor station variable and alarm conditions.
3- Type P3. Includes only the Interbay pumping
station with three gas engine-driven pumps having some storage
capability in the conduit upstream of the station. Control is
effected in the same manner as Type PI station.
4. Type P4. Includes only the Matthews Park pumping
station. This station has six electrically-driven pumps.
Electric power is supplied internally from diesel engine gen-
erators or from an external power supply. Control is accomp-
lished in the same manner as for Type PI stations.
5. Type P5. Includes pumping stations with four
electrically-driven pumps and having no storage capability
in the conduit upstream of the station. The equipment is used
only to monitor station variable and alarm conditions.
The foregoing classifications determined the instrumenta-
tion requirements in each station. It was decided at the time
the equipment was specified that the computer system manufac-
turer would assume responsibility for receiving and analog and
contact status signals at receptacles located on Metro's panels
near each TCU. Similarly, outgoing command signals would be
delivered by shielded cable to Metro's receptacles. (During
construction the interface point was moved to receptacles on
the TCU in order to provide a cleaner TCU unit.) The decision
meant that a system responsibility point was established with
Metro responsible for delivering the necessary signals at the
required voltage levels to the TCU. This interface allows
computer system trained people to be responsible for computer
equipment and facilities; and electrical and instrumentation
system personnel to be responsible for contact status or
analog signals from the equipment.
The system contractor/ Philco-Ford Company, built a
station: simulator to synthesize station analog signals or
contact closures. Concurrently, Metro constructed a TCU
simulator to check out the station signal prior to connection
with the TCU as explained in the calibration section.
80
-------
This common interface point for all stations permitted
Metro to prepare the stations to connect to the TCU's soon
after the CATAD contractor finalized his TCU design.
Integration of Local and Remote Control Functions
The first phase of modification involved system design
for control of the regulator and outfall gates. The second
phase involved purchase of electronic gate controllers for the
outfall gate controls (OGC's), and the third phase involved
construction work to install OGC's and auxiliary contacts and
relays.
The prime system design criteria were to allow the cen-
tral computer to control gate movement at any or all times/
and to allow each station to return to its own control system
in the event of telemetry link failure.
These criteria posed no problem for regulator gates as
there was no conceivable situation where continuous movement
from some computer-held gate opening position to another posi-
tion required to meet the fixed setpoint could cause damage or
inconvenience to the system or personnel.
The outfall gates, however, posed an entirely different
problem. The primary function of the computer control system
is to optimize use of storage in the large trunk sewers for a
particular condition. The computer-held level might be con-
siderably above the conservative normal setpoint level. During
high flow conditions the computer could hold the gate closed
resulting in a trunk level above setpoint. If the telemetry
link should fail, the excess water level would cause the gate
to open and begin immediate discharge, which might be entirely
unnecessary from the standpoint of trunk inflow. Given a few
extra minutes the regulator gate could discharge the excess
stored water into the interceptor. The concept of ramping the
setpoint signal was developed to solve this problem.
Since the existing pneumatic control did not lend itself
to ramping, a new electronic outfall gate control system was
designed. The final design established a reference setpoint,
set into the chosen electronic control system. When the com-
puter is in control, the central processor moves the controller
setpoint as required. If the computer link fails, local con-
trols compare the momentary controller setpoint with the refer-
ence setpoint, and utilizing pulses that are always available,
move the controller setpoint up or down to achieve balance
between the setpoint signal and the controller setpoint refer-
ence signal.
81
-------
The pulse rate is adjustable with each pulse representing
about one tenth of a percent change in signal, permitting
smooth ramping at any preset rate. The pulsing system operated
very smoothly/ but the pulses and associated relays caused elec-
tronic noise problems as discussed in Section VIII.
Design details of the system control instruments employed
in remote installations are presented in Appendix H.
Monitoring Functions
After the basic control system operating data was estab-
lished in the early stages of CATAD design, station character-
istics were evaluated. In addition to the basic data required
for computer control functions, provision was made to transmit
information which would aid station monitoring. Much of this
data was provided through station alarm contact status super-
vision. Additional data critical to early determination of
localized electrical system failures is also transmitted as
single point contact closures. Direct transmission of criti-
cal alarm information provides Metro operations personnel with
much information that had not been available previously. The
CATAD system will gradually displace the existing alarm system.
^A number of special analog measurements were required in
pumping stations including wet well level, pump speed, force
main pressure, pumped flow, pump program signal level, and
explosion hazard level. Special status points determine which
pump "is operating and which pumping mode is selected (lead,
follow or second follow). Force main valve position is also
monitored where applicable.
Accuracy
The accuracy of all existing measurements was carefully
checked by direct level measurements to bring instrument cali-
bration of the various stations into agreement with the system
datum levels as explained in the calibration discussion. The
specified 1/10 percent accuracy required of the telemetry system
is not achievable with most field instruments. However, mea-
surements were upgraded as much as practicable. Levels are
measured with extremely accurate electronic differential pres-
sure transmitters. Overall measurement accuracy is limited by
the accuracy of the bubbler system which can be affected by air
flow rate and water turbulence. The overall accuracy probably
approaches about two percent of full scale measurement.
82
-------
Some of the original pneumatic gate position transmitters
were unsatisfactory. Later station design uses a potentiometer
to provide position measurement. Although sufficient turns
were provided for 1/2 percent accuracy/ the dead band on both
the pneumatic and potentiometer systems was found to be about
10 percent, making the readings useless. After a study of the
gear train system, Metro suggested that the gate operator manu-
facturer make a slight mechanical change in the final elements
of the train. Potentiometers installed in later gate operators
have provided reasonably accurate position measurements. Over-
all accuracy is estimated at about three percent. Hydraulic
gates required a more complex device to determine the gate posi-
tion. In some installations that gate position transmitter is
operated by a steel tape attached to the gate. The tape is
wound on a spring-loaded reel which encloses the transmitter.
While this system is capable of better than one percent accur-
acy, some operational problems have persisted.
Accuracy of most analog measurements is considered within
three percent of full scale, except for the explosion hazard
reading which in many stations is far outside the accepted
limits of dependable information. Pump speed readings were
initially unreliable, but precise calibration equipment developed
by Metro's engineering staff provided the necessary standard to
allow accurate calibration of most existing tachometers.
Shielding and Grounding
Precautions were taken during design to provide pair
shielded cables for all analog signals transmitted to the TCU.
The systems contractor requested that all analog signals be
grounded at the TCU to shorten the total scanning time. This
request was approved by Metro and all instrumentation system
designs met this requirement. Shields were grounded at the
signal source end only. All enclosures were grounded to the
station ground.
Design Problems
One principal design problem encountered in the analog
instrumentation systems was the lack of internal isolation
(input-to-output electrical isolation) in many units. For
example, the high signal selector used in the outfall control
system is not isolated, requiring addition of isolation amp-
lifiers in the output circuit and one of the two input .
circuits.
Within a few years much of the complex control system now
assembled from standard industrial instrumentation units pro-
bably will be more effectively built to order from solid-state
component parts.
83
-------
CONSTRUCTION
Regulator and Outfall Station Construction
The first phase of regulator station construction was to
provide stations scheduled for new outfall gate controls with
electronic control systems to replace the existing pneumatic
systems (see Figure 25). The primary purpose was to obtain
FIGURE 25
OUTFALL GATE CONTROLLER
early recording levels to aid in programming the CATAD system
which was already under contract. Nine wall mounted outfall
gate controller units were purchased under Contract M23-67:
Supply, Delivery and Calibration of Electronic Control Equip-
ment and Accessories. Each unit contains a 3-pen recorder.
84
-------
A second contract provided installation of the outfall
gate controllers and modifications to both outfall and regula-
tor stations to provide signals to and controls from the tele-
metry control units. This contract, M10-69, CATAD Electrical
Interface Equipment/ covered installation work in nine stations.
Controls were installed in the Norfolk Regulator in 1972 when
the station was rebuilt. Controls were added to the existing
construction contract for Denny Way Regulator/Outfall station
by change order. Other regulator and outfall controls were
incorporated during design stages including controls for King
Street, Connecticut Street/ and Hanford Street No. 2 regulator
arid outfall stations.
Pumping Station Modification
Contract 70-6, CATAD System Pumping Station Interface,
was issued to provide central computer system control and mon-
itoring facilities and connections in Metro's pumping stations.
Six westside pumping stations were provided with control equip-
ment to permit their influent sewers to be utilized for storage
in conjunction with the trunk sewers of the Elliott Bay and
North Trunk systems. Monitoring facilities were provided for
the West Point Treatment Plant.
Seven Renton Division pumping stations were provided with
monitoring facilities only. These stations and the Renton
Treatment Plant report flow, pump speed and level indications
to the CATAD system.
The contract was unique because of the small ancillary
devices required to make proper connections to existing equip-
ment. The contract provided the instrumentation and electri-
cal work design for each station and required the contractor
to supply and install all devices. He was required to quote
a unit labor price for installing and a percentage markup for
purchasing unspecified relays, contact blocks and devices,
and for replacing and rewiring existing devices not applicable
to the new system. This work, closely supervised by the resi-
dent engineer, saved many man-hours of "as-building" existing
facilities prior to bidding. Precise "as-building" of each
station prior to issuance of the contract might have delayed
the modifications by more than a year. Field checks made
during design insured that no major long delivery items were
overlooked. The designer assumed that relay contacts were
available on existing relays as indicated in Metro's office
data. When actual conditions were found to differ^ it was
possible to perform the necessary modifications at a predeter-
mined cost.
All work in pumping stations then under construction was
included in the particular construction contracts.
85
-------
D DIAGONAL TR
(Closed)
FIGURE 26
AUTOMATIC MONITOR STATIONS
86
-------
DUWAMISH RIVER ESTUARY WATER QUALITY MONITOR
Operation of Remote Recording Water Quality Monitor Stations
Remote Water Quality Monitoring Stations are located at
five strategic locations on the Green-Duwamish River (Figure
26). These station samplers telemeter water quality data over
controlled intervals (usually every 60 minutes) to the central
computer where it is stored for processing. Monitor samplers
consist of pumps, lines, reservoirs, flow cells, analyzers,
probes, strip chart recorders and telemetry transmitters. The
sampling units are housed in concrete block houses (Figure 27)
FIGURE 27
BLOCKHOUSE FOR MONITOR
87
-------
at each sampling site to protect them from the elements and
vandalism. The lines, floats and pumps have to be placed so
that they are not readily accessible to the public (Figure 28)
FIGURE 28
MONITOR PUMP AND FLOAT
Surface pumps are attached to floats at least one meter below
the waterline. Bottom pumps are kept one meter from the river
bottom. These are located as close to the middle of the river
channel as possible to assure a steady flow of water to the
monitor station. These stations also include a mini lab and
work bench for manual chemical analysis and minor repair.
88
-------
Constant monitor temperature must be maintained for proper
equipment operation and to ensure good telemetry of data to
the central computer.
Analyzer readings are continually recorded on strip charts
to ensure data availability in the event of telemetry failure.
Data Transmission and Storage
Data is transmitted to the central computer over leased
telephone lines and stored on a permanent magnetic disk file
large enough to hold four days of data. The disk file has this
capacity so that the data can be handled automatically and
retained until it is practical for the computer operator to
copy the file onto cards for subsequent analysis and storage.
Data Compilation
Analysis of the monitor data is accomplished through a
series of computer programs. The first step obtains a listing
of all data for one month/ simultaneously correcting the DO
for conductivity and eliminating data whose test signal does
.not fall within the range 950-1050. The test signal proves
whether or not the remote station transmitter is correctly
adjusted.
Plots are then made of DO and temperature to assist the
operator in detecting malfunctions such as pump outages or probe
failures. These malfunctions are then compared with the regu-
lar weekly maintenance records of the remote sites. Unusable
data is removed from the file, and unstable data due to linear
probe drift is corrected so there are no discontinuities.
Another plot is then made of the corrected data and a statis-
tical summary is produced for regular water quality reports and
distribution to USGS and others currently involved in modeling
the Green Duwamish River estuary.
The corrected data is stored on magnetic tape for future
investigation of long-term trends in river water quality
(discussed in Section VIII) and as historical record.
Data Transmission/ Interface and Storage
Data is transmitted to the automatic water quality mon-
itoring equipment at the CATAD control center over leased tele-
phone lines where, originally, the data was converted to char-
acters printed on a hardcopy printer, and punched into paper
tape. As part of the CATAD contract, this equipment was inter-
faced to the CATAD system computer so that water quality data
could be input directly to the computer.
89
-------
Water quality monitor data are input to the computer via
the direct input/output channel using a specially-designed
interface. The interface processes messages character-by-
character into the computer. Transfer is initiated by raising
a ready signal in the water quality monitor equipment. The
interface equipment receives and stores the first character
in parallel form and interrupts the computer. In response/ the
computer executes a routine which reads the character into a
storage buffer. Each character in the water quality message
is handled similarly until the whole record (including all
data from one station) has been entered into the computer.
Data Use by CATAD
The river monitor data stored on the magnetic disk file
is now available for real-time computer program scanning. The
river data is scanned hourly by a program, arranging it into a
report which is printed out on the console logger (see Figure
29). Under each station name the top line of data refers to
river surface samples; the second line refers to bottom samples.
Column headings beginning with "C" are conductivity levels and
S.R.I, means solar radiation intensity, measured only at the
East Marginal Street Station.
The Water Quality log performs two functions:
1. It helps the console operator decide if overflow prior-
ities should be shifted to prevent discharging high
BOD concentrations to a river area already critically
low in dissolved oxygen.
2. It assists the monitor maintenance personnel in
locating station problems which are highlighted
with special code words.
There have been no critical river situations since this log
has been on line, but the system is capable of providing future
river environment protection.
Other on-line uses of the automatic river monitor data
are being considered. A special cathode ray tube display
would provide instant access to monitor data by either cen-
tral or remote console operators. An alarm lamp and message
printout would provide rapid warning and reaction to any mon-
itor condition or river situation which might require operator
decision or maintenance actions.
90
-------
06/1C/73 17CO V-ATER CTO LITY HOURLY LOG
C24CO C24COC D.O.
C460CC D.O.
FC1TO1-
rrr JT'
^F TT
TFf-T
C60.1
C6C.7
05P.6
KE NT
057.6
051 .1
C6C.4
0114
0392
000 1
00000
.CO 240
146CO
4C760
09 .24
07.92
09 .2P
0&.32
PB S.R.I.
06. P?
07.07 0.01
C7./6
COO 1
0102
37360
41 POO
00240
09 .7P-
10. & 7.
09 .44
08.00
07.23
FIGURE 29
CATAD RIVER MONITOR LOG
Monitor data directly linked to the control model was
one of the original project proposals. Overflow loading effects
would be predicted with a river model which would automatically
adjust overflow priorities in the CATAD control model. Since
tidal estuary modeling complexities are well known, it was
determined that this effort would better be handled by an
agency familiar with this type of activity. As part of a
cooperative agreement, Metro and the regional branch of the
United States Geological Survey will complete development of
the river model. Indicating the difficulty of estuary modeling,
this project was in progress prior to CATAD project initiation
in 1968.
91
-------
SECTION VI
OPERATING PROCEDURES
CENTRAL CONTROL FACILITY
This section summarizes routine maintenance and operational
tasks which keep the CATAD system running at the best possible
performance during a trying developmental period.
Specific details such as logs/ command procedures,, etc./ are
itemized in Volume V of this report.
During normal working hours/ the console operator is res-
ponsible for proper operation as discussed in the following
paragraphs.
Observations of Console Display Data
The console operator has seven cathode ray tubes (CRTs) for
visual display of station data. The console arrangement allows
display of one to seven different stations in a given gepgraphi-
cal segment. By selecting the various segments/ the operator
can display all system stations. The CRT station hydraulic dis-
play (see Figure 30) provides the operator with current infor-
mation about gate positions, trunk and interceptor levels/ control
setpoint positions/ and flow figures for regulator stations; in
addition/ it shows wet well level/ flows, setpoint position and
pump speeds for the pumping stations. Stations are scanned and
station information updated at regular intervals by the computer.
The console operator enters the scan interval at the console so
that the stations can be scanned at intervals of one minute, two
minutes, five minutes, or even ten minutes. The operator also
has two alternate CRT display formats available. The alarm for-
mat displays existing alarms at any monitored station in the
system (see Figure 31). When this format is requested, the
normal station data display is erased and station alarms are
displayed on the appropriate screens. Selecting another segment
(CRT display of other stations) automatically presents the alarm
display format for stations in that segment. The third format
displays total hourly rainfall in hundredths of an inch for
stations with rain gages (see Figure 32). The rainfall display
is automatically updated every ten minutes and reset to zero
each hour. This format simultaneously displays current pulse
counter rates, appropriate overflow volumes in million gallons
per day/ and values for explosive gas concentrations detected
in the station. The display format remains fixed until the
operator requests another format.
92
-------
DENNY WRY L«KE UNION
INTERCEPTOR LUL TRUNK
94 . 5'4
SETPOI-NT LEUEL SETPO
' 96,56 OCTU
UPSTRERM FLOW
- "'.* 25 . 0
DIMERTED ' F L. O k«
• : .- .8.8
OOWNSTRERM FLOW
3 8 » 5
REGULATOR GRT'E
' RCTURL 99.5
P,ES I^RED . 100.0
UNUSED 'STORRGE
OCTUftL 1
DESIRED 109.8®
TIDE l-E.MEL9SB24
TRUNK FLOW
STORED FLOW
OUERFLOW " ' -
FIGURE 30
STATION HYDRAULIC DISPLAY
93
-------
R « I- N I E R R U E N U E
PUMP S T R T I O H
R M T c T l_ ..•- (.
S T N E Q r
WRTER I
H R Z PI R D
K O U T
~OR ROTE
LIMIT
FIGURE 31
ALARM DISPLAY
94
-------
FIGURE 32
RAINFALL DISPLAY
95
-------
Two teletype units in the console room automatically log
data from the remote stations. One teletype records all alarm
and normal messages showing the station name, time the alarm is
received or cleared/ and the alarm description. Priority num-
bers have been assigned to certain alarm messages to indicate
the necessity for immediate or postponed action. The same tele-
type unit also records all command messages entered from the
console. The second teletype unit records hourly station data
logs. This unit also records a storm log showing rainfall and
overflow volumes. This log is printed only when rainfall is
recorded or when an overflow occurs.
The information from the CRTs and teletype units enables
the operator to monitor the remote stations/ and review them for
alarms. The teletypes operate 24 hours a day allowing the opera-
tor to review station data and alarms received when the console
is unattended. After review/ the operator determines which pro-
blems should be checked by maintenance personnel. The operator
will determine/ if possible/ whether problems should be checked
by treatment plant maintenance personnel/ the telephone company/
contract maintenance personnel/ or engineers. The operator can
often suggest possible solutions based on previous experience.
The operator can then test equipment operation to determine whether
repair is complete.
Data Storage
A variety of data forms generated manually and automati-
cally by the CATAD system provide historical and reference in-
formation. The console operator accumulates and stores the
following types of data:
1.
Printed logs from the teletypes — These logs provide
information on station operation and are a permanent
record of all control actions and alarms for the
stations monitored.
During storm situations/ magnetic tape is loaded to
accumulate station data. This tape provides a record
of the effects of various intensities of rainfall on
the system and the effectiveness of control strategies,
The operator collects and stores reference data on the
operational procedures and constraints for stations
within the system. These records provide information
to assist control decisions during storms; and provide
a reference for determining the extent of possible '
remote station problems.
96
-------
4. Manually prepared records indicate, by station, rainfall
and/or overflows during storms. These records are use-
ful for comparing various storms to determine control
action efficiency.
5. Maintenance manuals are provided to facilitate adequate/
effective repair when required. Equipment maintenance
records are kept in remote stations and at the central
control facility. These records describe solutions for
problems which may recur/ and help to determine whether
a piece of equipment causing frequent problems should be
modified or replaced.
Preventive Maintenance and Cleaning
The operator is responsible for general cleaning in the
console and computer equipment rooms. Frequent cleaning is
required to insure a dust-free computer room.
The operator maintains and files a preventive maintenance
checklist for central control facility equipment. The list is
filed when completed by maintenance personnel and is updated as
required. The operator also files reports on routine and emer-
gency maintenance performed on computer center equipment and
certain station equipment. The operator purchases spare parts
as needed/ and maintains the spare parts inventory.
Compiling Reports
The operator prepares reports on problems or situations
within the system. Since the CATAD system records data on many
of the stations within Metro, it provides a source of informa-
tion for solving problems or for justification of actions in-
volving those stations.
The operator also documents additions or modifications to
.the system such as new alarm messages/ procedures for reporting
station problems, or equipment additions such as the recently
added computer center fire protection system.
SUPERVISORY CONTROL DURING STORMS
The supervisory- control mode provides for direct operator
intervention in station controls during a storm situation.
During the first 16 months of around-the-clock operation, super-
visory control was the only alternative to independent station
operation in minimizing or preventing overflows. Supervisory
control requires an operator at all times to operate the console
during storms.
97
-------
The normal procedure for initiating and utilizing super-
visory control is discussed in the following paragraphs.
Alerting CATAD Operator to Storm' Situation
The operator is at the central control facility during regu-
lar working hours or is assigned on-call during nights and on
weekends. Treatment plant personnel monitoring a remote terminal
for alarm messages alert the on-call operator when the storm
begins.
Operator Begins Supervisory Control
The operator first loads a magnetic tape to store remote
station data during the storm. The operator then places the
system on a one-minute scan interval so that station data is
updated and stored on magnetic tape at one minute intervals.
He places stations in supervisory control.
Flow Control and Storage Utilization
The operator then reviews the CRT displays and alarm status
indicators for stations with existing or potential overflows.
Existing overflows are stopped/ when possible, by opening regu-
lator gates and closing outfall gates to divert flow from the
trunk into the main interceptor. Stations in local or independent
control close the station regulator gate when the interceptor
level reaches a designated setpoint, and keep the regulator gate
closed until the level in the interceptor goes below the setpoint.
The station then stores flow in the trunk behind the closed regu-
lator gate. The flow remains in the trunk line either until the
trunk setpoint level is reached and the outfall gate opens/ or
until the regulator gate opens. The console operator, through
supervisory control, can open and close regulator and outfall
gates to take advantage of additional storage capacity available
either in the trunk or interceptor and/ when required, the opera-
tor can command overflow at a selected station to relieve the
system of excess flow. The operator can also control certain
pump stations to increase or decrease the amount of flow going
through the main interceptor at those stations. The operator
puts the maximum flow through the station as soon as possible
to gain additional storage in the interceptor during the early
stages of the storm, and continues pumping maximum flow until
the treatment plant flow capacity is reached.
A number of considerations determine where storage should
be utilized, and when and where to overflow. Primary considera-
tions are:
98
-------
1. How much flow the treatment plant can handle — Plant
flow capacity can vary due to pump or engine problems,
sedimentation tanks out of service, or a high tide level
in Puget Sound affecting the amount of effluent that the
plant can discharge.
2. Location and intensity of rainfall — As a general stra-
tegy, the operator determines the area currently affected
by the storm and attempts to store in non-affected areas.
This procedure allows the maximum flow in the system to
be directed to the treatment plant from the affected area.
Storage in non-affected areas reduces the amount of flow
going to the treatment plant through the entire collec-
tion system, and permits increased flow from the area
receiving rainfall.
3. Tide levels — A high tide level in Elliott Bay or on
the Dusyamish waterway can prevent opening an outfall
gate when an overflow is required. The operator must
anticipate peak high tides and allow for available
storage during periods when some or all the stations
in the system cannot overflow.
4. The operator is provided with listings of upper and
lower interceptor, trunk, and wet well levels. Another
list ranks the stations in order from least harmful to
the most harmful overflow effects on receiving waters.
The operator utilizes these lists to determine storage
capacities, and to determine when and where overflows
should occur.
5. Specific station problems may arise which limit the
operator's ability to utilize supervisory control.
These problems may arise prior to or during the period
when supervisory control is attempted. The operator
may or may not be able to make allowances for them.
Specific problem details are covered in other portions
of this report.
Release of Stored Flow
When the rainfall stops and flows begin to decrease, the
operator can release flow stored in the system. The operator
maintains the maximum allowable flow to the treatment plant
without causing excessively high interceptor levels at stations
from which flow is being released. The operator attempts to
allow overflows only to stations at which the effects will be
least detrimental. The operator may, for instance, stop over-
flows at three stations and intentionally overflow another
station to maintain a balance of flow in the system.
99
-------
Return Stations to Local Control
Releasing stored flow may require the operator to man the
console for two to three hours after rainfall has stopped. The
CATAD operator returns stations to local control when levels
are sufficiently below setpoints to keep the regulator gates open
and outfall gates closed. The operator can leave the console
unattended with some stations still in supervisory control, when
those stations are diverting flow from the trunk into the main
interceptor at a rate which will not result in .dangerous levels
either in the interceptor or in downstream stations. The West
Point operator can return the stations to local control from
the satellite terminal.
AUTOMATED UNATTENDED OPERATION
The control mode alternative to ".supervisory" control during
storm situations places individual stations into the "automatic"
control mode. A given station can be placed into "auto" control
at any time, but only from the central console. Should the need
arise, the West Point operator can, at any time, return the
station to the "local" control mode.
The automatic control mode permits the computer to regulate
flow in the system on a 24-hour a day basis. The automatic con-
trol program logic is presented in Section VII of this report.
Auto control can remain active during dry weather flow periods
when it will not attempt to store any flow in the system. Regu-
lator gates are exercised during dry weather to maintain the
bottom of the gate close to the varying water surface. This
constant activity may be reduced later, but it is a good test
of control program algorithms and provides early indication of
mechanical problems permitting repair before a station must be
employed in a critical storm condition. When a storm situation
develops, increasing flows sufficiently, the system "automati-
cally" begins storage based on maximum flow determined for the
West Point Treatment Plant. As storm intensity increases, auto
control regulates and stores flow to pre-determined safe levels,
permitting overflows when and where they are needed. When flows
decrease, auto control releases accumulated storage to maintain
maximum safe flow levels in the system, thereby reducing or
stopping overflows.
Console operator intervention is limited to changing the
desired maximum flow figure for the West Point Treatment Plant
as^needed, and to changing the "rule curve duration" (time re-
quired for a station to achieve maximum storage utilization) for
100
-------
individual stations. Rule curve durations may be altered with
different storms depending on the storm location and intensity
and its effect on the system as a whole.
Auto control, therefore, has certain advantages over super-
visory control, including:
1. Auto control provides immediate reaction to storm situa-
tions. In supervisory control, the storm situation
must first be recognized by the on-duty treatment plant
crew so that the CATAD console operator can be signaled
by a radio device. Operational difficulties at the
plant can interfere with storm recognition and delay
the initial call to the console operator. On weekends
or nights, the operator must then get to the console
to begin control activities. Alertness and transporta-
tion factors affect the time delay before supervisory
controls actually begin.
2. Auto control reduces the number of "judgment" decisions
made by the CATAD operator. A supervisory control
operator must be trained to spot and recognize potential
problem situations such as dangerously high water levels;
whereas auto control stores only to pre-determined safe
limits which are stored within the computer memory banks.
3. Auto control reduces the chance of CATAD operator error
in control strategy due to fatigue, inexperience, or
lack of knowledge of the system.
MAINTENANCE EXPERIENCE
\
Maintenance tasks for the central control facility cover
all the physical areas shown in the floor plan of Figure 33.
Computer Equipment Room
Most central facility preventive maintenance is- performed
in this area, which contains the main computer and peripheral
equipment required for input and storage of programs and data.
It also contains interfacing equipment for the remote and
satellite station telemetry system. Preventive maintenance
tasks are itemized by a stored computer program which provides
a list for maintenance personnel. Maintenance task ^description
and frequency are taken from the equipment manuals for each
system component. Figure 34 shows a sample program output for
daily maintenance work.
101
-------
3
ft
tf
o
o
(^
fa
CENT
CAT
102
-------
M8NTHLY PREVENTATIVE MAINTENANCE SCHEDULE
MAG TAPE DRIVE (TBP) N* Dt S.
*-CLEAN TAPE TRANISP6RT
i-CHECK REEL H8LDD8HN ASSEMBLY
.-CHECK PH8T8SENSE HEAD ASSEMBLY
CHECK WRITE ENABLE ASSEMBLY
*
-•CHECK REGULATED P6WER SUPPLY
--CHECK CAPSTAN SfRVQ. * 8FF»SET ADJ./ TAPE SPEED* STAKT^STftP RAMP
•s-CHECK ALL TAPE M8TI8N FUNCTI8NS . .
MAO TAPE DRIVE (B8TT8M) N» 0. S« '
. -CLEAN TAPE TRANSPORT
f-CHECK REEL H8LDD8WN ASSEMBLY
-P. CHECK PH8T8SENSE HEAC ASSEMBLY '
CHECK WRITE ENABLE ASSEMBLY
--CHECK REGULATED P8'*ER SUPPLY
--CHECK CAPSTAN SERV8. « 8FF»SET ADJ.* TAPE SPEED* START-STOP »AMP
CHECK ALL TAPE M8TI8K FUNCTI8NS
LINE PRINTER* M8DEL 744C Nt D, S.
---VACUM C8MPLETE HAMMER ARLA —
, -CLEAN AND CHECK VERTICAL F8RMAT UM T DRUM
CARD PUNCH* MBDEL 7160 ~ . N« D« S»
---CHECK F8R BENT PUSHER FINGERS
FIGURE 34
» ..™-~> • •
DAILY MAINTENANCE WORK PROGRAM
103
-------
A good preventive maintenance program should result in fe'wer
equipment problems, and less "down-time" for the control system.
Periodically, equipment failures require repair work. Main-
tenance personnel must have ready access to spare parts and tools.
The availability of such items usually determines how long the
equipment remains "down." Maintenance personnel also need a
current set of manuals for both hardware and software problem
diagnosis. Computer availability has permitted Metro to com-
puterize an inventory program for both spare parts and system
documentation. Sample output from these routines appears in
Appendix K.
Equipment failures often result from either excessive wear
or heat. Temperature stabilization is achieved by an air condi-
tioning system independent of the building air system. This
system maintains constant temperature in the computer area while
providing extra air cleaning capabilities. A regular preven-
tive maintenance routine minimizes failures in this air supply
system. The computer equipment is further protected by cooling
fans in each unit to minimize the risk of overheating. The
computer center fire and environmental protection system monitors
the room temperature to initiate an alarm at 8OOP and to shut
down the central computer at 85°F. During unattended periods/
these environmental alarms are transmitted to the two main treat-
ment plants where the shift on duty has a call-out procedure for
each alarm. Figure 35 shows some elements in the environmental
protection system.
Computer Center Console Room
This area contains the console, mapboard, and two teletype
units. The amount and complexity of equipment in this area is
not as great as in the computer room. Most console room main-
tenance is performed on the teletype units. Monthly lubrica-
tion of the logging and event printers is required, as well as
periodic minor parts replacements. A spare unit is kept on
hand in case either "on line" unit should fail and require
extensive, lengthy repair.
Other maintenance in this area is limited to cleaning filters
for the console fan units, and periodic checks for unusual noises
or equipment failure. Replacement lamps are kept on hand for
the console and mapboard indicator lights.
An adequate stock of spare parts and written maintenance
procedures are the most important elements in keeping an on-line
process control system on-line. Continued operation and reduced
major failure costs easily repay the time and money invested in
a complete and well-supervised preventive maintenance program.
104
-------
**
^•ii^HS
FIGURE 35
ENVIRONMENTAL PROTECTION SYSTEM
105
-------
SATELLITE CONTROL FACILITIES
Surveillance
The two satellite facilities for the CATAD system are located
at each of the main treatment plants (see Figures 22 and 23). Two
facilities/ located in the treatment plants, assure that the sta-
tions in the system will be monitored and surveyed around the
clock. These facilities provide the operator with a cathode ray
tube (CRT) display keyboard to key in commands and a teletype
unit. The CRT displays the same data observed from the main
console, and simultaneously displays the station control mode
and shows alarm messages existing for that station. The tele-
type unit automatically prints alarm messages identical to those
printed on the units at the main console, allowing the operator
to identify station problems and to alert the CATAD operator or
maintenance personnel 'as needed.
Controls
/
Theory of operation for the two satellite terminals is
essentially the same. Both units permit the operator to key
in a coded message and receive a visual display, a printed log,
or a combination of both. Keyboard entry allows the operator
to enter the following basic commands:
1.
2.
5.
Display — This command displays a requested station
or a system of stations on the CRT. The operator can
also request that all data displayed be printed on the
teletype.
Alarm Review — The operator can display, and print if
desired, all current alarms for a single station or
for a system.
Local — The satellite operator can return a single
station or an entire system to local control from
either automatic or supervisory control. *
Test — The test command displays and prints (for
some tests) a pre-determined set of characters. Six
test sequences exercise the CRT and teletype unit to
detect operating problems.
Cancel — Cancels a:previously entered command which
the operator decides not to complete.
106
-------
6. Maximum Flow — Automatic control regulates the
amount of flow entering the West Point Treatment Plant
from stations within that system. Operating personnel
determine the maximum flow, in million gallons per day,
that the treatment plant can process at any given time
and alter the maximum flow value as needed. Automatic
control, in turn, releases or stores flow in the system
to achieve the maximum flow desired.
Maintenance .Experience
Satellite control facility equipment consists of a CRT, a
teletype unit, and the telemetry circuitry to transfer data to
and from the main computer. A preventive maintenance schedule
has been established to check and clean the CRT and teletype
monthly. Minor problems within the integrated circuitry of the
CRT have been easily resolved. The teletype unit has proven
reliable and has operated with few problems. The telemetry cir-
cuit, like others in the system, is utilized many times a day
and has been very satisfactory. If difficulties arise, the
telephone company can isolate the circuit address and perform
what is called a "loopback" to test the continuity of the circuit.
Loopback test results indicate the general location of the cir-
cuit equipment causing the problem and assist in determining
whether the telephone company or our own maintenance personnel
should investigate.
REMOTE STATIONS
The CATAD system is designed to monitor and control stations
in the Metro system. Stations are basically of two types: regu-
lator stations and pumping stations. Regulator stations utilize
a set of mechanically operated sluice gates to control flow from
the trunk line into the main interceptor. Pumping stations are
located at positions along the interceptor to lift the flow to
higher elevations.
Routine Procedures
During dry weather periods the operator is primarily con-
cerned with monitoring alarms from the various stations and
assisting maintenance personnel in checking problems. During
dry weather periods, a number of stations operate independently
of the CATAD control system in a "local control" mode (see Figure
36). A station in local control operates gates or pumps to
regulate flow to specified safe levels. If the flow exceeds
those pre-determined, fixed setpoints the station automatically
stores and diverts excess flow until the levels drop to a safe
point.
107
-------
STATION
CONTROLLERS
COMPUTER
Message
Transmission
Conditioning
Data
Display
Only
Alarm
Response
TOU
ACU
ecu
LEGEND
-COMMANDS
-STATUS 8 DATA
- INFORMATION
LOCAL CONTROL
FIGURE 36
LOCAL CONTROL
The CATAD operator can place a station into "supervisory
control" to take advantage of extra storage capacity in the
trunks and interceptor (see Figure 37). From the console the
operator can monitor levels in the various stations and adjust
gate positions or alter pump speeds to utilize maximum storage.
The console operator contacts the treatment plant main control
room to advise which stations should be placed in supervisory
control. The plant personnel, in turn, alert the various pump
crews as they radio in to check the stations. This practice
prevents the loss of CATAD control in the station and prevents
injury to any person already in the station.
The third station mode is "automatic control" (see Figure
38). In this mode the computer directly controls the gates in
the regulator stations. The"regulator gates are kept just above
the water surface in the trunks until the flow increases suffi-
ciently to require storage. The main treatment plant control
operators are advised of stations in auto control or in super-
visory control to prevent pump crews or other personnel from
causing a loss of CATAD control in the stations. A station can
be left in automatic control 24 hours a day, utilizing the
system to its best advantage in dry weather or storm situations.
108
-------
MODEL STAT
PROGRAM CONTR
•
MTyfiffrfc
Hi
j^COMI
t
Display 8
Command
Decision ^J
LEGEND
COMMANDS
—STATUS 8 DAT/
•4 INFORMATION
££ ..„„ MOTORS
~\
Message ^ .
Transmit and Shielding 8 \ .
Receive Conditioning i t
HJTER-** TCU ~** ACU
* * CCU
1
_i
OPERATORS
CONSOLES
-~^ 1
^
START STATION
HERE SENSORS
* SUPERVISORY CONTROL
t
FIGURE 37
SUPERVISORY CONTROL
CONTROL
JffiJCEJ. STATION • MOTORS
PROGRAM CONTROLLERS RELAYS ETC.
HP i
cow
.*-
Logging a
Override
Alarm
Response "
LEGEND
"COMMANDS
STATUS 8, DAT
•4 INFORMATION
| \ /
f Tronsinit'and Shielding 8 ] '
I Receive Conditioning A / ;
PUTER ^~ A£U
1CU or,
~4 ^ CCU
/
J
OPERATORS
CONSOLES
~~^L /
^X /
START STATION
^ AUTOMATIC CONTROL
A v;
P >
FIGURE 38
AUTOMATIC CONTROL
109
-------
Pump crews assigned to each station check the various
stations daily. Periodic tests are run on the air systems,
emergency power, and hydraulic systems within the stations.
Pump crews and maintenance personnel have been advised of pro-
cedures to follow upon entering a station so that automatic
control or supervisory control is not lost.
Calibrations
A variety of sensing devices is utilized in pumping and
regulator stations. The sensing devices require initial cali-
brations based on known reference data to insure accurate in-
formation about the stations and their equipment. The basic
sensors are:
1. Tachometers — Utilized on pumps to indicate their
speed. Tachometers are calibrated so that the speed
range, from zero to maximum speed, generates a pre-
determined electrical signal.
2. Explosion Hazard Detectors — Designed to detect
specific concentrations of various gases. The units
are tested by exposing them to known gas concentrations
and assuring that the required electrical signal is
generated by the unit. .
3. Air Bubblers— A system of air lines and bubbler tubes
are connected to detect pressure changes within the
bubblers. The units indicate levels in wet wells,
trunk lines, interceptors, and tide levels. Air
bubblers are initially installed at a fixed height
above a known elevation, such as a pipe invert, and
all data generated is referenced to that point.
4. Linear Potentiometers — Gate positions are indicated
by linear potentiometers which produce a known voltage
depending on wiper position on the potentiometer's
inner windings. The voltage is converted to gate
position in terms of percent-open, with a fully open
gate shown as 100.00%.
/•
5. Force Main Pressure Transmitters — FMP transmitters
are pressure-sensitive devices which indicate the pres-
sure from the pumps into a force main. The devices are
initially calibrated with a "pounds of pressure" refer-
ence and operate within a range preset in the unit.
Pressure change on a diaphragm in the unit results in
corresponding changes in voltage output.
110
-------
Most Metro system pumping and regulator stations were con-
structed before the CATAD equipment was installed. Some system
sensors were designed to perform within less critical tolerances
than required for the CATAD system. Some sensors have been modi-
fied, and in some cases, the tolerance limits expected were
enlarged to allow a margin for error. All sensors transmit data
within a fixed span of two to ten volts. Sensor malfunction gen-
erates an alarm message so that the sensor can be recalibrated or
replaced.
Maintenance Experience
CATAD equipment in each station has been divided into two
basic categories to divide maintenance responsibilities. The
first group covers the telemetry control unit and all equipment
contained within it. The second group covers all other station
equipment including the cables leading into the TCU. Responsi-
bility is divided because TCU maintenance personnel know little
about station operation, and plant maintenance personnel are
unfamiliar with computer electronics. Maintenance experience in
these two groups follows:
1.
Telemetry Control Unit — A preventive maintenance pro-
gram has been established to run frequent tests on the
basic components within each TCU in the CATAD system.
The checks periodically result in minor adjustments to
the modem, or frequency transmitting portion, in the
unit. Equipment problems arising within the TCU can
eventually be sensedxby the main computer. .Problems
may arise which require replacement of a power supply
or a micromodule within the TCU. Such problems cause
transmission of bad data, or loss of data transmission
altogether. The CATAD operator can alert maintenance
personnel so that checks can be made on the appropriate
TCU.
Generally speaking, the TCU loses the—command
status of a station when there is a changeicin AC power
to the unit. This can result when pump crews test the
emergency power system in the station or during a surge
in the normal power supply to the station. Emergency
power systems are tested weekly by the pump* crews.
Test procedure is as follows::
(a) Take the station off AC power. ii
(b) Start the emergency generator. The generator will
begin to run after a delay. st
(c) Test the generator and^ishut it off. -z^
(d) Return station to AC power.
Ill
-------
The TCU command relays drop out, causing a loss
of CATAD control, during the period between loss of
station AC power and starting of the emergency power
system. When the emergency generator system is started,
the CATAD operator is able to re-establish station com-
mand and resume normal operation.
Another occasional TCU related problem is loss of
telemetry between the central office and a remote sta-
tion. Telemetry loss can result from a problem within
the TCU or from telephone circuit difficulties. Test
devices built into the system allow the CATAD operator
and the telephone company to isolate the< locaion of the
problem, and determine who is responsible for repair.
Station Equipment — Treatment plant maintenance per-
sonnel perform scheduled preventive maintenance on all
stations within the system. In the process of perform-
ing these jobs, tests are made to detect possible pro-
blems in the station equipment. The CATAD system is
designed to monitor' critical station equipment and
generates alarms when the equipment is not working pro-
perly. . For the most part, problems are relatively minor
and can be resolved easily. Some specific station equip-
ment problems recur or are similar from station to sta-
tion. Once resolved, these problems are more easily
handled when they recur. Typical examples are:
(a) Faulty Linear Potentiometers — Regulator gate and
outfall gate positions are indicated from linear
pots. The wiper position on the windings produces
a position indication of 0 - 100 percent gate open-
ing. Periodically a pot winding will wear, result-
ing in an erroneous gate position transmission.
For example, the gate may be fully open (100 per-
cent) and the indicated position may be 85 percent
open. When a station is in local control, the
gate opens fully and automatically regardless of
the indicated position. Supervisory or automatic
control requires accurate .gate position indication
so that the gate can be regulated properly.
Automatic control issues commands to the re-
gulator gate to maintain a position just above the
water surface in the trunk. As the trunk level
drops during low flows, auto control issues a com-
mand to lower the gate position. Auto control
then checks to assure that the actual gate posi-
tion has moved to within two percent of the com-
manded position. A faulty potentiometer may indi-
cate no change in gate position and auto control
112
-------
reissues the command. The gate actually does
raise or lower depending on the command issued
even though the indicated position does not change.
This can create a critical situation if the regu-
lator gate/ following repeated commands to lower/
moves to a fully closed position/ stops all flow
into the interceptor and raises the level in the
trunk. A rising trunk level/ in turn/ can cause
an open outfall gate and an overflow.
, A revision has been made in the auto control
commanding procedure to prevent such an occurrence.
If after three command attempts there has been no
change in the actual indicated position/ the sta-
tion automatically returns to local control. The
gate opens fully and prevents overflow. The con-
sole operator can later return the station to auto
control.
(b) Faulty Setpoint Controllers — Two mechanical pro-
blems have arisen which have caused loss of set-
point control in both pumping stations and regula-
tor station. These problems are:
(1) Regulator stations — Setpoint controllers in
some regulator stations control the trunk
outfall gate/ and a level below the setpoint
keeps the gate closed. Supervisory or auto-
matic control procedure raises the trunk set-
point levels when and where possible to take
advantage of extra storage capacities during
storms. A difficulty in several stations was
caused by a faulty diode in the setpoint con-
troller. The system could lower the setpoint
level, but could not raise it. In each case,
replacing the diode solved the problem.
(2) Pumping Stations Pumping stations utilize
setpoint controllers to alter pump speeds
when the wet well level raises or lowers.
Both supervisory and automatic control util-
ize setpoint control to directly adjust pump
speeds during storm situations. Increasing
the pump speeds results in lower elevations
increasing storage capacity, and decreasing
the speed begins storage utilization' earlier
than would occur in local control. A set-
point controller problem can cause a loss of
pump speed control for a station in local,
supervisory, or automatic control. The pro-
blem has thus far been detected by a pulse
113
-------
counter rate alarm. The pulse counter rate
determines the. rate at which the setpoint
changes are made. The computer monitors the
number of pulses or counts within a ten-minute
period. When the pulse rate goes to zero/ the
setpoint is not moving and cannot be moved.
Replacing the pulse counter solved this problem.
114
-------
SECTION VII
PROGRAM DEVELOPMENT
GENERAL '
CATAD system programming, like most real-time systems, falls
into two principal•categories:
1. Systems programs
2. Applications programs
Systems programs comprise the general routines for programming
and operating the computer and general-purpose peripheral input
and output devices, as well as special.routines required for the
hardware operation and data management functions of the specific
hardware provided for data acquisition, data display dnd control.
Applications programs comprise the routines related to the spec-
ific-process, including scaling, limit checking, alarming, stor-
ing measured variables, computing dependent,variables from mea-
sured variables, logical process analysis, defining control
sequences, and formatting logs and data displays. Figure 39
shows the organization of the major software elements.
e
SYSTEMS PROGRAMS
Systems programs include the following principal categories:
1. Operating System
2. Operating System Replacement Routines
3. Process Device'Input and Output Interrupt Handlers
4. Real-Time Clock Routines
5. Executive System
6. Debug Facilities
7. Programming Languages
The computer manufacturer generally provides standard soft-
ware for a general-purpose data processing computer system which
includes bulk memory devices and conventional input and output
peripheral devices. The CAIEAD system supplier has developed
extensive software for the computer, including a highly sophis-
ticated operating system intended for real-time application, as
well as an assembler, a Fortran compiler and debugging aids.
From time to time, the supplier releases new software versions
which correct errors in previous versions, or extend system
capability.
115
-------
Public Library
Monitor
FORTRAN Routines
Special CATAD
Routines
Debug
Exe c utive
Station Data Files
Language
Processors &
Batch Programs
Power Fail
Memory Protection
Buffered I/O
Real-Time Clocks
Process Devices
Interrupt Handlers
Executive Levels
and Sublevels
Background
FIGURE 39
BLOCK DIAGRAM - CATAD SOFTWARE
116
-------
Various CATAD program functions required significant operat-
ing system extensions. All extensions have been developed with-
out altering the vendor-furnished software. Thus, future operat-
ing systems released by the vendor may be incorporated into the
CATAD system.without changing the revised operating system or
reprogramming the extensions.
Operating System
The basic operating system used by the CATAD system computer
is the supplier's real-time batch monitor (RBM). The operating
system provided by the computer supplier provides the following
system functions:
1. Multiprogramming
2. Public library subroutines
3. Background checkpoint/restart
4. Overlay loader :»; ;
5. Disk management system
6. Peripheral device input and output
7. Accounting and elapsed-time routines
8. System initialization and creation '
9. Real-time tasks
10. Control commands
Multiprogramming— The operating system permits convenient execu-
tion of multiple foreground real-time control tasks and background
job batch processing. Foreground tasks are protected during exe-
cution of undebugged background jobs by the system's memory-
protection hardware, which prevents foreground area alteration
or transfer of control by a background program to a foreground
area unless specifically allowed.
Public Library Subroutines — The system provides a public library
of reentrant subroutines which may be used by both foreground and
background jobs. Subroutine reentrancy permits a foreground and
background task to be interrupted during execution of the subrou-
tine by a higher-level task which may execute the same subroutine.
The public library for the CATAD system includes most of the
important Fortran subroutines, including floating point arith-
metic subroutines.
Background Checkpoint/Res tart — The checkpoint/restart feature
permits a background job to be interrupted during execution by
a foreground task and stored on the disk. The background area
is placed.in a protected status and may bevused by the foreground
task. This feature is very important to the CATAD system, which
requires the entire core memory allocated to background jobs for
execution of the foreground real-time model task.
f
117
-------
Overlay Loader — The overlay loader permits a large disk-resident
task to be segmented into a group of smaller subtasks which can
fit into a limited amount of core memory. The segments are stored
on the high-speed head-per-track disk system and can be loaded
very quickly into core memory. Since the buffered input/output
system is used for segment transfers from the disk, other tasks
may be processed while the segment is being loaded.
Disk Management System — The disk management system permits use
of the disk system for storage of disk-resident programs and
temporary and permanent data files. Files may be stored either
as sequential or random-access files.
Peripheral Device Input and Output — The peripheral device input
and output routines provide the means for data transfers between
the computer core memory and peripheral devices attached to the
buffered input/output channel. All input and output to the com-
puter system's standard peripheral devices utilize two operating
system routines: M:READ and MrWRITE. These routines initiate
data transfers which, once started, will continue to completion
without further program intervention. During data transfers,
foreground tasks at other interrupt levels or background jobs may
be processed. The task initiating the data transfer may request
an interrupt at the completion of the transfer, which will resume
processing the initial task unless a higher-priority task is
then processed.
Accounting and Elapsed-Time Routines — Routines provide back-
ground ^ob accounting and elapsed*.time measurement. Starting a
job/ the routine logs the start time, user name and account
number. The accumulated elapsed time is logged at the end of
each activity. At job completion, the date and time, the total
background job time, the time used by foreground tasks during
the job and the accumulated idle time within the job are logged.
The total background time for the job is recorded in the account-
ing file on the disk.
System Initialization and Creation — A system generation routine
provides the means for creating a system for a specific hardware
configuration and system application. Many features of the operat-
ing system require dedicated core memory and are only included at
the option of the user. The selection of optional features to be
included in the operating system is made during system creation.
Real-time Tasks — The operating system provides routines to
implement special hardware features provided for real-time systems.
Operating system routines are provided by the computer mahufac*-
turer to implement the following special hardware features:
1. Power on/off
2. Memory-protection violation
3. Memory parity error
118
-------
Control Commands — The operating system provides routines for
processing computer operator commands which initiate loading and
execution of programs.
Operating System Replacement Routines
Special routines were prepared as replacements for standard
operating system routines furnished by the computer manufacturer.
These replacement routines either corrected deficiencies in the
standard routines available at the time or provided extended
capability as required by the executive system and the foreground
and debug systems. Replacement routines were prepared for the
following operating system functions:
1. Power on/off
2. Memory-protection violation
Power On/Off — The power on/off interrupt handlers are designed
to perform these major functions during a power shutdown and
restart sequence:
1. Halt and then restart all active input and output.
2. Save and restore .all hardware register status and
resume execution where left off.
Upon detection of an impending power loss, the power-off
interrupt handler is initiated, and all register contents are
saved for restart. A check of all input/output channels is
made, any active channel is halted, and a wait loop is entered.
A power-off receiver is then executed. It reads and stores the
time for the external real-time clock.
When power returns, the power-on interrupt is triggered,
arid all interrupts are disarmed and disabled. Those interrupts
having connected tasks are then armed, enabled and checked to
see if they were active at the time of power loss. If the task
was active, a special restore routine is connected to the in-
terrupt and triggered. All input/output channels active at power
loss are then set to end-action-pending and error status so that
retries of the transfer can be initiated by the originating in-
terrupt levels. The protection registers are then restored,
and .the power-on message is bid. After the power-on receiver
.is run, the protection-violation and parity-error interrupt levels
are run to completion at the power-on interrupt level if either
or both were active at power loss. The last of these routines to
be run then exits from the power-on interrupt level. A counter
is set and incremented by an internal clock to provide a delay
before exiting for full system recovery.
119
-------
The power-on receiver for the CATAD system is designed to
reset the protection registers for foreground debug if required.
It also triggers all interrupt levels where clock or I/O comple-
tion interrupts might have been lost in the power failure.
The clock time is read from the external real-time clock
and stored and a bid is placed for a lower level program to take
appropriate action depending on the duration of the power outage.
Memory-Protection Violation — The memory-protection hardware
causes a unique hardware interrupt any time an instruction in
unprotected memory attempts to modify or transfer to protected
memory or attempts to execute an input, output or control instruc-
tion. Background programs, and foreground programs using the
foreground debug system, are executed in unprotected memory. The
interrupt initiates a protection-violation task which determines
whether the violation is for any of the following, permissible,
purposes:
1. A valid request for input or output on buffered input/
output channel using operating system routines.
2. A valid request for use of public library subroutines.
3. A valid request for direct inpu*t or output to a process
or direct control instruction device by the foreground
task executes under the debug system.
If the violation does not satisfy any of these criteria,
the task causing the violation is aborted either by the inter-
rupt handler itself or via the foreground debug system.
All requests for program segment overlays and data transfers
to core storage are checked to ensure that they do not extend
beyond prescribed limits.
All requests for input are checked to ensure that buffer
limits fall within prescribed areas. Requests for public library
subroutine execution are checked to ensure that program control
will return to prescribed areas.of core memory after completion.
Process Device Input and Output Interrupt Handling
Core-resident interrupt-handling routines are provided for
each process input and/or output device as follows:
1. Remote station telemetry
2. Satellite terminals
3. Printers
120
-------
4. Console push buttons
5. Water quality monitor
6. Watchdog timeout
Each of these devices is attached to the computer direct
input/output channel/ and all input and output to the devices is
under program control. Each device interface is provided with a
separate hardware interrupt to the computer. The device inter-
face unit signals the computer when data is ready for input to
the computer or when a data output transfer has been completed.
A fast response is necessary for devices which input data
to the computer. Once the data transfer has been initiated by
the computer, the transfer" proceeds, one word or one byte at a
time, until the transfer is complete. The computer must therefore
accept each data word or byte before the next byte arrives and
displaces the preceding word or byte. The transmission by remote
station telemetry of a single message requires a minimum of 26
milliseconds, and of a single character from the satellite ter-
minals, a minimum of 8.33 milliseconds. This precludes any pro-
cessing of the data by the interrupt handler using a disk-resident
routine because the maximum disk access time is about 34 milli-
seconds.
The interrupt handlers therefore store input data in a buffer
and after data transfer is complete, initiate processing by an
application.
Remote station telemetry — The primary task of the remote sta-
tion interrupt handler is to respond to interrupts generated by
the central station telemetry interface upon either completion
of the successful serialization of each word of data to be trans-
mitted to a remote station or upon reception of a word of data
from a remote station. In addition, error recovery procedures
have been added to the handler. This allows multiple attempts
(usually three) to obtain a successful transmission of data to
or from a remote station without requiring intervention by the
application program, which initiates all scan or contact output
sequences and processes the results of remote station activities
after data transfer completion.
A remote station activity will be terminated and restarted,
if any retry attempts are to be permitted, when any of the follow-
ing error conditions are detected:
1. 200-300 milliseconds have elapsed between interrupts
generated by the central station telemetry interface
during a scan or command sequence
121
-------
2. A remote station other than the desired station has
responded to the .Central station telemetry interface
3. Bad data has been received from the remote station.
When a successful scan or command has been completed, or
no more attempts to achieve a successful scan or command ban be
allowed, the remote station is turned off and a program is
initiated to process the results of the scan. If the unsuccess-
ful operation was a command to place a remote station under local
control, an emergency command sequence is used which comprises:
1. Attempt to turn on remote station
2. Start command sequence whether or not station
acknowledges receipt of request
3. Transmit command to remote station.
Satellite Terminals — Separate hardware interrupts for the West
Point and Renton Satellite Terminals are used for processing both
input from the terminals and output to the terminals. Input to
the computer from the device interface is asychronous, an inter-
rupt being generated upon the completion of each character trans-
mission. Output from the computer to the device interface is
synchronous, the interrupt being triggered internally by a rou-
tine initiated by the 2 KHz real-time clock.
The interrupt handler for each satellite terminal completely
directs all input or output once it is initiated by the applica-
tions program. The applications program specifies the input
buffer into which data is to be stored or the output buffer from
which data is to be transmitted, and no further action is required
by the applications program.
During transmission of a series of characters to a satellite
terminal, after transferring each character to the device inter-
face, the routine sets the 2 kHz clock to trigger the interrupt
handler after a delay of between nine and tweive milliseconds.
At this time the driver outputs the next character. When a res-
ponse is expected from the remote terminal, the clock is set for
a delay of 90 milliseconds, and the driver waits for an interrupt
from the hardware indicating that a character has been received.
If the character is not received from the hardware within 90
milliseconds, the clock triggers the terminal interrupt handler,
and the driver terminates the transmission and signals an error
condition to the executive program.
122
-------
The interrupt handler provides three modes of operation;
two for data output and one for data input. The "poll" provides
the means for data input, causing transmission of specific infor-
mation displayed on the CRT at the satellite terminal to the
central station. The "erase/write" and "row address write"
modes may be used for data transmission. The "erase/write"
clears the screen and writes the message starting in the upper
left corner of the screen. The "row address write" does not
erase the screen, but writes the message on the screen starting
with the first position on the indicated row.
Although the two satellite terminals comprise entirely
different hardware configurations, the Renton Satellite Terminal,
being programmable, has been made to emulate exactly the fixed
logic of the West Point Terminal. Interrupt handlers for both
satellite terminals utilize a common reentrant input and output
control subroutine resident in the Public Library but with separ-
ate areas containing data and other information, including the
hardware address of the device, the address of the data buffer
and its size, an indication whether data is being transmitted or
received, and whether the message is to be printed at the satel-
lite terminal.
Printers — A single hardware interrupt is provided from the
device interface for the events and logging printer. The inter-
rupt handler completely directs output of messages to both tele-
types once the message has been stored in a buffer by an applica-
tions program. The applications program specifies the beginning
of the buffer in which the message is stored and the number of
words in the message (two characters per word). All further
action is handled by the interrupt handler.
Console Push Buttons — The interrupt handler for the console
push buttons receives an interrupt from the operator's console
each time a push button switch is closed. The handler reads
the status of the two words defining the matrix and determines
which switch was closed. (If more than one push button was
depressed simultaneously, there is no response to the command.)
The handler then requests execution of the appropriate applica-
tions program operating under the executive system.
Water Quality Monitor Interrupt Handler ~ The water quality
monitor interrupt handler accepts character input from the
water quality monitoring system to the device interface on the
computer input/output channel and places the characters in a
core buffer. Each time the device interface receives a new char-
acter from the water quality monitoring system, it triggers the
computer water quality interrupt and initiates the interrupt
handler.
123
-------
The interrupt handler reads the character from the device
interface and stores it in a 40-word buffer. Checks are made to
see if the buffer is full or if the last character was a carriage
return. If either of these conditions is true, an applications
program is bid to store the data record on a disk file.
Watchdog Timeout Interrupt Handler •— The watchdog timeout inter-
rupt handler is initiated by detection of a malfunction in the
operation of the direct input/output channel by the watchdog
timeout hardware.
The task searches for the read or write direct instruction
which caused the fault. When it locates the instruction, it
encodes a message containing the location and effective address
of the instruction and prints the following message on the com-
puter operator's console:
DIG FAULT: LOCN:XXXX EFAD:XXXX
Ur
Real-Time Clock Routines
Real-time clock routines utilize the hardware real-time
clocks for initiating applications programs at designated inter-
vals, for synchronous output to hardware devices and for batch
processing elapsed-time accounting. The real-time clock routines
are executed by the Counter = 0 interrupt-handling routines,
which set the clock counters to trigger the Counter = 0 interrupt
at multiples of the basic clock interval. The Counter = 0
interrupt-handler tasks and the intervals between execution are
shown in Table 11.
Clock
Table 11. PROGRAM CLOCK FUNCTIONS
Function
Clock 1 Operating system elapsed-time
accounting
Clock 2 Synchronous output to satellite
terminals
Clock 3 Applications program real-time
initiation
Clock 4 Unused (Counter = 0 interrupt
level used for non-clock purposes)
Program
Interval
(milliseconds)
1000
100
124
-------
Clock 1 — is used by the basic operating system for date and
time and for elapsed time measurement for background : job account-
ing.
Clock 2 — is used to trigger the satellite terminal interrupts
for synchronous output and for timeout on asynchronous of char-
acters to or from the device interface.
Clock 3— is used for initiating all clock dependent real-time
programs, for monitoring output to various devices/ and in the
implementation of the program timeout feature of the watchdog
hardware. Output to various devices is monitored by checking
to determine whether a completion interrupt or an input request
has been received within a prescribed time after issuance of
the output command. The principal functions of Clock 3 are as
follows:
1. Issues watchdog hardware control command
2. Remote station telemetry timeout
3. Printer timeout
4. Debug program timeout
5. Maintains 24-hour time-of-day clocks and initiates
programs at one-second intervals.
Executive System
The executive system provides multiprogramming capability
for applications programs and directs input or output operations
by these programs. The executive program also includes facilities
for on-line testing and debugging of applications programs. It
comprises a group of reentrant subroutines which are callable by
applications programs written in either assembly language or in
FORTRAN. .The executive system also performs functions in connec-
tion with the foreground debug system and a system for implement-
ing the program hang-up detection feature of the watchdog timeout
hardware.
Multiprogramming — The executive system provides for concurrent
execution of ten applications programs. Programs are processed
in a priority sequence. The program at the highest level of
priority will be processed first until it is completed or until
it is suspended for an input or output operation.
A subpriority scheme is also provided in which 16 related
or unrelated programs may be assigned to each main level of pro-
gram priority. When more than one request ("bid") is present
for execution of a program at a main level of priority, the pro-
gram at the highest level will be executed first. The execution
of a program at any level of main priority must be completed
before .any other program at that same level of priority can be
executed, regardless of its priority level.
125
-------
Two basic procedures are common in utilizing core storage
for multiprogramming:
1. Dynamic relocation or loading a program at execution
time into any area of core memory which is available
at execution time.
2. Loading each task into a specific area of core memory
assigned to that task.
The CATAD system utilizes the second procedure. The main
levels of program priority are controlled by ten hardware prior-
ity interrupts provided for each main priority level of applica-
tions program (see Figure 40). Programs at each main level of
priority are allocated a specific area of core storage for execu-
tion. The levels of subpriority of various application programs
are under control of an executive routine.
The areas of core memory allocated to each main level of
priority are contiguous, each level occupying an integral multi-
ple of 256 words (one page). The first page for each level con-
tains temporary storage and information used by the executive
routines for execution of programs on that level. The other page
or pages allocated to each level are used exclusively for execu-
tion of the programs assigned to that main level. The program
may be loaded and overlayed by any number of segments in sequence
or may load a subroutine and transfer to it for better utilization
of the available core storage. The storage allocated to the main
level of lowest priority includes the entire area of core allocated
to background jobs. The memory page which contains the information
used by the executive program for this priority level is contiguous
with the area allocated to the next-highest level of priority.
However/ execution of a program at the lowest level requires the
use of the operating system background checkpoint/restart feature.
The main levels of priority are numbered from 0 to 9 and
sublevels are numbered from 0 to 15, with level 0, sublevel 0,
having the highest priority.
Each main level of priority has associated with it a dedi-
cated area in core for COMMON data storage. The COMMON area for
each level is used for input-output buffers, transferring data
between programs, saving data between runs of a program, etc.
In addition, there are' areas in COMMON which contain general
information not limited to a specific level. A program may
access any information in COMMON but may modify only a limited
number of cells (limited by debug considerations) outside of its
own area in COMMON.
126
-------
Main Storage
14592.
26366
36864
Monitor
Executive and
Executive Subroutines
Systems Programs
for CATAD System
Executive Levels
(Applications
Programs Area)
COMMON
Non-Resident Foreground
Background
49151
Executive
Sublevel
Tasks on
Disk
FIGURE 40
CATAD EXECUTIVE STRUCTURE
127
-------
Each applications program (which may perform multiple
tasks) is assigned to a ma'in level of priority and to a sub-
level of priority within the main level. In order to execute
an applications program, a bid is placed for that program either
by calling the EX:SUBID subroutine in the executive giving the
level and sublevel of the program, or by performing the necessary
functions directly. The EX:SUBID subroutine can be called only
by an application program which is executed within the executive.
The EX:SUBID subroutine sets a bit corresponding to the program
sublevel in a word associated with that level and triggers the
hardware interrupt connected to that level. When that interrupt
becomes the highest-level hardware interrupt waiting, the execu-
tive is entered and determines the highest-priority sublevel
which has been bid and initiates the corresponding program.
If the program has remained in core since last executed, no
segment loading is required and control is transferred to the
program immediately. Otherwise, the EXrLDSEG subroutine would
be called on to load the appropriate program segment.
The EXrLDSEG subroutine, which may be called by any sublevel
program as well as being used by the executive, loads the re-
quested program segment using information generated by the operat-
ing system overlay loader. After the required program has been
loaded into the program area, control is returned either to the
sublevel which called EX:LDSEG or to the entry point for the pro-
gram segment.
When execution of a sublevel program has been completed,
it transfers to a routine EX:FINIS, which looks for any current
bids at that sublevel; any such bids are handled in order or
priority. If there are no bids pending, the program exits
through the operating system routine M:EXIT, which performs the
exit sequence and turns off the hardware interrupt.
Because each sublevel runs to completion when initiated, a
long program could delay the processing of a high-priority pro-
gram at another sublevel for an excessive period of time. To
alleviate this program, the running program may bid for itself
and exit. If a higher-priority sublevel at the same main level
of priority is awaiting execution, it will be initiated. If no
higher-priority sublevel has been requested, the program which
was temporarily terminated .will be resumed.
Direction of Input and Output — Subroutines in the executive
system direct input and output to foreground devices on the
buffered input/output channel. Input or output subroutines are
provided for the following devices by the executive system:
1. Fixed-head disk
2. Computer operator's console
128
-------
3. Satellite terminals
4. Logging and events printers
5. CRT displays
The input and output subroutines perform the following
functions:
1. Transform the given arguments from the calling sequence
into the necessary form for the particular driver used
2. Check to see if the requested device is busy and queue
or place the request next in line for execution if it
is busy
3. Initiate the transfer by calling on the operating system
service routine or interrupt handler associated with the
device
4. Suspend the program requesting the input or output while
in progress
5. Check the status upon completion of the input or output
6. .Automatically retry most operations if the first try
was unsuccessful
7. Print an error code if all tries are unsuccessful
8. Unsuspend and restart the sublevel which requested
the input or output.
All calling sequences for input or output are compatible with
programs written in FORTRAN. Programs written in assembly lan-
guage use a FORTRAN-compatible calling sequence. The executive
routine takes the arguments from this calling sequence and trans-
forms them into the form required for the device by the operat-
ing system or by the interrupt handler.
To permit the maximum amount of processing while input or
output functions are being performed, all requests for input or
output operations are queued. Queuing allows better utilization
of input/output devices and ensures the fastest possible response
time for all foreground programs using a particular device. If
a device is busy when input or output is first requested by an
applications program/ an indicator is set. When the current
input or output operation is completed, the indicator is checked
and the highest-priority program requesting the device is then
activated.
129
-------
After an applications program initiates an input or output
operation via the executive subroutine or while waiting in a
queue/ the program is suspended and is resumed when the input
or output operation is complete. Programs at a lower level of
priority may be executed during the input or output operation •,
subject to interruption by a higher-priority task.
The executive input/output routines check each input or
output operation upon completion and/ if unsuccessful/ the opera-
tion will usually be repeated a maximum of three times. If the
operation is still unsuccessful, an error message will be printed
on the computer operator's console. The error message indicates
the time/ level/ sublevel and the location from which the input
or output operation was requested/ and a code indicating the type
of error.
Fixed-head disk and CRT display-error messages are output
each time a failure occurs because both are infrequent and be-
cause the disk is critical to system operation. Teletype and
satellite terminal error messages, which are more common/ are
printed once and will be reprinted only after the failure has
cleared and recurred.
A supplement to the checks made by the operating system has
been found necessary when reading and writing data to the disk.
Occasionally/ a single byte will be lost when transferring data
to or from the disk. This particular failure is detected by
inserting a check word as the last word on each disk sector. A
special routine verifies the check word after reading the data;
and if the check word is not present/ the bad sector is stored
for later analysis/ a message detailing the failure is printed
on the computer operator's console, and the sector is restored
from a backup disk file.
Additional system security is gained by having two tele-
types for the logging functions. Normally one teletype is dedi-
cated to alarm and events printing and the other to hourly logs,
demand logs, and alarm review; but if one teletype should fail/
all logging functions are sent to the other teletype automatically.
The following subroutine calls with arguments provide FOR-
TRAN-compatible calls for input/output to foreground devices.
EX:RADRD (FILE/BUFFER/WORDS/SECTOR)
Reads the specified number of words into the buffer
starting at the given sector of the disk file to which
the operational label is assigned.
130
-------
EXtRADWT (FILE, BUFFER, WORDS, SECTOR)
Writes the specified number of words from the buffer
starting at the specified sector of the disk file.
EX:CHKRD (FILE, BUFFER, WORDS, SECTOR)
Reads the specified number of words into the buffer
starting at the specified sector of the disk file.
The last word is incorrect, the sector is saved, an
error message is printed, and the sector is restored
from a backup file specified in a table.
EXrMESSG (BUFFER, WORDS)
Writes the specified number of words of EBCDIC characters
from the buffer to the keyboard printer.
EX:CRTWT (DEVICE, BUFFER, WORDS)
Writes the specified number of words of control and ASCII
characters to the console logger teletype.
EX:WPTRD (BUFFER, SIZE, WORDS, STATUS)
EX:RTNRD (BUFFER, SIZE, WORDS, STATUS)
Reads characters input by the operator at the West Point
or Renton Satellite Terminal and packs them two ASCII
characters per word in the specified buffer. The number
of words filled is returned with an error status number.
If the number of words filled exceeds the buffer size
specified, error status is returned. The status value
details the type of error that occurred, with zero in-
dicating a successful operation.
EX:WPTWT (BUFFER, WORDS, PRINT, MODE, STATUS)
EX:RTNWT (BUFFER, WORDS, PRINT, MODE, STATUS)
Outputs the specified number of words of ASCII char-
acters from the specified buffer to the West Point or
Renton Satellite Terminal and designates the transmit
mode and whether output by the satellite terminal
printer is requested. The status value details the
type of error which occurred, with a zero indicating
successful completion.
131
-------
Program Hang-Up Detection
The executive system incorporates a procedure for detecting
a program hang-up and implementing the program timeout feature
of the watchdog timeout hardware (see Figure 41). The program
timeout feature, if implemented, requires that a control command
be issued to the watchdog hardware within a timeout interval of
1024 milliseconds. Failure to issue the control command within
the prescribed interval causes closure of a relay which can be
connected to the control circuits of an external device to acti-
vate or deactivate the device. In the CATAD system, the relay
is connected to the remote station telemetry circuits in such a
way that closure cuts off the signal carrier to remote stations,
which transfers the stations to local control via carrier detec-
tion hardware.
To monitor the system for program hang'-ups, the control
command to the watchdog timer must be initiated by a program
at the lowest level of priority. The watchdog program at a low
priority level cannot initiate the control command if there is
a hang-up in a program at a higher level which will not release
control of the processor. Since the execution of a number of
applications programs will require substantially more time than
allowed by the hardware timer, a method was developed using soft-
ware for effectively increasing the timeout interval to three
minutes. The program hang-up detection scheme detects hang-ups
at any level of priority above executive level eight. Executive
levels eight and nine contain routines which may exceed the three-
minute timeout.
Program hang-up detection is implemented via three related
programs which are executed at three different hardware priority
levels. Each program must be executed within a prescribed time
interval.
1. A program at the highest sublevel of executive main
level seven sets a core location used as a watchdog
clock to -1800 arid resets a watchdog indicator to 0.
2. A routine at the highest level and sublevel of priority
in the executive system, which is initiated at one-
second intervals by the real-time clock, sets a watchdog
indicator and requests execution of the program at exe-
cutive level seven at one-minute intervals.
3. A program initiated by a hardware interrupt which is
triggered by the 60 hz real-time clock, checks the watch-
dog clock every 100 milliseconds and, if non-zero,
increments the clock and issues a control command to
132
-------
SOFTWARE
60 HZ Real-Time Clock
HARDWARE
Counter 3
(Interrupt #4)
100 Millisecond
Send RESETS
Signal to
Watchdog
Hardware
(Interrupt #12)rJMiow Hardware j
' to Time Out
• C End ^
Increment
Watchdog
Counter
1 -,
Executive
Level 0
Sublevel 0
(Interrupt #22)
Executive
Level 7
Sublevel 0
(Interrupt #30)
FIGURE 41
PROGRAM HANG-UP DETECTION FLOWCHART
133
-------
the watchdog hardware. If the timer has decremented to
zero, no command is issued, and the hardware will time-
out, closing the relay..
Time out of the watchdog will, therefore, occur under any
of the following conditions:
1. The 100-millisecond clock program does not run within
the watchdog hardware time interval
2. The watchdog routine at executive level seven does not
run within three minutes after requested by the watch-
dog routine at executive level zero
»
3. The watchdog routine at executive level zero does not
request execution of the watchdog routine at executive
level seven within three minutes of the previous request,
Foreground Debug
The foreground debug routines permit the testing and debuqq-
ing of applications programs under real-time conditions while
tested and proven features of the CATAD system continue in normal
operation. The debug system is enabled by the hardware memory-
protection feature of the computer which provides positive protec-
tion to designated areas of core memory without affecting the
execution of programs in unprotected areas.
Because applications programs are executed under the execu-
tive system, the design of the debug system is closely related
to the executive system, and components of the debug system are
included therein.
The foreground delaug system comprises routines which are
incorporated into various parts of the system as follows:
1
2
3
4
5
Memory-protection violation task
Real-time clock routines
Public library routines
Executive system routines
Special non-resident debug routines
Memory-Protection Violation Task — The memory-protection vio-
lation task is initiated when a program being executed in unpro-
tected core memory attempts to modify or transfer control to a
™^f£ arf* ofJ?ore memory or attempts to perform an input or
output operation directly. The memory-protection violation task
nas been described previously.
134
-------
In addition to the normal protection functions, the pro-
grammer may define a list of allowable direct control or direct
input or output operations by the program being debugged so that
a real-time program may be tested for correct input and output
operation. ,
When a foreground applications program is being debugged,
all core memory within the area allocated to programs at the
main level of priority of the program being debugged.and assoc-
iated COMMON storage is unprotected. All the remaining core
storage is protected, including the memory allocated to back- _
ground. Background is protected to ensure that the program being
debugged would not adversely affect either background or fore-
ground programs. Therefore, any active background program is
suspended and saved on the disk until the foreground debugging
process is completed, at which time the background program is
restored from the disk and allowed to continue. While the appli-
cation program is being debugged, it is the only program running
in an unprotected status. ;,; .
Real-time Clock Routines — Two elapsed-time clocks have been
set up to limit the amount of time that the sublevel under debug
can run continuously. One clock monitors the total elapsed time,
including input/output, that the program being debugged runs.
The other clock keeps track of the computation time exclusive
of input/output, that the sublevel runs. If either clock times
out, it generally indicates that a programming problem, such as
an endless loop or a parity error, has resulted from an attempt
to access unavailable core memory, which causes the machine
fault interrupt task to abort the level.
Public Library Routines — Three public library subroutines are
provided for modifying a limited number of cells outside the
unprotected program and COMMON areas. These subroutines set a
word to a given value, set a bit within a given, word, clear a
bit within a given word, and test a bit within a given word.
In addition to allowing modification within the unprotected
program and COMMON areas for programs being debugged, these
subroutines check the address of the word to be modified against
five prescribed modification areas. The limits on these addi-
tional areas are set to encompass only those core memory loca-
tions which may be modified. Each of these routines uses an
executive routine to check the return address of the subroutine,
and verify that the addressed word is within either the program
COMMON area or one of the modification areas.
135
-------
Executive System Routines — Executive routines direct the exe-
cution of the applications program which is being debugged.
Executive routines make the following validity checks for
foreground programs being executed under the debug system:
1. Check arguments of calls to public library or executive
system subroutines for buffer limits outside of areas
prescribed by debug parameters entered by programmer.
2. Check requests for transfer of disk files by the program
against a list of allowable files.
3.
4.
Check running time of program against limits as follows:
30 seconds maximum excluding suspension for input or
output, 3 minutes maximum total running time.
Check the location to which control will be transferred
against prescribed limits after completion of a segment
overlay.
5. Check the number of words to be output via an executive
routine against prescribed limits.
6. Check requests for execution of a protected foreground
program against a list of permitted requests.
Executive routines will abort a program being executed under
the debug system if an error is detected, and will initiate an
error message via an executive routine.
All requests by foreground programs for execution of another
applications program are checked to ensure that the sublevel of
the requested program has not been suspended under the debug
system.
Special Non-Resident Debug Routines — The debug system includes
a number of special-purpose routines which are not included in
the normal CATAD system, but are loaded on request in an area of
core allocated to non-resident routines. These routines perform
the following functions:
1. Set up parameters for programs to be executed under
debug system, including (a) level and sublevel, (b)
beginning and last addresses of program and associated
COMMON areas, (c) allowable core memory modification
areas, (d) allowable devices and disk files, (S) allow-
able control actions, (f) allowable program requests.
2. Activate and deactivate debug system.
136
-------
3. Terminate debug mode .and place program in normal mode.
4. Suspend program being debugged and deactivate debug
system.
5. Provide for loading applications programs in the system
while it is on-line.
6. Provide dump of program registers for program committing
the last protection violation; i.e., the foregound pro-
gram being debugged.
Programming Languages
The specification for the CATAD system required two language
processors to be available on the central computer: a macro
assembler to be used primarily for system programming and other
real-time programming, including all real-time I/O, and a FORTRAN
compiler used both for programming some of the real-time tasks
as well as for background batch processing. Important features
specified for the assembler were input/output pseudo instructions
and the macro instruction capability. Features of the FORTRAN
compiler required by the specification had to satisfy ANSI
X3.9-1966, FORTRAN with a few exceptions to permit compilation
of Metro's existing library of programs and the capability for
interfacing assembly language with FORTRAN. Although the ori-
ginal FORTRAN system consisted of the manufacturer's standard
"Basic FORTRAN" compiler with a specially developed preprocessor
designed to meet the requirements of the specification, shortly
after delivery of the system, an ANSI FORTRAN compiler became
available.
The ANSI FORTRAN compiler can process in-line assembly lan-
guage within a FORTRAN program, though the real-time control
program does not presently use this capability. The interface
in use between the two languages is the mechanism of subroutine
calls to assembly language from FORTRAN, and vice versa. Several
FORTRAN-callable assembly language subroutines have been placed
in the public library, including FORTRAN math routines and rou-
tines to manipulate bits within a word, to convert codes, and
to perform executive services.
Assembler — Principal elements of the three-pass assembler are
symbolic machine instructions, macros, and assembler directives.
These elements are arranged in a hierarchy, with assembler direc-
tives being the most basic. The assembler directives permit
both the generation of machine code for output and the descrip-
tion of macros to the assembler. Macros, in turn, are used by
the assembler to specify machine instructions in symbolic form,
which alone or nested in other macros provide the bulk of
symbolic coding for the CATAD system. Among the special-purpose
137
-------
macros written to expedite programming for CATAD are macros
describing an extended set of conditional branches and macros
which allow text to be prepared in computer-compatible EBCDIC
code but assembled in peripheral-compatible ASCII code.
The assembler frees the programmer from concern about
addressing modes. For storage-addressing machine instructions
or directives, the assembler automatically chooses the appropri-
ate one of several machine-defined addressing techniques. De-
pending on the contents of the location counter/ the base regis-
ter and directives in effect, the assembler may invoke non-
relative, self-relative, base-relative, or indirect addressing
in the object code generated for a statement,
PORTRAIT Compiler — The use of FORTRAN in CATAD real-time pro-
grams is limited to subroutines which perform extensive floating
point calculations on array variables: specifically flow, stor-
age, and auto-control calculations in level eight and the model
in level nine. The ANSI compatible compiler is a one-pass pro-
cessor with an option for listing the output of a compilation
in assembly language form and for conditioned compilation of
designated FORTRAN source statements. All floating point calcula-
tions are performed in three-word extended precision. Integers
occupy one word. The FORTRAN library loaded with FORTRAN programs
contains subroutines defining arithmetic operations for floating
point variables, as well as routines defining all the intrinsic
and basic external functions of the language. All FORTRAN lib-
rary subroutines used by real-time routines are reentrant and
are permanently resident in the public library. This makes them
available for use by both foreground and background programs;
thus saving a considerable amount of core space.
Real-time FORTRAN programs do not communicate directly with
station data files on the disk system, so they contain no FORTRAN
input or output statements. Data is transferred to and from real-
time FORTRAN programs either through argument lists or via the
COMMON area, which is accessible for reference by both FORTRAN
and assembly language programs and is accessible from all levels.
138
-------
REAIi-TIME APPLICATIONS PROGRAMS
All CATAD system real-time- applications programs are executed
in a multiprogrammed mode under the control of the executive system.
The priority assignment of each applications program is shown in
Appendix D.
Real-time programs are described here in terms of their func-
tional application rather than by their priority assignment.
Following is a list of principal functional applications:
1. Executive Clocks
2. Remote Station Telemetry Control
3. Contact Status Monitoring
4. Analog Data Monitoring
5. Alarm Processing
6. Data Storage Routines
7. Console Routines
8. Satellite Terminal Routines
9. Logging Routines
10. Hydraulic -Calculations
11. Supervisory Control Routines
12. Automatic Control Strategy
Priority Assignment
Generally program priority has been assigned on the basis
of urgency, time dependency, frequency of execution, logical
sequencing of execution and program duration.
Tasks which relate to the security of the system, e.g.,
central station contact status monitoring, environmental pro-
tection system alarming, and remote station command control,
are at the highest level of priority. Time dependent tasks,
such as the executive system clock, telemetry system, control
and satellite terminal poll control, are at this same level.
Tasks which are performed frequently and at regular inter-
vals, such as remote station data acquisition, are assigned higher
levels of priority than tasks which are performed on a random
basis such as console or satellite terminal operation.
The logical sequence of execution has also been considered
in priority assignment. Thus, remote station data acquisition,
remote station status monitoring, analog scaling, alarm process-
ing and the hydraulic calculations which use these data are
executed in decreasing order of priority.
Tasks which require the longest time for execution, such
as hydraulic calculations and the real-time model, tare executed
at the lowest levels of priority.
139
-------
Executive Clocks
Executive clocks initiate: real-time tasks at prescribed
intervals. Each executive clock consists of a core memory loca-
tion which is incremented at cine-second intervals when activated
by an executive routine initiated by the realrtime clock inter-
rupt handler. A clock is activated by storing a negative integer
at the core memory location. The clock will timeout, i.e.,
increment to zero, after the number of seconds specified by the
integer number have elapsed.
Four executive system clocks are provided for each remote
station which perform the following functions:
1. Initiate station periodic scans
2. Initiate station data scans after a time delay follow-
ing a station control change command
3. Initiate station data scans during regulator gate
position changes after program-controlled time delays
4. Initiate station data scans during setpoint changes
at pump stations or regulator stations or during
outfall gate position changes after program-controlled
time delays.
The interval for periodic scans may be altered by the con-
sole operator either on a station-by-station basis or on a
system (West Point or Renton) basis.
Executive clocks provided for each of the two satellite
stations are used as time delay clocks in terminal control.
The executive routine also checks the minutes and seconds
clocks and initiates the following tasks at the prescribed
intervals shown below:
Function
Satellite terminal polling
Watchdog timeout program request
Hourly log data freeze
Interval
10 sec.
60 sec.
60 min.
(on hour)
Water Quality log
60 min.
(15 min. after hour)
140
-------
The executive routine which is executed at one-second inter-
vals also initiates the .following tasks each second:
1. Monitor central station environmental system
2. Blink appropriate lamps on console and map display
Remote Station Computation Intervals
Remote stations are scanned at regular intervals'. The in^
tervals for data scanning and computation depend upon storm water
flow conditions. Data at each remote station is scanned at pri-
mary interval and various computations are made at multiples of
the primary interval.
At each primary interval, the following computations dia-
grammed on Figure 42 /are made for each remote station: •
1. The reference voltage is checked against limits
2. Readings of analog sensors are checked against
sensor limits and out of limits conditions alarmed
3. Analog values are checked against control limits and
alarmed if out of limits ""7"
4. Analog values are digitally filtered to dampen wave
action or ,surges and eliminate electrical noise which
has been superimposed on the signals
5. Status points are checked for change and if an
abnormal condition exists, an alarm is output to
an events printer
Secondary computations are made following the completion
of every fifth data scan. The secondary interval computations
which are diagrammed on Figure 43 include:
1. Computing
gate
flow through each regulator or.outfall
2. Computing flow through each pumping station
3. Computing inflow to each regulator and pump station
and upstream storage volumes
4. Checking storage volumes against operating rule
curves and initiation of control action when
required.
141
-------
REFERENCE
VOLTAGE
ANALOG
INPUT
SCALING
SENSOR
LIMITS
OPERATING
LIMITS
DIGITAL
FILTERING
DATA
INPUT
1
STATUS
INPUT
PULSE
COUNTER
INPUT
STATUS
CHANGE
PUMP
STATION
OPE^ATJNG
I
PULSE
RATE
LIMITS
ALARM a DISPLAY
FIGURE 42
PRIMARY INTERVAL COMPUTATIONS
142
-------
PRIMARY
INTERVAL
COMPUTATION
STATION FLOWS
RATING CURVES
INTERCEPTOR
FLOWS
TREATMENT
PLANT „
CAPACITY
INTERCEPTOR
MAXIMUM ALLOWED
FLOWS
TO TERTIARY INTERVAL
COMPUTATION EACH SIX
SECONDARY COMPUTATION
INTERVALS
TIDAL
CONDITIONS
FLOW AND STORAGE
REGULATION
FROM TERTIARY
INTERVAL
COMPUTATION INTERVALS
IF ACTIVE
OUTFALL
GATE SET-i
POINT •
PUMP STATION
SET POINT
REGULATOR
GATE POSITION
iCALC.
CONTROL
ROUTINE
'CONTROL
ROUTINE
CONTROL
ROUTINE
FIGURE 43
SECONDARY INTERVAL COMPUTATION MODULE
143
-------
The tertiary interval computations comprise the detailed
flow routing and storage regulation calculations -diagrammed on
Figure 44 to project inflow hydrographs at each remote station
from measured ra'infall amounts, and adjust rule curves in accor-
dance with the previously described procedure.' Tertiary inter-
val computations are made at six secondary intervals.
Remote Station TeleMetry Control
Remote station data acquisition and control programs per-
form the following functions:
1. Remote station transmission initiation
2. Command starting
3. Command control
4. Telemetry data storage
Remote Station Transmission Initiation — An executive routine
which is a part of the same program as the executive clock rou-
tines initiates three types of remote station scans in strict
order of priority as follows:
1. Contact change requests
2. Command scan requests
3. Periodic scan requests
All scan requests at each level of scan priority are com-
pleted before any requests at a lower level of priority are
implemented.
Command Starting — All commands to cause a change in control
mode or to cause gate movements or setpoint changes are effected
by opening or closing contacts at the remote stations. Changes
in control station mode (e.g./ from local to supervisory control),
gate position and setpoint commands, whether originated exter-
nally by the operator or internally by *a program within the exe-
cutive system, are directed to the command and starter routine.
It is important to channel all requests through this program to
ensure that necessary validity checks are performed and to pre-
vent interference between two commands at a station. Therefore/
the routine checks commands for validity and schedules commands
to be started in order of priority.
Before a command start request is processed/ a check is
made to ensure that there are no contact changes in progress at
the station. A contact change is considered to be in progress
from the time that it has been initiated by the command starter
until it has been determined by the command scan handler that
the change is either successful or incomplete. After either
determination has been made, the command handler will request
the command starter to check for additional command requests.
144
-------
PRIMARY
INTERVAL I
COMPUTATIONS!
MEASURED
RAINFALL
10- MINUTE INT.
UNIT
HYDROGRAPH
SANITARY
FLOW
ESTIMATED
RUNOFF
TRUNK a INTERCEPTOR
FLOWS FROM SECONDARY
INTERVAL COMPUTATIONS
CHANNEL
ROUTING
FLOW
ROUTING
WATER
QUALITY
STORAGE
REGULATION
OVERFLOW
PROBABLE
RULE CURVE
ADJUSTMENT
RULE CURVE
TO SECONDARY
INTERVAL
COMPUTATIONS
FIGURE 44
TERTIARY INTERVAL COMPUTATION MODULES
145
-------
The following command req.ues.ts are valid:
1. Station control change, (e.g., auto/ supervisory or
local control)
2. jOutfall control change (for outfall only)
3. Regulator gate move request
4. Setpoint move request
5. Outfall gate move request
After a control mode change has been requested and trans-
mitted, all gate or setpoint movements currently in progress
are halted because only the command enabling contact data (local
or remote) are transmitted to the remote station. This occurs
even if the station is already in remote control and a new re-
mote control command is requested.
The following conditions must be true before a gate or set-
point movement will be accepted and started:
1. No telemetry or reference voltage alarm may be in effect
2. No gate/setpoint sensor alarm may be in effect
3. Requested movement is more than the control deadband
4. Requested position is positive
5. Requested position is within limits
6. Gate or setpoint movement is not currently in progress.
If all the above conditions are satisfied, the command starter
flags the remote station as having a contact change in progress,
sets up the contact output bits for transmission to the remote
station and requests the remote station telemetry control rou-
tine to transmit the command contacts. If the command cannot
be started because all the conditions are not satisfied, a rou-
tine is requested to update displays and print appropriate
messages.
Command Scan Handler — When a remote station has been commanded
to change from one control mode to another, or a gate position
or setpoint change has been initiated, command scans are issued
via the remote station telemetry control routine. The command
scan handler verifies that the requested contact changes have
been implemented and monitors the progress of the gate or set-
point movement.
If the contact change was requested to change a station
control mode, indicators are set to cause any command in pro-
gress to be aborted, and the new station control state is stored
with other station data.
146
-------
If the contact change was unsuccessful, the command handler
requests the station driver.,to'try to repeat the .contact change
command a maximum of three times. Otherwise, the remote station
telemetry control is requested to process any pending scan re-
quests, and the command starter routine is requested to initiate
any unprocessed command requests.
If the contact change command starts a gate or setpoint
movement, checks are made to ensure that the correct relays have
been set and the appropriate station executive clock is initial-
ized to trigger a command scan at the time that the routine pre-
dicts that the gate or setpoint will reach 50 percent of the
requested movement.
After completion of the command scan, the following checks
are made:
1. Proper operation of remote station telemetry system
2. Proper operation of analog-to-digital converter
3. Proper operation of sensors
4. Proper position of control status contacts
5. Proper rate of movement of gate or setpoint
against limits.
If any of these conditions are not satisfied, the command
handler sets up a command to terminate the operation and re-
quests the remote station telemetry control routine to issue
the command and a "command not complete" alarm is initiated.
If the gate position is within specified control deadband limits,
the command is similarly terminated. If all the checked condi-
tions are satisfied and the requested position is not within
the deadband, the command is allowed to continue and a clock is ,
initialized to trigger a command scan when the gate or setpoint
is predicted to reach 50 percent of the remaining travel and the
above procedure is repeated. If the command is to close a gate
completely, the movement is allowed to continue until the torque
limiter on the gate motor trips the gate close auxiliary relay.
Telemetry Data Storage —; Complete data from a remote station
scan is stored in 24 core memory locations, each containing 10
bits of station data as follows:
Contact status — 8 words
Pulse counts — 4 words
Analog signals —• 12 words
After the remote station interrupt handler has received all
24 words or has tried three times and failed, the interrupt handler
initiates an applications program, which stores the data in a
station raw data file, or, if the scan was unsuccessful, initiates
a station telemetry alarm.
147
-------
Contact Status Monitoring
*
Contact status is monitored at the remote stations and at
the central station by applications programs.
*
I
Remote Station Change Detection — A routine at executive level
one compares the current status of contacts at the station with
the proper status. If an alarm is detected, the routine updates
the alarm status table in the station data file, which is used
by the alarm processor. The routine also indicates the proper
status of lights on the console alarm and status panel.
Central Station Status Monitoring— An applications program at
executive level zero reads the status of central station contact
inputs which monitor the central station environmental protec-
tion system, including the following:
1. Products of combustion detectors (8)
2. High-temperature relay
3. Halon system trouble
4. Fire indicating unit trouble
5. Six volt DC battery trouble.
Detection of a change in status initiates a request for an
alarm program which is also at level zero. If the change is
from a normal condition to an alarm condition, the alarm program
initiates transmission of an alarm message to the central station
alarm printer and to both satellite terminals.
Analog Data Monitoring
Analog data monitoring routines perform scaling, limit check-
ing, and digital filtering of analog signals from remote stations.
Data Scaling and Limit Checking — The analog to digital converter
is checked by a routine at executive level zero which compares a
transmitted reference voltage against prescribed limits.
Two other executive system routines at executive level one
(one for regulator stations and one-* for pump stations) convert
analog voltages to levels, pump speeds and gate positions and
compare these against prescribed limits. A check of pump speeds,
number of pumps running and pump station mode is made at another
priority level. In the event that any of the analog sensors are
outside of prescribed limits, or the pump operation does not check
against programmed pump operation, the routines update the appro-
priate status word in the alarm status file in the station data
148
-------
file for the. alarm processor. The routines, also update the table
in the station data file showing ,the proper status of lamps on
the console alarm and events panel for use by. the console routine.
The scaled values are stored on the converted station data
file.
Digital Filtering — Analog smoothing routines read the results
of data scaling routines from the converted station data file;
certain analog values are smoothed using a first degree regres-
sion on the five most recent values. The filtered analog values
are used for hydraulic calculations and for display on the opera-
tor fs console CRT displays and are stored on the station flow
data file.
Pulse Monitoring
Pulse counters at remote stations record pulse counts from
pulse generators and from rain gages. Each pulse counter is a
9-bit binary counter holding a maximum of 511 counts, rolling over
to zero on the next count after the counter is full. Current
and previous readings of the pulse counter routines determine
the number of pulses received within a ten-minute interval,
correcting for roll over if necessary. •
An electrical spike or power failure may cause the pulse
counter to reset to zero resulting in an erroneous reading. Such
an occurrence can be detected by a reasonability check of the
incoming pulse counts.
Pulse Generators — Pulse generators are incorporated in the con-
trol circuits at a number of remote stations. The operation of
these pulse counters is essential to the proper operation of the
control system and they are monitored to detect any malfunction.
The pulse rate is then compared with allowable limits, and if
outside limits, the alarm status is stored in the alarm file in
the station data file and in the light panel file in the station
data files.
Rain Gage Monitoring — There are currently six rain gages in
use in the CATAD system. They are located at the East Marginal,
Denny Lake Union, Matthews Park, Kenmore, Renton Treatment Plant
and Kirkland remote stations. The rain gages are bucket-type
which tip and dump whenever 0.01 inch of rain is collected. Each
tip generates an electrical pulse which is accumulated in the pulse
counter.
Rain gage pulse counts are transmitted from each station
that has a rain gage whenever the station is scanned. After the
scans are made at ten-minute increments past the hour, these
readings are saved in computer memory. These data are subsequently
149
-------
accessed by a separate program task/ checked for reasonableness
and saved on disk files for use by. both real-time and background
tasks.
Each file record contains one station's, rainfall for the 12
hours preceding the current time of day. The succeeding 12 hours
are set to zero and may be used for predicted rainfall in project-
ing rainfall data. The file record is circular; thus the midnight
reading is the 144th item on the list and the reading at time
0010 is the first item in the list. The rainfall readings are
collected in real-time at ten-minute intervals throughout the
day and saved in the file. Thus, the file will have a 12'^hour
history and space for 12 hours of projected rainfall whenever
it is used.
Alarm Processing
The alarm processing routine is executed at executive level
three and performs the following functions:
Alarm Count Checking — To avoid excessively frequent alarm messages
being logged, some alarms have to be repeated successively before
an alarm message is output. An alarm counter file keeps the count
of the number of successive recurrences of alarm conditions. If the
count is less than a maximum, the alarm request is reset to normal;
if the count is equal to or greater than the maximum count, however,
the request is allowed to stand. The maximum count is generally
one or two.
Alarm Status Change — If an alarm status change is detected after
the current alarm status has been processed by alarm count check-
ing, whether from normal-to-alarm or a return-to-normal, an entry
is made into the alarm table. Each entry in the table is three
words long, comprising six items of data: the station number,
an index to the alarm in the alarm message file, an indication
of a return-to-normal or alarm message, the time of the alarm,
the alarm value (if appropriate) and a decimal factor. This
table provides enough information for the alarm printing programs
to construct and print the message.
Alarm Message Output — The alarm table provides the means for
transmitting alarm messages to the satellite terminal alarm pro-
grams and the events logger alarm program. Within certain limits,
each program can define alarm as required. Since a new alarm
can be added to the table very rapidly, it is important to allow
loggers time to print it without holding up further scans.
150
-------
There are four pointers to the table:
1. The Reiiton Remote Terminal Index
2. The West Point Remote Terminal Index
3. The Events Logger Index
4. The Alarm Table Index
The Alarm Table Index points to the most recent entry to
the table. The other pointers indicate the last entry .processed
by each logging program. If any one of the three alarm logging
programs gets to be 30 entries behind, an alarm table overflow
message is output to the satellite terminal or logger.and the
oldest alarm will not be transmitted to the device.
Data Storage
An efficient procedure for data base management is essen-
tial to real-time processing. The principal data files are as
follows:
1. Station telemetered data
2. Station physical data
3. Station pipe data
4. Water quality data
5. Operating Data Storage
6. Rainfall data
Rainfall data files are discussed elsewhere in this
report.
Station Data Storage — Telemetered data from remote stations is
stored on a new disk file after each step in processing. Although
consuming more file space than a single data file, this procedure
allows asynchronous processing of data from the same station at
different levels of executive system priority.
If programs at different main levels within the executive '\
system were simultaneously but independently processing data
from the same station data file, the files would be located temp-
orarily in separate areas of COMMON storage allocated to each
level. If the program at each level processes a part of the data
and sets up various indicators independent of the other programs,
each program in storing its processed data back on the one disk
file would write over and destroy the changes made by the other
level. To avoid this, a rule has been set up so that two or more
levels may not write to the same disk file.
151
-------
The hierarchy of station files is as follows:
1. Raw data file — contains raw station data from remote
station telemetry
2. Converted data file — contains scaled analog values,
updated table of alarms and updated console lamp status
3. Plow data file — same as 2 above with computed flows,
storage amounts, further updated table of alarms and
updated console lamp status
4. Alarm data file — same as 3 above with pulse counter
data, further updated table of alarms after modification
by alarm counter and further updated console lamp status
5. Hourly log file — same as 4, stored away each hour for
the use of the hourly log program
6. Demand log file — same as 4, stored away when a demand
log is requested.
An alternative procedure (used by the contractor) required
that each step in processing had to be completed before the
next processing step was initiated. However, if one processing
step requires substantially more computer processing time, this
step becomes a bottleneck to computation.
Physical Data Files — A physical data file for each station
comprises station characteristics used in station hydraulic cal-
culations as well as the results of these calculations. The
contents of the physical data file are summarized on Table 12.
Pipe Data Files — Pipe data files contain data on the storage
element associated with each station as well as intermediate
and final results of the storage calculation. The contents of
the pipe data files are as follows:
1. Pipe data - dimension, slope and head loss coefficients
2. Storage rule curve data
3. Storage calculation - intermediate flow and level data
4. Calculated storage volumes.
Water Quality Data Storage — An applications program initiated
by the water quality interrupt handler takes and stores the
contents of the core buffer filled by the handler on a disk
file.
152
-------
Table .12. CONTENTS. OF PHYSICAL DATA FILES
Pump Stations
Local station control program data
Pump performance tables - head vs. discharge at
constant speed
Pump cavitation limits
Head loss calculation data
Force main calibration constants
Bypass weir data
Flow data - calculated and maximum
Raw analog sensor historical data for digital
filtering
Regulator Stations
Geometric data - invert elevations
Gate dimensions
Gate coefficients
Interceptor and return line piping data -
dimensions/ slope and head loss coefficients
Bypass data - overflow weir or gate data
Flow data - calculated and maximum
Raw analog sensor historical data for digital
filtering
Operating Data Storage — An executive system routine stores
station data files of selected remote stations or the entire
system on magnetic tape files for use in background analysis
or system operation. Storage of operating data is initiated
or terminated from the computer operator's console rather than
the system operator's console since this is essentially a back-
ground oriented task and requires demounting and mounting of
magnetic tape reels.
Console Routines
All operator's console and map display board command and
display functions are implemented under program control of the
computer.
1. Lamp blinking
2. Console Control
3. Alarm and Status Display
4. CRT Display formatting
153
-------
5. Station push button handler
6. Command push button- handler
7. Se'gment push button handler
8. Digital data entry
One hundred and fifty push buttons, about 200 lights and
7 CRT screens form the operator's console. From the computer's
standpoint, all push buttons differ ^nly in that they have diff-
erent hardware addresses. Programs are therefore necessary to
differentiate between different types of push buttons and to
initiate the appropriate operation. Decoding of push button
switch closures is done by a core-resident hardware interrupt
handler which initiates the appropriate push button processor.
In addition to processing push button requests, program's
are needed to write CRT formats, update display data, and
initiate control commands. Checks must be made to ensure that
the operator observes specified rules in initiating commands.
A number of miscellaneous tasks are required to process push
button requests, turning on or off appropriate lamps, changing
displays and printing messages.
Lamp Blinking — Lamps on the operator's console and map display
blink (one second on; one second off) to indicate the occurrence
of an unacknowledged alarm or to designate on the map display
the station last selected on the station select panels to which
a supervisory command would be being transmitted. The lamps are
blinked by a routine executed once each second which turns on or.
off the lamps for which a blink request has been made by the
alarm processor or the station select push button handler. Blink
requests are made for:
1. Station push buttons when an alarm occurs at the station
until the station push button is depressed by the
operator
2. Segment push button lower lamp (red cover) when an alarm
occurs at a station within the segment until the push
button is depressed
3. Alarm acknowledge push button when alarm occurs until
the push button has been depressed
4. Map display station symbol red segment when alarm
occurs at station until corresponding station select
push button on the operator's console is depressed
5. Map display station symbol white segment when corres-
ponding station select push button on the operator's
console is depressed.
154
-------
Console Control Panel — The console control panel comprises the
following items:
1. Digital data entry push buttons
2. Numeric projection-type display
3. "Invalid Command" indicating lamp
4. Push buttons as follows:
Lamp Test
Map Display Lamp Test
Console Power.
Alarm Acknowledge
Cancel Entry
Enter
Following entry of any command the program checks the com-
mand for a valid entry sequence and input data limits. If the
command does not satisfy the program requirements, the program
turns on the "Invalid Command" lamp. If the command is accept-
able it is printed on the events printer. The operator may check
the printed command against his intended command and either press
the "Enter" button to activate the command or the "Cancel Entry"
push button to abort the request.
The "alarm acknowledge" push button blinks when a new alarm
is detected. Depressing the push button turns off the lamp and
the audible alarm.
The lamp test push button (and lamp test push button on each
panel) and the console power push buttons are the only non^program
controlled push buttons on the console.
Alarm and Status Display — Alarm and status indications appear
on both the map display and console panels. The following alarm
indications appear continuously on either the console or map
display.
1. The bottom lamp of the segment selection push button
(red) is turned on
2. The red quadrant lamp of the station symbol on the
map display is turned on for any alarm
3. The amber quadrant lamp of the station symbol on the
map display is turned on for a high water level alarm.
The green quadrant lamp of the station symbol is turned on
when the station is in either of the two remote control modes
(supervisory mode or automatic mode).
155
-------
The alarm and status panel on the console .is a selectable
display. Indicating lamps on the alarm and status panel display
the current status at the station selected when the last station
push button was depressed.
The status of all alarm and status lamps on the console
and map display is updated after each station scan.
CRT Displays — The seven console CRT units display alphanumeric
data from a maximum seven of remote stations under computer
control. The Metro interceptor system has been broken down into
segments based on geographic and hydraulic considerations and
data from remote stations within each segment are displayed
simultaneously. The operator selects a particular segment for
display on CRT units by depressing the appropriate segment
select push button. The CRT displays are related to the map
display by turning on a lamp within the map display symbols for
the remote stations included in the current CRT display.
The CRT displays have a maximum capacity of 768 characters
in 24 rows of 32 characters per row. The display capacity is
not sufficient to display all data from each station so that
the operator can select from any of three data groups for the
CRT display (see Table 13).
Table 13. CATHODE RAY TUBE DISPLAYS
Display Type
Hydraulic
Alarm
Miscellaneous
Data
Data
Water levels
Gate positions
Pump speeds
Flows
Storage rates and volumes
Current station alarms
Rainfall
Overflow rate
Pulse generator rate
Explosion hazard levels
The display controller provides capabilities for blinking
individual characters in the display. The blink feature is used
in the hydraulic data display to indicate sensor values or com-
puted values which are in an alarm status. Examples of the
different display modes are shown in Figures 30, 31 and 32.
156
-------
The display mode is controlled by a console command sequence
in which the mode is entered via. the digital data entry panel.
Additional data formats may be: .added conveniently by the pro-
grammer.
The display formats are stored on disk files. To write
the formats on the display/ all seven screens are erased sim-
ultaneously and then filled with appropriate formats for the
stations in the segment most recently selected. The formats
are only written when a new segment is selected or the display
mode is changed. Outputs to the displays which are attached to
the computer buffered input/output channel are made via an operat-
ing system routine which initiates the data transfer from a core
memory buffer to each display which is performed by the channel
controller without further program action. An interrupt signals
completion of the output to the display.
The CRT displays are updated after each station data scan.
Station Push Button Handler — The station push button handler
actually performs or initiates the following functions when
triggered by the push button interrupt handler after the opera-
tor depresses a push button on the station panel other than a
system push button.
1. Turns off the push button lamp previously on in the
same station panel
2. Turns on the lamp in the depressed push button
3. Causes the alarm and events panel to display the
status of the station associated with the push
button
4. Terminates blinking of the push button lamp if
blinking
5. Turns on white lamp of map display station symbol and
sets up blink request
6. Initiates update of CRT displays and status and alarm
panels.
Segment Select Push Button Handler ~ The segment select push
button handler actually performs or initiates the following
functions when triggered by the push button interrupt handler
after the operator depresses a push button on the segment
select panel.
157
-------
1. Turns off the upper, push button lamp (white) previously
on in the segment select panel
2. Turns on the upper push button lamp in the depressed
push button
3. Terminates blinking of the lower push button lamp if
blinking
4. Causes data for the stations within the segment to be
displayed on the seven CRT displays in the current
format mode
5. Turns off white lamps in map display which are on
(except lamp that turned on by station select push
button)
6. Turns on white lamp in map display in symbols for
stations within the segment selected.
Command Push Button Handler — The command push button handler
actually performs or initiates the following functions when
triggered by the push button interrupt handler after the opera-
tor has depressed a push button on the command panel.
1. Turns on the lamp in the push button which was
depressed
2. Checks the entered command for validity, and if
valid initiates printing of the command (where
applicable) and if invalid turns on the "invalid
command."
A tabulation of the type of checks made for each command is
shown on Table 14.
Digital Data Entry Push Button Handler — The digital data
entry push button handler routine reads the push button which
was depressed and causes the appropriate number to be displayed
on the adjacent numerical display. As each digit is entered,
the numbers in the display are shifted one position to the left.
The digital data entry processor saves up to six consecutively
entered digits, and keeps a running sum of the digits entered,
assuming that the digits form a decimal number. If seven digits
are entered consecutively, the "invalid command" light is turned
on and the console is locked out by program until the cancel
entry button is depressed.
158
-------
Table 14.. CONSOLE COMMAND CHECKS
Command Push Button
Valid
Command
for
Station
Type'
Auto Control
Supervisory Control
Local Control
Regulator Gate Position x
Outfall Gate Position x
Setpoint Position x
Diverted Flow x
Maximum Flow x
Pump Discharge x
Alarm Review
Demand Scan
Data Log
Rule Curve Duration
System Time
Scan Interval
Pump Sequence x
Test
Valid
Command
for
System
x
x
X
Segment
Containing
Station
Currently
Displayed
x
x
X
X
X
X
X
Digital
Data
Entry
Within
Limits
x
x
x
x
x
x
x
X
X
X
X
X
X
X
Satellite Terminal Routines
The satellite terminal applications programs were written
to provide the operators at the treatment plants with the
following capabilities:
» '
1. Receiving system alarms
«
2. Monitoring remote station data and status
3. Placing a selected station or system into the local
control mode from either the supervisory or auto-
matic control modes
4. Entering treatment plant data for real-time control
. and for producing reports.
The seven programs/ which access the two satellite ter-
minals, run under two dedicated levels of the executive. The
programs perform the following functions:
1. Poll the satellite terminal periodically to see if the
operator has entered a request. If a request is pre-
sent, this program will decode the request and bid for
the correct program to handle the request.
159
-------
2. Put a designated station or system into the local con-
trol mode
3. Send all alarm and normal messages to the satellite
terminal
4. Display at the satellite terminal a summary of all
alarm conditions at a specified station or system
5. Format and send to the satellite terminal a display
of the data pertaining to a specified station
6. Allow the operator to enter digital data which can be
used to generate 'reports on treatment plant operation.
This program is also used to enter real-time operating
data.
7. Sequence through various tests to check the functions
of the satellite terminal.
The programs were originally prepared for the West Point
Satellite Terminal. However, since the Renton Satellite Terminal
emulates the West Point Terminal, the same programs can be used
with only very minor differences. These differences result from
the physical differences in screen sizes - the West Point display
screen is arranged as 24 rows by 64 characters and the Renton
display screen is arranged as 22 rows by 92 characters. Figure
45 shows a typical West Point console display.
Terminal Polling — This routine is initiated every ten seconds
by the executive clock. If another routine on this level is
not in an uninterruptible condition/ the program will immed-
iately poll the satellite terminal. If the terminal operator
has not entered a data transmission request and the time in
minutes has changed, the date and time will be transmitted to
the remote terminal and, on the hour, it will also be printed.
If only the "ENTER" button at the remote terminal has been
pushed (distinct in that the word count is zero), the poll
program sends an "ENTER COMMAND" message to the remote terminal.
Indicators are set so that no other program will output to the
remote terminal for 30 seconds which will give the operator
time to enter a command.
If the operator has entered a command, the program will
read the command, decode it, check its validity, set flags and
bid for the program which will perform the requested command.
If the entered command is invalid, a message will be sent to
the operator informing him of his error and requesting that he
reenter the command.
160
-------
.
ii,f
tt.i
.
- : -• -.-:- - -
HOttlW
^^^SlUtUS---
TCU SWITCH IH REHOTI*
FIGURE 45
WEST POINT CONSOLE DISPLAY
Control Command — The control command routine permits the
operators at the treatment plants, which are manned 24 hours,
to restore a station or system of stations to local control.
This operation might be performed when the central console is
unmanned and a malfunction occurs at stations in automatic or
supervisory mode.
To put a single station in local mode, the following
..command is entered fay the operator: LOCAL, STATION/ station
name (minimum of four alphanumeric, characters). (Underlined
characters are required; others are optional.) To put an
entire system in the local mode, the command would be:
LOCAL, SYSTEM, system name (minimum of four characters). The
poll routine, when it decodes the local command, indicates each
station which is to be put in local mode and a request is then
placed for the control program.
161
-------
The control program echoes the command for verification by
the operator before it issues the command. If the operator veri-
fies the command, then the appropriate indicators are set and a
request is placed for the command program to put the station in
local control. A local control message is printed at the satel-
lite terminal and on the events printer.
Alarm Routines — The alarm routine is bid whenever any alarm or
return-to-normal messages are entered into the alarm table. Each
new alarm table entry is decoded and a complete message constructed.
The message is then transmitted to the remote terminal where it
is displayed and printed. If-the status from sending the message
is bad, it is resent, up to a fixed number of times, until the
status is good. If the status is still bad after all retries
have been made, a message is printed on the events logger indicat-
ing that the alarm message could not be sent to the remote terminal,
After each alarm message is sent, a bid is made for the poll pro-
gram (so the operator can enter a command) and this program. Re-
bidding for the alarm program allows for several alarm messages
to be output in sequence without completely tying up the level.
Alarm Review — The alarm review program allows the operator at
the satellite terminal to request a display on the CRT and option-
ally print a review of all current alarms at a specific station
or within an entire system. The commands used to initiate this
display are:
ALARM REVIEW,.STATION, station name, PRINT (optional) or
ALARM REVIEW, SYSTEM, system name, PRINT (optional)
The poll routine decodes these commands, sets requests for
the station, or stations, for which an alarm.review is desired
and initiates the alarm review program.
The alarm review program sends out a heading including date
and time and searches for a station with a request for an alarm
review. The table of alarms in the station alarm data file is
analyzed to see if any conditions are currently in alarm and if
no alarms are present, the program continues on to the next sta-
tion. If alarms are present, a message is constructed comprising
the station name, the time of the last scan, and a list of current
alarms. The alarm review program sets a request for its own exe-
cution and when reinitiated processes requests for alarm reviews
for other stations.
162
-------
Station Data Display — The station data display program provides
the operator at the satellite terminal with the ability to moni-
tor field data at any pump or regulator station. The data dis-
played are the same as the hydraulic data displayed on the console
CRT. In addition, explosion hazard, hourly rainfall data, and
lines showing current station alarm and status are included in
the same display.
To display the data from a specific station, or to display
the data from all stations in an entire system sequentially, the
operator enters the command:
DISPLAY, STATION, station name, PRINT (optional) or
DISPLAY, SYSTEM, system name, PRINT (optional).
The PRINT command optionally requests that the displayed
data also be printed. The poll program decodes the command, sets
indicators for the station or stations to be displayed and re-
quests execution of the display program.
The display program checks each remote station until it
finds a request for a display. It then reads the appropriate
display format and the station data file from the disk. Data
values are converted to ASCII format and placed in the display
format. If any of the display values are in an alarm status,
they are set off by brackets. The display characters are then
packed, removing extra blanks, and transmitted to the satellite
terminal. An additional line including explosion hazard and
hourly rainfall is created and transmitted. Additional programs
create tables of current alarms and status at that station and
transmit the alarms to the remote terminal to appear at the bottom
of the display. If there is a request for the display to be printed,
it is coded into the last message sent. When the display of the
current station is finished, the display program sets a request
for its own execution and when reinitiated processes requests for
displays at other stations. If there are no further requests,
the program exits.
Data Entry — The data entry program enables the satellite ter-
minal operator to enter real-time data for use in control and to
enter treatment plant data which can be used to generate daily
reports. The format of the data entry command- is
DATA, xx
where xx is a two-letter code which indicates to the program what
type of data is to be entered. When the poll program decodes this
command, the data entry program is requested.
163
-------
The data entry program verifies that ,the given two-letter
code is allowed/ and, if allowable, acquires information regard-
ing the data to be entered from a disk file corresponding to the
given code. The information includes the number of data items
to be entered, the locations in which the data are to be stored,
the minimum and maximum value for each item of datum, and a string
of characters describing each item. The item description is
transmitted to the remote terminal where the operator enters the
corresponding value and initiates transmission to the central
station. The value received at the central station is decoded,
checked for entry errors and echoed for verification. If the
value is verified by the operator, it is checked against the given
limits and if the value is within limits, it is stored in com-
puter main or secondary storage. The program is repeated for
each item of data until all data in the sequence have been
entered.
Test Routine — This program executes a series of tests or a
single test designed to check the operation of the satellite ter-
minal and its interface with the central computer. The format of
the test command is:
TEST, n
where "n" is the number of the desired test function as described
below:
0 (or no "n" given) - perform all test in sequence
1 - send full screen with bit pattern 01010101 (*'s)
2 - send full screen with bit pattern 10101010 (U's)
3 - send full screen of vertical bars (bit pattern 0111 1110)
4 - send full screen of dashes (bit pattern 00101101)
5 - perform comprehensive test.
Tests 1 and 2 are designed to check the computer linkage to
the terminal. Tests 3 and 4 check the vertical and horizontal
linearity of the display. Test 5 checks nearly all functions of
the device and involves transmitting and receiving the full char-
acter set, comparing the transmitted and received characters and
printing the full character set to check the operation of the hard
copy printer. Test number 0 will perform tests 1 through 5 in
sequence with a ten-second delay between each test.
Logging Routines
These programs execute all logging functions and most events
messages printed on the logging and events printers. The only
messages which are not output by the logging programs are command
messages and certain system malfunction alarms. The logging rou-
tines construct and print the following messages.
164
-------
1. LOGRALRM prints alarm and return-to-normal messages as
they occur
2. LOGRHOUR prints an hourly log of all stations in each
system
3. LOGRHADL prints a summary of any storm activity in
the last hour
4. LOGRDMND prints a log of the current values at a given
station or in a given system at the operators request
5. LOGRDADL prints a summary of storm activity since the
beginning of the hour at a given station or in a given
system at the operator's request
6. LOGRALRV prints a summary of all active alarms at a
station or in a system depending on the request
7. LOGRWQAL (LOGRWQHD) prints the data acquired through
the water quality monitoring equipment.
To allow the highest priority logging programs to run as
soon as possible, each logging program except the alarm message
program exits after printing one or two lines and requests its
own execution, thereby permitting requested higher priority rou-
tines to be executed before completing the remainder of the log.
LOGRALRM - Alarm Logging — This program formats and prints the
alarm and normal messages on the events logger as they occur.
The alarm message is built up from the information in the
alarm table to include an alarm or normal indication, the time
the event occurred, the station name, a standard alarm message
which indicates the priority of the alarm and for a sensor alarm
the converted analog value. After the complete message is con-
structed, it is printed on the events printer. The program prints
alarms until all messages in the alarm table have been printed.
LOGRHOUR - Hourly Log — On the hour, the program which last
updates the station data file stores a duplicate of the data on
another file which remains frozen throughout the hour. When
data for all stations have been frozen, a request is placed for
each station and the hourly log program is bid. The hourly log
program accesses the data so that the printed log shows data for
all stations from the same scan. If for some reason the data for
a station does not correspond to the time of the log, for example,
if the telemetry has failed, the time of the last successful scan
is printed on the log. A sample hourly log is shown in Figure 46.
165
-------
C6/16/73 C2CO V FOI KT SYS HOURLY LOG -
TRKI.VL TRKSFT TIDE OUTFOS
I NTLVL I NTSET RFGFOS
DF2TFR RS
136.98 143.30
135 £9 137.75 28.7
L CITY RS
15C.3C 157.06
113.87 1CC.P
1 V V DF WY RS
101.57 109.96 102.68 -C.I
94.64 96.56 09.9
LOG BE IKY RS
ICC. 93 0.3
96.56 99.9
FT NT RF
ccw: RS
LAN2FR RF
2 HA WORD
1 fc?NFCRD
105 .15
97.76
102.01
97.64
102.66
9F.65
RS
ICC. 90
9F.73
FS
105 .17
102.14
1C4 . 39
102.40
105
101
105
1C2
1C 5
102
1CF
.96 102.09
.35
.98 102.94
.75
J.5 102.03
.75
.05
-0.1
46.4
C
95
. -C
70
0
34
-0
.0
.3
.7
.6
.4
.0
.1
- REGPLATOR STATION
OVRFLO TRKFLO STOFLO
DIVFLO UPS FLO DNSFI.O
0.0 10.0 -1 .1
11 .1 11 .1
O.C
6.4 26.5
O.C
13.5 20.0 3f'.3
0.0 4.5 T.O
4.5
0.0
4.9
0.0
12.3
O.C
C.9
0.0
5 .1
0.0
4
22
10
17
14
36
6
15
.6
• 2
.4
.9
-C.3
27.1
-1
.9
30. 2
U NUSTO
EXPhA 7
0.18
4.9
0.04
0.02
5. 9
0.05
.3
.6
.9
.0
5
25
1
21
.4
.6
.8
.1
C.27
1 .04
FP* B30M RS
103.55
99 .04
105
100
.91 102.10
.40
0
54
.3
.3
MCEICANRS
102.17
100.36
105
101
.93 102.05
J55
—C
47
t>XRFOLK RS
CKFLA N RS
HAIEOR RS
V KIGh RS
PTI1 ECl'IH
106.37
106.26
101 .55
1C1 22
ice. 51
ice. 26
116.41
107 .40
RS
ICC. 47
9f.26
110
106
lor
103
109
.50 1C3 .58
.97
.01 102.^1
.21
105 .96
.13
0
IOC
.4
.0
.1
.0
0.6
50.6
0.1
50.3
C.C
108.37
99
1C2.41
.5P
99
.C
C.4
34.0
0.2
2.6
0.0
3.3
0.0
1 .6
0.0
5 .2
0.1
2.1
0.0
0.4
o.c
2.5
3,
10,
4.
10.
1
3
2.
.3
.0
,0
,7
.2
. 3
5
0.
12.
0.
14.
2,
R.
2
'
3.
C.
2.
5
7
7
0
o
4
j
p
C
0
0.01
0.23
O.lfi
FIGURE 46
HOURLY LOG
166
-------
A feature of the logs is that if a particular item does not
apply to a station/ the space for that item is left blank so that
only meaningful data items appear on the log. The log is construc-
ted to allow a great deal of flexibility in the log format and to
simplify changes.
\
LOGRHADL .- Hourly Storm Log — After data from each station is
logged by the hourly log program, a request is made for the storm
log program. A sample hourly storm log is shown in Figure 47.
The storm log is only printed when there is significant informa-
tion to be recorded.
LOGRDMND - Demand Log — This program prints a log of data at a
station or in a system at the request of the console operator.
To initiate a request for a demand log, the operator pushes the
desired station or system push button, the demand log push button,
and the enter push button in sequence which initiates the demand
log program to print the data for all requested stations.
The log and printing format is identical to the hourly log.
LOGRDADL - Demand Storm Log — For each station printed on the
demand log, a request is. made for that station's storm log. If
there has been any activity and there is significant data at any
of the requested stations, that data will be printed after the
demand log.
LOGRALRV - Alarm Review — The alarm review program prints a sum-
mary of all alarm conditions present at a station or system as
requested by the console operator. The operator presses the
desired station or system push button, the alarm review push
button, and the enter push button in sequence which initiates a
request for the alarm review program.
The alarm review program first sends out a heading including
date and time and then it searches for requests for an alarm review
at a station. When a request is found, the data for that station
are brought in. The station name is printed, and if no alarms
are present, a "NO ALARMS" message is printed. For each alarm
present, the alarm message is printed.
After completing the alarm review for a station, the program
requests its own execution, and repeats the procedure searching
for other requests for alarm review. If.other requests are pre-
sent, they are processed; otherwise, the program terminates.
167
-------
C 6/^5/72 C-4CC FTOPJ LOG
CFJTtFP F£
LfK F.FKfY RF
r.oc DF IPY PS
riNf* PS
CCVN FS
i^ N:-FR KS
2 K? rFrpD FS
IK? ?f CRD PS
r: FS
F VtPC FS
CKFL? b' ES
EFWOK F£
C.C1 C.06
o.cc o.oc
C.C1 0.06
0*04 O.O6
.OVRVOL OVRTOT
C.O 0,00 0.76
5,4
3.0
0.0
o.c
0.0
22,7
0.0
." 5..,? ; "
C.06
o.cc
e.oc
o.oc.
0.00
0.55
0.00
0.04
1 .24
0.03
c . 34
0.23
0.05
O.ol
1 .29
0.14
0.0 O.CO
0.0 C.OO
FIGURE 47
HOURLY STORM LOG
0.02
168
-------
LOGRWQAL - Water Quality Hourly Log — This program prints an
hourly log of the data received by the Water Quality Monitor System
which automatically scans five-remote water quality stations (Renton
Junction, East Marginal, 16th Avenue South, Spokane Street and Kent)
each hour. This interval may be reduced or demand scans may be
initiated which require that the first scan for a station during
the hour be determined.
At fifteen minutes past the hour, the executive clock program
requests the Water Quality Log. The program causes the console
data logger to skip to a new page and print the titles including
date and time and' the column headings (see Figure 29) . The water
quality data file is then searched, starting from the back and
working forward, for the first record for the current hour. Each
record is accessed in sequence from that point for each successive
station. The record is checked for.proper format, and the char-
acters are converted to ASCII text. If there are any errors in
the format or if any characters are unrecognizable, the station
name, the message "BAD DATA," and the converted record are printed
with no reformatting. Otherwise, the characters are moved into
the output buffer in a format that will line up with the printed
headings and the station name and reformatted data are printed.
If no data are available for a station, the station name and the
message "NO DATA" are printed. This program is repeated for each
successive station until data for each station have been printed.
Hydraulic Calculations
One of the major functions of the CATAD system is to convert
analog data from the remote stations into trunk flows, interceptor
flows and pump station discharges so that the flows can be used in
storage calculations and for automatic control. Since these
calculations are comparatively lengthy, they run at a low priority
level to avoid delaying other more time dependent tasks.
Most of the analog data from remote stations is used for flow
calculation. At regulator stations, analog data are received on
gate positions and water levels; at pump stations, data are re-
ceived for setpoint, wet well level, pump speeds and force main
pressure. These data are sufficient for the calculation of flows.
*i
Flows are calculated at regular intervals defined by the scan
interval currently in effect at the station; flows are calculated
every fifth scan on the even five or ten minutes after each hour.
So that no two executive levels write on identical disk
files, newly calculated flows are not displayed on the console
until the next regular or command scan of the station.
169
-------
Regulator Station Flows — Trunk and interceptor flows, diverted
flows and overflows are calculated at each regulator station.
Trunk flows are determined indirectly by totaling computed
amounts of diverted flows, overflows and stored flows. Inter-
ceptor flows are computed from measured interceptor levels
based on steady, uniform flow in the interceptor using Manning's
equation.
Since backwater in the interceptor may extend upstream of
a regulator or pump station, interceptor flow is calculated for
both uniform and non-uniform flow conditions. Non-uniform flow
is calculated by the difference in level between two stations.
The lesser of the uniform and non-uniform flow amounts is used
as the interceptor flow.
Diverted flows and overflows are calculated from measured
values of gate position and trunk, interceptor and tide levels.
Four hydraulic conditions (shown in Figure 48) have been con-
sidered in computing gate flows:
1. Non-uniform, open-channel flow (regulator gate only)
with the gate out-of-flow and diversions controlled
by losses at the return line entrance and by steady
uniform or non-uniform flow in the return line.
2. Non-uniform flow (regulator gate only) with the gate
in the flow and the quantity of diverted flow
controlled by losses at the gate and by steady uniform
or non-uniform flow in the return line downstream.
3. Partially- or fully-submerged orifice or weir flow
with flow controlled by the head difference at the
ofifice.
4. Free orifice or weir discharge.
The downstream head at regulator gates or weirs for
conditions (2), (3) and (4) depend on the level of steady
uniform or non-uniform sub-critical flow in the return line
downstream and on mementum considerations.
The level of steady uniform or non-uniform flow in the
return line below the gate is determined by the level for
steady uniform flow in the return line or a calculated back-
water level from the interceptor. The level .at the downstream
side of the gate is computed from a relationship based on
momentum of the gate efflux and of the steady flow below the
gate. Depending on these conditions, the emerging jet may be
partially or fully submerged or downstream flow may be swept
by the efflux from the gate producing an hydraulic jump
downstream.
170
-------
>/X\\ //A\ //A\\///\V/AV/A\ //K<
NORMAL DEPTH DOWNSTREAM
2. SLUICE GATE WITH DOWNSTREAM CONTROL
1
/A /7A\/yA\//A\//A//A\//A\ //A
£
y/A\ //A\ //A\ //A\/7X\\//A\\ //A\\/7X&/7AX /7A\//CVvy/A\ //AX//AN///
3. SUBMERGED SLUICE GATE
4. SLUICE GATE WITH FREE DISCHARGE
FIGURE. 48
SLUICE GATE FLOW REGIMES
171
-------
Since the quantity of diverted flow through the regulator
gate depends on downstream conditions at the gate and these down-
stream conditions depend on the quantity of flow in the return
line, the diverted flow is calculated by an iterative procedure.
Only conditions (3) and (4) apply to the outfall gate, the
downstream level being the measured tide level.
Pump Station Flows — Pump station discharge is computed by
three procedures:
1. From pump characteristic curves;
2. From differential pressures in force mains;
3. From flow meter measurements.
Pump flows may be computed on the basis of measured wet well
levels and measured pump rotational speed from model or prototype
characteristic curves. Since the pump discharges to a free water
surface, the discharge head may be defined as the difference in
elevation of the impeller center line and the free surface. The
elevation of the free surface varies somewhat with flow but may
be computed from the station geometry. The suction head may be
determined directly and the discharge determined for the pump
speed and head from pump characteristic curves.
At pump stations with long force mains, friction loss in the force
main may be used to indicate flow quantity. Since the pump discharges to
a free surface, a single pressure measurement at the lower end of the force
main measures the head loss. At stations which have been calibrated as
discussed elsewhere in this report, flow is calculated from the calibration
curve which is a function of the single observed pressure value. This cal-
culation is in addition to the calculation based on pump characteristic curves.
Flow meters in the West Point and Renton Sewage Treatment Plant
effluent lines measure the discharge of the influent pumping stations. Since
there is an appreciable time delay between the influent and effluent flows,
the plant has a considerable levelizing effect on flows. Therefore, flow
meter measurements are useful only for long-term comparisons of influent and
effluent flow measurements.
Storage Calculations — The volume of flow stored in the trunk during any time
interval is computed by measuring level changes in the trunk at the regulator
station. Since the regulator station is at the lower end of the storage
reach, a level water surface is assumed for simplification.
Referring to Figure 49, the cross-sectional area at Section A between
depths DI and D2 at the beginning and end of the interval .respectively may be
computed from the trunk sewer geometry. The storage reach length LI at the
beginning of the interval is the distance to Section B at the intersection of
a level surface through DI with the surface of the flow at normal depth Dn on
pipe bed slope Sb and is determined from the equation LI = (DI - Dn)/Sb.
Similarly, the storage reach length L2 at the end of the interval to Section C
'172
-------
is" L2 = (02 - Dn)/Sb. The cross-sectional area of the pipe at
Section B between depth Dn and Dn + (02 - DI) and at a section
midway between Sections A and B may be determined from the pipe
geometry. The storage volume between Sections A and B and
between B and C may be determined by the prismoidal.formula.
The average storage rate during the interval may be determined
from the storage volume and the length of the interval. In
addition to computing storage volume during the interval, the
remaining storage between the current and maximum trunk levels
is also computed. This establishes the time remaining until all
trunk storage has been filled at current storage rates.
The average inflow rate during the interval is the algebraic sum of the
outflow rate through the regulator and outfall gates and the storage rate
during the interval.
Since the actual change in depth during a ten-minute interval is usually
less than 0.1 foot, the computed storage volume is sensitive to transient water
surface disturbances. To eliminate the effect of such disturbances, observed
surface level values are smoothed by regression techniques based on five or
more preceding measured values.
During transient flow conditions immediately after opening or closing
a gate or resulting from a rapid change in inflow, actual conditions may depart
materially from the assumed level water surface in the storage reach. Although
storage rate computations based on an assumed level water surface may result in
substantial errors in calculated storage, such errors tend to average out as
flow stabilizes. Calculated storage volumes are smoothed by regression tech-
niques to damp out transient effects on calculated flows.
=BED SLOPE
FIGURE 49
STORAGE VOLUME MODEL
173
-------
Station Flow Control
Control of flow in the CATAD system is implemented by move-
ment of regulator and outfall gates/ trunk and pump station set-
points. Gate and setpoint movements described in Section V may
be initiated by the console operator by direct position control
or indirectly through use of flow control commands. To assure
the safe operation of the system, numerous checks are made on
console command requests to ensure that only valid commands are
accepted and executed.
Direct Gate or Setpoint Control — A request for a gate or set-
point movement initiates the command starter program. Any gate
or setpoint position may be requested and a movement may be re-
quested whether or not the gate or setpoint is presently in
motion. The command starter accepts all requests, checks the
validity of the request and decides when and whether to proceed
with the command. When a command is already in progress, action
on any new command is delayed until the first command is complete.
Requested Flow Control — Routines are included for producing a
desired rate of flow at a regulator or pumping station. One rou-
tine determines the regulator gate position required to divert
the specified flow and a second routine determines the pump sta-
tion setpoint required to discharge a specified flow at the station.
Every time station flows are calculated, stations with active
requested flows are checked to see if sewage levels have changed
enough to require a new gate/setpoint position.
Regulator Station Requested Flow — A request for a specified
flow at a regulator station is converted to a regulator gate posi-
tion. The program calculates the maximum value of the diverted
flow based on the current trunk level with the gate at the same
level, or fully open if the trunk level is above the top of the
gate opening. If the requested flow is greater than the computed
maximum, the requested gate position is at the level of the trunk
or fully open, whichever is lower. If the requested flow is less
than the maximum, a gate position is assumed and the flow calcu-
lated. If the calculated flow is greater or less than the requested
flow, a smaller or larger gate opening is assumed and the proced-
ure repeated until the calculated flow is within prescribed limits
of the requested flow. A requested.flow of zero is interpreted
as a request to close the gate completely if the regulator station
is under supervisory control. The program will accept any posi-
tive value for the requested flow. A request for a negative flow
is alarmed as an "invalid command." If the regulator station is
in an automatic mode, the rate of opening a gate when releasing
stored flow is limited; and if the interceptor is full, the regula-
tor gate will be closed.
174
-------
Pump Station Requested Flow — A request for a specified flow at
a pump station is converted to request for a specified setpoint.
The pump station control equipment will increase or decrease the
speed of a specific number of operating pumps (operating mode)
in proportion to the wet well level within a specified band (see
Figure 50). When the level rises above this band/ an additional
pump will be turned on. When the level falls below this band, a
pump will be turned off. The pump speed may not be proportional
to the wet well level throughout the band, but may maintain a full
speed for a range of levels at the upper end of the band arid mini-
mum speed for a range of levels at the lower end of the band. The
proportional controller for the new mode of operation then increases
or decreases pump speeds as the level changes within a new propor-
tional band. The number of modes is limited by the total number
of pumps and the requirement that at least one pump be in opera-
tion at any time. •
The pump station setpoint controlled by the computer is input
to the pump controller in place of the wet well level permitting
indirect control of station discharge via the pump controller.
Pump controller operation is simulated by the program which com-
putes the pump discharge at an assumed setpoint level based on the
number of pumps on and the computed pump speed at this level. If
the computed flow is greater or less than the requested flow, the
assumed setpoint is lowered or raised accordingly until the com-
puted flow is equal to the requested flow within prescribed limits.
- The program can determine the maximum and minimum flows for
the current operating mode, and if the requested flow is greater or
less than these flows, a pump must be turned on or off as required.
f
If the pump station is in supervisory control, a flow requested
by the operator is checked against the computed maximum and minimum
flows for the current mode. If the requested flow is outside these
limits, a message is printed on the console showing the maximum or
minimum flow and the corresponding setpoint. If the operator
wishes to change modes, and turn a pump on or off, he may enter
a direct setpoint request which moves the setpoint above or below
the mode change level of the pump station controller. He may then
reenter the requested flow.
If the pump station is in automatic operation, the program
automatically changes pump station operating modes without opera-
tor intervention.
175
-------
INTERBAY PUMP STATION
160 400
140
120 300
too
S I
3E oc
*••» •*-*
£ 80 § 200
5 Q.
< (0
» i
0 2
60
4O 100
20
o o
PUMP SPEED/WET WELL-
I PUMP RUNNING
PUMP SPEED/WETWELL
2 PUMPS RUNNING
DISCHARGE/ WETWELL
2 PUMPS RUNNING
DISCHARGE/WETWELL
I PUMP RUNNING
9OOO 9200
9400 9600 9800
WETWELL (FT. X 100)
FIGURE 50
TYPICAL PUMP CONTROL PROGRAMS
10000
176
-------
AUTO 'CONTROL STRATEGY
Automatic control is a software control state which provides
for unattended control of remote stations by the CATAD central
computer. Whereas supervisory control permits requested flow/
gate and setpoint commands to be entered by the console operator/
only foreground programs may initiate commands under automatic
control. Special control routines/ running at executive level
eight, determine appropriate requested flows and setpoints for
each station in auto control. These flows and setpoints are
affected by various conditions including interceptor flow, trunk
level, trunk storage, tide level and conditions at downstream
stations.
Objectives
The primary objectives of the automatic controls are:
1. To use the maximum storage in tributary trunk sewers
to control flows in the interceptor sewers
2. To use the maximum capacity of the interceptor sewer
to carry combined flows to the treatment plant
3. To limit flows into the treatment plant to the
maximum plant capacity
4. To ensure safe and orderly system operation.
The procedure for regulating storage in trunk sewers is
based on the use of desired storage versus time curves ("rule
curves"). These curves establish the rate at which flow should
be stored in each trunk sewer which/ in effect, determines the
rate of diversion into the interceptor.
Appropriate rule curves are calculated at regular intervals
by the real-time model of the interceptor system. The curves are
affected by the following parameters:
. 1. Direction of storm movement
2. Probable precipitation amounts
3. Probable duration of precipitation
4. Water quality criteria governing overflows.
The operating strategy used for automatic control is based
on general hydraulic considerations and actual experience in manual
operation of the system under supervisory control. Actual experi-
ence demonstrates that maximum use of available storage is
achieved by maximizing the flow through the interceptors subject
to the limitation that total flow into the treatment plant must
177
-------
not exceed the maximum plant capacity. It is only when allowable
interceptor flow has been reached at one or more points in the
interceptor that storing is commenced in the trunk sewers. Stor-
age rates are controlled by storage rule curves determined by flow
routing using the real-time model.
The automatic control procedure was conceived and implemented
with system safety and integrity the paramount concerns. Automatic
control will run, and run safely, without benefit of the hydraulic
model, and can operate safely with arbitrary storage rule curves.
The rule curves are considered tools for optimization, subject to
the limitations imposed by physical and safety constraints.
Remote station control is achieved by using facilities avail-
able to the operator under supervisory control. They comprise:
1. Control of flow diversions into, the interceptor sewer
at regulator stations by direct control of regulator
gates
2. Control of flow at pump stations indirectly through set-
point control (except at Kenmore P. S. )
3. Control of maximum storage in trunk sewers indirectly
through control of high-level setpoints.
Functions 1 and 2 are controlled by the use of the "requested
flow" function currently implemented for supervisory control.
Function 3 uses the "setpoint" command implemented for supervisory
control.
Software Control Modes
Three software modes have been set up to ensure orderly
station control. They comprise:
INACTIVE - the normal dry weather condition
ACTIVE/STORING - during which the trunk is being stored
in accordance with the station operating rule curve
ACTIVE/RELEASING - during which trunk storage is being
released subject to limitations imposed by the per-
mitted interceptor maximum flow and regulator gate
opening rate limit.
Stations are transferred from mode to mode as conditions
permit. To avoid unstable transfers between modes, transfers can
only be from INACTIVE to ACTIVE/STORING to ACTIVE/RELEASING to
INACTIVE and not in the reverse order.
178
-------
Conditions for transfer from mode to mode are as follows:
INACTIVE to ACTIVE/STORING
either: Interceptor flow greater than 100 percent of
maximum allowed interceptor flow.
or:
or:
Any station downstream is active/storing.
Trunk level is within one foot of local setpoint.
ACTIVE/STORING to ACTIVE/RELEASING
Interceptor flow less than 90 percent of maximum interceptor
flow and no downstream station is active/storing.
ACTIVE/RELEASING to INACTIVE
Trunk level is two feet or more below local setpoint.
Description of Operation
Under dry weather conditions, all stations are in inactive
mode: all flow at regulator stations is transmitted without any
storage taking place, and all pump stations are operated to simu-
late local control operation. For wet weather operations, sta-
tions under CATAD control have been classified into four groups,
as described below.
Regulator Stations — Regulator stations fall into two categories:
1. Interceptor regulators, where sewage from a combined
trunk sewer can be diverted into an interceptor or
overflowed to salt water
2. Trunk regulators, where storage exits in a sewer but
overflows either cannot occur or are restricted because
discharges would be to fresh water.
See Table 15 for specific station classifications.
Interceptor Regulators — When flow at an interceptor regulator
station reaches 100 percent of maximum flow, it is placed in
active/storing mode and the diverted flow into the interceptor
is controlled to increase storage in the trunk as specified by a
storage rule curve for the station. If flow in the interceptor
exceeds the maximum flow, diverted flow will be reduced, thus
violating the rule curve. Whenever a regulator station is placed
179
-------
Table 15. CLASSIFICATION OF
STATIONS FOR AUTOMATIC CONTROL
REGULATOR STATIONS
Interceptor
Trunk
Eighth Avenue South
West Michigan Street
Harbor Avenue
CheIan Avenue
Norfolk Street
Michigan Street
Brandon Street
Hanford Street #2
Lander Street
Connecticut Street
King Street
Denny Way Local
Denny Way Lake Union
Dexter
Lake City Tunnel
University (Future)
Montlake (Future)
3rd Avenue West (Future)
Hanford Street #1 a
PUMP STATIONS
West Marginal Way
East Marginal Way
Duwamish k
Interbay
\
Henderson Street
Rainier Avenue
Belvoir Place
30th Avenue Northeast
Matthews Park
Kenmore
a This station is not a true regulator, and its only use
is for emergency overflow.
k Provides data only; no command capability.
180
-------
in active/storing mode/ all stations upstream are also placed in
active/storing mode and begin to store. The rule curves at the
stations are set up, however, so that upstream stations tend to
store slower than downstream stations. Once placed in storage
mode, regulator stations will store up to 100 percent of allowed
storage and then attempt to maintain that storage until conditions
permit transfer to active/releasing mode. This can only be done
when interceptor flow at the station is less than 90 percent of
maximum allowed flow and no stations downstream are in active/stor-
ing mode. Once a station is placed in auto/release mode, it
attempts to divert as much flow into the interceptor as allowed
by ibhe maximum interceptor flow for the station subject to a gate
opening rate check until all storage is released. When most of
the storage has been released, the station is placed in inactive
mode and the regulator gate follows the trunk level.
Most regulator station trunk setpoints for local control are
well below the level of the highest spring tides but are above
mean high water (four feet above mean sea level). The setpoint
provides storage in the trunk during the periods when tides above
the setpoint would cause backflow of seawater if the outfall gate
were open. Most regulator stations also have a maximum safe set-
point above which local backwater flooding might occur. This level
is 12 to 18 inches above the local setpoint.
To ensure available storage in the trunks in the event that
a spring high tide is predicted to exceed the maximum safe set-
point, the trunk setpoint is dropped to the local control posi-
tion between the time that the tide level rises to within two feet
of the local setpoint and the time that the tide begins to fall.
When a station is in the inactive control mode, the setpoint is
always maintained in the local position, regardless of tide.
Trunk Regulators — These stations only begin to regulate flow
when the pump or regulator interceptor station at the end of the
trunk enters the active/storing mode. When the storage mode is
entered, these stations store as specified by a storage rule curve.
When the controlling interceptor station enters the active/releas-
ing mode, these stations also begin gradually to release storage
in a controlled manner to avoid sudden surges which might upset
control stability in other parts of the system, and ultimately
resume local control operation where the flow is not controlled
at all.
Pump Station — In interceptors carrying combined sewage flows,
interceptor capacity is limited. However, some interceptors in
the north end of the system, which carry only sanitary flows,
have spare capacity, and therefore storage capacity. In addi-
tion, pump stations on combined trunk sewers have, under most
conditions, spare storage capacity upstream. To provide best use
of these varying types of pump stations under CATAD control, they
181
-------
have been divided into two groups for the purpose of automatic
control:
1. Interceptor stations, where as much flow as possible is
transmitted to the treatment plants to maximize inter-
ceptor flows
2. Trunk stations, which have available capacity, either by
virtue of being on a combined sewer, or because only
sanitary flows are handled. See Table 3 for classifi-
cation of stations.
Interceptor Pump Stations — As soon as wet weather is detected
by significant increase in interceptor flow, these stations are
controlled to run at the maximum station discharge, subject to
limits imposed on pump speed to avoid cavitation damage to the
pump impellers. Under this control philosophy, the level of
sewage in the pump station wet well is drawn down as low as
possible, making available as much interceptor capacity as possi-
ble. If the maximum allowed discharge at the station becomes lower
than the actual station flow, the station setpoint (which controls
the pump speed) is reduced to the wet well level so that the sta-
tion no longer attempts to draw down the level of the wet well,
and instead, only reacts to changes in flow after they occur.
Trunk Pump Stations — These stations simulate local control opera-
tions until a designated controlling station downstream (pump or
regulator) enters storage mode. When this occurs, the station^
begins to store on a rule curve until maximum allowed storage is
reached at the station or the controlling downstream station enters
active/releasing mode. When the downstream station enters releas-
ing mode, it is a sign that spare capacity exists downstream and
the trunk pump station also begins to release storage in an orderly
fashion by gradually pumping down the wet well until normal local
control simulation operation can resume.
Operator Controls
The console operator has three control functions available
to him under auto control:
1. Automatic Control Function. The operator is free to
enter stations into automatic control or remove them
from auto control as he may desire. Commands in pro-
gress are halted when a station is put in auto control,
and in most cases, may not be resumed until the fifth
regular scan after the contact change occurred.
182
-------
2. Rule Curve Duration. When a station is placed in active/
storing mode, it begins to store along a rule curve. The
time taken to store a trunk from 0 percent to 100 percent
is the rule curve duration. A large rule curve duration
produces a slow rate of storage; reducing the rule curve
duration produces a faster rate of storage. This func-
tion only affects regulator operation when a station is
in the active/storing mode.
3. Maximum Flow Function. This function provides the opera-
tor with a means of entering the maximum plant inflow
permitted at the West Point Treatment Plant. Normally,
this function reflects the maximum plant inflow as ascer-
tained from the West Point plant operator. In special
situations, however, reduction of the plant maximum flow
can be used as a control tool to regulate flow to the
plant.
REAL-TIME MODEL
Control Plan
The real-time model is at the top of the control hierarchy
which comprises four principal levels of implementation as
follows:
' - 1. Local control (hardware)
2. Supervisory control
3. Automatic system control
4. Optimized automatic system control.
Each level is superimposed upon the next lower level and is im-
plemented via hardware controllers or program routines at the
next level. Control at each level may be exercised independent
of controls at higher levels thereby permitting orderly program
development.
The real-time model is the device for implementing optimised
.automatic system control. The model which is executed periodi-
cally under program control is used to develop an operating plan
for effective utilization of available storage within Metro's com-
bined sewers and separated.sanitary sewers. Storage control is
affected using a technique similar to that used for sewer basin
.flood management, in which available storage in the main and tri-
butary system is utilized during the period of storm runoff based
on an operating rule curve for each storage element, which defines
planned amounts of stored runoff as a function of time. These
operating rule curves are developed on the basis of computer
studies in which historical floods are routed through a system
183
-------
model. Routing studies are generally made by batch processing
on a general purpose computer. Similar techniques are used in
the CATAD system except that an operating plan to regulate runoff
by effectively using available storage in the storm water collec-
tion system is based on actual and projected rainfall patterns.
The plan is implemented by adjusting the storage rule curves which
determine rates at which the automatic control program will store
runoff at various control structures.
Historical flows are used for setting up initial operating
rule curves which are then adjusted dynamically by real-time analy-
sis of observed rainfall and runoff data using a system model.
The results of the real-time analysis are implemented by adjusting
the rule curves in accordance with actual rainfall and runoff
patterns.
The procedure for model analysis and rule curve adjustment
is as follows:
1. Storm hyetographs are constructed by combining measured
precipitation with an estimate of future precipitation
based on various criteria including seasonal factors,
the U. S. Weather Service forecast, the direction of storm
movement, and the pattern of actual precipitation
2. The hyetographs are converted to inflow hydrographs
entering the Metro sewer system by the unit hydrograph
method
3. Combined sewage flows are routed through the trunk sewers
down to the major storage control structures
4. Combined flows are routed through storage control
structures and down to the treatment plant as specified
by initial or improved storage rule curves and as limited
by physical and safety constraints
5. A list of flow excesses (i.e., possible overflows) which
were not absorbed by the interceptor system are used to
adapt rule curves to reassign storage to:
a. make the best use of available system storage, and
b. assign overflows in accordance with water quality
constraints.
184
-------
For the purpose of converting rainfall to runoff, tributary
areas have been divided into about 40 subareas. To facilitate
flow routing, the sewer system has been subdivided into a network
of about 70 pipes. Principal control features include: 12 regu-
lator stations which actively control flow entering the intercep-
tor sewer; and 10 pump stations, some of which have a potential
for storage upstream.
Rainfall Prediction
The manufacture of hyetographs for use in developing runoff
obviously can be profitably improved in th£ light of actual re-
sults. In the initial stage of model testing, therefore, only
actual real-time rainfall data is used to construct runoff hydro-
graphs using the unit hydrograph method. It is anticipated that
provision will shortly be made to extrapolate actual rainfall
data into an entire predicted precipitation event by using fre-
quency, time-duration and intensity-duration analysis based upon
seasonal factors and storm direction to predict future rainfall
in a precipitation event.
There are presently six rainfall gages being monitored by
the CATAD system. Of these, .three are located in the tributary
areas of combined sewers.
Runoff Model
Rainfall hyetographs are converted to inflow hydrographs
entering the Metro trunk sewer system by the unit hydrograph
method. The unit hydrographs were developed for each tributary
subarea in the system using a method described in Eagleson (25)
in which unit hydrographs are synthesized from various data
including the shape and size of the tributary basin. The unit
hydrographs have been simplified to a triangular shape for use
in the runoff model, (see Table A~3> Appendix A,).
Sanitary flow is represented by a small additional amount of
equivalent rainfall applied to each subarea. A suitable sinu-
soidal pattern has been developed for the sanitary equivalent
rainfall to represent daily peaks and troughs in the system's
dry weather flow characteristics.
Routing Through Storage Reaches
The treatment of storage reaches in the model is related to
the treatment of regulator stations, operating rule curves and
interceptor capacity. The handling of storage by the Storm Water
Management Model, which assumes that a storage length has finite
volume but zero length, is insufficient for the CATAD model.
This approach implies that there is no lag between the time a
wave appears at the top end of a storage reach and the time that
185
-------
it arrives at the downstream end. Flow transmission times are
significant for long reaches and a number of trunks in the Metro
system with slopes shallower than 0.001 are several thousand
feet in length.
The approach chosen for routing flows through 'storage reaches
is intended to be compatible with the method used for storage
computations in the CATAD system. The inflow hydrograph is
routed through the storage reach on the basis of the continuity
equation/ i.e., the outflow from the reach during each interval
will be equal to the inflow to the reach less the amount of flow
stored during the interval. The change in' storage during the
interval is determined by the operating rule curve so that for
a computed rate of inflow to the reach the outflow rate is deter-
mined by the rule curve. If the computed trunk outflow is within
the regulator gate diversion capacity/ and within the residual
capacity of the interceptor, the trunk outflow hydrograph from
the storage reach becomes the inflow hydrograph to the interceptor
sewer. If, however, the computed trunk outflow exceeds the regu-
lator gate diversion capacity or the residual capacity of the in-
terceptor, the inflow hydrograph to the interceptor is limited
to the gate capacity or the residual interceptor capacity. The
remainder of the trunk outflow comprises a hydrograph of excess
flow.
When storage in the reach is equal to the maximum storage
in the trunk, all inflow must be directly transmitted through
the reach with no net storage increase and outflow will equal
inflow.
Control Elements
There are three types of control structures which have signi-
ficant effect on the flows in the system:
1. Bump stations
2. Regulator/outfall stations
3. Inverted siphons.
Special routines have been written to describe the effect
of these structures on routing of-flow. Regulator stations,
which are central to the utilization of trunk storage, have
been simulated to closely approximate behavior of automatic con-
trol.
In real-time, the operation of a regulator station is affected
by three constraints:
1. The storage rule curve
2. The maximum allowed flow in the interceptor
3. The maximum capacity of the regulator gate and the pipe
connecting the gate with the interceptor.
186
-------
Either the gate capacity or the maximum allowed interceptor flow
will form an upper limit on diverted flow.
In simulation, the maximum interceptor flow,is calculated
for each station before the routing is performed. The maximum
flows are calculated from the current capacity of the West Point
Treatment Plant, and hydrographs of unregulated flows entering
the system. In the event that physical limitations interfere
with the ability of the station to follow the operating rule
curve, the difference between the computed diverted flow and the
maximum diverted flow is considered to be excess flow, or poten-
tial overflow.
Rule Curves
The operating rule curve relates the rate of storage of sewage
in a trunk sewer upstream from a regulator station as a function
of the elapsed time from the beginning of the storm.
Basic operating rule curves have been developed from experi-
ence in operating the system under automatic control and from
background studies. These curves are being expanded into a
family of curves for various types of storms as a result of rout-
ing actual precipitation events. The primary difference in rule
curves for various types of storms is the duration of time allowed
for filling available storage. Since the number of potential
storm patterns is essentially limitless, generalized rule curves
for individual stations based on typical storms may not apply
to a specific storm in which the geographical distribution of
rainfall may differ considerably from the assumed storm. Never-
theless, the curves provide a point of departure from which to
make more precise analysis using the hydraulic routing model.
Figure; 51A shows that too strict an adherence to rule curves
could caus^ some problems in the case of rapidly occurring storms.
During the limited automatic control phase, programming has been
developed which provides a modified rule curve control as shown
by Figure 51C. This^means the program will follow the rule curve
until the rate of inllow greatly exceeds the storage rate of the
control curve; then the computer will select the inflow rate
itself- and completely fill all storage in the trunk line before
any overflow is permitted. (Note: The diversions and sewer
configurations being discussed here refer to the diversion type
regulator of Figure 5A).
Part B of Figure 51 shows the idealized control curve. The
upper hydrograph shows that as inflow to the sewer increases,
the programs will divert all this flow to the interceptor for
treatment until the interceptor capacity at the point of diver-
sion is met. At this point, commands will be issued to begin
187
-------
o
oe
o
u
I
Id °
v LU
M± ^
O
O
o
o
CD
in
B
H
Ul
co
w
1
o
I
EH
H
i
EH
a
o
o
o
EH
CO
HI U
X
b.
o
o
09N ' 31VH
MOIdNI NMOIS
o
10
0350 30V801S -13d
3AMOD
188
-------
reducing diversion rates and storing excess inflow within the
trunk sewer. The commands force the station to follow the ideal-
ized rule curve which, in the time period between time 1 and 2,
makes use of 100 percent of the available trunk storage, holding
the excess flow for later release to the interceptor.
Figure 51C shows what happens when inflow exceeds this
idealized rule curve. The total hydrograph is larger, than in
the idealized case, and the design storage which can be.managed
by the programmed rule curve is insufficient to retain all the
inflow from this larger storm without overflows. The control
strategy is shown in the lower graph where the programmed rule
curve is not followed, but the actual storage rate is determined
by the rate of inflow in the top graph. Between time 1 and time
2, all storage in the trunk line is utilized. At time 3, after
the overflow has taken place, the interceptor can accept addi-
tional flow and storage if released in controlled steps. This
prevents excessive wave action within the interceptor from the
rapid release of stored water from many stations.
Actually, the overflow shown in Figure 51C possibly would
not take place because the limited automatic control program
orders any upstream stations to begin active storing as soon
as a station with heavy storm inflow exceeds its capacity to
divert to the interceptor. Referring to the top hydrograph on
Figure 51C, the program tries to increase the diversion capacity
of the interceptor (raise>the double horizontal line) by storing
flows at upstream stations. Thus, the area above the double line
would be reduced to an amount equivalent to the storage volume
available in the trunk line. In this manner, the cross-hatched
area, representing overflow volume, is reduced or eliminated.
Rule Curve Adjustment Using Optimizing Model
The result of routing flows through the interceptor system
is .a determination of stations with excess trunk flows and of
stations with available storage. From these results, it may be
determined whether there is net available storage or excess
trunk flow in the system. In the first case, it may be possible
to adjust station rule curves to eliminate overflow entirely.
In the second case, station rule curves can be adjusted to re-
assign storage from trunks with potential overflow to trunks
with available storage to reflect station overflow priorities,
based on the relative impact on water quality of overflows into
Puget Sound and overflows into enclosed waters such as the Duwa-
mish River or the Lake Washington Ship Canal. This may be done
by appropriately increasing or decreasing storage rates in the
various trunk sewers. Experience has shown that overflows
naturally occur first at stations that overflow into Puget Sound,
and only rarely at stations that discharge into confined waters.
The practical effect of the model analysis will, therefore, be
189
-------
to divert storage from stations at the downstream end of the inter-
ceptor system to stations having under-utilized storage capacity
in the upstream reaches of the interceptor system. In instances
where precipitation is local, model analysis will provide relief
for portions of the system overburdened by runoff by storing dry
weather trunk flows at the stations .which handle separated sewage.
Related Background Programs
A collection of background programs has been developed to
support implementation of the CATAD system. The program listing
and documentation appears in Volume II of this report. Brief
descriptions of some of the more useful routines are included
here. Programs are used for data analysis and reduction or to
diagnose features of the hardware and software.
Backwater — Backwater consists of a group of FORTRAN routines
designed to compute flow profiles in circular or horseshoe-shaped
cross-sections under conditions of steady varied flow. The pro-
gram contains specific subroutines to handle overflow weirs and
other special structures in the North and Elliott Bay interceptors.
There are routines to handle the Fremont inverted siphon, the
Canal Street relief, emergency bypass structure in the North
interceptor; and lowhead crossings at Lander and Hanford Streets
in the Elliott Bay interceptor. Program use of special routines
is controlled by parameters on input data cards describing the
physical features of the sewer system.
Inflow from unregulated trunks and overflows from relief
structures are calculated where appropriate. Inflow from unre-
gulated trunks is computed as the sum of anticipated sanitary
sewage flow and surface runoff calculated by the rational for-
mula. Rational runoff coefficients for use in the rational
formula (equation 1) were estimated by considering the nature
of the areas in question, the existing tributary sewer system,
and observed flow rates where available. Sanitary sewage flows
were estimated from tributary population estimates.
Q = CiA
C - Rational runoff coefficient
i = Rainfall rate
A = Tributary area
Flow profile computations are carried out by the so-called
standard step iterative method"as described by Chow (27). Pro-
vision was made to add turbulence-induced head losses at channel
expansions, contractions and changes in alignment.
190
-------
Flow profile computations are limited to subcritical flow
(flow depth greater than critical 'depth) but can handle critical
control sections that may occur. Each reach is tested for super-
critical flow and its occurrence initiates a search upstream in
the sewer for a control section from which profile computations
may proceed.
Program input consists of data cards describing the sewer
system physical features, and of operator input via teletype.
The operator inputs external solution parameters including rain-
fall data, regulated trunk flows, and flow rate and water surface
elevation at the downstream end of the system via teletype in
response to program-generated messages. Upon completion of any
set of computations, the operator can initiate a new run with new
external parameters via the teletype.
BACKWATER was designed to serve three functions. First it
was designed to aid real-time inflow regulation to the North and
Elliott Bay interceptors via the CATAD system in the supervisory
mode. The operator can enter real-time rainfall data and pro-
posed regulator inflows to get a prediction of interceptor res-
ponse. The program can also be used with historical data to gain
system behavior experience. The steady flow assumption limits
this application. Second, the program may be used to simulate
the behavior of proposed sewer designs under various flow condi-
tions. Third, the program may be used to simulate the response
of existing sewer systems to physical modifications. Effects on
flow profile patterns can be examined prior to undertaking major
modifications.
Palmer-Bowles Flume Rating Curves — Metro has a large system flow
measurement program to determine locations of excessive infiltra-
tion and to provide information useful in final development of
mathematical models applicable to the CATAD system. Initial work
in this study included programs required for rating the Palmer-
Bowles flumes which perform this measurement in manhole locations
around the Metro drainage area (see Figure 52). Manual flume flow
calculations at various depths involved many hours of work with
"Areddi" curves to establish a rating curve for the flume. Since
this effort was required for more than 100 different measuring
flumes in this study, a computer program was applied.
The program accepts basic physical data from the flume and
upstream sewer line. The data includes sewer diameter, shelf
height, flume side slopes, and elevation of vertical side slopes.
The program produces a computer rating curve on the digital
plotter (see Figure 53).
191
-------
6 MAX.
LOCATE WITHIN 6" OF
TRANSITION
FLOAT WELL
(Made of V. C.
pipe)
PLAN VIEW OF MANHOLE
8"
SECTION
FIGURE 52
INSTALLATION OF PALMER-BOWLES FLUME
192
-------
1.25-
1.09-
0.94^
0.78-
H-
LU
LJ
U_
T-- 0.62-
CL
UJ
Q
O.47-
0.31-
0.16
0.00
0.00
PALMER - BOWLUS FLUME
DIAMETER I5.O IN.
BASE WIDTH 5.0 IN.
SILL HEIGHT 1.2 IN.
SIDE SLOPE 2.00
UPSTREAM O.OO250
0.78
1.56 2.34 3.13
FLOW, CFS
FIGURE 53
FLUME RATING CURVE
3.91
193
-------
Monitor Data Retrieval — Since 1966, Metro has continuously col-
lected water quality data from remote sites on the Duwamish River.
Before the Sigma II computer was available, data was transmitted
via telephone lines to the central office, punched in paper tape
and typed on a dedicated, electric typewriter.
The paper tapes were collected and cards produced for data
analysis.
The CATAD contract provided for the existing water quality
monitor to interface with the computer and bypass the paper tape
phase which would become a backup to computer collection. Data
is now concurrently transmitted to the paper tape and the computer.
Software writes the data on a large capacity random access disk
(RAD) file. The file holds four days of data.
A data retrieval program takes the data from the RAD, formats
it and punches cards for subsequent data analysis.
The RAD data is in a non-standard code which must be con-
verted to EBCDIC prior to punching. ' Transmission is subject to
failure and the data must be edited to assure that the punched
cards accurately represent the measured parameters. The most
frequent transmission problem is character duplication or char-
acter loss in transmission. Since a number of characters in the
record are known (e.g., blanks, decimal point and end of record
mark) these poorly transmitted records may be corrected or
eliminated.
When the cards are punched, the file is cleared and the
pointers are reset so the next data is properly located on the
file.
Force Main Calibration — A pressure transducer was installed in
the force main at each pumping station to determine flow through
the station. The salt velocity method was used to measure discharge;
i.e., the pressure in the force main was stabilized at a measured
level and then salt was introduced at the pump. A recording con-
ductivity meter was installed at the end of the force main and
synchronized with the time of salt injection. The resulting time
vs. conductivity recording was digitized and a computer program
determined the center of mass and, therefore, the transport time
at that pressure. Force main pressure was then changed and the
process repeated. Hydraulic aspects of this study are discussed
in Section VEII. Here, program considerations are summarized.
194
-------
Knowing the geometry of the force main and, therefore, its
volume, it was possible to compute the flow at the measured
pressure.
At the conclusion of a day's sampling, data was digitized
and an attempt was made to fit the data with a curve of the form:
Q = Ap + Bp2 + Cp3
where A, B and C are constants determined by the method of least
squares, and p is the dynamic head. Plots were generated to
determine how well the generated curve fit the measured data and
to determine where more data might be needed.
Computerizing the data analysis permitted investigators to
decide whether more data was required at a particular station
before calibration equipment was removed.
Examples of the Flow versus Dynamic Head curves are included
in Appendix F.
Process Data Storage — Hourly log data for each remote station
in the system was judged too sparse to determine the station's
reaction to changing physical parameters. The magnetic tape drive
offered a means of obtaining great quantities of data, rapidly
sorting out unwanted information and displaying the selected data
for detailed analysis. During storms, a policy was established
whereby the data collected from all regulator and pumping stations
was saved after each scan on magnetic tape for a historical record
and for data analysis. The program called PROCESS DATA STORAGE
is stored in the disk library. To begin data storage, the tape
is mounted, the drive readied and the operator types Q MAGBGN on
the programmer's console. It replies by asking "what stations?"
The operator types ALL or a unique code for a single station.
By repeating the process, multiple stations can have their scan
data stored on tape. A similar process is followed to terminate
the data storage program.
Special control and/or hydraulic problems of a rapidly chang-
ing nature at any remote station can be singled out for a rapid
scan rate as short as two seconds. Collected data is stored on
the system magnetic tape for later analysis.
Data saved includes: the data displayed on the operator's
console CRTs, analog inputs from each station, all contact status,
the analog values converted to engineering units, and the contents
of the rainfall pulse counters. This data is in addition to static
data such as conversion factors, analog value ranges, maximum pulse
counter rates, etc.
195
-------
Process Data Retrieval — The data on magnetic tape can be subse-
quently displayed in two forms: plots of data value vs. time,
and tables of data values. Most often, the data is plotted using
a program called PROCESS PLOT.
A card is read indicating which station is to be plotted.
The required station is read from the magnetic tape and saved on
the bulk memory unit (magnetic disk or RAD). A parameter card
is then read indicating the date the tape was created, a code
designating which variable is wanted and its factor, name and
units. The specified variable is then read from the RAD and
plotted. Another parameter card is read and the process repeated
until all the required variables for the station are plotted.
Examples of plots generated by this program are included in
Appendix G.
Figure 54 shows how the data gathering and displaying pro-
grams in analyzing remote station control problems. In this
case, pump mode controllers were functioning improperly and pumps
were cycling on and off at a damaging rate. Faulty gate setpoint
controllers, pump tachometers and other instrument malfunctions
have also been investigated and corrected with CATAD acting as a
centralized data gathering facility for the entire Metro system.
There are obvious cost advantages for a centralized system com-
pared to installing multi-pen recorders at each remote site or
a "traveling" precision recorder set up at a station to analyze
a specific problem.
Process Data Tables — When exact process data values are needed
rather than a graphical representation, the data from the magnetic
tape is reported in tabular form. Data for more than one station
may be listed. This program helps to assess the effects of sta-
tion interaction.
Program input includes: station number, number of scans to
skip between reported record, variables and contact status to
list. The option to skip records reduces the length of a table.
If a station were on a one-minute scan for 24 hours, the table
would have 1440 entries. With scans on a five-minute interval,
the program could skip the intervening scans to give a 288 entry
table.
Diagnostics — To perform hardware maintenance while operating
foreground control programs, a series of programs were written
to diagnose hardware malfunctions. The computer manufacturer did
not have on-line diagnostics for Sigma II peripherals. A special
diagnostic program was written for each unit. Two of the more
complex programs are discussed below.
196
-------
10.00-
UJ
LJ
U.
I
X 8.0O-
O.
UJ
CC
111
6.00 H
4.00-T 1 1 1
0.00 2.00
—1—
4.00
NO. OF PUMPS ON
6.00
T
8.00
TIME
IO.OO
12.00
A. CYCLING PUMP
EAST MARGINAL WAY PUMP STATION
-30
"'I
-00-
14.00
16.00
2345
TIME HOURS
B. EXCESSIVE GATE MODULATION
LANDER ST. REGULATOR
FIGURE 54
EOJIPMENT PROBLEMS
197
-------
Magnetic Tape Diagnostic — This diagnostic was written to detect
three typical magnetic tape unit problems. The error conditions
tested are:
1. Unsuccessful read operation
2. Unsuccessful write operation
3. Compare errors.
An unsuccessful operation is defined as MiREAD or M:WRITE
(XDS supplied monitor service routines) returning with the accumu-
lator not equal to zero after an initiate and wait request for
input or output (I/O). This may indicate an unrecoverable I/O
error, incorrect record length or other error conditions documented
in the RSU reference manual under I/O completion codes.
A compare error occurs .when a buffer with known contents is
written to the tape and subsequently read back and the read
buffer is not identical to the written buffer.
\
The diagnostic writes 6000 records each 180 words long con-
taining a known data pattern. The tape is then rewound and the
6000 records are read and compared with what was written. If the
cycle was successful, it is begun again.
On the other hand, if a compare error is detected, the con-
tents of the bad buffer are written on the operator's console
with the data and time the error occurred. The tape is then *
rewound and read again to determine whether a read or write
problem exists. The tape is then written again.
All three error -types have been detected on both units.
Compare errors are the most serious in that there is no way of
detecting the error unless contents of the record are known.
(If the contents are known, there is no reason to read it from
the tape!) This could be especially serious if data that can't
be regenerated is saved on tape.
Calcomp Plotter Diagnostic — The plotter was bought directly
from the manufacturer rather than supplied by Philco. A diag-
nostic to test the controller and the plotter drives the plotter
to trace a circle and report transmission errors. Certain con-
troller components were heat sensitive and caused plotter mal-
functions; i.e., the plotter would lose its origin and draw.
eccentric circles. When the faulty component was replaced, the
problem cleared up and the equipment provided a year of mainten-
ance-free operation. However, it now appears that plotter opera-
tion is causing bad RAD transfers, thereby compromising the
integrity of the real-time system.
198
-------
Core Diagnostic — As part of the field acceptance test/ a diag-
nostic was developed to determine whether foreground programs
I were inadvertently writing in the background core area. The
program fills the background with a known pattern and constantly
checks the pattern for alteration. No problems have ever been
detected by this diagnostic.
Fremont Inverted Siphon Hydraulics — The Fremont inverted siphon
shown in Figure 55 was originally constructed as a twin-barrelled
structure with 60-inch and 48-inch conduits without provision
for directing flow preferentially to either conduit. In 1972,
the forebay and afterbay structures were modified to direct dry
weather flows preferentially to the 60-inch conduit (see Figure
56). The modifications, undertaken to provide better cleansing
conditions in the siphon and the upstream sewer, included con-
struction of an ogee shaped outlet weir and an inlet weir for the
48-inch conduit.
The Fremont siphon modifications complicate hydraulic profile
calculations. Four distinct flow regimes exist with corresponding
control conditions. In three of these, the depth of flow upstream
of the siphon depends on the energy and hydraulic grade line eleva-
tions downstream. Hydraulic profile calculations through the siphon
also require distribution of the total flow rate between the two
barrels to equalize head losses. Profile calculations involve a
trial and error procedure requiring a knowledge of the hydraulic
and energy grade line elevations downstream and the total flow rate
through the siphon. A computer program was constructed to perform
the trial and error solution for upstream depth given the above
variables.
A water surface sensor was installed as part of the GATAD
system in the Canel Street relief and emergency bypass structure
approximately 180 feet upstream of the siphon forebay. This
sensor controls upstream regulator stations to eliminate ^overflows
of combined waste water at the relief structure. This sensor cannot
determine flow rate in the interceptor or flow depth downstream of
the siphon because of the complicated siphon hydraulics and back-
water effects. A second water surface sensor was installed in the
junction structure where the Central trunk joins the North inter-
ceptor. This structure, located near the intersection of Third
Avenue W. and W. Ewing S'treet, is approximately 180 feet downstream
of the siphon afterbay structure.
The water surface sensor in the Central trunk junction cham-
ber will control the future Third Avenue W. regulator station
which is part of the overall CATAD system. The sensor also serves
two other purposes. First it provides a control point regulat-
ing inflow to the interceptor from the Interbay pumping station,
which controls the hydraulic grade line elevation at the central
199
-------
FORE BAY
STRUCTURE
FREMONT INVERTED
SIPHON
BO'SIPHON
4»* SIPHON
AFTERBAY
STRUCTURE
—CANAL STREET RELIEF ft
EMER9ENCY BYFA8S
STRUCTURE (C.S.»)
JUNCTION CHAMBER
. SANDCATCHER
^-OVERFLOW (FUTURE
REGULATOR STRUCTURE)
AcATAD WATER
FIGURE 55
PLAN OF FREMONT INVERTED SIPHON & VICINITY
200
-------
-*-^ ^ l-^4
Jj C
' ' If
^ .
^~ 'i H
si
' ; <.
ec , \ .
. — -*
' H!V .
m^
vo
in
fa
EH
I
fa EH
CO H
H O
H H
EH >
H
h"l Q
H^»
£-4
P <3!
PM J3
Q Scj
W 04
H H
fa CO
H
Q Q
Q W
O >
W H
201
-------
trunk junction chamber. Second/ it provides a means of computing
flow rate through the Fremont inverted siphon. The water surface
sensors at the Central trunk junction chamber and the Canal Street
relief and emergency bypass structure indicate the change in water
surface elevation through the siphon. This information may be
used, via rule curves or other methods, to compute the total
siphon flow without time-consuming trial and error flow balancing
through the siphon conduits.
202
-------
SECTION VIII
IMPLEMENTATION - OPERATION PROBLEMS
Considerable delay was experienced in implementing CATAD
System Controls. Problems fall into the following major categories:
1. Programming Management
2. Problem Definition
3. Hardware Problems
4. Programming Problems
Electrical noise was an especially troublesome hardware prob-
lem and is covered in greatest detail.
PROGRAMMING MANAGEMENT
Specification Requirements
The CATAD System Control procurement specification required
the system vendor to provide complete systems programs and some
application functions. Metro retained responsibility of developing
applications programs required for CATAD system automatic control.
Many major control systems suppliers provide standardized
hardware using a specific computer; general purpose peripheral
input and output devices; and various items of data acquisition,
display, and control hardware. These suppliers also have standard
system programs for these devices. From a cost standpoint, the
user should utilize these standard programs as much as possible.
To use standard software, programming specifications must be
predominantly functional. This procedure, unfortunately, gives a
supplier who has not developed standardized hardware and software
a great deal of latitude in developing specific system programs,
and limits the owner's control of programming techniques used by
the contractor.
CATAD specifications were functional for the most part, but
defined some very specific program requirements:
1. A multiprogram priority executive system.
2. All input and output to computer peripheral equipment and
to each specific item of special purpose hardware to be
asynchronous :and handled via a unique subroutine,
3. A system for debugging foreground programs on-line.
The requirement that input and output to each external device
be under control of the executive system through a unique routine
203
-------
associated with that device is a practical necessity in a real-time
system and serves the following objectives:
1. Permits queuing of multiple requests for use of a device.
2. Facilitates diagnosis and correction of hardware or program
bugs.
3. Facilitates program changes to accommodate hardware changes.
Implementing Specification Requirements .
Effective implementation of a computer directed real-time
control system requires a systematic plan for program development.
The plan should provide a logical development sequence from basic
computer software to a comprehensive system.
The CATAD system contractor had not previously developed special
software for the computer supplied for the CATAD system so he started]
with operating systems supplied by the computer manufacturer. The
contractor prepared task-oriented programs, then attempted to in-
tegrate these individual tasks into a complete system. Each task
comprised all input and output to the hardware device as well as
the applications programs related to the device; all of which
were attached to a hardware interrupt.
Applications programs were disk resident or used disk files
requiring frequent transfers between the disk system and core memo-
ry. Hardware interrupt assignments were made without regard to
time dependency of certain tasks. As a result tasks which might
work satisfactorily when run independently, conflicted when run con-
currently.
The contractor resolved many of these problems with consider-
ble assistance from Metro's consulting engineer. Although resolution
of these problems allowed the contractor to ship the system from
the factory to Metro's site, it did not provide a truly usable
system. The contractor's programs, while functionally acceptable,
were cumbersome to use; imposed excessive overhead on the system;
and were difficult to debug.
A logical development sequence that would prevent such con-
flicts was demonstrated in subsequent reprogramming, which was
developed in the following sequence:
x
1. Preparation of overall system design
2. Preparation of system programs
3. Preparation of applications programs
The system programs described earlier in this report were not
furnished by the contractor but were reprogrammed by Metro's con-
sulting engineer to meet the original system specifications.
204
-------
PROBLEM DEFINITION
A major problem encountered during CATAD system implementation
was the contractor's inability to isolate problems to either the
hardware or the program. Following the endemic industry pattern,
hardware personnel pointed to programs and programmers pointed to
the hardware. CATAD system development experience indicates that
problems can generally be traced to four sources in roughly the
sequence shown:
1. Program bugs
2. Inadequate hardware documentation
3. Hardware malfunctions
4. Hardware design deficiencies
A major program development problem was the apparent failure
of contractor personnel to isolate problem sources by systematic
plans of attack using the computer as a tool for problem diagnosis.
HARDWARE DIAGNOSTIC ROUTINES
Problems which require the greatest time to resolve often re-
sult from hardware malfunctions or documentation deficiencies.
Such problems are difficult to isolate, and hardware personnel doufc^t
their existence.
The most effective hardware problem isolation technique is a
special hardware diagnostic routine which can provide either:
1. Raw values of computer input
2. Actual hardware response to computer output
This information can be compared with hardware documentation
to ascertain whether the hardware is performing correctly.
HARDWARE PROBLEMS
A number of hardware problems were encountered during the
contractor's original implementation as well as later. The prin-
cipal equipment affected by hardware malfunctions were:
1. Disk Storage System
2. Interrupt System
3. Peripheral Equipment
4. Power Fail Detection
5. Remote Station Telemetry
6. West Point Satellite Terminal
7. Renton Satellite Terminal.
205
-------
Many hardware problems encountered while the system was
located at the contractor's facilities may be attributed to the
environment/ which exposed the system to contamination by dust
through doorways open to the outside and to overheating in the
absence of air conditioning. Many problems have abated signifi-
cantly or disappeared since the equipment was moved to the cen-
tral control facility in Seattle.
Other computer system hardware problems occurred at the con-
tractor's facility during periods when preventive maintenance was
postponed to avoid interference with programming. Regular main-
tenance was resumed when equipment failure delays began to cost
more time than preventive maintenance.
Disk Hardware
Many problems were traced to the disk system initially fur-
nished to the contractor. The disk has a substantial complement
of spare tracks (about 20 percent). A number of bad tracks and
other problems were encountered on the original disk/ which was
replaced. The newly designed replacement disk failed shortly
after delivery. A second replacement was required.
The redesigned disk has subsequently proven highly reliable.
Computer Interrupt System
Problems have been encountered with random interrupts in the
computer interrupt system. Although random interrupts are not
difficult to correct/ isolating resulting program execution pro-
blems may have been very time consuming.
Peripheral Equipment Failures
The most frequent hardware problems involved electro-mechani-
cal peripheral devices/ principally the card punch and card reader.
Other problems were experienced with the graphic plotter .and mag-
netic tape units which were interfaced to the computer buffered
input/output channel by other than the computer manufacturer.
This is a very complex interface with much critical timing. Pro-
blems have continued with these interfaces.
Card Reader — Considerable problems were encountered with the
card reader as originally furnished to the contractor. Replacing
the read facilities with a fiber optic system effectively resolved
the difficulty. This caused considerable delay during early pro-
gramming .
206
-------
Card Punch — Card punches are notoriously troublesome and the
punch supplied for the CATAD system was no exception. Heavy usage
in the poor environment at the contractor's facility resulted in
frequent malfunctions. Considerable skill is required to effect
satisfactory repair. The entire punch mechanism has recently
been replaced.
Keyboard Printer — All the working parts of the keyboard printer,
which is critical to programming/ were replaced while at the con-
tractor's facility.
Plotter -- The plotter interface initially presented major pro-
blems and interfered with disk transfer operations. The original
problem was traced to an improper connection. However, periodic
disk transfer problems appear to occur only during plotter opera-
tion. A mechanical plotter failure requiring return to the fac-
tory for parts replacement occurred within two years of the
installation.
Magnetic Tape Unit — The magnetic tape unit contained a number
of design defects when installed. Although these defects were
largely corrected, the units have not operated dependably and
frequent tape transfer errors have been observed. Since the
magnetic tape units are used for disk storage backup and for
system generation, this problem has been a considerable annoyance.
Power Fail Detection Circuits — Power fail detection circuitry
is designed to interrupt the computer When voltage drops below a
specified threshold. The computer then stores the contents of
volatile hardware registers in non-volatile core memory available
for restoration when the voltage rises above a specified thres-
hold, higher than the power off threshold. The power fail detec-
tion threshold was initially set below the level at which the
computer circuitry was reliable and the power fail routines operated
improperly.
The problem was resolved by raising the power fail circuitry
threshold as high as possible without causing the circuits to
resppnd to normal transients.
207
-------
West Point Satellite Terminal
Many hardware problems were encountered with the West Point
Satellite Terminal at the contractor's facility. A persistent
problem was associated with adjustment of delay line row write
circuitry which was finally replaced by a solid state memory.
Other West Point Satellite Terminal design deficiencies were
corrected by field wiring changes. Many of these problems were
uncovered through the cooperative effort of hardware personnel
and programming personnel of Metro's consulting engineer (see Pro-
gram Problems).
Renton Satellite Terminal
The equipment for the Renton Satellite Terminal was procured
subsequent to the equipment at the central station and under a
separate contract. Numerous problems were encountered in imple-
menting this terminal both in hardware and software.
The satellite terminal equipment comprised the computer, CRT
terminal, keyboard printer and one end of the data link with the
central terminal. Programming furnished by the supplier included
only a real-time operating system and diagnostic routines to test
the terminal hardware.
Field acceptance tests on a system of this nature are gen-
erally made so that, if possible, the testing includes only the
hardware and software furnished by the supplier. The central
processor, the CRT terminal and the keyboard printer were, there-
fore, tested using the supplier's standard diagnostic routines.
The serial data link was tested using a local feedback loop in
which a second communication adapter was installed and the output
of the modem transmitter module was input directly to the modem.
receiver module. This arrangement is equivalent to a four-wire
full duplex link and permitted testing of Renton terminal hard-
ware independent of central station equipment and leased telephone
circuits. From these tests, the equipment appeared to operate
satisfactorily..
t*
The hardware and software at the central station were designed
to communicate with the West Point Satellite Terminal. Therefore,
until the Renton Terminal software to emulate the West Point Ter-
minal was available, the complete data link could not be tested and
numerous hardware problems associated with the data link were dis-
covered after applications programming was begun.
It is difficult to separate discussion of hardware problems
from software and documentation problems since there is much over-
lapping. For this report the problems are separated on the basis
of the source of the problem as finally uncovered.
208
-------
The principal problems traced to hardware defects were as
follows:
1. Defective core memory blocks (2/512 bytes)
2. Serial data link
a. Adapter connections
b. Modern connections
c. Modern logic
d. Modern relay noise interference with automatic I/O
For these hardware problems to fall under the normal warranty
on the equipment, it is first necessary to demonstrate that, in
fact, a hardware problem actually exists. Since the equipment had
apparently performed properly during acceptance tests, the vendor
will generally assume that subsequent problems are the result of
program bugs or in the equipment of others.
Core Memory— Two blocks of 512 bytes of core memory were dis-
covered to have failed after the core memory had successfully
passed core diagnostic tests after installation. The problems re-
sulting from such a failure do not indicate their cause, and con-
siderable effort is necessary to assign the cause to hardware or
software.
Serial Data Link - The serial data link adapter provides the inter-
face between the communications modem and. the computer. Prelimin-
ary diagnostic testing of the data link adapter was done with the
system in a "stand alone" mode; that is, input and output were done
directly using hardware I/O instruction and bypassing the auto-
matic input-output system. This mode is more suitable for check-
out since it isolates the device from possible problems in the
automatic input-output hardware or in operating system software.
Problems were encountered in both receiving and transmitting via
the serial data link adapter.
The normal sequence of operation is as follows:
1. The central station equipment in the "rest" state main-
tains the transmitted frequency in a logical "1" state.
2. The modem at the Renton Satellite Terminal translates
the transmitted frequency as a logical "1" and presents it
to the serial data link adapter which is in the "rest"
state.
3. The central station equipment transmits a character which
consists of a logical "O" (start bit), eight character
bits and a logical "1" (stop bit) and returns to the "rest"
state.
4. The appearance of the start bit at the satellite terminal
receiver causes it to begin clocking on the new character.
209
-------
It was discovered first that the modem was wired for a seven-
bit character including parity rather than an eight-bit character
including parity as specified. This was changeable through a
strapping option.
After correction of the character format, the serial data
link still could not transmit or receive. It was then found that
the modem was inverting the transmission and translating a logical
"0" as a logical "1" and vice versa. In the "rest mode" the SDL
adapter at the satellite terminal was translating the "rest" sig-
nal as a continuous logical "0" which was interpreted as a start
bit followed by eight "O" data bits, followed by another start bit,
etc. The condition was corrected by interchanging two wires be-
tween the receiver and the control modules in the modem. Although
the satellite terminal was then able to receive, it could not trans-
mit. This problem was not resolved by a simple wiring change but
required considerable testing using an oscilloscope and a detailed
study of logic diagrams.
The modems can be operated in a two-frequency or three-fre-
quency mode at the option of the user. In the two-frequency mode,
the frequency is shifted between the mark or space conditions, i.e.,
"1" or "O" state, at the transmission frequency of the modem (base
rate). In the three-frequency mode, the frequency is shifted up
or down from the center frequency for the mark or space condition
of each cycle of the transmitter and returns to the center frequency
for one cycle. For the three-frequency option, a separate logic
component controls the up or down shift. For the two-frequency
mode, one of these components is removed. It was found that the
wrong component had been removed. To correct the condition the
other identical component, which is not required, was moved to the
correct location.
After the data link had been modified to operate successfully
in the "stand alone" mode, it then would not operate under the real-
time operating system using the automatic input-output facility.
After considerable hardware diagnosis using an oscilloscope and
checking of software, the problem was traced to noise induced by con-
tract bounce in a relay in the modem which caused an oscillation in
the clear-to-send line from the modem to the data set adapter. This
noise was causing a bit in the status register associated with the
device to oscillate, which resulted in a series of erroneous inter-
rupts to the computer. This modem relay is used to turn the trans-
mitter on or off in a two-wire circuit. However, since a four-
wire circuit is provided, the transmitter may be left in the "on"
state at all times. The relay was, therefore, strapped in the "on"
position and corresponding changes made to the software.
210
-------
PROBLEMS
Detailed discussion of all programming problems encountered
in implementing the CATAD system would be voluminous and contribute
little to the state of the art. Some of the problems that were
especially difficult to isolate, indicate some of the subtleties .
which may be encountered. Their resolution may be informative to
others.
,Principal programming problems related to the following
areas:
1. Hardware documentation
2. Disk operating system
3. Power on/off
4. Executive and debug systems
5. West* Point Satellite Terminal
6. Renton Satellite Terminal
7. Remote station telemetry
Hardware Documentation Problems
Experience shows that a major source of programming problems
is inadequate definition of hardware operating characteristics
by the designer. This particularly relates to actions required
when an abnormal operating sequence is detected.
Operation must be defined for all possible combinations of
external events and for all possible sequences of these events.
These possible combinations and sequences of events must be defined
for both normal and abnormal operation of the device hardware.
Abnormal operation may include misoperation or malfunctioning of
a device directly or indirectly controlled by the computer. For
example, the computer may attempt to transmit a message to the
West Point Satellite Terminal while the satellite operator is
attempting to enter data. Often a system bug may surface as a
result of an unforeseen combination or sequence of events (which
the program cannot handle properly) months after the program has
been thoroughly checked out. This can be diagnosed if a single
routine controls the device.
Disk Operating System Problems
To build a real-time control system around a computer sup-
plier's standard software, at least one member of the programming
team must become intimately familiar with the software details.
This is especially true in applications such as the CATAD system,
where software features were tested to their limits.
211
-------
A major problem, was encountered in the disk operating system.
Real-time data acquisition and control systems commonly acquire
data via one program and perform appropriate control calculations
via a second program. These tasks may be of different length,
and are processed independently. The computer acquires data from
one station and stores it on a data file while concurrently pro-
cessing data from another station on the file. The disk operating
system did not permit access to data from the same file by two
programs running concurrently. A program attempt to transfer a
file between core and disk storage after a lower level program had
initiated a transfer of the same file caused a system hangup.
(Note: This problem has been corrected in subsequent versions of
the operating system.) The contractor did not analyze the problem
systematically to determine the origin. The source of the problem
was uncovered through detailed analysis of the operating system
by Metro's consulting engineer who also proposed the solution. The
solution consisted of a procedure for queuing all data transfers
to or from the disk.
The newly designed computer operating system contained the
usual complement of bugs which caused some delays.
Power On-Off Tasks
The power on and power off tasks in the operating system
provided for a simple sequence of events; a power interruption
followed by a restoration of power. However, actual power fail-
ure may comprise a sequence of events in which a circuit breaks
open and automaticequipment at the switching station attempts
to reclose the breaker. Thus the power on task may be interrupted
while in progress by a second power failure. If the program is
not prepared for this sequence of events, the system fails to
operate.
The CATAD system power on task waits for about one minute
before restarting the system after a failure to ensure that the
restoration is permanent.
Executive System
The principal executive system program development problem
was to incorporate the executive system with minimum modifica-
tion to the computer operating system structure. This permits •
incorporating computer operating system improvements without ma-
jor executive system changes. Thorough analysis of the operating
system was necessary to achieve this objective.
An early executive system version developed by the contractor
required a large number of transfers from the disk system to in-
itiate a task, resulting in excessive system overhead. The present
system minimizes disk transfers, significantly reducing overhead.
212
-------
Remote Station Telemetry
Contact status input circuitry is installed for either 40
or 65 contact inputs. External devices may be connected to only
a portion of these inputs. It was assumed that if no 'external
device was connected, the input would appear to be in the open
state. When this assumption was'used in programming, spurious
alarms occurred for non-existent devices because the hardware cir-
cuitry made a non-existent contact input appear to be closed. The
data acquisition system program now masks out all non-existent con-
tact inputs in status input data before checking for status changes,
West Point Satellite Terminal
West Point Satellite Terminal programming problems related to
dual use of the terminal for two-way message communications with
the central terminals initiated by the satellite terminal opera-
tor; and for transmission of random alarm messages initiated by
the central computer to the satellite terminal. A series of cen-
tral terminal alarms tied up the satellite terminal so that the
operator could not enter emergency commands or request data dis-
plays. Under other conditions low priority data entry by the sat-
ellite terminal operator could preclude transmission of critical
.alarm messages from the central computer. If, in central compu-
ter polling or message transmission, the satellite terminal failed
to receive the normal end of message code sequence, the satellite
terminal would hang up. The terminal would accept neither mes-
sages from the central station nor any data entry at the satellite
terminal. The condition could be cleared only by turning off the
satellite station power supply and turning it back on again, re-
'initializing the logic. The logic could not be reinitialized from
the central station. These problems were resolved by a combin-
ation of hardware and program changes.
The first problem was resolved by a program change giving the
satellite terminal operator control of the data entry keyboard for
a.fixed time interval after he initiates a message transmission.
The data entry procedure follows:
1. The operator depresses the "enter" button.
2. The central computer detects the request on its next poll
and sends a message to the satellite terminal requesting
the operator to enter his command or data request.
3. The operator has 3.0 seconds to enter a coded command or
data request. During this time, no alarm messages will
be transmitted and the operator is in control. If the
operator does not transmit a message within the alloted
time, any alarm messages waiting for transmission will be
transmitted and the operator will have to reinitiate his
entry procedure. To prevent the operator from interrupt-
ing an alarm message by depressing data entry keys at the
213
-------
satellite terminal, a hardware logic change was made to
the satellite station equipment. This change disables
the entire keyboard when a beginning of message code is
received from the central terminal and reactivates it when
an end of message code is received.
The terminal hangup problem was corrected by a hardware logic
change which enables the logic to be reinitialized by a start of
message code. When the computer fails to receive a satellite ter-
minal response indicating that the terminal has received a data mes-
sage or a poll after three attempts, the computer transmits a start
of message code to reinitialize the terminal logic and reinitiates
the transmission.
Renton Satellite Terminal
The initial function of the Renton Satellite Terminal was to
simulate the hard-wired West Point Satellite Terminal. It was not
practical to require the supplier to furnish the system software
since the operation of the satellite terminal involves interaction
with hardware and software (which is non-standard) of the central
station equipment. With divided responsibility for central sta-
tion and"satellite terminal hardware and software, the difficulty
of determining problem sources would be extremely complex.
The programming of the satellite terminal was undertaken by
Metro's consulting engineer, who was programming the central sta-
tion. With correctly functioning hardware, a proven real-time
operating system, and adequate hardware and software documentation,
this should have been a relatively simple programming task. How-
ever, as a result of deficiencies in hardware, supplier furnished
software, and documentation, the job became a nightmare. The in-
teraction between hardware and software problems is discussed
under the heading of hardware problems.
The software problems fall in two general categories as
follows:
1. Operating system problems
2. Documentation problems
Operating System — The real-time operating system was furnished
as a binary paper tape. When the tape was loaded into the com-
puter, the system failed to operate properly. It was found that
the clock interval used for timing by the operating system was
inconsistent with the basic hardware clock interval. The paper
tapes furnished by the supplier were for a new model of the com-
puter, which was not available at the time of the bidding. This
new model is generally software compatible with the system fur-
nished to Metro, but has a different clock cycle. The problem
was temporarily corrected by a program patch until a correct tape
was received.
214
-------
The routine included in the system for operating the serial
data link was not applicable to the CATAD system. The routines
furnished were designed for communication over a two-wire half
duplex data link, rather than the four-wire full duplex link which
was furnished. Mode switching required for a two-wire link is in-
appropriate for a four-wire link. In a two-wire link the trans-
mitter carrier must be turned on before transmission is begun by
energizing a relay and turned off after transmission is complete
by de-energizing the relay. A command to turn the transmitter on
or off results in an interrupt when the transition is complete.
The transmitter on-off relay had been strapped in the "on" state to
eliminate contract bounce as discussed in connection with hardware
problems. Therefore, no interrupt was received when the transmitter
was commanded to turn on or off as expected by the routine. The
routine was, therefore, revised for operation over the four-wire
circuit which, in effect, simplifies the routine. At the same time
the routine was modified substantially to revise the input sequence
to facilitate use by applications programs.
Documentation — The CRT display and the keyboard printer would not
respond to commands under the real-time operating system although
they would operate using the supplier furnished diagnostic errors in
the documentation which gave incorrect addresses for both devices.
A major problem resulted from errors in the basic documentation
of machine language codes for three logical operations and one input-
output operation. The supplier's documentation was used to create a
cross-assembler to enable the programmer to assemble programs for the
Renton terminal minicomputer on the central station computer. This
makes the more convenient and faster peripheral equipment at the
central station available for assembly of programs for the minicomputer
which has only paper tape for program input and output. By using the
cross-assembler the programmer may prepare his programs on punched
cards, which facilitates program editing, list his assembled programs
on the high speed line printer and punch his object programs on paper
tape for entry in the minicomputer. The errors in documentation
resulted in the generation of incorrect program code by the cross-
assembler. The documentation errors were discovered after much time
and effort and the cross-assembler revised accordingly.
FLOW CALCULATIONS
CATAD system planning anticipated that flows calculated from
manufacturers' pump unit performance curves might prove unreliable,
si^ice sewage pumps operate in a harsh environment, pumping abra-
sive solids and corrosive solutions. Abrasion of pump impellers,
solids deposited in pump unit intake sections, and particularly
recirculation due to abrasion of the impeller housing wear rings
were expected to degrade pump unit efficiency relatively quickly.
Actual operating experience indicates that this degradation takes
215
-------
a period of years and that pump efficiency is not likely to drop
more than 20 percent before worn wear rings are replaced and
other necessary maintenance is performed. Experience with
calculated discharges for the pump stations in the system leads
to similar conclusions. Manufacturers' pump unit performance
curves may be used to calculate reliable flows provided that
critical analog sensors, particularly pump speed sensors, are
reliable.
Because flows calculated using performance curves were
considered dubious, force main pressure sensors were installed
at many pump stations to provide alternative means of calculating
station discharge. An extensive program was undertaken to cali-
brate the pressure gages using the Allen Salt Velocity Method.
Plows generated by these gages have not been entirely reliable
because of rapid fluctuation of analog pressure values. Pressure
variation of 2 percent of full gage span can alter the calculated
flow by as much as 10 to 20 percent. Force main pressure values
are filtered before being used to calculate flow.
Suction head at the pump stations is measured by a wet well level sensor.
There is no provision for direct measurement of discharge head at any station,
although discharge head is needed in flow calculation. Initially, the static
discharge head was assumed constant, defined by the elevation of the pump
invert or of the crest of a discharge weir at the downstream end of the force
main. This assumption was inadequate, and an iterative procedure was used to
calculate depth at the downstream end of the force main. A depth was assumed
at the end of the force main and an assumed discharge head was obtained. The
computed force main discharge was compared with the computed pump discharge
and depth calculations were revised until the computed pump discharge and the
force main discharge agreed.
Extensive testing of storage calculations revealed two deficiencies in
the flow calculation procedures for regulator stations. No allowance has
been made for the effect of interceptor backwater affecting the tailwater at
a regulator gate and no transition had been provided between fully-submerged
and free discharge conditions. A backwater allowance was added, and a method
of calculating the degree of gate submergence was developed.
Trunk Inflow and Storage Calculations
Considerable effort has been expended to find a workable method of
calculating storage in trunk sewers upstream from regulator stations and
inflow to the storage reach. The problem results because upstream conditions
(i.e., inflow into the storage reach) are computed from observed downstream
conditions. Since inflow changes occur independent of downstream conditions,
they appear downstream after a delay equal to the wave travel time. Unsteady
flow equation solutions must, therefore, assume a fixed relationship between
inflow and depth at the upper end of the storage reach since neither is
measured directly. This assumption is. satisfactory unless the inflow rate of
change is altered, which may result in an unstable solution caused by the
time lag between the 'inflow change and the corresponding downstream level
change.
216
-------
Three development methods were attempted to solve the
calculation of storage:
1. A direct finite difference solution of the unsteady flow
equations;
2. A modified finite difference solution of the unsteady
flow equations, using simplified steep-fronted waves in
an idealized storage reach to represent changes in flow;
3. A method based solely on a continuity relationship with
a simplified water surface profile.
Method 1 was tested using various inflow predictions.
Solutions were rather lengthy and showed instability when large
flow changes were imposed at the upstream or downstream.ends of
the storage reach. Method 2, which relied on modeling surges in
the storage reach, gave incorrect results when compared with actual
test data. Method 3, a direct approach, requires little iteration
and gives results acceptably close to test data. However, the
computation exhibits a tendency to over-react to large flow changes
as did Method 1, requiring heavy damping of results. Method 3,
which' is currently in use for real-time storage and inflow compu-
tations., only satisfies the equation of continuity (i.e., outflow
= inflow + change in storage). . -
Pump Station Control
Pump station control experience indicates that special care must be taken
with smaller pump stations in the system, where major wet well level changes
can occur in five minutes or less. A station in automatic control is normally
checked every five scans to determine if control action is needed. Under dry
weather conditions, control action is normally initiated every ten minutes. It
became apparent that this is too long for some smaller pump stations, which can
pump their wet wells dry in ten to fifteen minutes. To ensure safe small
station control, each station is assigned a maximum scan interval which cannot
be exceeded wher the station is in automatic control as well as a specific
number of regular station scans between control checks. Each station is
controlled on a schedule appropriate to its operating characteristics.
Flow Routing
The critical part of the system hydraulic model is routing flows through
open channel sewers. This task requires a solution, or an approximate solution,
of gradually varied unsteady flow equations. A number of investigators have
applied themselves to solving unsteady flow equations and have attempted to
alleviate the two major problems confronted; ensuring a stable solution, and
avoiding lengthy solution time. Routing model development also faced the
difficulties of assuring a stable unsteady flow analysis under conditions of
gradually varied open'channel flow and of selecting a solution which can route
30-40 point hydrographs through a 70-element pipe system within the constraints
of real-time control.
217
-------
Three approaches to solution of the unsteady flow equations
have been tried:
1, Direct finite difference solution
2, Method of characteristics solution
3. Modified finite difference solution
Initially, a direct finite difference solution was attempted.
It proved to be unusable. A solution using the method of charac-
teristics was developed successfully and provided acceptable
results in test cases. It was extremely time-consuming, howevert.
requiring subdivision of pipe reaches and exact definition of
boundary conditions at numerous intermediate points in the sewer
transport system. A modified finite difference solution developed
for the Storm Water Management Model was tried and found accept-
able. This is not a true solution of the unsteady flow equations,
as it does not reconcile depths between pipe reaches. It has the
advantage of swift solution and the ability to handle long reaches.
It cannot be used for storage reaches where backwater exists be-
cause flow/depth relationships are defined by a modified Mannings
equation where depth is a direct function of flow. In such cases,
a normal routing is made to the end of the storage reach includ-
ing travel time through the reach. Actual travel time is less
through the deeper storage reach than for the assumed normal flow.
The inflow to the storage, reach is then routed through the reach
based on continuity alone, the amount stored being assumed to
conform to the storage rule curve. The amount stored is limited
by a family of curves relating inflow to available storage.
ELECTRICAL NOISE EFFECT ON SOLID-STATE SYSTEMS
Although solid-state digital equipment susceptability to
electrical noise — that is, random or spurious signals — has
long been evident, preventive measures are difficult to apply in
an industrial environment. Regulator stations are small and
scattered through an industrial area where electrical interfer-
ence is severe. The original design isolated the digital equip-
ment in its own separate metal enclosure to provide some shielding.
All incoming cables were shielded and analog cables had addition-
al shielding on each pair. In addition, status wiring was run
only a few feet in these cables. All status points were fed from
nearby interposing relays.
Random station noise problems were noted from the time of
the first installations. However, computer system operational
problems prevented orderly assessment of noise effects. Even nor-
mal electronic instrumentation equipment installation for CATAD in-
terface at Harbor Avenue regulator station presented extreme and
unusual instability problems. Since the station is located in the
area occupied by a moderately large steel mill, the mill's activ-
ities were immediately suspected. Measurements made with the
engineer's oscilloscope showed a familiar voice-modulated envelope
218
-------
on every terminal and conductor in the station. The source was
an A-M broadcast station whose transmitter was located nearby.
The problem cleared up a few days later when a bulldozer acci-
dently knocked over the transmitter tower.
After the CATAD system began regular reporting, it became
apparent that much noise was generated by relays and contractors
within each remote station. Relays operating frequently, such
as those associated with the pulse generators, and the pulse
generators themselves, were thought to cause many communication
failures.
A task group was set up in February, 1973, to evaluate this
problem and recommend solutions. Preliminary checks of several
stations showed that some telemetry circuits presented more com-
munication problems than others. Oscilloscope techniques showed
that many problems —• false alarms and lack of accurate command
capabilities— were caused by adjustment of the modem receiver
bias at both the central station and the remote stations. The
adjustment was extremely critical and slight maladjustment caused
erratic loss of transmitted data. After these adjustments were
made, the stations performed acceptably, but not ideally. The
task group obtained authorization to use outside consultants with
suitable test equipment to evaluate the noise problems and recom-
mend steps to alleviate false triggering of the TCU's input cir-
cuits. The consultant chosen was the Electromagnetic Interference
Group of The Boeing Company in Seattle. They made measurements
over a two-week period and submitted a detailed report with recom-
mendations for limiting the effects of external noise on the TGU
and for improvements within the TCU. (28) Their report was favor-
ably reviewed and the task group's final report was submitted to
Metro. The task group had already recommended TCU modifications;
shielding modem input leads within the TCU and frequent bias ad-
justment checks by the maintenance contractor. Additional Boeing
Company recommendations are presented in steps. First, they suggest
addition of a power line filter; second, addition of ceramic capa-
ci t.ors to several sections of the logic busses within the TCU;
and third, the correction of timer induced transients (pulse gen-
erators) by modifying or replacing the existing pulse generators and
associated relays. They suggest that these steps be taken prior
to considering more complex and costly measures.
The task group feels that digital equipment susceptability
to random noise — even that generated by a miniature relay — is
the major weakness of such systems. Present day state-of-the-
art permits digital equipment in almost any location. Metro's
noise problem is not unusual, but quite common. Much of the re-
search in this area has been directed to designing solid-state
input and output components and high noise immunity logic systems
that are less susceptible to random voltage spikes. (29, 30) .
The availability of input/output isolation systems incorpor-
ated in basic equipment led the task group to recommend that fu-
ture TCU's be specified with new designs using the latest isola-
tion techniques.
219
-------
LAKE CITY TUNNEL SURGE
The Lake City tunnel facilities were first utilized for
storage in 1971 after completion'' of a tunnel regulator gate sta-
tion. These facilities are shown in schematic form in Figure 5.
In 1972, system modifications provided a diversion from the
Green Lake trunk sewer into the Lake City tunnel "so the Municipal-
ity could repair a section of 138-inch trunk sewer lying downstream of the
MATTHEWS PARK
PUMPING STATION
FUTURE
REGULATORS
GREEN LAKE TRUNK
DIVERSION STRUCTURE
LEGEND
SANITARY
COMBINED
LAKE CITY
TUNNEL
REGULATOR
FLOW
FLOW
FIGURE 57
LAKE CITY TUNNEL SYSTEM SCHEMATIC
diversion. While the diversion was in effect, hydraulic transients
in the Lake City tunnel system were so great that, on at least two
occasions, sewage overflow occurred at the north portal of the Lake
City tunnel. These hydraulic transients took place during extremely
high flows in the entire sewerage system.
220
-------
Facility Description
The Green Lake diversion structure consists of a modified
existing caisson structure on the Green Lake trunk sewer and a
48-inch diameter conduit connecting to an existing access struc-
ture on the Lake City tunnel (see Figure 58). The existing Green
Lake trunk sewer caisson was divided with a wall .containing a
36-inch x 36-inch sluice gate installed in the flow line of the
trunk sewer. The caisson wall was extended vertically with the
top forming a weir crest over which trunk sewer discharge in ex-
cess of the capacity of the 48-inch interconnecting conduit could
return to the downstream 138-inch conduit.
This bypass or diversion structure was constructed to facili-
tate repair of 6900 feet of 138-inch existing trunk sewers. All
reclamation work on the downstream facility was done from inside
the 138-inch conduit. A continuous water stage recorder installed
in the Green Lake trunk sewer caisson monitored water surface el-
evation and provided an indication of flow diverted to the Lake
City tunnel. Green Lake trunk diversion to the tunnel commenced
November 1, 1972, and remained in continuous operation until
March 2, 1973, when the discharge was returned to 138-inch conduit.
Diverted flow averaged approximately 18 cfs and varied from 7 to
41 cfs during dry weather. With the onset of storms, the diverted
flow increased markedly with peak rates sometimes exceeding 180 cfs
diverted to the Lake City tunnel.
The .Lake City tunnel system under the operating conditions
during late 1972 and early 1973 consisted of five primary elements
(see Figure 59). These elements are:
1. A drop structure just downstream of the Lake City tunnel
regulator providing a means for lowering the flow in the
Lake City tunnel to the North interceptor system.
2. The Lake City tunnel regulator.
3. 'The Lake City tunnel consisting of approximately 17,400
feet of 96-inch conduit transporting sewage between the
Lake City tunnel regulator and the Matthews Park pumping
station.
4. The Green Lake diversion structure approximately 6000
feet upstream of the regulator.
5. The Matthews Park pumping station and its force main.
Three of these elements have static operating characteristics,
while (1), the regulator and (5), pumping station, operational
characteristics are controlled by the CATAD system.
221
-------
w
w
CO
oo
in
PS w
CO H
D so
H « H CM
PM < CO
^ 04
g
S H
W Q
W
O
222
-------
m
W
PS
§
H
O
O
CJ
o
oo
o
ro
.8
o
O
223
-------
Operation Criteria
The Lake City tunnel regulator and the Matthews Park pump-
ing station provide storage during periods of high flow in the
Lake City tunnel system and upstream of the Matthews Park station.
CATAD control of the Matthews Park pumping station is described
in Section V. CATAD controls the regulator station as a directly
operated gate, also described in Section V, with the following
exception. When the regulator station is returned to local/auto-
matic control, the gate is programmed to open fully and remain.
Controls are adjusted to provide safe return to the fully-opened
position without hydraulic overload to the drop structure down-
stream of the regulator.
When the Green Lake trunk was diverted to the Lake City
tunnel, CATAD console operators were appraised of the expected
flow increases. They were advised to expect flows in the neigh-
borhood of 140 mgd and to maintain a storage elevation no greater
than 161 feet, upstream of the regulator gate. They were also
requested to limit gate flow under controlled operating conditions
to 200 cfs.
Hydraulic Transients
Continuous water stage recorder charts monitoring the Green
Lake system during the diversion to the Lake City tunnel indicated
that on several occasions transient conditions occurred in the in-
terconnected systems. Ten incidents on separate dates can be in-
ferred from the charts. The chart records indicate hydraulic
transients of considerable force and duration with repetitive and
periodic characteristics. One overflow incident at the north por-
tal was not accompanied by indications of transients on the diver-
sion chart. Probable causes of surging and oscillations of the
water surface elevation include: exceeding recommended limitations
of storage elevation in the Lake City tunnel; a closed or virtually-
closed gate at the downstream end of the tunnel; a substantial
volume of liquid diverted to the tunnel from the Green Lake system,
probably at widely-varying large rates of flow; and inadvertent
loss of pumping rate computer control at the Matthews Park pump-
ing station. These severe incidents each occurred during periods
of protracted storm conditions.
Available data tends to correlate the events although there
are many discrepancies. The compressed time frame of the diver-
sion structure water stage recorder chart and inaccurate hour
setting of the charts present difficulties. Operational records
on the Matthews Park pumping station are incomplete. In some
instances, only hourly logs were available for review.
224
-------
Conclusions
It now appears that the principal factors contributing to
the transients in the system were significant changes in the rate
of flow directed to the tunnel from the pumping station, the Green
Lake system, or both concurrently without proper regard to re-
maining storage volume. The potential for significant surge or
other transient phenomena is less when the pumping station is the.
single major source of flow.
Gate movements affect velocity change, but are much less likely
to induce transient conditions. In a full conduit, the rate of
gate movement is such that a pressure wave generated at the gate
requires about 9% seconds to travel the length of the conduit and
return to the gate. In a partially-filled conduit, a wave-reflected
free-water surface is closer to the gate, reducing the round trip
distance and time interval. Gate opening area change either in
the opening or closing direction possible in approximately 10 seconds,
would normally effect little water velocity change in the conduit
upstream of the gate. Some pressure change will occur. ,
Lack of gate movement control, particularly in the opening
direction, significantly affects flow conditions at the drop
structure entrance. Hydraulic model tests of the drop structure
and approach channel determined the discharge characteristics with
and without a gate installed in the approach channel. With storage
in the conduit, discharge under the partially-opened gate is at a
very high velocity, and the water prism assumes a position of
severe superelevation at its contact with the chamber spiral boun-
dary. Extreme turbulence is present. Tests indicated that a gate
opening and head relation should limit discharge rate to about
200 cubic feet per second through a partially-opened gate. Tests
without a gate installed (equivalent to the gate in the fully-
opened position) under uninterrupted flow conditions and without
storage indicate that normal progressive discharge rate changes
have no appreciable effect. Conditions are very stable and the
performance effective up to a discharge rate twice the capacity
of the tunnel.
Pump Cavitation
Artificial pump speed versus wet well limitations were added
to computer command programming in the fall of 1972 after an op-
.erator's inadvertent entry of an erroneous pump speed request
caused extreme pump cavitation at one of the pumping stations.
Automatic lockouts provided in the control system to override the
CATAD controls on low and high wet well conditions were not Suf-
ficient to prevent the cavitation.
To better understand the causes of the cavitation problem,
a general description of the pumping station flow control system
and the selection of the pumping unit follows. After initial
225
-------
flows in the sewer at the station site and the ultimate flows to
be handled by the station are determined, the sewer is sized to
maintain a minimum velocity of two feet per second. Once the
pipe diameter and gradient of the station influent sewer are de-
termined, velocity depth and quantity depth relationships for all
depths of flows are known. These relationships are graphically
represented as curves "A" and "B" in Figure 60. They are valid
as long as a free-flowing regime is maintained in the pipeline.
It^is impossible to strictly maintain this theoretical relation-
ship, but it is sufficient to assume that pump capacity varies
directly with speed. This is assuming three pumping units of
similar characteristics are required to accommodate the maximum
flow conditions at the station, and the wet well pumping program
would be as shown in Figure 60.
Pump purchase specifications should include the net positive
suction head (NPSH) available for the pump along with the other
operating criteria. The NPSH available is a function of the sys-
tem in which the pumps operate and can be calculated for each
installation. The station pump flow control curve sets one of
the criteria for calculating the NPSH. The available NPSH should
be greater than the required NPSH of the pump. When there is in-
sufficient available NPSH, pump cavitation occurs.
Pump command limitations provide that no pump under computer
control shall have a speed greater than that allowed a pump in
the operating mode 1 (see Figure 60). These limitations provide
practical control to prevent pump damage, but enable CATAD pump-
ing control programs to use the storage available in the inter-
ceptors. Actual operating data on one station served by a 102-
inch interceptor allows a storage volume greater than 2h million
gallons to be developed through CATAD programs which draw the
influent interceptor below normal flow depth ranges.
FIELD EQUIPMENT CALIBRATIONS
Force Main Calibration
Force main calibration at selected Metro pumping stations
was undertaken to develop a computer formula for accurate indica-
tion of flow at selected locations. This calibration effort also
provides a reference index for checking force main deterioration
and deterioration of pump impellers in the pumping stations.
The work was divided into three basic areas:
1. Field Work
2. Computations
3. Permanent Equipment Selection
226
-------
CM
UJ
• *
UJ
• to
U.
CL
0
k
C
/~
\
> <
-
^
N
\,
r
*
- U-
CJ
o
ii
CJQ-
T> (
3 i
.
\
X,
>
D
3
"*^,
\
\
f
*
O
U.
.5
(CO.
s
0 <
\
V
/\
0 I
3 <
\
V
\\
\
o
3
V
V
y
\
*
6 <
»-
0
0
Ul.
"o.
UUJ
OQ
\
\
\
\
\
\
o <
3 <
1-
- ft: -
Ul
ce
UJ
UJ
1
\
\
\
.
\
\
M
3 <
N.
\
\
A
3
to
§ *
CM z UJ
< * Ul
— so in
l Ij OT
-J. ": <
_i
*! u. O
0 \£
CD <
0 UJ >
*~- ^ • o
o < **
* u. r!
.0 o !*J
» z z
0 = 5 Q
a h- ==
* i o: Z
O5O <:
1 °-
ro 0 ^
f_i £ 0
^ n!
0 U-
o
JO H±d3Q
m
Hid30 0 IV
o aod i3S iius si
_.._ SIH1
3AI10V ION SI HI 300W
v>
to =
UJ 5
O.ZUJ
(0300
Ul
QNV
H3M3S NOdO SQN3d30
SNIiNnH lN3A3«d 01
dV1U3AO lN3)3lddnS
1 Sdwnd z
I n 3QOW
Ul
JUJ
^UJ
u.
CO
T3A3T TI3M 13M
1S3WOT IV MOTd 0 01
9NIQNOdS3MaOO Q33dS
I 3QOW
o
cr
Q.
o
tc
«-
z
o
o
Ul
Ul
_l
Ul
o
VD
o
o
H
H
Q
W
H
ft
CO
9
H
1
EH
3
O
U
w
227
-------
Field work included developing a method for testing of force
mains; developing equipment for force main calibration; and running
actual field tests. Computations involved test data review and
fitting to fixed mathematical formuli; and selecting the best
formula for the computer program. Equipment selection involved
selecting transmission equipment, installing the equipment in the
field and calibrating permanent equipment at each station.
Force main calibration test stations were chosen with two
objectives: first, the service areas should provide representative
flows so that their results could be interpreted to indicate flows
in other areas of the system; secondly, stations such as the Inter-
bay station, could be used to check flow calculations and backwater
curves run through a large section of interceptors. Stations
with force main transmitters are listed in Table 16.
Table 16. FORCE MAIN CALIBRATION STATIONS
West Point System
Kenmore
Matthews Park
30th N.E.
Interbay
Renton System
Juanita Hts.
Kirkland
Bellevue
Sweyolocken
N. Mercer Isl.
S. Mercer Isl.
The Wilburton and Heathfield pumping stations were included
in the original list.
The Wilburton station was deleted when analysis showed un-
usable force main pressures. This force main discharges into the
middle of the Wilburton siphon, causing a varying dynamic head on
the force main which cannot be calibrated in proportion to the
flow through the station. The Heathfield station was deleted
after initial calibration efforts. Preliminary analysis of sta-
tion data obtained indicated up to 30 percent difference in the
repeatability of calibration runs. Additional runs after cali-
brating equipment was rechecked verified the discrepancies. It
is believed the vertical alignment of the force mains in this
station creates a partial air block in the force main line, caus-
ing an uncalculable restriction in the main. Since this station
is scheduled to be replaced within ten years, it was decided to
eliminate the station and not try to correct the force main
problems. Renton system maintenance personnel reported there are no
problems maintaining flows for the area that the station services.
Various force main methods were reviewed including: the
use of flumes or weirs; total count method using radioactive iso-
topes; and the salt velocity method. Flumes and weirs were dis-
carded because of capacity problems; the ratio of minimum to
228
-------
peak flows; and installation problems at a location which could
isolate the flow of the pumping station. Force main discharge
locations into the receiving interceptors did not readily lend
themselves to installing flumes or weirs.
The best method might be the radioisotope total count meth-
od. However, the Municipality's stations are designed with pump-
ing capacities in excess of today's flows. The availability of
water for testing a station is a problem. Low flow rates can be
calibrated by storing in sewers upstream of the station. Occa-
sionally, an extremely heavy rain in the service area produces
an individual station flow close to present connected capacity.
Calibration of high flow rates must be done during periods of
storm flows. Radio-active isotope use requires that a radio-
logical safety officer be available during the test period.
Since some test schedules depended on the weather, a method_util-
izing Municipality personnel would avoid keeping a radiological
safety officer on call.
The method selected is a variation of the "Salt Velocity
Method of Water Measurement" by C. M. Allen and E. A. Taylor. (31)
The first force main calibration tests were made at the Kirk-
land station. Instrumentation, primarily pneumatic, was borrowed
from the instrument shops to test the validity of the proposed
program. Three calibrated test gages were purchased. These tests
showed the^ need for more accurate equipment and additional instru-
mentation was obtained.
Conclusion and Recommendation
The force main calibration program provided the CATAD system
with a convenient method of developing primary flow meters in
existing pumping stations where flow meters were not provided in
the original installations. The program developed calibration
methods for the force mains and mathematical formuli for computer
programs. Because of the significance of this newly developed
method of flow analysis, and possible applications in other cities,
detailed description of the methods and equipment are described
in Appendix I.
The following procedures are recommended:
1.
Force main calibration should be rechecked in a contin-
uing program. Each force main should be recalibrated at
least once every five years.
Force main pressure transmitter range elements should_be
changed to fit the actual operating ranges found within
the stations. The unit originally installed at the Ken-
more station has a minimum span of approximately 40 psig,
Units with a minimum span of approximately 5 psig were
substituted.
229
-------
3. Flow meters should be considered for variable speed
pump controls. If practical, flow meters, rather than
force main pressures, .should be used to provide remote
flow indication through the CATAD system.
Pump Speed Calibration
Accurate representation pump speed proved difficult. Attempts
to calibrate pump speeds using a standard strobe light to stop mo-
tion on some particular point of the pump shaft showed that each
technician came up with a different set of values, varying 10 per-
cent or greater. Even least squares analysis of the test data pro-
duced unacceptable values. When this data reached the CATAD con-
sole, the results did not correspond to other known station data,
generating the many invalid pump speed alarms.
The pump speed analog value is generated by a d-c tachometer
generator coupled to the pump shaft, producing a known voltage per
rpm (see Figure 61). A pump speed indicator is located on the
pump station control panel. The CATAD equipment parallels a drop-
ping resistor and a voltage-to-current converter with the local
speed indicator to provide a signal to the TCU. The resistance is
adjusted to produce the desired 2-10 volt output signal in the TCU.
Most of the station pumps are driven by electric motors and the
SPEED GENERATOR
STATION SPEED INDICATOR
VOLTAGE TO
CURRENT
TO TCU
FIGURE 61
PUMP SPEED ANALOG CIRCUIT
230
-------
10-volt level adjustment is slightly greater than the synchronous
speed of the motor.. Calibration problems of the pump speed ana-
log signal were discussed with the Metro electrical design depart-
ment. After a search of the literature of equipment available,
they decided that no equipment on the market could provide the
desired pump speed calibration accuracy. They developed a retro-
reflective photosensor head which, with a frequency counter, pro-
vides an accurate representation of pump speed. Equipment use.d
for pump speed calibration is listed in Table 17.
Table 17. PUMP SPEED CALIBRATION EQUIPMENT
Component
Computing Counter
Retroreflective
Photo Sensor
(Schematic in Appendix C)
Tripod
Characteristics
Monsanto Model 107A
Metro Electrical Engineering
Department Constructed
Heavy-Duty Photographer's Tripod
with Tilting Head
The computing counter was selected because its ability to
indicate rpm with an update at least every four seconds in the
low frequency range of 100-500 rpm. The device can also be used
as a frequency meter for other electronic work. The retroreflec-
tive photo sensor was job-built by the Metro electrical engineer-
ing department to specifically meet the pump speed calibration
need. The unit was miniaturized so that it could be readily fo-
cused through pump frame handholes. Final size is approximately
1 and 3/4 inches high by 2 and 1/4 inches wide by 4 and 1/4 inch-
es long. The unit contains a source light and photoelectric cell
to pick up the reflected light which indicates frequency as well
as the electronics required to give the frequency impulse for
.transmission to the computing counter. The electronics are
driven by a 9-volt battery, and the source light is driven by a
3-volt battery, providing ;70 to 80 hours of usage. The unit oper-
ates successfully up to approximately four inches from the shaft.
A tripod is used to mount the sensor and hold it steady adjacent
to the equipment. An electrical conduit extension shaft permits
the tripod to suspend the sensor over larger pump casings. Co-
axial cable connects the sensor to the counter.
The pump shaft is reconditioned with a band of black paint sprayed
onto the rotating shaft. The pump is stopped and a longitudinal white stripe
of tape placed on the shaft to provide a reflective surface. The sensor is
mounted adjacent to the shaft and connected by cables to the counter in the
punp control room. The pumo is started, and the counter read while the pump
is held in various steady-"state conditions.
231
-------
The test data was then subjected to solution by the least-
squares analysis. A typical calibration table prepared for a
station is shown in Table 18, CATAD console checks on various
stations with the counter in place has shown less than approxi-
mately two percent error in pump speed versus actual speed read at
the station with the proper values entered into the computer pro-
gram*.
The pump speed retroreflective photosensor and computing
counter are a valuable maintenance aid for the Metro system main-
tenance department. It has been well received by the mechanics
working within the system. Municipality maintenance forces have
built a second sensor and acquired a second computing counter
for regular pump speed calibration.
A photograph of the calibrating equipment in use is shown on
Figure 62.
FIGURE 62
CALIBRATION EQUIPMENT IN USE
232
-------
Table 18- PUMP SPEED CALIBRATION DATA
1
7/11/72
Pump No:
Date:
V/I No. : v/i Hie
Pin No.: 5 G/H
Analog No. ; 7
Volts RPM
3.63 407.63
4.46 593.43
5.45 809.49
6.35 1009.4
7.11 1175.6
7.78 1322.9
8.70 1525.5
1730.5
9.63
Minimum Pumping
Speed 600 RP.M
Maximum Pumping
Speed 1745
Pump No: 2_
Date: 7/11/72
V/I No.: V/I 112C
Pin No.: 5 J/K
Analog No.; IQ
Volts RPM
3.5:8
4.50
5.41
6.33
7.14
9.75
8.54
7.43
9.61
9.61
400.24
601.85
800.03
1000.5
1177.4
1744.0
1483.4
1243.1
1738.9
1741.0
Minimum Pumping
Speed 600 RPM
Maximum Pumping
Speed 1745 RPM
Pump No
Date:
V/I No.
Pin No.
Analog
Volts
3 .54
3.76
3.99
4.46
5 . 36
6.27
7.05
7.07
7.36
7.91
8.62
. : 3
7/11/72
: V/I 113C
: 5 L/M
No.: 11
RPM
400.34
450.06
500.94
600.74
800.52
1000.04
1174.5
1177.0
1241.0
1365.0
1520.0
Minimum Pumping
Speed 600 RPM
Maximum Pumping
Speed 1745 RPM
233
-------
Level Sensing
Difficulty providing reliable water levels as an indication
of sewage flow appeared early in the calibration of equipment
feeding the CATAD system. Some sewage flows appeared to be going
uphill and tide levels at stations a few miles apart appeared to
differ by two feet or morel When the stations were originally
constructed, no attempt was made to adjust the controls of one
station relative to the levels in any other station; all stations
were calibrated on a "stand alone" basis. This complicated in-
dication of water surface profiles required for CATAD system
operation.
The normal Metro system level sensor is a pneumatic bubbler
located in the interceptor or trunk sewer with back pressure read
by a differential pressure transmitter. Transmitters have approx-
imately 300 inches maximum range with elevation and suppression
of transmitter zero and adjustable span. In most instances, level
indication locations within the system were sufficient to provide
the needed water surface indication. Pre-CATAD level sensor ad-
justment had generally been done with little care. Adjustment
within 10 percent of desired values provided adequate station oper-
ation. Adjustment and calibration used inexpensive gages of
questionable accuracy, which did not provide an accurate indica-
tion of water surface profiles. Sensor calibration for the CATAD
system was undertaken to provide accurate indications of water
surface profiles to a known datum base. Most stations are rela-
tively new and good as-built drawings were available to provide
sufficient information relative to bubbler tube elevations in the
trunks and interceptors. In a few instances, reference elevations
were brought in by survey crews. The primary level sensor calibra-
tion equipment was a Wallace and Tiernan series 65-100 portable
pneumatic calibrator and a shop-fabricated hook gage for accurate
determination of water surfaces at the time of calibration. All
sensors were calibrated to a fixed datum base; calibration ranges
were established; and sensors are now regularly calibrated to this
criteria.
FIELD PROBLEMS
Gate Position Study
The gate position study was undertaken after universally poor
gate position transmitter resolution was found in the gear-driven
gate operators, as discussed in Section V. Modifications of the
gear train final elements improved the reading somewhat, but
this remains one of the most critical readings in the system. The
manufacturer's position transmitter drive was not intended for
high resolution tracking.
234
-------
The Municipality began a search for a better transmitter and
system. One unit obtained for inspection operates on a tape reel
principle similar to the units used on some of Metro's hydraulic
gates. This model has a multi-turn potentiometer with an accuracy
about 1/100 percent of full range. No field testing of the unit
has been attempted. Such a unit, although accurate/ may not
stand the severe environment of the gate operator rooms. It does
not carry the UL or RM label for application in hazardous areas.
Presently available models would require excitation by an intrin-
sically safe system or would require special mounting within the
explosion proof gate operator housing.
Explosion Hazard Study
The explosion hazard alarm is monitored by the CATAD system
to make Metro immediately aware of the condition. Before the
CATAD installation this alarm was grouped with other priority 1
alarms on Metro's telemetry system. CATAD system analog trans-
mission capability allows the degree of explosive hazard to be
transmitted as an analpg signal along with the alarm status.
The standard alarm monitor units provide a proportional
voltage output for recorder monitoring. This signal is ampli-
fied and transmitted to the TCU as an analog point.
The explosive gas hazard must be monitored because the Seattle
portion of the Metro system in particular is subject to gasoline
and chemical dumps. Gasoline usually comes from gasoline truck
accidents (overturns) which the fire department washes down the
storm sewers. The vapors from these spills as well as from ille-
gal industrial solvent dumping present an extreme hazard in man-
holes and stations downstream from the point of entry. Little
methane hazard is present in the Seattle system because of the
relatively short sewage retention time.
Transmitting the station explosion hazard alarm only would
provide adequate early warning that some particular preset vapor
concentration had been reached in a station wet well. Transmit-.
ting the analog signal permits the console operator to assess the
severity of the problem and its rate of change. This informa-
tion helps to decide whether an operator should be dispatched to
the site and whether the treatment plant should prepare for diver-
sion or retention of the impending hazardous inflow.
Metro's operations group noted monitoring equipment problems
over the years (particularly with sensors) and undertook an in-
vestigation of equipment on the market prior to installation of
monitors in selected stations not presently equipped. The study
brought to light information regarding explosive gas monitoring
that is not available in the literature.
235
-------
The degree of explosion hazard is a qualitative measurement
based on lower explosive limit (L.E.L.). This measurement varies
widely with the type of gas and is a measure of the concentration
of a particular flammable gas under controlled conditions. In a
sewer system, the type of hazardous gas is not predictable and
the surrounding atmosphere is not normal. At best, the reading
is a relative value. Metro had followed the manufacturer's pro-
cedure and calibrated the monitor heads on methane in the past.
This calibration method reduces the monitor output signal for
gasoline by 40 to 50 percent of full scale L.E.L. reading. The
monitor should be calibrated using hexane or heptane gases. As
a compromise, heads were calibrated on a 2 percent methane gas
and transmitter gain set for 100 percent L.E.L., which resulted
in a normalization of the calibration point to equivalent pro-
pane. Propane is not used for calibration because it is un-
available in the air dilutions required.
The investigation found that some heads installed in exhaust
air ducts were totally ineffective because the high volumetric
flows in these ducts dilute the vapors. A study is underway to
relocate and increase the number of sensing points and to locate
the sensor heads closer to the sewage surface. A contract will
be let soon to install equipment in selected stations in order
to follow the flow of hazardous discharges through the system,
providing an accurate measure of the arrival time of the hazard-
ous material to the West Point plant where it must be contained
or bypassed.
Dexter Avenue Regulator Station
The Dexter Avenue regulator station (see Figure 63) was
constructed on the Central Interceptor in 1971 permitting storage
of combined wastewater in the 84-inch sewer upstream of the sta-
tion to minimize overflows to Lake Union from a side channel
overflow structure immediately downstream of the station.
The benefits of the Dexter Avenue regulator station were
described in "Maximizing Storage in Combined Sewer Systems." (6)
During a 1969 monitoring period prior to construction of the
regulator station, overflows to Lake Union occurred on eight
separate occasions. Analysis showed that the regulator station
would have eliminated all but one of these overflows.
The regulator station was provided with a modulating regu-
lator gate and an emergency bypass gate as shown earlier in
Figure 5D. Regulator gate control modes are shown in Figure 64.
Under dry weather flow conditions (Mode 1) the regulator gate
lip is set at a fixed elevation above the sewer invert. During
moderate storms when flow exceeds the capacity of the 48-inch
sewer downstream of the station (Mode 2) the regulator modulates
to store wastewater upstream and release a quantity of wastewater
236
-------
upstream and release a quantity of wastewater equal to the down-
stream sewer capacity. Under peak storm water conditions when
the upstream storage capacity is exceeded (Mode 3) the gate modu-
lates to maintain storage at the maximum level to minimize over-
flows. The bypass gate is provided with a tripping mechanism
backed up by an electronic tripping device. Once activated, the
bypass gate moves to a fully open position and remains open until
it is manually reset.
Under peak storm water inflow conditions (Mode 3) the station
operates under the following setpoint elevation conditions:
Bypass gate trip
Regulator gate open
Opening neutra
Closing neutral
Regulator gate closed
4- V»=3 1 __. "3
T~-= 1 /I
J. d J. rr '
1
ueau
band
width
i
FIGURE 63
DEXTER AVENUE REGULATOR STATION
237
-------
IS
inu.
I!
Ilil
§1 iffl
£""
§3=
in
vo
§
3
X!
w
Q
« CO
D ptj
O Q
H Q
P4 S
o
u
w
I
g
EH
a
g
238
-------
When the water surface elevation rises to the regulator gate
open position Celevation 2}, the regulator gate begins to open at
a constant speed to release wastewater from storage. The regu-
lator continues to open until the upstream water surface drops
to the opening neutral position (elevation 3) at which point the
gate motion ceases. Likewise, in Mode 3, the regulator gate be-
gins to close when the upstream water surface drops to the gate
close position (elevation 5) and continues to close until the
water surface rises to the closing neutral position. The neutral
positions may overlap 'to place the closing neutral position
above the opening neutral position. If the upstream water sur-
face rises above the bypass gate trip position (elevation 1), a
float mechanically initiates bypass gate opening.
On two known occasions, the bypass gate was inadvertently
tripped by high upstream water surface elevation. The resultant
surge overflow to the Lake Union outfall backed up in connected'
storm sewers releasing objectionable combined wastewater on lo-
cal streets. A study was initiated to determine the cause and
to recommend optimum setpoint conditions. The station was visit-
ed to determine in-place setpoint conditions. It was discovered
that, contrary to initial station design, the regulator gate open
setpoint was set at an elevation above the bypass gate trip posi-
tion. This situation (which was immediately corrected), virtual-
ly insured bypass gate opening in Mode 3 conditions.
A computer program was constructed to simulate regulator
gate operation under all flow conditions. This program simu-
lated station behavior under various setpoint conditions with a
triangular-shaped inflow hydrograph. The program was used to
calibrate Mode 3 setpoint conditions to maximize storage while
minimizing the possibility of inadvertently opening the bypass
gate. The program also tested the effect of bypass gate
opening rate on the subsequent rate of overflow into Lake Union.
Additional drawings of this regulator and discussion of the
simulation program are presented in Appendix J. The success-
ful application of this technique using available background
computer time on the real-time computer led to its use for analy-
s i s of other difficult hydraulic control situations in the
Metro collection system.
SPECIAL MAINTENANCE REQUIREMENTS
The interface boundary between the system assigned to the
CATAD contractor's maintenance and that for which Metro assumed
full responsibility was carefully chosen. The reason for limit-
ing the contractor's area of responsibility was discussed in
Section V under "Design Considerations."
239
-------
During preparation of the first CATAD interface contract for
the Elliott Bay Interceptor, the design engineer suggested that
Metro build a simulator to properly check out the stations before
acceptance. The number of multi-pin plugs to be checked point-by-
point indicated a large possibility of error and excessive engin-
eering time required for checkout. Metro assigned its consultants
to provide two units, one for pumping stations and one for regu-
lator stations. Investigation indicated a saving in common equip-
ment and space if a single unit simulated the telemetry control
unit (TCU) operation for both regulators and pumping stations.
A size limitation was imposed by the necessity to use the simula-
tor in a "package" pumping station (Kenmore) with a manhole cover-
type entrance. Portability placed a weight limitation on the unit,
The simulator was built with analog indicators and status lights
plus command pushbuttons. The unit measures 18 x 20 x 10 inches
and weighs about 30 pounds. A reversible legend plate fits over
the face of the unit. One side of the flip plate identifies
functions for a pumping station and the other side identifies
functions for a regulator or outfall station. The internal wir-
ing nomenclature (pin identification) for all receptacles is en-
graved on the face under the flip plate. Figure 65 shows the
simulator in operation.
FIGURE 65
METRO TELEMETRY UNIT SIMULATOR
240
-------
The CATAD contractor, Philco-Ford Company, had build a
station simulator1 providing signals to check the operation of
their TCU's and their control system prior to equipment delivery.
The manufacturer and Metro realized that the two simulators would
be extremely valuable in locating and correcting problems in the
remote stations whether on the TCU side of the interface or within
Metro's station equipment. Because of the nature of the system,
some malfunctions are difficult to trace. A faulty plug or broken
wire in the interface connecting cords can be very difficult to
find. Separating the two units easily determines whether the
problem is on the TCU side or on the station side. This would
apply even if one technician were responsible for all system
maintenance. Where two different maintenance groups exist, as in
the CATAD system, the problem is even more important because each
group can be expeditiously freed of the other's problems. The
station simulator appears in Figure 66.
FIGURE 66
CONTRACTOR'S PUMP STATION SIMULATOR
241
-------
All CATAD stations in Metro's system are pre-checked for
proper operation and connection and are operated directly from the
simula,tor prior to TCU connection. All alarm and status points
included in each station are checked prior to plugging in the TCU
cables.
The simulator detects all errors which might subject expen-
sive solid state TCU input components to damaging voltages due to
improper station equipment installation.
OPERATIONAL CONSTRAINTS
As pointed out by Grigg (11) and Tucker (32), local and muni-
c i p a 1 sewerage agencies have great difficulty developing a fully
controlled sewage collection system because of many factors whiqh
tend to constrain development. Metro encountered many of these
road-blocks, but a dedicated, hardcore group of staff and operating
personnel continued with the project, believing in the goals of
the control system. These people hammered at the multitude of
problems to forge a workable, if not quite successful, control
system. The remainder of this section describes significant con-
straints which were overcome and those which continue to limit
effort.
Sewer System Layout
West Point Maximum Flow - CATAD system control is applied to the
portion of the Metro area which routes sewage to the West Point
Treatment Plant. Since the West Point Plant must process all sew-
age added to the interceptor network by CATAD regulator stations,
the regulators must not divert a total flow which exceeds treatment
plant capacity. Maximum treatment plant capacity is dictated by
design and operational constraints. Design hydraulic capacity of
the plant is approximately 325 mgd, but a more workable flow of 250 mgd
is normally considered maximum. Operating conditions could de-
crease the maximum plant capacity. Emergency or preventive main-
tenance might require that a primary sedimentation tank be removed
from service, reducing plant capacity.
The treatment plant maximum flow limitation must be evaluated
by the CATAD console operator when the system is in supervisory
control mode. The automatic control system handles the limita-
tion by providing maximum flow data to the control program through
a keyboard entry from the central or West Point operator's console.
The maximum flow restriction at the West Point Treatment Plant
rarely has an adverse effect on CATAD system operation. The
designed maximum plant capacity is sufficient to handle a normal
balance of wet weather flow from the interceptors. An unusual
condition might exceed plant capacity, but this would be very
infrequent.
242
-------
Preventive maintenance which would adversely affect plant
flow can often be scheduled for dry weather periods. This re-
quires coordination between plant operations scheduling and the
control system supervisors. This situation is not an insurmount-
able obstacle. The only situation which might have serious adverse
effect on the CATAD system through limitation of plant flows is
failure of some piece of plant machinery which makes a portion of
the plant unusable. Normal plant maintenance is oriented to mini-
mizing such problems. A good plant maintenance program controls
the problem.
Interbay Pumping Station - Flow reaching the West Point Treatment
Plant travels one of two possible routes. Flow from the southern
collection area must pass through the Interbay pumping station;
from the north, if low must pass through the Fremont Siphon (see
Figure 67). Maximum Interbay pumping station capacity is approx-
imately 120 mgd. This limitation is often a key factor in determining
control for storm conditions. Even though the West Point Treat-
ment Plant may not be handling its full capacity, it could be nec-
essary to initiate a storage plan at Elliott Bay regulator sta-
tions since Elliott Bay Interceptor flow has reached the maximum
for the Interbay station.
Fremont Siphon - The Fremont Siphon flow limitation is similar to
the Interbay flow limitation. The Fremont Siphon limits flow
from the northerly area of the system. The siphon's maximum flow
is approximately 168 mgd. Greater flow would cause overflow into the
Lake Washington Ship Canal, an important fresh water body. Fresh-
water overflow avoidance has high priority. Thus, a storage
scheme must often be implemented upstream of the Fremont Siphon
to regulate flow at the siphon itself.
Flow limitation conditions at Interbay or the Fremont Siphon
may occur independently of a maximum flow restriction at the West
Point Treatment Plant. A heavy, localized storm involving only
the north area or only the south area might exceed the capacity of
the appropriate station without reaching the treatment plant limits,
Flow Regulation of the North Area - The northern portion of the
CATAD control area is dominated by pumping stations. Only one
regulator station is now operating. Future plans include four
more storage regulators along the North Interceptor (33, 12). The
existing regulator station, which controls the Lake City Tunnel,
provides a large storage volume for flow coming from the most
northerly parts of the system. A second major storage volume is
available upstream of the Matthews Park pump station which serves
the same drainage area as the Lake City Tunnel Regulator. The
rest of the north area is either controlled by pumping stations
with very small storage volumes or does not include CATAD control
stations. Thus very little control can be exercised over a rela-
tively large percentage of the North Area flow. The Lake City
243
-------
WEST POINT TREATMENT PLANT
DESIGN DWF= 125 MGD
MAX. = 350 MGD
*«v
*-s\
Vs\<
r«, -2\
ui IBL\
-~*&\*
BALLARD INTERCEPTOR
DWF = 5MGD
MAX. = 29 MGD
INTERBAY P.S
DWF = 45 MGD
MAX.= 120 MGD
FREMONT SIPHON
DWF = 71 MGD
MAX. = 168 MGD
LAKE CITY TUNNEL
REGULATOR
#UNDER SURCHARGED CONDITIONS
WILL PASS 400 MGD
FIGURE 67
FLOW LIMITATIONS AFFECT CONTROL STRATEGY
244
-------
Tunnel and Kenmore Interceptor storage must be used to avoid exces-
sive flows at the Fremont Siphon. Storm situations can exceed their
capability to limit the Fremont Siphon flow sufficiently to prevent
overflows. Additional stations in the north area will considerably
enhance CATAD control capability.
Environmental Conditions
Storm Types - The CATAD control system is used most often during
Seattle's winter rainy season. Spring and fall storms require use
of the system much less frequently than during the winter. Summer
storms, with very few exceptions, are too brief to require special
controls. However, overflow control during the summer recreational
season is most important.
In addition to the seasonal storm condition variation, differ-
ent storm types can be identified according to their impact on the
CATAD control system. The most difficult is a long duration, gen-
eral rainfall which is quite common during the winter months. These
storms produce nearly continuous rainfall for up to 4 or 5 days.
Intensities as high as 0.20 to 0.25 inches per hour are possible,
but the average storm intensity is in the area of 0.06 inches per
hour. Four to eight hour periods with continued rainfall of 0.10
to 0.15 inches are common in these storms. Such storms put heavy
demands on the CATAD system due to both large runoff volume-and
area-wide involvement. The large runoff volume will tend to main-
tain high levels in all areas, decreasing the limited storage vol-
umes to zero. The general geographic coverage of the storm pre-
cludes selective use of storage in low rainfall areas to ease the
impact in other areas. This type of storm is the most likely to
cause overflows simply because total hydraulic requirements exceed
system capacities.
The more common winter storm lasts from 8 to 24 hours with in-
tensities similar to the longer storm. These storms are also com-
mon in the spring and fall, with shorter storms in the spring and
longer ones in the fall. These shorter duration rainfall periods
display a characteristic common to the Puget Sound area: storm
activity, especially during the winter, generally approaches from
the southwest. With total rainfall duration of less than one day,
the rainfall tends to be less constant even though measurable pre-
cipitation may be occurring in all areas. Storm movement across
the basin tends to introduce areas of high and low intensity rain-
fall according to varying patterns of cloud cover and variable
land elevations (see Figure 68). The variable nature of this storm
type allows the CATAD system to control runoff more effectively
since storage can be alternate 1 y built up and released as "rain-
fall segments" of a storm traverse the area. Also, the storm's
directional nature helps provide advance warning of its impact by
involving outlying areas of the system first as it moves into the
area. Data gathered from the first rain-gaging stations to be af-
fected by the rainfall can often give a good indication of area-
wide effects.
245
-------
KOTEi
NUMBERS REPRESENT AVERAGE
RAINFALL INCHES/YEAR
FIGURE 68
ISOHYETAL MAP OF SEATTLE
246
-------
Storms up to eight hours long can be expected in any season.
Short storms may vary widely in total rainfall volume and maximum
intensity. Most storms approach from the usual southwest direc-
tion and can be handled by the control system in much the same
manner as longer duration storms. From late spring to early fall,
an occasional storm may approach from another direction. This
usually means a short period of intense rainfall, with little im-
pact on the sewer system. Most short duration storms can be com-
pletely contained by the control system with the possible excep-
tion of a very high intensity local storm where local trunk and
interceptor capacities are overwhelmed by runoff volume.
Tide Effects — High tides in Elliott Bay and the mouth of the
Duwamish River may make it impossible to overflow at selected sta-
tions since the tide level is above the trunk sewer level. If an
outfall gate were opened, salt water would flow into the outfall
increasing the total amount of water in the system. This situa-
tion cannot be permitted. Tides high enough to prevent the use
of outfalls along Elliott Bay and the Duwamish River are uncommon,
but they do occur; most frequently during the winter months.
Winter high tides are also aggravated by wind conditions. A
southwest wind during a winter storm may build a higher water
level along the northeasterly shoreline of Elliott Bay. This
condition converts a moderately high tide into one with signifi-
cant impact on the control.system. The original assumption was
that it would be extremely uncommon for a tide high enough to
limit overflows at many stations to occur simultaneously with a
heavy storm requiring many overflows. / However, during the winters
of 1971-1972 and 1972-1973, there were significant periods (total-
ing about ten days) when high tides and high storm inflows caused
trunk levels to be run higher than normally considered safe. If
this condition becomes aggravated in any way, it may become nec-
essary to develop a means of removing excessive storm water vol-
umes from the trunk and interceptor system to prevent flooding low
elevation sewer connections.
System Limitations
The CATAD system controls a sewer network which, for the
most part, predated CATAD. The sewer network was not designed
with computer control in mind. Some sewer network construction
and operation factors do not optimally integrate witlrv the CATAD
concept. Some of these restrictions may be unavoidable, but others
might be avoided by designing for special system requirements.
The most favorable condition is to design, the sewer system and
the control system as a single integrated package; however, the
number of locations where this approach is possible is very small
since the sewers usually precede the control system. However,
for areas using computer control, it is mandatory that sewerage
system expansion plans be formulated to take advantage, of control
technology.
247
-------
Regulator Stations with Small Storage Capacity — Three Metro
system regulator stations are notable for small storage capa-
cities: Harbor Avenue, West Michigan Street and Denny Way Local.
The Harbor Avenue and West Michigan Street regulators control
trunk lines with rather severe gradients. In all these stations,
the small storage capacity of the trunk lines causes two problems.
First, an intense but localized rainfall affecting one of the
stations creates runoff conditions which exceed the station
capacity and forces overflows regardless of available interceptor
space. Secondly, the small storage capacity of the lines makes
them useless in lessening the impact of a storm in another area.
If the system uses storage at one or more of the stations, the
storage is almost immediately exhausted, creating the double
problem of managing the rest of the system while releasing stor-
age in the exhausted area.
Pump Stations with Small Storage Capacity — Certain CATAD con-
trol system pumping stations also have small storage volumes
available. Among these are the East Marginal Pump Station in
the southern area and the 30th Avenue and Belvoir Pump stations
in the north. Storage volumes available at these stations are
so small that it is impractical for an operator to attempt to
use this capacity in the supervisory mode since the wet wells
could easily rise to a flooding condition before the operator
could respond to the condition. In any case, the storage
volume gained would be insignificant in a total storage strategy.
This limitation is equally true in automatic control mode.
Overflow Weirs Causing Reduced Storage — Most regulator stations
are protected against a critically high trunk level by an over-
flow weir. Weirs are constructed to act as a fail safe device
so that overflows will occur when the trunk reaches a designated
maximum level, even if the outfall gate does not function. This
protection mechanism may not work in a high tide condition.
Emergency weir overflow lines are normally protected by a flap
gate which will not allow high tide level waters to flow back
into the sewer system. The overflow weirs are usually not a
factor in the CATAD control system operation. There are, how-
ever, infrequent occasions where an emergency weir location lower
than the actual maximum trunk level prevents use of a small in-
crement of storage in a trunk. • This lost storage is usually
very small, but in some cases, could be quite valuable.
Two important overflow weirs occur in the system separate
from regulator stations. One of these is in the Elliott Bay
Interceptor upstream of the Interbay pumping station; the other
is at the Fremont Siphon under the Lake Washington Ship Canal.
248
-------
The purpose of the Elliott Bay Interceptor weir is to limit
the total flow reaching Interbay and protect the station from
dangerously high flows which might cause flooding. This weir
is positioned so that it does prevent the use of the total in-
terceptor capacity for storage. It is undesirable to allow
overflow directly from Elliott Bay Interceptor at this location
since primary settled sludge from the Renton Treatment Plant
as well as separated sanitary sewage flows are directed to the
West Point Plant through this interceptor. An overflow would
consequently have a more serious effect on water quality than a
stormwater diluted overflow directly from a trunk line. Control
system operation may use interceptor capacity only up to the
level of this weir.
The Fremont Siphon weir limits the maximum flow through
this structure. Overflow at the Fremont Siphon is directed into
a fresh water body which is very undesirable. Storage upstream
of the siphon must be manipulated to protect this location.
Stations Contributing Exceptionally Large Flows — The Hanford
Street trunk joins the Elliott Bay interceptor at the Hanford
II Regulator Station, The Hanford Street trunk carries flow
from a major area of south Seattle; the drainage area served by
the Hanford Street trunk is much larger than other trunks meet-
ing the Elliott Bay interceptor. This fact makes the Hanford
II regulator a major factor in the operation of the Elliott
Bay interceptor. Maximum interceptor flow is governed by the
120 mgd capacity of the Interbay Pumping Station. Under storm
conditions, the Hanford Street trunk may carry over 50 mgd, so
that this trunk line tends to have a disproportionate effect
compared to the other twelve regulated trunks feeding the Elliott
Bay interceptor. This imbalance causes control strategies for
many regulators on the Elliott Bay Interceptor to be dictated by
conditions within the Hanford Street trunk line. If the volume
of this trunk were introduced into the system as more numerous,
small connections, more variation in control strategies might
be available.
249
-------
Unsupervised Inflows — Larger trunks connecting to interceptors
are controlled by regulators, but many smaller trunks are not.
These unregulated connections make demands on interceptor capa-
city for which there is no real data gather facility. In periods
of general rainfall, this is not a significant handicap since
the contribution of the unregulated trunks can be estimated on
the basis of known regulated flows. However, even though the
flow in such an unregulated trunk may be estimated, it cannot be
controlled. In periods of unevenly distributed rainfall, esti-
mation of such flows is difficult.
Uncontrolled and unmonitored inflows to the interceptor
system hinder precise control system operation. If all inflows
were monitored and controlled, the CATAD system could effectively
predict and dictate flow conditions throughout the interceptor
system. As the percentage of monitoring and control stations
decreases, concessions must be made to approximation, which is
often barely adequate. For operation of a CATAD-like system,
the number of unmonitored and uncontrolled inflows to the system
must be minimized.
Critical Elevations — Operating experience with the CATAD sys-
tem supervisory control mode has indicated three locations where
storage use is impeded by low connections to trunk sewers. At
the Eighth Avenue South, Chelan Street and Brandon Street regu-
lators, sewage backup of low areas in nearby residences and bus-
inesses connected to the corresponding trunk sewers occurred
during some storms while trunk levels were below their designed
maximum. Evidence indicates that existing connections at levels
below the trunk line design are not protected by one-way valves.
Consequently, maximum storage levels at these stations must be
reduced below the original design. These limitations are not
now critical since the storage volumes in these trunks are not
heavily used for most storm conditions. It is possible that
these connections will have to be corrected to handle expanded
requirements in the future.
250
-------
OPERATING EQUIPMENT PROBLEMS
The CATAD control system depends on reliable, accurate opera-
tion of sewage handling equipment in regulator and pumping stations
and treatment plants to implement its control strategy. Scheduled
and emergency loss of operating equipment affected the CATAD sys-
tem efficiency during early operation.
Regulator Stations Out of Service - For several months during early
CATAD system operation, three major Elliott Bay regulator stations
were unavailable, partially or completely, for CATAD use due to
station modification. The Brandon and Michigan Street regulators
were completely out of service while being modified. The Hanford
Street II outfall gate was removed from service for an extended
period.
Removal of the Brandon and Michigan Street regulators from
service prevented control of a large portion of upstream storage
on the Elliott Bay Interceptor, dictating more drastic control
measures at downstream stations. Greater overflow volumes
occurred due to loss of upstream storage volume.
Loss of outfall gate control at Hanford Street II was not a
serious control problem. Station modification plans called for
loss of regulator gate control as well as outfall gate control.
However, the Hanford Street I station on the Hanford Street trunk
provides a gate for diverting sewage into a stdrm sewer which
serves as an overflow structure. Relative trunk level elevations
at the two stations allow storage held behind the regulator gate
at the Hanford Street II station to be relieved by overflowing at
the Hanford Street I station. Thus, the storage capacity of the
trunk line could be used without causing excessively high levels
as long as the Hanford I regulator was available for control.
This scheme was used satisfactorily while the Hanford Street II
outfall was out of service.
Gate Sensors - Regulator and outfall gate positions are sensed
with multiple turn wire-wound potentiometers mechanically linked
to the gate operating mechanism. Potentiometers and mechanical
linkages have failed completely or developed intolerable amounts
of variability as usage increases. Failures of this type cause
loss of automatic control of a regulator or outfall gate for
a few hours or days until maintenance can correct the situation.
In critical periods of intense rainfall, maintenance personnel
are usually on call to correct such situations, so that downtime
exceeding a few hours in critical situations is uncommon. Even
a few hours, however, can cause a large increase in overflow
volume.
251
-------
Pump Problems - All flow from the Elliott Bay Interceptor must be
pumped through the Interbay Pumping Station. Three engine driven
pumps are available at the Interbay station with a maximum of two
pumps operating simultaneously to; provide the station's design flow
of 120 mgd. During the first year of CATAD operation, numerous prob-
lems were encountered with the engines driving the pumps at this
station. The engines are fueled by either natural gas or propane.
At various times speed governor malfunctions, carburetion difficul-
ties and other engine problems caused pumps to go off line, dras-
tically reducing station capacity. When an engine drops off line,
an operator at the station must either manually restart that engine
or start the third engine to replace the one which failed. Since
the station is manned only on a regular 7:00 a.m. to 4:00 p.m.
Monday through Friday shift, engine failure during high flow con-
ditions can be disastrous.
For much of the wet winter season of 1972-73, engine problems
plagued the Interbay station,. For many months only two pumps were
available since maintenance of the third pump was delayed waiting
for scarce replacement parts. It was not always possible to run
both pumps at their designed maximum capacity due to problems
with an overspeed limiter on one of these pumps. Several times,
while the station was running at maximum available capacity, a
pump dropped off line. When pumping capacity is suddenly halved,
the station wet well level immediately jumps to a flooding con-
dition initiating total station shutdown by automatically closing
an influent control gate. When this occurred, the CATAD console
operator immediately called out maintenance personnel to attempt
to restart the station. In the interim, available storage capacity
was used in an attempt to prevent overflows until the pumps at
Interbay station were restarted. This strategy was often inef-
fective since a major portion of the total system's storage was
already filled by high storm flow conditions. Large overflows
often resulted. The critical location of this station makes
improved maintenance and operations procedures mandatory to keep
the station at full strength in future winter periods if the CATAD
system is to perform in an optimal manner.
Lake City Tunnel Surge - The major storage volume for the north
section of the CATAD control area occurs in the Lake City Tunnel
controlled by the Lake City Tunnel Regulator. This facility pro-
vides an exceptionally large volume of storage: up to 6.5 million,
gallons. During December, 1972, the Green Lake Trunk, which
crosses the Lake City Tunnel, was diverted into the Tunnel to
allow maintenance to be performed downstream on the Green Lake
Trunk. The increased tunnel flow decreased total storage time
available in this facility. During high intensity storms in
December, 1972, and January, 1973, tunnel storage was used up to
the maximum safe value indicated by available system data.
Twice, serious surge conditions occurred in the Lake City Tunnel
sending large volumes of sewage traveling up the tunnel and out
252
-------
through the influent structure at the upstream end of the tunnel.
The violent expulsion of a large .quantity of sewage from the tunnel
damaged the influent structure and washed out a gravel roadway
leading to the station. To prevent further damage, including pos-
sible tunnel damage due to excessive internal pressures, storage
in the tunnel was suspended until an investigation of the situa-
tion could provide an hydraulic explanation of the facility's
behavior. Preliminary analysis indicated the surging condition ,
was initiated by movements of the regulator gate at the down-
stream end of the tunnel while high head conditions existed due to
the large volume of storage. Diversion of Green Lake trunk flow
into the Lake City Tunnel may have been a contributing factor. The
diversion was terminated in March, 1973, and storage has been
allowed to levels considerably below the original maximums since
that time. Final operating guidelines will not be implemented
until a full study of the surge conditions is completed. Until then,
a; serious restriction in control strategy exists because of the
limited tunnel storage volume.
West Point Treatment Plant Operation Problems - Two conditions at
the West Point Treatment Plant required lower than normal limit
on plant flow and, in storm conditions, caused earlier overflows
because of more rapid usage of storage.
First was the loss of some primary sedimentation tank facil-
ities due to machinery malfunction. Some situations make it
necessary to completely drain a tank to make repairs. This de- .
creases the total flow rate which the control system can send to
the treatment plant with obvious adverse effects.
The second problem at the West Point Treatment Plant is over-
heating in effluent pumping engines. The plant must have pumps to
discharge its effluent to Puget Sound under high tide conditions.
At low tide, gravity flow handles the discharge. The problem of
high tides coincident with high storm flows was discussed earlier
in this section.. This problem is further complicated when the
engines driving the effluent pumps at the treatment plant have
been subject to overheating .problems through periods of prolonged
use. When a portion of the effluent pumping facilities must be
shut down under excessively high tide conditions, the control
system is faced with the dilemma of high storm flows, low treat-
ment plant capacity and no overflow capability.
Communications
CATAD control system operation has inserted a new group of
people into the standard operating procedures of the Metro sewer
system. Prior to CATAD all operating aspects of regulator and
pumping stations and other facilities were handled by personnel
based at the main treatment plant where the collection facilities
terminated. Operations and maintenance personnel were centered
253
-------
at the West Point and Renton Treatment Plants to support facilities
related to each plant. The CATAD control system involves the
facilities of both areas. No specific plan was made to introduce
CATAD system operation into the ^standard operating procedures of
the Metro sewer system. Metro has been largely dependent on a
consultant for engineering services pertinent to the physical
sewerage facilities. No specific plan was made to provide lines
of communication with these consultants for the CATAD facility.
Inadequate planning in these areas has introduced some difficulties
in CATAD control system operation.
Controlled Facilities Engineering Data - To operate any controlled
facilities in either a supervisory or automatic mode, it is neces-
sary to have an operational profile of the facility. This informa-
tion includes pipe inverts and sizes, gate inverts and sizes, pump
characteristics, etc. It was discovered soon after the initial
control system installation that such data was often missing,
inaccurate, or difficult to locate. The problem became critical
when the first phases of supervisory control were initiated. At
that time operators often found that there was not enough data
available to indicate limits of storage, flow rates, etc., that
could safely be attempted at a given facility. Time has alleviated
this difficulty as more information is found. However, new con-
trolled facilities being added to the system present many of the
same problems as the initial implementation. Difficulty obtaining
accurate physical data about stations indicates a deeper problem
of recording "as built" data on installed facilities. Lack of
this data hinders operation of the control system. In some cases
stations were not used even though they were physically ready for
control because the necessary operational data was not available.
Operator Records - To evaluate CATAD system operation and provide
improved operating procedures, it is necessary to have a history
of operation. The record-keeping procedure was especially im-
portant for early development of supervisory control procedures.
The need for good record-keeping procedures at the operator's
console was not foreseen, and procedures have developed slowly.
As the automatic control mode takes over system operation, such
records become less important, minimizing future problems due to
lack of such data. As the automatic control system replaces the
supervisory control procedure, the need for more than one console
operator diminishes. During the supervisory control period, two
operators attempted to cover all control requirements for the
system. The operators replaced each other on a rotating shift
basis during storm periods, making it necessary to pass opera-
tional information from person to person to maintain the contin-
uity of a control strategy. No specific procedures were designed
to implement this information transfer; consequently, operating
errors occurred. The overall impact of these errors was small in
terms of overflow volume. For a system permanently operated in a
supervisory mode, detailed communication procedures must be estab-
lished and checked to enhance operator efficiency.
254
-------
Interface Between CATAD and Operations Personnel — The CATAD con-
sole operator now has the most detailed knowledge of Metro sewer
system operation excluding the treatment plants. The CATAD con-
sole operator observes the operation of each facility within the
CATAD system and calls out operations and maintenance personnel
to check and/or correct any problems he finds. Some communica-
tion difficulty has occurred between the CATAD operator and opera-
tions and maintenance personnel due to the essentially different
"vocabulary" familiar to each. Regular maintenance and operations
personnel deal with the physical gear at the regulator and pumping
stations. The CATAD operator deals with data telemetered from
the remote station and attempts to interpret the operating condi-
tion of station equipment from the displayed data. The communi-
cation problem arises from the fact that a malfunction detected
by the control system and alarmed on the basis of apparent symp^
toms often has a cause in a completely different area. The con-
sole operator uses a selected set of computer data to make deci-
sions. It is difficult to give personnel involved with mechani-
cal/ hydraulic, electrical and other physical gear a good explana-
tion of a problem to aid in their maintenance effort. Field crews
generally know little about the operation of the computer control
system. They may not understand the vocabulary which the console
operator uses to describe a problem or comprehend the procedures
involved in getting data from the station to the CATAD- system.
It would be valuable to cross-train console operators in basic
maintenance skills,, especially instrument operation, and give
maintenance personnel a basic understanding of the control com-
puter "s operation.
Supervisory Control Console
A basic plan for the operation of the CATAD system from the
operator's supervisory control console was laid out in the ori-
ginal system specification. Console operating experience^has led
to many minor operating procedure revisions. Some significant
console problems deserve mention in a discussion of system operat-
ing difficulties. »
Events Printer — Console room hardware design provided enclosures
for the events and logging teletypes to reduce the noise level
from these devices. However, the noise covers became quite imprac-
tical. The cover design significantly lowered the noise level but
also made it difficult to see the printing on the.teletypes. One
purpose of the events printer is to log a verification message for
each operator command before it is entered. Using this facility,
the operator should visually verify each command on the events
printer before entry. The noise enclosure makes the operator move
away from his normal operating position to read the messages on the
events printer. This has led to an operating procedure where the
operator does not verify his commands, resulting in occasional
errors of operation, especially during the busiest storm periods.
255
-------
Lack of Operating Data — Original design of the operator's con-
sole called for a single station data display format on the CRTs.
This single format provided a display of the basic analog values
being scanned for each station in the CATAD system. However, one
significant piece of data, the accumulated rainfall, was not avail-
able except by special request for a printed log. Rainfall data
is quite important to an operator performing supervisory control
of a portion of the system since it indicates future runoff loca-
tions and quanitities with which the operator must deal. This
problem was solved by providing a group of display formats, each
individually selectable by the console operator, which allow a
greater spectrum of data to be displayed. This development has
made data available on the CRT displays which originally was
available only on one of the teletype printers. The operator's
task has been centralized at the control console with less depen-
dence on the peripheral printers.
Command Completion Difficulties — A major hindrance to early super-
visory operation of the sewer system was the lack of accurate data
on the operating characteristics of station equipment. The most
common problem was inaccurate information on the rate of regulator
and outfall gate motion. With inaccurate data, it was difficult
to get correct results from gate movement commands since the
expected gate performance did not agree with the actual situation.
As these problems were eliminated by collecting more up-to-date
data, further difficulties in the hardware telemetry system affected
the system's ability to successfully complete an•„operator's commands.|
These difficulties and their solutions are discussed elsewhere in
this report. From the console operator's standpoint, the net f
effect of all the commanding difficulties was the necessity to
reenter commands, often three or four times, to achieve the de-
sired results. This slowed implementation of the desired control
strategy and made the operator's hectic job even busier.
Pump Station Setpoint Restrictions — After a period of supervisory
operatzon of the system, the consulting engineer realized that the
commanding procedure allowed the console operatbr to request pump
operation which might result in pump cavitation. It was decided to
prevent the operator from requesting pump operation that might be
detrimental to the mechanical integrity of the sewer system. Ori-
ginally, this limitation would not allow the operator to request a
higher pump speed than would have been used by the station func-
tioning in the local control mode. This would severely limit con-
trol system performance since higher than normal pump speeds are
often used at the beginning of a storm period to pump as much
sewage as possible out of an interceptor to create the greatest
storage. The approach finally adopted consisted of checking a.
requested pump setpoint against safe limits established by a study
of the operating properties of the individual pumps. This solution
allows the greatest control of pump operation while protecting
against serious damage.
256
-------
Flow Commanding — In addition to the basic command procedure of
requesting gate positions and pump setpoints from the operator's
console, an additional function allows commands entered as re-
quested flows. This procedure allows the operator to request the
flow quantity in mgd which he wishes to have diverted through a
regulator gate or pumped by a pumping station. Programming pro-
vided by this capability checks the gate position or pump setpoint
periodically and issues commands necessary to maintain the desired
flow rate. The flow commanding facility eases the operator's chore
by making it unnecessary for him to continually monitor water level
changes at all the stations and adjust for variations. Programming
problems were encountered during the development and early imple-
mentation stages of the flow commanding algorithms. These problems
involved erroneous operation or complete failure of flow commanding.
The supervisory operator was often required to fall back on the more
basic command procedures. Flow commanding problems were eventually
eliminated and this facility is now integral to both supervisory and
automatic control schemes.
Satellite Console
Through the initial CATAD development period, the satellite
console at the West Point Treatment Plant has been the control and
observation point for the system when the central console facility
is unmanned. A special section describes the operation of this
satellite terminal. Reliance on this terminal has caused some
difficulties for the CATAD operator in maintaining the supervisory
control scheme.
Operator Training — West Point Treatment Plant personnel who
monitor the satellite console have training concentrated in treat-
ment plant operation. No effective special training sessions have
been provided to instruct the console operators in the operation
of the remote terminal or to provide them with background concern-
ing the facilities the CATAD system monitors. Treatment plant
operators have been forced to learn about CATAD system operation
by using the system. This is not the best approach since these
individuals perform the major portion of CATAD system monitoring.
Knowledgeable operation of the CATAD system requires understanding
functions of the various regulator and pumping stations, back-
ground in the physical constants pertinent to each station's opera-
tion, and knowledge of the operational methods of CATAD monitoring
and control facilities. Few, if any, of the West Point operators
have sufficient training or experience to develop this level of
understanding. Sickness and personnel turnover also contribute
to the variability of operator experience. The satellite terminal
operators lack of background makes the CATAD console operator's
job vastly more difficult. One of the CATAD console operators is
available on a 24-hour on-call basis to handle problems related to
the CATAD facility. The operator on call depends on the personnel
257
-------
monitoring the remote console to observe and recognize significant
alarm conditions and notify him of the need for action. Many times,
important conditions which should have resulted in immediate opera-
tor notification have gone unresolved for a number of hours. Even
more frequently/ the CATAD operator receives unnecessary calls at
all hours because the treatment plant operators cannot discriminate
significant conditions from unimportant ones. Consequently, the
CATAD operator has a much heavier responsibility. He must be able
to interrogate the West Point operator and aid him in the operation
of the remote facility to get the information necessary to make a
knowledgeable judgment about required action. This commentary is
not condemnation of the treatment plant operators. Rather, it
points to the need for support personnel to provide adequate init-
ial and regular refresher training to all console operators.
Limited Operations from the Remote Consoles — Even with adequate
manning of the remote consoles, problems would still exist. The
remote consoles provide a monitoring facility, but not a control
station. The monitoring capability is limited also, since data
available at the central console is not available to the remote
consoles. A major omission from the remote console displays is
rainfall data, which is nearly essential for making intelligent
evaluations of system operation.
The remote consoles cannot truly be called control consoles
since they only allow one command; removing a station from super-
visory or automatic control and returning it to the local control
condition. The lack of command facilities again makes the CATAD
console operator's job more difficult. Whenever the operator i&
notified by the treatment plant personnel of a problem that
occurred outside of the normal shift of the console operator, the
operator must return to the central console control .facility to
perform remedial action. He cannot request that the remote con-
sole operator perform the required functions from his terminal.
All monitoring and control functions of the satellite con-
soles will eventually be implemented with additional software.
At that time the satellite consoles will duplicate, at a slower
speed, all central console functions. Improved system reliability
and ease of use should result.
Console Operators
The success of the CATAD system depends on the console opera-
tor in two ways. First, the operator must be sure that all necessary]
functions to achieve a storage strategy are operating correctly at
all times. Secondly, whenever supervisory control operations are
to be performed, a console operator must run the system. Introduc-
ing the automatic control scheme lessens the amount of operator
258
-------
time required on 24-hour call basis to handle emergency condi-
tions. A number of factors have limited the console operators
during the period of CATAD operation to date.
Training Time — The console operator requires knowledge of many
phases of the operation of the sewer network, control stations and
the computer facility. A training period of approximately one
month should introduce an operator to the basic layout and func-
tions of the control console and the standard procedures used by
console operators. However, experience with operators working at
the console indicates that even after six or more months, they
may still be learning about standard console operation and espec-
ially about the remote facilities. This long learning time is
simply explained by the huge volume of information to assimilate.
For an operator to be effective at supervisory control of the
sewer system, he must not only know the maximum limits of storage
and other related parameters appropriate to each station, but he
must also have an intimate awareness of available storage volumes,
gate movement times, runoff delays from initial rainfall, sewer
flows as a function of rainfall volumes, and many other character-.
-------
As the rainfall pattern changes/ the operator must dynamically
alter his storage and control strategy. Rainfall alone/ of course,
is not sufficient data on which to base a control strategy. The
most important data item for control is trunk sewer flow. However/
if the operator waits for the storm flow to reach regulator stations
to get an accurate measurement of the flow/ his control actions will
most likely start too late. The operator must anticipate flows and
the time of their major impact through the system to implement an
effective control strategy. In the CATAD system an operator may
be expected to implement a control strategy involving fifteen or
more stations, continuously monitor these stations and modify their
gate positions or pumping rates to make best use of the available
storage resources. Experience indicates that fifteen stations under
supervisory control is the maximum which even the best qualified
operator can handle. A work shift of eight to twelve hours of such
activity is extremely demanding.
Reaction Time to Off Duty Callout — Storm conditions may begin
when the central control console is not manned. When this happens,
the on-call CATAD operator is notified by the West Point operator
who is monitoring the satellite terminal. If the supervisory con-
trol mode is to be used for storm control, the operator must have
enough notice to allow him to reach the control console and init-
iate an effective strategy before the major flow conditions are
upon him. During early operation of the system, the problem of
getting adequate notice to the operator caused some possibly
unnecessary overflows to occur. Experience with the CATAD system
has generally alleviated this problem.
Working Conditions — CATAD console operators are subject to per-
iods of very high activity alternated with periods of low activity.
The most intense requirements occur during the wet winter season;
however/ during the development period of the system, the operators
have also been required to respond to off hours system failures
detected by monitoring at the West Point terminal. From early 1972
when the supervisory control program began, Metro has operated the
control system with two men available as console operators. One
of these men is a console operator; the other a programmer who acts
as console operator when required by storm conditions or other cir-
cumstances. During the wet winter months of 1972-73/ it was not
uncommon for each operator to work over 40 hours of overtime in a
month with maximums of over 70 overtime hours a month. If super-
visory control is to be a normal operating mode, it would be wise
to employ at least four operators to work a scheduled rotating
shift. Where automatic control is replacing the intermediate step
of supervisory control, the two operators .have been able to handle
the interim period although at times the demands on these employees
have- been far beyond reasonable expectations. In addition to long ,
working hours/ the console operators are subject to considerable
mental pressure since it is their responsibility to minimize over-
260
-------
flow volumes. A few times during the 1972-73 winter, the condition
occurred where high tides and high storm flows combined to force
excessively high trunk levels. In these cases, an operator must
do all in his power to avoid flooding residential and commercial
sewer connections. The overall operation of the CATAD system
during the rainy winter period puts considerable pressure on the
individual operator.
Intra-Agency Cooperation — It is the responsibility of the CATAD
console operator to observe the operation of the sewer system and
its control components and notify the required maintenance per-
sonnel when problems are detected. Maintenance personnel are not
always available and when a scheduling conflict occurs, the main-
tenance director decides which functions have priority. The CATAD
control system is a recent addition to the Metro sewerage opera-
tion, and many more "veteran" aspects ofthe operation are accus-
tomed to being higher on the list of maintenance priorities. The
fact that the CATAD system has been installed and is operating in
an area separate from the standard plant and system operations has
hindered easy integration of the CATAD effort into the total Metro
sewerage operation. In areas where it is necessary to receive the
cooperation of other operating and maintenance personnel, the fact
that CATAD has been an isolated and often misunderstood effort has
certainly caused some difficulties. Any agency attempting to in-
troduce a CATAD-type system would be well advised to educate every-
one who would have to have contact with the system to minimize
misunderstandings arising from ignorance. Kudos as well as the
criticisms should be evenly distributed among those playing a part
in the total system to achieve a maximum acceptance and efficiency.
261
-------
SECTION IX
WATER QUALITY STUDIES
INTRODUCTION
The Seattle Metro comprehensive water quality monitoring pro-
gram started in 1963. Additional stations and parameters were
added in 1967 to assess the impact of combined and separated sewer
overflows within the Metro drainage basin (Figure 69). The CATAD
demonstration project to improve receiving water quality by reduc-
ing overflow frequency and volume included detailed analysis of
physical, chemical and biological data from both combined and
separated sewer overflows. This data must be correlated with the
corresponding meteorological parameters: temperature, rainfall
intensity, duration and frequency.
The two major water quality data collection objectives for
this report were:
1. To show the dynamic water quality effects of Metro's
sewage interception, regulation and separation program.
2. To establish a water quality parameters baseline relative
to CATAD demonstration project implementation.
_Data analyses and conclusions are presented in Section X. This
section considers water quality data which indirectly relates to the
sewage collection system and to the immediate and overall receiving
water quality. Related water quality studies, completed or currently]
underway, which might bear significantly on combined sewer overflows
are discussed.
COLLECTION SYSTEM STUDIES
General
For this demonstration grant, regulator station studies focused
on combined sewer overflows, primarily in Elliott Bay and on the
Green-Duwamish River. Because of the potential impact of the large
sewer separation program described in the interim report (12) , spec-
ial consideration is given to the water quality changes at those
regulators serving areas which have undergone separation (Appendix
A) . ^ The water quality impact of large paved areas such as freeway
drainage, as developed in several recent studies within King County,
is included to relate stormwater runoff to the combined sewer over-
flow loading.
262
-------
DIAGONAL T.P.
(Closed 1989)
307
LEGEND
• CHEMICAL ANO OR BIOLOGICAL STA.
O BACTERIOLOGICAL ' STATION
O CONBINEO STATION
A AUTOMATIC MONITORS
FIGURE 69
WATER QUALITY MONITORING STATIONS
ASSOCIATED WITH THE CATAD DEMONSTRATION GRANT
263
-------
Overflow Sampling Locations and Parameters
The overflow regulation objective was to minimize water qual-
ity degradation by reducing overflow frequency and volume. Over-
flow sampling was divided into physical parameters relating to the
regulators/ meteorological conditions, and water quality loading
at each regulator site listed in Appendix A and illustrated in
Figure 8.
The physical parameters include flow data; overflow volumes
and frequency; rain intensity, duration, and frequency as well as
average, maximum, and minimum daily temperatures.
The chemical data is based on chemical oxygen demand (COD),
nitrite-nitrate nitrogen, ammonia nitrogen, total soluble phos-
phate, and solids analysis. Solids analysis includes total settle-
able solids, suspended solids and volatile solids.
The biological data includes biochemical oxygen demand (BOD)
and bacteriological analysis. The membrane filter technique is
used for total coliform, fecal coliform, and fecal streptococci.
These tests follow "Standard Methods for the Examination of Water
and Wastewater" (34) and have also been mentioned in the interim
report (12).
Regulator Sampling
The CATAD sampling program began in November, 1969, with seven
sampling sites. Six additional sites were added in April, 1971. At
the end of June, 1973, the sampling program was reduced and data
evaluation begun.
Sampling devices were either the Composite or Sequential
Sewer-Test, Vary-Sampler, manufactured by Sirco Controls of Vanr-
couver, British Columbia, Canada. Recording devices were two- or
three-pen strip chart recorders manufactured by either Foxboro,
Inc., Foxboro, Mass., or the Fischer & Porter Co., Worminster,
Pennsylvania.
The major difference between the samplers is the method of
sample collection. The composite sampler collects given amounts
of sample at predetermined times after the initial gate opening
and stores all samples in a single 12-liter glass bottle. Although
individual samplers vary, approximately one liter is collected each
time the pump is activated. Completely filling the bottle triggers
a float switch, ending the sample sequence.
264
-------
In the sequential sampler, samples are stored in separate
one-liter bottles. Sample quantity is, therefore, fixed at one
liter for each time division on the program tape.
Collected samples are stored in the sampler refrigeration unit
as described in the interim report (12). Temperature in these
units is maintained at 4°C to minimize biological activity or
freezing which could alter the constituents of the sample.
Sample timing is accomplished by a rotating "Programmer disk"
(see Figure 70) that opens and closes a microswitch supplying power
FIGURE 70
PROGRAMMER DISK
to a vacuum pump.' "Programming" the sampler consists of changing
the position of ridges on an embossed adhesive tape. Sample se-
quences vary due to a size difference of the programmer disk fur-
nished with newer model samplers. Sequences are listed in Table
19.
265
-------
Table' 19'. SEQUENTIAL 'SAMPLER TIMING
Time the Numbered Sample is. Taken (minutes)
Stations
13 to 26
30 to 40
G.
G.
1
0.'a
0.
2
+15
+10
3.
+30
+20
4'
+45
+30
5'
+60
+40
6'
+90
+50
7
+120
; +60
8'
+180
+ 90
9
(§60
(§60
a As outfall gate opens
Regulator stations utilize similar multi-pen color coded strip
chart instruments to record: wastewater level in the sewer trunk,
tide level at the outfall gate, and position of the outfall gate.
Tide level is not recorded where it does not influence the over-
flow volume discharged. Representative loading figures for storm-
related combined sewer overflows depend on successful operation of
recorders, samplers, and on accurate laboratory analysis.
A technician is assigned to sample collection, sample analysis,
some minor maintenance and inspection, and coordination of suppor-
tive maintenance and data processing activities. Samples are col-
lected after most heavy rainfalls. When time permits, the sampling
equipment is left recharged and operational.
Routine inspection scheduling is determined by the reliability
of the equipment at each station. Regulator samplers are checked
once a week (more frequently if required). During maintenance
visits, samplers are tested under conditions approximating an
actual overflow condition. Problems are usually minor: sample
tube blockages, microswitch adjustments, and fuse replacements.
Recorders are marked and dated during inspection visits.
Recorder malfunctions and major sampler breakdowns are handled
by an instrument technician from the West Point Treatment Plant
maintenance section. Most tasks involve cleaning recorder pens and
repairing faulty electrical components in the samplers. Mainten-
ance crews performing other duties at the regulator stations per-
iodically mark recorder charts and record outfall gate movements
and sampler counter operations.
Metro's Water Quality Laboratory performs chemical analysis
of the overflow samples. The laboratory is geared to handle large .
numbers of similar samples, so few difficulties were encountered
during the study. High sample concentrations required a separate
set of glassware to* prevent contamination of other water quality
samples.
266
-------
Overflow samples are tested for eight different parameters
using Standard Methods (34). Settleable solids are determined
using an Imhoff cone, method #224F. Suspended and volatile solids
are determined using Gooch crucible - glassfiber filter filtration,
method #224C.. Total suspended matter utilized method #224D.
Biochemical oxygen demand (BOD) is determined by method #219, using
a 5-day incubation period. Several dilutions are used to account
for great variation in samples from different stations. Chemical
oxygen demand (COD) is determined by method #220 using a 40-minute
refluxing period and deleting the addition of Ag2 and 804 to the
concentrated H2SO4. ••
Until February 19, 1971, a hydrazine reduction method was used
to determine nitrite-nitrate nitrogen (35). It was then determined
that the cadmium reduction method (36) gave better results. This
method was used for the remainder of the study, both in its manual
form, and adapted to the Technicon Auto-Analyzer. A few drops of
dilute CuS04 are added to sewage samples to remove any H2S which
would reduce the efficiency of the cadmium filings. Ammonia nitro-
gen is determined using the Nesslerization Method, I132B. Twenty-
five milliliters of unfiltered samples are analyzed with a spectro-
photometer rather than Nessler tubes. Phosphate is analyzed by
the single reagent digestion method, #223B. Although listed as
the Hydrolyzable Phosphate Test, this procedure gives values very
close to the total phosphate concentrations. During the testing
period, inter-laboratory standardization was initiated to check
reproducibility of analytical results. This verified the quality
of the laboratory results. Some difficulties were encountered in
reproducing ammonia-nitrogen results with the Nessler procedure.
^ Flexible scheduling allows maximum use of available space,
equipment and manpower. Metro's Water Quality laboratory is able
to analyze large numbers of samples arriving at unpredictable
times, with njinimum conflict.
Data handling is usually confined to computing chemical and
physical data, transferring overflow volume coordinates to work-
sheets, and logging pertinent data. Computer division personnel
punch cards for both water quality parameters and overflow volumes.
Overflow volumes are determined from the chart records of
trunk sewage level, tide level, and outfall gate opening. The
computer program treats the gate as an orifice that may be sub-
merged by tidal action (37). Overflow volume is the rate of flow
multiplied by duration. The program accounts for scale factors
between chart readings and water levels within the sewer (see
appendix of Reference 12).
267
-------
Sewer Separation Study
Seattle's $76/000,000 bond issue to partially separate all
combined sewer systems tributary to Lake Washington included a
cooperative study between Metro and the City of Seattle to deter-
mine the quality and quantity of overflows with separation, as
well as the impact of CATAD computerized storage management. Of
the four stations sampled (see Figure 71) Cooper Street represents
a small (103 acres) and Sand Point a larger (183 acres) drainage
area served by a separated system. Windemere and Henderson repre-
sent large combined systems from 448 and 539 acres, respectively.
Equipment and primary analysis are discussed in the interim report
(12). Most data collection utilized a vacuum type, clock-motor
drive sequential sampler and a seven day chart, clock-driven flow
level recorder.
The quality and quantity of the separated stormwater and com-
bined overflow streams were measured over a two-year period at
zauaple sites in four different drainage basins. Quality measure-
jaents included total and fecal coliforms and fecal streptococci,
total phosphate, ammonia and nitrite-nitrate nitrogen, settleable,
suspended and volatile solids, BOD and COD. All analyses were
conducted in accordance with Standard Methods with the exception
of ammonia (38) . Quantity data was computed from stage height
recording. Rainfall data discussed in Section X was obtained
from the City of Seattle's rain gage network described in Section
IV (see Figure 11).
Fr.eeway Drainage Study
The objectives of this study for the Washington State Highway
Commission were to define the quality of the stormwater runoff
from highway surfaces and to estimate the environmental impadt of
such runoff (39) . "»
Two freeway drainage studies were conducted. The first, on
an elevated portion of Highway 1-5 (see Figure 72) during 1970-71,
was chosen because of the lack of interference from other sources
of drainage. The second study on the 1-90 corridor from North
Bend to Lake Washington (see Figure 72) examined the impact of
additional traffic and road surfaces in the area.
Parameters measured in the 1-5 study consisted of antecedent
rain, suspended and settleable solids, COD, BOD, nitrite-nitrate
nitrogen, total phosphate, free ammonia, and oils which were mea-
sured by the Soxhlet extraction process (40). Most of this data
was presented in the interim report (12). During the 1-90 study,
traffic volume and its impact on amounts and types of residual mat-
erial deposited were added. Airborne dust from background sources
not attributable to vehicle emmissions was also measured.
268
-------
LEGEND
13 SEPARATION STUDY SAMPLE
SITES
FIGURE 71
STORM SEWER SEPARATION SITES
269
-------
Sections of Freeway
where samples were taken
FIGURE 72
FREEWAY DRAINAGE STUDIES
270
-------
The 1-90 study was divided into three phases. The first con-
centrated on the Mercer Island/South Bellevue area and Lake Wash-
ington. The second phase examined a typical highway construction
site at Eastgate; and the third phase investigated the vicinity of
North Bend, a semi-rural area crossed by a major highway.
Sampling crews were alerted to impending rainstorms through
the cooperation of the National Weather S.ervice-Federal Aviation
Administration radar surveillance unit in Auburn, Washington.
Runoff volumes from diverted drains were measured in 25-gallon
containers at 15 minute intervals. Samples taken from more than
one drain were composited every 15 minutes. 10 to 15 samples were
cpllected for each storm of 4 to 6 hours duration.
Samples were separated into three portions: .the original
sample, the filtrate (from a 934 AH Reeve Angel glass filter),
and the solids retained on the glass filter. Analyses were per-
formed on the original sample for turbidity, conductivity, pH,
settleable solids, total suspended solids, volatile solids, oil,
COD, total coliforms and fecal coliforms. Analyses of the solids
from the filtered samples included Kjeldahl nitrogen, total phos-
phorus, lead, zinc, copper, cadmium and chromium. Analyses of
the filtrate were BOD, ammonia, nitrate-nitrite nitrogen, hydro-
lyzable phosphorus, lead, zinc, copper, cadmium and chromium.
The suspended solids analysis performed on the original sample
included both the settleable solids and volatile solids. Oil con-
tent was determined by the Soxhlet extraction with a Freon solvent.
An atomic absorption spectrophotometer was used for heavy metal
analysis. Heavy metals separation was accomplished by particulate
digestion with a nitric perchloric acid mixture.
Data Calculations
Significant data calculations included: '
1. Mean Concentrations — This value was determined for
each constituent by summation of the total loading quan-
tity calculated for each sample interval and dividing it
by the total volume collected. The overall mean was cal-
culated by averaging the means for individual storms.
2. Loading Figures — The daily loading figure (mg/m2/day)
was calculated by totaling the individual loadings
corrected for 100 percent runoff, for each sample interval
and dividing by the total area of the freeway drained and
the number of days following the previous rainfall. The
loading rate based upon traffic intensity was calculated
271
-------
as a linear function of the drainage area (mg/m/vehicle).
It was determined by dividing the total loading corrected
for 100 percent runoff by the length of the area drained
and the number of vehicles using the highway during the
previous period.
The significance of the 1-90 and 1-5 studies to CATAD and
runoff from the Metropolitan area lies in the fact that over 40
percent of the City's surface area is presently paved over for
automotive related uses.
Urban Storm Drainage
Urban runoff data collection began in March, 1973, for six
discrete basins of various land use; to determine pollution wash-
off characteristics in the Seattle area. To evaluate land use
impact on overflow specifically in the CATAD control area, three
of these basins were considered: the 800 acre multiple residential
area View Ridge; 75 acres of commercial development with a high
percentage of paved surfaces in South Center, Tukwila; and 100
acres of the Central Business District of Seattle (see Figure 73).
The View Ridge area has both single family and multiple family
residential land uses. Most residential roof drainage downspouts
empty into sanitary sewers. However, the results of monitoring
this area, when compared to smaller areas, are quite valuable.
The South Center complex in Tukwila is open daily and even-
ings and experiences large volumes of automotive traffic. This
area has a well-defined, separate, easily monitored storm drainage
system. It represents a typical large commercial land use area.
Water quality parameters measured were: BOD, COD, dissolved
oxygen (DO), ammonia nitrogen, nitrate-nitrite nitrogen, total
nitrogen and phosphorus, suspended and settleable solids, dissolved
solids, total and fecal coliform, chlorides, heavy metals (Hg,
Cu, An, Pb, Cr, Cd, As, Fe), temperature, pH conductivity and
turbidity.
Manual sampling consisted of sequential samples taken every
15 minutes as it began to rain. The alert system cooperating with
the NWS-FAA radar surveillance and bellboy paging system developed
during our freeway drainage studies were used in this study. Samples]
were collected using a bucket made from a Nalgene #2302 two-gallon
aspirator bottle with the top removed at the shoulder. The spigot
on the bottom of each bucket was fitted with rubber tubing and a
pinch clamp. The bucket was lowered on a rope into the manhole,
allowing sampling without entering the manhole. Where flows were
272
-------
LEGEND
VRI VIEW RIDGE
VR2 VIEW RIDGE
CBD CENTRAL BUSINESS
DISTRICT
SC4 SOUTH CENTER
SC4
FIGURE 73
»
URBAN STORM DRAINAGE AREAS
273
-------
heavy and the bucket did not tip and fill upon hitting the water
surface/ an additional line attached to the spigot allowed the
bucket to be tipped for filling.
One complete sample consisted of filling two one-gallon poly-
ethylene jugs for general chemical tests (one sterile bacterio-
logical bottle, one dissolved oxygen bottle), and recording the
time and temperature. The spigot permitted aseptic filling of
bacteriological sample bottles and collecting dissolved oxygen
samples with minimal aeration. All samples were stored on ice while
in the field. When possible, upon arrival at the sample site, a
background sample was collected before the initial stormwater
surge.
First flush samples were picked up one to three hours after
the surge and returned to the laboratory where they were immedi-
ately preserved and analyzed. Additional samples were picked up
at approximately three hour intervals. Since most storms lasted
five to six hours, the second sample pickup was often delivered
by the field crew when they returned to the laboratory.
Receiving Water Quality
The comprehensive receiving water quality program has been
described by Gibbs (41) and Gibbs and Isaac (42), and results
are available in Metro's six month reports to the Water Quality
Monitoring Review Board. The two principal receiving water bodies
relative to this demonstration grant are the Green-Duwamish River
and Elliott Bay. Onshore and offshore manual and automatic monitor-
ing stations are indicated in the map in Figure 69.
Manual Water Quality Monitoring
The Elliott Bay offshore stations are sampled twice monthly
from the U. S. Geological Survey's 40 foot survey vessel RV Hydor
and occasionally from a 16 foot outboard boat. Onshore sampling
is made weekly. Samples are analyzed for total and fecal coliform,
dissolved oxygen (DO), temperature, salinity, and transparency,
using a secchi disk.
Nutrient data have been collected at four stations on a pri-
mary productivity study in Puget Sound from 1967 to the present.
The productivity station 174 is very near Station 181 in Elliott
Bay (see Figure 69).
In the Green-Duwamish River (the name changes from the Green
River at the confluence of the Black River to form the Duwamish
River below) demersal fish population data was collected between
1966 and 1971 using a J.6-foot otter trawl boat. Total and fecal
coliform, ammonia nitrogen, nitrite-nitrate nitrogen, and total
274
-------
hydrolyzable phosphate were manually sampled from the river.
Studies On the benthic deposits of chemical oxygen demanding (COD)
matter in the sediment were made for this demonstration grant with
data available for the lower river between 1963 and 1973.
Automatic Water Quality Monitoring
Due to estuarine system complexities/ Metro recognized that
manual sampling to determine necessary water quality parameters
at all stations under various tidal conditions in the Duwamish
Estuary was impractical. Consequently, five automatic monitors
were installed at strategic river locations. The multiple objec-
tives of these monitoring stations were:
1. Demonstrate compliance with State water quality standards
2. Maintain a continuous surveillance of water quality con-
ditions as they relate to the migration of anadromous fish
such as salmonids
3. Document improvements in water quality conditions predicted
as Metro's construction and operational program was
initiated.
This monitoring program is being carried out in cooperation
with the U. S. Geological Survey (USGS). The regional USGS branch
is developing a complex water quality model of the Duwamish Estuary
and relies on these automatic monitors for some of its primary
calibration data.
Surface river water is pumped to the automatic monitor sta-
tions where DO, pH, conductivity and temperature are continuously
recorded. In addition, river bottom water is monitored at the two
downstream stations for conductivity, DO, and temperature. Data
collected at the five locations shown in Figure 26 of Section V
is telemetered to the computer and stored for processing.
The physical description of the automatic monitoring stations
is discussed in Section V. Data transmission, storage, compila-
tion, and input to the CATAD control system are discussed in
Section VII.
Station performance is maintained weekly, or oftener depending
on the status of data collected by the computer; i.e., extreme
temperatures indicating pump failure or very low DO reading,
usually meaning probe failure. During high river flow and heavy-
sediment load, lines are often plugged with sand and silt,, requir-
ing more frequent cleaning. Every three or four weeks, reservoirs
are cleaned of aquatic growth, algae, barnacles and mussels.
Appendix P lists routine monitor cleaning procedures which assure
good data continuity.
275
-------
Temperature, conductivity and pH analyzers are calibrated
monthly using standards. Winkler dissolved oxygen (DO) samples
are collected and analyzed in the monitor "mini-lab" weekly,
depending on probe condition. Titrated values are compared with
analyzer readings when samples are drawn. Minor probe drift
adjustments are usually required.
276
-------
SECTION X
DISCUSSION OF RESULTS
INTRODUCTION
The previous sections of this report have concentrated on physi-
cal aspects and development of the CATAD project. Two big questions
remain. What does it accomplish? How much does it cost?
t
These questions are answered in the remaining two sections of
this report. This section discusses CATAD project performance as
reflected by water quality improvements and changes in overflow
quality and quantity discharged to the receiving waters. The sec-
tion is organized in the same sequence as Section IX which outlined
the scope of the various water quality studies comprising the total
project.
RECEIVING WATER
Receiving water quality is examined by reviewing the automatic
monitor data from the Duwamish River and the bacteriological data
from nearly all water bodies bordering the city. Small scale
studies of river bottom sediments and aquatic life have not been
sufficient to justify any major conclusions.
Monitor Data Analysis
Comprehensive data generated by the automatic monitor system
provides an excellent opportunity to view long range effects of
interception and regulation in the biologically and chemically
critical Duwamish Estuarine environment. The estuarine environ-
ment has particularly critical dissolved oxygen (DO) problems
because the saltwater wedge beneath the fresh surface waters con-
tinually transports settleable organic solids from the surface
back upstream where they accumulate until periods of very high
flushing occur. Consequently, in heavily loaded estuarine environ-
ments, the greatest fluctuations, hence the greatest stresses, occur
near the head of the major salinity wedge between the bottom and
surface DO concentrations. The two automatic monitoring stations
best located for monitoring this condition in the Green-Duwamish
Estuary are the 16th Avenue South station at the surface and bottom,
and the Spokane Street location at the surface (Figure 74). The
shaded portion is the area of critical DO fluctuation. Spokane
Street bridge surface DO values were also included because they
might better reflect the impact of fresh water storm drainage
overflows on immediate DO suppression.
277
-------
Sj
CO
o
o:
o
x
CO
3
H
1X4
£5
O
H
I
o
H
1
CO
u_
o
Ill
_1
z
278
-------
CA.TAD regulator system operation is based on the premise of
overflowing first into Puget Sound and then, as necessary, into
the upper estuary. Since interception and regulation in the Duwa-
mish Estuary, progressively higher minimum and average DO values
have been observed. • • /
The maximum and minimum weekly values are plotted for the
years 1969 and 1972 at the Spokane Street monitor for surface
values, and at 16th Avenue South monitor for surface and bottom
dissolved oxygen (Figure 75).
The improved DO trend at both the 16th Avenue South and Spo-
kane Street monitors was apparent between the years 1969 and 1972.
The hydrological data in Figure 76 shows significant river flow
increases in 1971 and 1972, but not in 1973. In Figures 77 and 78,
plots of surface and bottom minimum and maximum temperatures be-
tween 1969 and 1972 show little change. This data might indicate
that improved DO conditions in the river resulted from higher
flushing; however, when the low runoff 1973 monitor data is com-
pared with 1969 data, it is nearly always higher than the same
time period in the earlier year (Figure 79).
To clarify the DO trends, portions of the 1969 and 1972 data
with apparent linear trends (May 12 to June 25 and September 15 to
December 15) were selected. This data was statistically compared
using analysis of variance. The results of these analyses indicate
a statistically significant increase in DO with time (see Table 20),
Bottom Sediments
Benthic deposition in the lower Green-Duwamish Estuary has
been studied since 1963. In this dredged section of the lower
river, significant decreases in sediment chemical oxygen demand
(COD) have been recorded between December of 1968 and 1971 (see
Eigure 80). This was partly attributed to diversion of the Dia-
gonal Treatment Plant flow to the West Point Treatment Plant and
to nearby land filling which produced observable quantities of
sand erosion settling in the mid-channel sampling site. Since that
time, however, further COD reductions have been recorded through
most of the lower estuary for December. Contrary to the continuing
decline, the June, 1973, observations averaged slightly higher.
This may reflect the difference between summer and winter flushing.
Additional urban storm drainage data is needed to understand the
role of stormwater as a possible flushing mechanism in the lower
estuary.
279
-------
C3
in
, H
Q
I
O
CO
CO
H
Q
tf
•\
K
CO
I I N\J I
280
Q
\
O
-------
vo
r-
M
'S'J'O MOHJ AHH1NOW IMV3IN
281
-------
-MAXIML M
to.o-
5O.O-
-MINI/ 1UM
\ /
,—'
37.5
JAN FEB MAR APR MAY JUN JUL AUG SEP NOV / DEC
1971
50,0
JAN FEB MAR APR MAY JUN JUL AU6 SEP OCT NOV
FIGURE 77
GREEN-DUWAMISH RIVER TEMPERATURE RANGE
AT 16TH AVENUE SOUTH, SURFACE
282
-------
•M. \XIMUM
•um /MUM
1969
FIGURE 78
GREEN-DUWAMISH RIVER TEMPERATURE RANGE
AT 16TH AVENUE SOUTH, BOTTOM
283
-------
— o o
-------
en
^
EH
^i
Q
S
w
$
X
O
Q
^
O
CO
CO
H
Q
O
W
CJ
s
rt.
H
§
^
fa
0
CO
H
CO
3
2!
•a!
P^
o
EH
H
&
O
s
U
H
EH
5*
|
<
«
H
#
ffi
CO
H
rtj
3»
Q
1
O
CM
0)
i~H
•a
EH
CM
•
in
ft
co
•
en
in
II
fa
a)
M
id
3
tr
CO
a
id
0)
g
O O
• •
CT» rH
in
CO
a)
rl
id
3
O1
CQ
MH
O
0
3
CO
o in
en r-
m r^
0
•a
(U
rl
fa
m
o
to
.C -H
•P SH
VO p|
rH CQ
en
in
rH
•
in
^*
ro
•
CO
II
fa
in rH
• •
01 rH
in co
cn^'
o
rH
rH CM
O\
0) C
-P -P
0) -H
m is
m
rH
O CM
id I-H
M-)
rl O
3 -P
CO
in
(U rH
C en
Q) * —
0)
r* jj
t> C
«> -H
rH &
CM
•
in
'ft
U3
•
^*
CPt
II
co en
• •
in o
oo
CO CO
in o
oo r--
H 00
C
<0 G
< tri
Xi-S
•P M
rH CO
rH
•
in
^
•
CM
CO
••*
||
fa
O CO
• •
rH 0
CM
rH
/
o r-
• •
rH in
CM CM
i—l
rH CM
en
^^
0)
VD -H
rH &
CM
en
,
CM
CM
II
fa
in en
0 0
CM
in r~
o en
CM <£>
H 00
r-
Between
Within
^
in
fl) CM
id «>
in O
3 -P
CO
CM
• rH
co m
0)
fi en
id C
O M
a a
co co
in
rH
in
A
VD
VO
OO
||
fa
oo o>
• •
CM O
oo en
CM O
H CM
en
Between
Within
in
O CM
Id rH
<4H
rl O
3 -P
CQ
in
• H
•P\
co en
0)
Spokan
Winter
285
-------
X33J - HldBd «3Aia
S 8 S
o
oo
H
C.OI x 6i|/6uj- N39AXO 1VOIH3HO
286
-------
Aquatic Life
The Green-Duwamish River uniquely exemplifies the feasibility
of maintaining adequate water quality for fisheries resources in a
highly industrialized metropolitan area. Data from the Washington
State Department of Fisheries, Green River Hatchery, shows that
adult salmon returns to the hatchery have been more than adequate
to maintain significant runs of Chinook and Coho salmon in the
river system (see Table 21). Although the Coho salmon data de-
Table 21. ADULT SALMON RETURNS3
YEAR '
1967
1968
1969
1970
1971
1972
CHINOOK SALMON
5,038
8,114
6,650
10,714
9,887
6,828
COHO SALMON
12,736
50,856
45,414
70,868
21,174
9,076
To Green River Fish Hatchery
clined in 1972, scientists from the Washington State Department of
Fisheries attributed this decline to:
1. A poor wild run of salmon placing greater demands on
hatchery runs by sport, commercial and Indian fisheries
2. Reduced hatchery production because of budget cutbacks
3. Changes in fishery management to harvest as many of the
surplus fish as possible.
A comparison of trawl catch results in the Duwamish River be-
tween the pre-sewer outfall diversion period (1967-68) and the
post-sewer diversion period (1970-71) suggests an increase in the
English sole population (see Table 22). The reasons for the de-
creased catch between 1970 and 1971 are not completely known, but
may be partly attributed to natural population fluctuations. For
example, the sand sole Psettichthvs melanostictus. a bottom fish
not previously captured in the Duwamish Estuary, was a -common
species during the post-diversion peridd. This species diversifi-
cation suggests that the increase in dissolved oxygen enabled this
species to survive in the Estuary.
287
-------
Table 22. TRAWL CATCHES OF ENGLISH SOLE
IN GREEN-DUWAMISH RIVERa
YEAR 1ST AVENUE 'SOUTH
1967 8.7
1968 15.5
1969 10.9
1970 27.8
1971 11.8
a Average for each year
16TH AVENUE SOUTH
7.0
2.4
9.2
45.5
11.4
STATION KW
0.4
0.4
3.9
8.8
3.2
Results from these and other studies demonstrate that removal
of raw and treated waste discharges to the lower Duwamish River
Estuary has improved water quality thereby ensuring continued
growth and survival of salmon and other aquatic life.
Bacteriological
Bacteriological analysis results presented in the Interim
Report (12) recorded striking improvements in total coliform counts
between 1968 and 1970 in Elliott Bay. The improved quality was
attributed to the diversion of several large raw sewage outfalls
to the West Point Treatment Plant. When the 1971 and 1972 results
are graphically compared with the previous median total coliform
counts (see Figure 81), all the 1971 and 1972 results are below
the water quality standards for inshore Elliott Bay. The improv-
ing trend between 1970 and 1972 is self-evident. Comparisons of
median total coliform counts before and after CATAD control became
operational are shown in Table 23. Seven of the nine stations corn-
Station
Number
230
231
232
233
234
235
236
237
238
Table 23. MEDIAN COLIFORM COUNTS
ELLIOTT BAY SHORE STATIONS
Median Total Count per 100 mis
4/71 to 3/72 4/72 to 3/73
(pre-CATAD) (post-CATAD)
160
180
200
490
560
310
360
51
24
97
90
160
390
380
140
180
48
35
Months that
counts were
lower
10 of 12
9 of 12
7 of 12
7 of 12
5 of 12
9 of 12
9 of 12
5 of 12
7 of 12
288
-------
IIO.OOO
240,000
110,000
ItO 00
100,000
10,000
JO
E
O
o
oe
LJ
Q.
O
O
DC
O
O
O
<
o
1,000
100
WATER QUALITY
STANDARD IN
ELLIOTT BAYy
184 188 230
STATIONS
233 235 236
FIGURE 81
238
ELLIOTT BAY COLIFORM LEVELS AND STANDARD
289
-------
pared showed improving trends. Continued monitoring will show
whether this is a real improvement or just a small difference in
annual counts.
In the east and west waterways of the Duwamish River, median
total coliform counts are hardly improved (see Table 24).
Table 24. MEDIAN COLIFORM COUNTS - DUWAMISH ESTUARY
Station
Number
301
302
303
304
305
Waterway
East '
East
East
East
West
Median Total Count per 100 mis
4/71 to 3/72 4/72 to 3/73
(pre-CATAD) (post-CATAD)
370
640
950
2600
2300
720
800
880
1400
2200
Months that
counts were
lower
6 of 12
6 of 12
6 of 12
7 of 12
7 of 12
The reduction in maximum monthly counts (see Table 25) at
Stations 304, 305, 306 and 307 in the Duwamish, upstream from major
shipping activity, show a distinct improvement, particularly during
summer low flow months when CATAD better utilizes storage capacity.
When coliform count at Station 305 in the lower Estuary is compared
with count from the upstream Station 315 in the Green River above
the area affected by CATAD, decreased estuarine contamination can
readily be seen between 1968 and 1972 (see Figure 82).
Low monthly median total coliform offshore counts in 1971 and
1972 (see Table 26), do not fully reflect an improving trend since
only 27 of 48 months are lower. The offshore data may reflect
background variation not directly related to the CATAD overflows.
The Lake Washington Ship Canal, Lake Union, and Portage Bay are
all heavily used recreational areas. While regulator overflow
sampling was not planned for this part of the system, several
significant changes have taken place. Cleaning the Montlake siphon-,
Fremont siphon reconstruction, and construction of the Dexter Avenue
regulator have all occurred in the last three years. The total coli-
form counts between 1970 and 1972 for 14 stations during either
April to September or July to September, show substantial improve-
ments (see Figure 83). These improvements partly reflect the 1972
CATAD efforts to utilize storage in the large Lake City tunnel,
thereby relieving hydraulic loading on the Fremont siphon and avert-
ing possible back-up and overflow from Montlake siphon upstream
relief points.
290
-------
:*
I
W
a
§
K
H
>
H
»S
w
w
w
En
2
D
O
in
CM
CM
en
rH
r-
o
CO rH
en
rH
CM
en
rH
VD
o
CO rH
f^
en
rH
CM
f^
• en
CO rH
rH in
g 0
1 CO rH
o r-
o en
rH rH
In CM
O r-
MH en
•r( rH
O o
U CO rH
r-
i — I en
Id rH
+J
O
EH CM
g en
3 rH
6 co
•H O
X CO rH
rd r-
S en
rH
CM
r^
en
rH
CM
o
CO rH
en
rH
CM
r-
cn
rH
i
r*~l
o
CO rH
t^.
en
rH
C-j
4J
fi
S
O
O
CM
CO
O
o
r-H
in
O
O
CO
CO
o
o
CO
in
o
CO
vo
o
o
CO
vo
o
0
CO
co
0
o
•a*
VD
o
0
rH
rH
0
0
rH
CO
0
o
co
rH
o
o
rH
CO
o
CO
r-
o
o
en
CM
r-H
•H
M
<
o
vo
00
o
o
en
CO
o
o
vo
rH
o
o
o
CM
o
0
en
rH
o
o
in
o
0
in
CM
o
o
r*^
*3*
0
o
o
CO
0
o
co
CO
0
vo
vo
o
o
o
co
CD
o
VD
rH
0
O
VD
rH
^1
id
s
CD
o
co
CM
O
O
VD
CO
O
O
CM
CO
O
0
^j*
vo
o
0
co
CM
O
O
rH
in
o
0
CM
•^
o
o
rH
r^
o
o
CM
CO
0
O
[*^
rH
O
O
*a*
CO
0
o
-a1
rH
o
o
en
CM
o
o
CM
rH
0)
3
1-3
CD
0
rH
in
O
O
CO
CM
O
O
r^
co
o
o
rH
CM
O
0
en
o
o
o
VD
i-H
O
O
VD
"*
O
O
CO
in
o
o
CO
CO
o
0
o
o
rH
0
O
en
r-
o
CM
o
o
•e^*
CO
O
o
fN,
i-H
I>1
rH
rj
^
O
o
co
in
o
o
o
o
VD
O
o
o
f*^
o
o
. co
rH
O
O
O
• CM
i
O
o
o
o
CM
O
O
O
VD
CM
O
0
rH
rH
O
o
rH
CO
o
o
CM
CM
O
VD
-a*
o
o
o
in
CD
*sf
"31
o
*3*
r^
4J
CO
3
tn
•<
o
0
m
CO
o
o
o
CM
O
O
rH
en
0
o
^
*"•
o
0
CD
CM
t
O
o
CM
i-H
O
O
0
CO
rH
O
O
en
CO
o
o
CO
CM
o
o
in
o
o
rH
CM
o
o
o
CM
o
o
co
CM
O
o
CM
rH
0)
g
0)
-P
Q-i
Q)
co
o
o
o
CO
CD
O
CM
••*
0
O
in
CM
CD
o
vo
•*
o
o
CO
CO
o
o
o
H
o
o.
co
o
o
en
t^-
o
0
CO
CM
O
O
53
rH
rH
0
O
en
rH
o
o
53
CM
o
o
rH
**
o
o
o
CM
CM
0)
o
0
o
o
o
i-H
*3*
0
O
CM
CM
0
O
o
rH
rH
O
o
VO
co
o
o
CO
"31
o
o
vo
i-H
o
o
0
CM
0
o
CM
in
o
0
r-
rH
o
0
en
CM
o
o
vo
rH
CD
o
CM
CM
O
Q
VD
CM
O
O
in
i-H
M
fl)
g
0)
O
o
o
CO
CM
0
O
CM
rH
O
O
CM
CM
O
0
CM
rH
O
O
CO
CO
O
O
CM
rH
O
0
VO
O
O
00
*sf
o
0
o
VO
0
0
CO
CM
O
O
CO
oo
o
o
CO
in
o
o
CO
r-
o
o
in
<*
0)
g
(D
U
V
Q
o
o
CM
CO
0
O
rH
CM
O
O
'CO
CM
O
0
O
CM
O
VO
VO
O
o
in
CM
o
0
VD
CM
O
O
-a*
CM
O
O
CM
rH
0
o
CO
CM
0
O
^3*
rH
O
O
t^
CM
O
O
p^
rH
O
O
0
CM
^4
id
0
c
m
^
f S £J
CO O
co ro
rH
o o
o o
VD CM
rH CM
O O
CM O
CO f-
CO
0 0
CD O
CM 0
CM CM
O O
^* O
r-~ ^
CM
o o
o o
CM in
CM CM
O O
O O
cn r-- .
rH •-)
O CD
o o
in •*
CM in
0 0
O CO
r- VD
rH
O O
CO O
in VD
in
0 0
o o
o co
CM
o o
0 0
O CM
CM cn
o o
0 0
[^ ^J<
rH
0 0
o o
CO O
CM r-
r?
id
3 .G
M O
Q M
cu id
pn S
291
-------
CM
00
111
O
H
O
o
CO
292
-------
in
O"\
rH
H
•a1
rH
01
CO
rH
P!
to
ffl
S3 CO
CO
P rH
g
,_.
O •*
CO
M rH
^
<
W co
CO
H
CM
CO
, —
P_
OC
SB
&H
13
i
o
rH
en
I
O
CM
O
in
o
CM
o
I
o
o
0
CM
O
in
o\
vo
1
CO
CM
*4*
in
1
m
in
o
CM
o
CM
O
CM
1
O
CM
rH
•H
p
<
O
o
CO
rH
I
O
CM
O
CO
rH
1
O
CM
O
in
1
o
CM
o
in
in
o
CO
CM
*3*
CTl
1
CO
CM
*3*
O
I
CO
CM
O
rH
rH
1
r-
in
o
CM
I
o
CM
to
tn
3
0
CM
|
0
f-^
CM
O
CM
1
O
o
r-
o
CM
I
o
CO
in
rH
CM
1
o
0
CO
rH
O
«^i
rH
1
0
CO
rH
O
CM
|
rH
*sl*
O
CM
1
in
vo
o
CM
1
O
CM
<1)
3
&
w
o
[^
CM
|
O
CO
rH
O
CTl
i — |
1
CO
CO
CM
*3*
1
O
in
CM
o
in
in
i
o
••*
CM
G\
CO
1
CO
CO
o
CM
rH
|
in
p^
CD
in
1
o
CM
O
CM
1
O
CM
rt
<1>
•8
-p
0
o
o
o
vo
CM
|
CO
CO
o
co
rH
1
VO
o
o
rH
^f
|
O
CO
o
o
a\
CO
1
0
o
CO
rH
O
in
cr>
1
CO
vo
o
VD
CM
|
0
vo
CO
CM
|
*a*
CM
CO
in
rH
I
o
CM
rl
0)
5
O
o
o
in
|
o
in
CM
CO
^3*
I
0
rH
co
in
r^
I
o
r-
CM
o
CO
^*
I
o
o
*d*
i-H
0
r^
I
0
CO
rH
CM
CM
|
O
CO
CM
in
vo
1
••*
••*
co
CO
1
CT>
**
H
co
1
CO
in
CM
CM
|
in
vo
o
CM
1
o
CO
&
§
a
0
r~ CM
1 l
o o
in CM
co r~
CO •*
1 1
o o
-* CM
r?
a
2 "o
(!) (0
h 2.
293
-------
100,00 fl-
,000-
1000
100
10
\
\
— '
^-"
512 5
\
\
\
X
X"
SHIP
4 5
./
•*
X
X
CANA
6 5
f
\
\
\
\
\
\i
/
/
/
V
\
\
\
\
\
\
7 5
ILY-S
1971
V
\
\\
\
,
1 PRIL
I97t
8 5
EPT
^
S
ST
•SEPT
LAK
!0 5
/
/
/
/
/
/
/
r
«••»
VNDA
/
/
/
/
E UNI
r 5
1
r
X.
^
RD
\
\
\
\
\
)N
!8 5
\
\
\
\
\
LY-S*
I97O
^~
/
/
F
r 5'
/
fs
~PT
•^
^
\
/
/
ORTA
16 5
-—.^
v
\
\
\
\
iE BA
i8 5
\
\
\
\
\
\
V
1
\
\
\
^
f
W 5'
\
\
\\
\\
^.
"^^
\Z 5
r
FIGURE 83
MEDIAN COLIFORM COUNTS,
LAKE WASHINGTON SHIP CANAL
294
-------
RELATED STUDIES
The CATAD project concentrates on combined sewer problems re-
sulting from excessive additions of storm runoff, so it is essen-
tial to study such effects separately. First, since rainfall patterns
and intensities play a major role in overflow characteristics, storm
analyses are evaluated. Much of the urban area is related to auto-
motive uses, so runoff quality typical of streets, highways and
parking areas is studied next. Finally, the effects of sewer separa-
tion are examined to distinguish changes in overflow quality or
quantity from tributary areas which were partially separated during
the course of this study.
Storm Studies
The evaluation of CATAD system performance through the demon-
stration grant period required a thorough analysis of rainfall
patterns. Rainfall data from six gages in the Elliott Bay-Duwamish
River region was collected and analyzed for the period November 1,
1969, through July 1, 1973. The data included one storm of 0.47
inch for a 60-minute period which, according to Figure 84, is a
five-year storm. Average annual rainfall for the period was 38.25
inches, within the average of 30-40 inches (see Table 27). So it
can be assumed that the collected rainfall data is representative.
Total storm data, summarized in Table 27, shows that in the study
period, storm activity for both total number of storms and inches
of rainfall, was relatively normal with the following exceptions:
1. 1970 had a drier than normal summer season.
2. 1971 had more storms, and was somewhat wetter
during summer but well within the norm.
3. 1973 had unusually light rainfall up to July 1
although the number of storms was normal.
295
-------
20.0
15.0 =
8.0 =j
6.0
4.0
O
X
a
CO
X I. O =
o
== O.8
= o.<
to
z
til
I-
z
_l
_J
£
z
).4
0.2
« 0. I =
.08 -
.06
.04
.02
X
NOTE
FREQUENCY ANALYSIS BY METHOD OF
•EXTREME VALUES, AFTER GUMBEL
15 20 30 40 50 60
MINUTES
DURATION
FIGURE 84
3 456 8 10 12
HOURS
18 24
CITY OP SEATTLE RAINFALL FREQUENCY,
INTENSITY, DURATION
296
-------
Storms
Inches
Storms
Inches
171
37.16
144
40.38
87
18.22
Table 27. TOTAL ANNUAL STORM DATA3
96
20.46
82
22.75
87
9.33
Slimmer0
Season
11
1,54
22
3.64
16
2.38
a Note: Annual, average 30-40 inches/year (43)
Summer average 2-6 inches/year
k January to June inclusive
c June to August inclusive
Average monthly rainfall patterns for the Seattle area are
shown in Figure 85.
This section defines summer or recreational periods as those
months averaging two inches of rainfall or less. This recreational
period extends from May through September. A computer program was
developed to analyze rainfall data for total accumulative rainfall
maximum rates and average rates for various ranges as shown by
sample tables presented in Appendix K. Although the demonstration
grant sampling phase began in November, 1969, regulator and sampl-
ing unit construction was not complete until April, 1971. Compari-
sons of present overflow data with earlier data are inaccurate due
to this missing information. However, the storm study includes all
valid rainfall information for the entire period November, 1969, to
July, 1973.
Table 28 compares storms during the demonstration grant period,
compiled by averaging together data from all operating rain gages.
The table shows no unusual change in storm patterns over the demon-
stration grant period. Total accumulative rainfall amounts, over
the total period and for post-supervisory and post-auto control
periods, show an average of 80 percent of storms is under 0.3 inch
with a slight increase during post-supervisory control periods.
Maximum rainfall rate distribution is even closer, with a variation
297
-------
18-
16-
14-
12-
10-
8-
6-
4-
2-
JAN FEB MAR
JUN JUL AUG SEP OCT NOV DEC
FIGURE 85
AVERAGE MONTHLY PRECIPITATION
CITY OF SEATTLE
298
-------
,of only three percent for all rates of 0.07 inch per hour or less
during any of the three study phases. Average rainfall rates for
storms of 0.05 inch per hour or less are essentially constant over
the entire period and for each control phase.
During summer storm periods only, the number of storms which
accumulated 0.3 inch of rainfall or less varied no more than seven
percent over the entire period and for all study phases. Similar-
ily, maximum rainfall rates and average rainfall rates were closely
related for the entire study period and the individual control
phases. The last column in the table indicates any significant
changes in the percentage of summer storms versus total storms for
the different project phases. Although a significant increase in
summer storms is indicated during the post-automatic control phase,
the data is biased by the fact that two of the three months were
defined as summer months.
An alternate rainfall data review is shown in Figure 86. The
percentage of summer storms versus total storms for each range of
total rainfall, maximum rainfall rate, or average rainfall rate is
plotted for the entire study period and for each control phase.
This figure indicates a greater percentage of storms occurring in
the lower ranges of total rainfall and maximum or average rainfall
rates during the supervisory and automatic control phase of the
study. This apparent distribution change during the last fifteen
months of the study as compared to the total forty-five month per-
iod occurred simultaneously with an apparent increase in percentage
of summer storms. The 20 percent increase in summer storm activity
shown in Table 28 is partly offset by the lower number of more in-
tensive storms which occurred after implementation of CATAD super-
visory and automatic controls."
299
-------
a
s
o
1
s
U
IN
(1)
H
•o
4
1
)
4
c
a
}.
it
(U
(H
4-
CO
j.
1C
cu
X
(U
it
K
•
CJi
^
(!)
_[.]
rc
•
X
id
v/s
£
•H
O
a
•H
V>
H
S^
C
H
in
o
W
J5
—
C
•H
O
v;
J3
O
«
d
g
0)
4J
id
•
I
a)
4J
id
X
a
«
H
id
H H
id
P
o
v/S
•o
o
•H
M
04
I
ro
O!
O
Ol
01
,_{
CTi
cn
CO
00
OO
OO
^1*
r^
00
•^
CO
r^
^^
c^
Ul 1
id cn
0) vo
00
01
ro
ro
01
en
CO
10
CO
CO
!>.
CO
CO
r»
<£>
01
X
Q «*
E~t 1
6cn
OJ\
SH
m
cs
0
lO
00
CN
^*
a\
*&
CO
o
cn
CO
CO
t^-
t**
OJ
CO
n
§CN
*N^
EH r*^
irf
l '
4J 03
w r^
O\
CN
^j*
r-
^
VD
in
CT\
o
CO
o
CO
m
cn
r-
r~"*
m
OO
ro
f^.
O "sx
4J r^-
p
< i
4J ro
w r-
OX
(Xt ^
,T3
: o
i 'H
I )-l
!2
X
i a
i^J
4J
U
Ol
•a
5
o
01
M
M
3
300
-------
sS 20
I 1 f 1 1 1 I 1 1 I I I I I I I 1 I
0.01 0.02 0.03 OO1 005 0.06 OX>7 O.08 009 0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18
MAXIMUM RATE RANGE, (inch per hour)
IOO-
90-
t£ 8
UJ
i«H
W
CO
z so-4
tr
| 50-
o:
O 40 •
_] 30 -
-POST-AUTO
0.10 0.20
0.30 O.40 O.50 O.60
TOTAL RAINFALL RANGE tin)
0.70 0.80
POST - AUTO
POST-CATAD
PRE-CATAO
.01 .02 .03 .04 .05 .06 .07 .08 .09 .10 .11 .12 .13 .14 .15
AVERAGE RATE RANGE, (inch per hour)
FIGURE 86
SUMMER RAINFALL DISTRIBUTION
jOl
-------
It was somewhat difficult to analyze storm patterns and their
effect on overflows for this demonstration grant. Storm patterns
might be divided into three groups:
1. An area-wide or uniform storm which is nearly the
same for all time periods at each analyzed gaging
station
2. A traveling or frontal storm which proceeds from
one direction to another
3. A pocket or localized storm which occurs or is
much heavier at one or two gaging stations.
Table 29 shows storm pattern variations for the project control
phases and seasons. Only storms of 0.05 inch or greater are in-
cluded. All control phases show a pattern change from frontal
winter storms to localized summer storms. The post-CATAD and
post-auto phases both indicate an improved ratio of the theoreti-
cally more-controllable local and frontal storms compared to the
less-controllable uniform storms.
Highway Runoff
Metro assisted the Washington State Department of Highways
in preparing environmental impact statements for interstate free-
way 1-90. Samples were drawn directly from roadway runoff before
it entered either storm drain or combined sewer. The findings of
these efforts are summarized below.
Suspended solids, BOD, COD and oil concentrations appeared
relatively high while nutrient levels were comparable to those
found in urban drainage samples. Lead and zinc were the major
heavy metals and were found principally in the solids portion of
the sample.
An example of pollutant concentration and loading found in
the runoff from the 1-90 corridor is presented in Tables 30 and 31.
This data was obtained from the South Bellevue interchange during
Phase I of the study. Differences in pollutant concentration and
loading for different geographical areas may have been due to diff-
erences in road surface, weather, type of nearby soils which may
be washed or windblown onto the roadway, types of vehicles, and
average vehicle speed.
302
-------
o
H
H
Pi
EH
§
Oi
o
EH
01
Ol
oo ^r 01 VD
dP dP
00 VO VD 00
rH rH rH
N-* ^-«
dP dP
IJD "41 01 in
00 "* i-H 01
t^- dP Ol dP
rH OO •«*
OO dP dP
rH r** ^i* OO
rH 01 01
LO dp "^ dP
V0 00 in rH
rH OO OO
01
r— oo
X. Q r-
<1 EH r->
rtj U 1
0 cy> i
1 VD -P 01
0) X w c^
M ' — 1 O X.
Pi p— j P[ ^i
dP
01 O
H 00
dP
rH f~
*-*.
o\o
01 OO ,
H
r^ OP
H 00
vo
VD dP
^J*
01
Ol dP
OO
~~"
OO
f^.
OX.
•P r-
f '
-P oo
en r~
o x»
a, -*
'H
H
(d
rl
(D
-P
nJ
0)
M
Cn
rl
o
o
c
•H
in
o
m
o
o
•P
CO
td
303
-------
Table 30. STORMWATER RUNOFF QUALITY
SOUTH BELLEVUE INTERCHANGE
PARAMETER
Temperature, °C
PH
Conductivity , micromhos/cm
Turbidity , JTU ' s
Settleable Solids, ml/1
Suspended Solids, mgs/1
Volatile Solids, mgs/1
Total coliform, organisms/100 mis
Fecal coliform, organisms/100 mis
COD, mgs/1
Oil, mgs/1
BOD, mgs/1 (s)
Ammonia, mgs N/l (s)
Nitrate-Nitrite, mgs N/l (S)
Kjeldahl Nitrogen, mgs N/l (I)
Phosphorus, mgs P/l (s)
Phosphorus, mgs P/l (I)
Lead, mgs/1 (s)
Lead, mgs/1 (i)
Zinc, mgs/1 (s)
Zinc, mgs/1 (i)
Copper, mgs/1 (s)
Copper, mgs/1 (i)
Cadmium mgs/1, Total
Chromium, mgs/1 Total
MAXIMUM
8.2
7.8
4500
175
2.7
1496
396
14,000
2,200
433
96
42
0.44
1.85
4.20
0.12
1.10
0.52
4.50
0.64
0.749
0.20
0.16
0.016
0.01
MINIMUM
4.0
6.4
48
24
0.1
42
12
20
20
30
0.0
5
0.05
0.30
0.09
0.02
0.03
0.10
0.10
0.005
0.010
0.01
0.01
0.004
0.01
MEAN
6.0
7.1
500
84
0.4
246
57
650(D
so'1'
123
14.9
13
0.22
0.75
0.50
0.05
0.26
0.17
0.99
0.059
0.111
0.034
0.018
0.004
0.01
(1) Median Concentration
(S) Soluble
(I) Insoluble
304
-------
Table 31. LOADING FACTORS
SOUTH BELLEVUE AREA
Parameter
Suspended Solids
Volatile Solids
COD
Oil
BOD
Ammonia as N
Nitrate-Nitrite as N
Kjeldahl Nitrogen as N
Phosphorus as P
Phosphorus as P
Lead
Lead
Zinc
Zinc
Copper
Copper
(S)
(S)
(S)
(I)
(S)
(I)
(S)
(I)
(S)
(I)
(S)
(I)
Loading Factors
mgs/m^/day
202
47
135
12.1
9.0
0.12
0.44
0.96
0.04
0.24
0.12
0.72
0.02
0.10
0.01
0.02
mgs/m/vehicle
25.8 X 10~2
6.1 X 10~2
17.4 X 10~2
15.5 X 10-3
11.6 X 10-3
15.8 X 10~5
55.4 X 10~5
12.4 X 10~4
5.3 X 10-5
3.2 X 10~4
1.6 X 10~4
9.2 X 10~4
2.6 X 10~5
13.2 X 10~5
13.2 X 10"6
26.4 X 10~6
(S) Soluble
(I) Insoluble
305
-------
Dispersion coefficients and dilution factors measured as a
part of this study were used to recommend the best stormwater out-
fall sites for Lake Washington. These coefficients showed that
the pollutants were dispersed rapidly in the receiving waters to
levels within Washington State's established standards (44) or
close to accepted criteria where specific standards are unavail-
able (45, 46). This dispersion was further confirmed by measur-
ing receiving water quality during and after a typical runoff
period.
In addition to measuring storm runoff pollutants from exist-
ing ^ freeway facilities and projecting these values to future
facilities and traffic increases, an attempt was made to evaluate
the impact of highway construction. As expected, the major pollu-
tants measured were solids and oil, with solids coming from large
areas where ground cover was removed and oil from new asphalt
paving. Streams adjacent to this type of highway activity are
definitely effected by roadway drainage. Construction effects
can be minimized by adopting improved pollutant control techni-
ques (39).
In general, no significant water quality problems were indi-
cated from the 1-90 stormwater runoff discharge into major receiv-
ing water bodies, particularly Lake Washington. Cumulative effects
were not investigated. Preventive measures such as sweeping and
vacuuming should help alleviate problems caused by the two major
freeway surface pollutants; solids and oils.
Sewer Separation and Urban Storm Drainage
The City of Seattle's combined sewer separation project in-
creased the volume and frequency of stormwater discharge into Lake
Washington. Evaluation of the separation program impact on water
quality indicated the coliform organism concentration had dimin-
ished by a factor between 60 to 140 times, and a lower fecal to
fecal strep ratio, indicative of non-human sources, was evident
(12). Stormwater nutrient concentrations were lower than in the
combined sewer, showing less potential for contributing to eutro-
phication. The nutrient concentration reductions varied by the
following factors: 5 to 8 times for total phosphate, 1.3 to 2
times for nitrate, and 7 to 11 times for ammonia. Oxygen demand
and solids were 2 to 5 times lower (12).
Despite the increased stormwater flow to the lake, total load-
ing of the parameters under study indicated a significant loading
reduction achieved by separation. The discharge factors in pounds/
acre/year after separation for the various constitutents were as
follows: BOD - 19.4; COD - 88; suspended solids - 72; NH4 - 0.63;
NO3 - 0.46; and P04 - 0.70. This represents a decrease of 37 to
64 percent compared to the combined sewage overflow.
306
-------
To determine the role of stormwater from combined sewer
systems on water quality, preliminary data from Metro's urban
storm drainage studies was reviewed. From a single storm in
March, 1973, samples were collected from a residential (VRl),
commercial (CBD-7), and parking lot area (SC-4). Water quality
data from those samples are presented in Appendix L. In the area
draining a large parking lot at the South Center shopping mall
(south of the Norfolk St. Regulator), hydrolyzable phosphate,
dissolved oxygen and total coliform lagged the peak rainfall by
15 minutes. Nitrate, lead, zinc, iron, oils, turbidity and solids
peaked at the same time as the flow. Suspended solids, BOD,
cadmium, chlorinity and Kjeldahl-nitrogen peaked 15 minutes before
peak rainfall; while some parameters such as ortho phosphate and
ammonia dropped and then stayed low. In the case of the residen-
tial area (VR2-1), the lag was not recorded for some parameters
such as nitrate, turbidity, oils and COD; but the others, like
the hydrolyzable phosphate and ortho phosphate, peaked before the
rainfall did, suggesting that these parameters are initially con-
centrated (perhaps in the standing water in the pipes) and then
readily flushed from the system.
Copper, lead, zinc, oils and chlorinity were highest from the
parking area (SC-4), while turbidity, nitrates, phosphates and
coliform were much higher in the residential area. COD's were
similar for both areas.
The peak patterns for the water quality data from the central
business district were more complex. The extremely high bacterial
counts, phosphate and all metals but lead, suspended soilds, chlor-
inity, BOD and oils, clearly substantiate the potential of runoff
from this combined system to degrade water quality (Appendix L).
Rainfall during the analyzed storm was greater in the central area
than the others observed, but many of the relatively high parameters
were measured during a period of relatively light rain preceding
the main storm. The capacity of the CATAD system to absorb light
rains from the central business district where the greatest regulat-
ing effort is concentrated is consequently a significant step towards
abating the water quality degradation which would otherwise occur
with every rain. Establishing parameter loading values will depend
on numerous storm intensity and antecedent rain periods being re-
corded and the resultant water quality data.
Tentative loading values for three storms in the 100-acre
central business district are recorded in Table 32 along with
values from one storm in the 150-acre separated residential commun-
ity of Lake Hills. Curves of increased parameter loading with
increased rainfall are suggested for relative loading of nitrite-
307
-------
Table 32. LOADING (Ibs/storm)
Station
Date
Rain (Inches)
Suspended Solids
BOD
COD
N02-NO3-N
Hydrol. P04
CBD
3/10/73
0.33
309.00
61.40
233.60
0.72
0.33
3/16/73
0.06
49.60
23.60
87.50
0.14
0.26
6/6/73
0.08
146.80
94.60
191.80
0.49
0.77
Lake Hills
3/16/73
0.08
157.20
16.80
1.22
0.49
nitrate nitrogen, suspended solids and COD (see Figure 87). From
the table, phosphate appears 1.6 times higher for the central busi-
ness district during the one storm comparison. Nitrite-nitrate
nitrogen/ on the other hand, was 2.35 times higher from the resid-
ential area.
When these observations from the recently initiated urban storm
drainage study are repeated in subsequent storms, it will be poss- ,
ible to anticipate loading changes from regulator overflows result-
ing from further separation.
Water quality data alone from the Norfolk Street Regulator
between 1970 and 1972 indicates slight increases for settleable
solids, suspended solids, volatile solids and phosphate, while BOD
and COD concentrations appear similar and the ammonia and nitrite-
nitrate nitrogen decreased (see Figure 88). This area was appro-
ximately 87 percent separated by 1972. The Norfolk drainage area
did not respond with markedly larger nitrite-nitrate nitrogen, or
phosphate decreases as the residential Lake Hills data suggests.
Perhaps additional data from storm drainage studies will better
demonstrate the effects of land uses and topographical configura-
tions on storm runoff characteristics.
REGULATOR STUDIES
Paralleling the receiving water studies is a program of sample
collection and data reduction related only to overflow events. In
the early stages of this demonstration grant project, the require-
ment for precise and complete data was recognized and the semi-
automatic equipment described in Section IX was installed. Over-
flow quality and quantity information provide a basis for evaluating
308
-------
500-
1-0
CHEMICAL OXYGEN DEMAND
0-1 0-2
TOTAL RAINFALL (Inches)
FIGURE 87
STORM DRAINAGE LOADING
0 3
309
-------
ISO -,
--. 100-
v.
D>
E 50H
1500 -i
-;; 1000 -
•v.
O>
-§ 500
1970 1971 1972
SETTLEABLE SOLIDS (Increase)
1970 1971 1972
COD (Same or decrease)
9000 -i
1970 1971 1972
•SUSPENDED SOLIDS (Increase)
9.0 -,
,-. 6-0 -
3 3-0-
1970 1971 1972
AMMONIA-N (Decrease)
9000-,
= 6000-
>s
o»
E 3000-
1
1970 1971 1972
VOLATILE SOLIDS (Increase)
1-5 -
tr 1-0 -
v.
IO-SH
1970 1971 1972
NITRATE - N (Decrease)
1500 -,
^ 1000-
E 500-
1970 1971 1972
BOD (Same or decrease)
3.0-,
_ 2-0 -
1970 1971 1970
PHOSPHATE (Increase)
FIGURE 88
NORFOLK REGULATOR - LONG TERM OVERFLOW QUALITY
310
-------
receiving water loading as related to changes observed in the
ongoing area-wide monitoring and for measuring performance of
individual regulators and the entire system. A limited sampling
program began in the summer of 1968. System-wide data was not
available until April, 1972, the starting point for most of these
studies.
Regulator Water Quality Analysis
Initial regulator data analysis included a general evaluation
of each regulator station's performance relative to the area con-
trolled, samples collected and relative frequency of overflows
recorded. This data is presented in Table 33. The four most
reliable station/sampler combinations were selected from Table 33
to represent the entire study area and smaller subareas which had
undergone sewer separation.
Some stations seldom overflowed while others overflowed fre-
quently with either high intensity and short duration, or moderate
intensity and extended duration. This dissimilarity was attributed
to the amount of storage available (see Figures 89 and 90). Sequen-
tially sampled water quality and flow quantity were available for
Denny Way-Lake Union, Connecticut Street, Hanford, Chelan Street,
and Michigan Street regulators. The Hanford Street regulator serves
a mostly hilly, residential area, 42 percent of which has been
partially separated (27). Denny Way-Lake Union and Connecticut
Street serve the central business district. Chelan Street and
Michigan Street serve the flat south industrial area. The areas
served vary in size from 1.17 square miles at Michigan Street to
6.3 square miles at Hanford Street (Appendix A).
The average of sequential water quality parameters vs. time
during overflow periods is presented in Figures 91 - 95. These
sequential sample plots illustrate unique loading profiles for
each of the stations. The correlation between water quality para-
meters is most apparent for the Connecticut Street Regulator
(Figure 91). The ammonia and nitrite-nitrate nitrogen, total
phosphate, COD, settleable, volatile and suspended solids all peak
after more than 10 hours while the BOD loading appears highest
initially and gradually declines.
Sequential averaged data for Chelan Regulator (Figure 92) is
characterized by two initial peaks within the first hour for
ami|pnia, nitrite-nitrate nitrogen, total phosphate and BOD. Solids
and COD, with delayed peaks, follow the trend of the Connecticut
regulator.
311
-------
i s
I -i
I!
$
H *O O)
o co o •»
CO O) CO i-I
3
CO ^3* OO O\ -rH rH
CO I""- OO OO
•-i rH rH rH
rH OJ rH rH Ol rH
33 33
o
3
o
3
5
•g,
o
CO
o
ro
§
5 5
\St
H r-l
r, J-.
•O1 N
r,
VD
R
O O
^ ^
••a- m
rH
^
in
rH rH rH
^o ID m
I I I
I
312
-------
INTERCEPTOR FLOW
o
o
o
o
o. .o
so
o
o
*
c.
into
u_
20
cc.
UJ
>
Oo
o
LflNDER
tn
gfe
§I
U_
§.g
10 Q_
UJ
O
gUJ
O. .OZ
UJ
10UO.OO 1080.00 1120.00 1160.00 1200.00 1240.00 1280.00 13?0.00
TIME
MINUTES AFTER MIDNIGHT
HflNFORO 1
INTERCEPTOR FLOW
(NOT OPERATING)
^Odd.00 1040.00 1080.00 1120.00 1160.00 1200.00 1240.00 1280.00
TIME
MINUTES AFTER MIDNIGHT
FIGURE 89
OVERFLOW CHARACTERISTICS OF LARGE STORAGE STATIONS
313
-------
DENNY LOCAL
INTERCEPTOR FLOW
(NOT OPERATING)
°10dO.OO 1040.00 1080.00 1120.00 1160.00 1200.00 1240.00 1280.00 132*0.00
TIME
MINUTES AFTER MIDNIGHT
KING
INTERCEPTOR FLOW
•V ¥ ¥-
O>
.o
?tb5b.OO l'OMO.00 1080.00 1120.00 1160.00 1200.00 1240.001280.00 132*0.00
TIME
MINUTES AFTER MIDNIGHT
FIGURE 90
OVERFLOW CHARACTERISTICS OF SMALL STORAGE STATIONS
314
-------
o
_J<=>
o
o
o
o
ZJ«C\J'
o
Q_
o
o
o
co
o
o
o
COoo"
o
o
10.00 20.00
SEQUENCE
^.00 10.00 20.00
SEQUENCE
10.00 20.00
SEQUENCE
10.00 20.00
SEQUENCE
Q
Sg
o
o
O
OQ
o
o
o.
CM
O
CD
O.
=!•
en
cn§
10.00 20.00
SEQUENCE
1.00 9.00 17.00
SEQUENCE
10.00 20.00
SEQUENCE
10.00 20.00
SEQUENCE
CONNECTICUT STREET
FIGURE 91
CONNECTICUT STREET OVERFLOW QUALITY VARIATION
315
-------
|O
0°
cnoo"
UJ
en
o
o
o
o_
o
CO
o
to
COca'
O
o
e\j
o
o
o
o
"1.00 8.00
SEQUENCE
^.00 1.00 8.00
SEQUENCE
1.00 8.00
SEQUENCE
1.00 8.00
SEQUENCE
og
o
o
Qcvi
O
03
o
o
o
CM
CO
o
=f
CD
CO
cng
CHELRN
^3.00 U.OO 8.00
. SEQUENCE
^3.00 1.00 8.00
SEQUENCE
1.00 8.00
SEQUENCE
1.00 8.00
SEQUENCE
FIGURE 92
CHELAN OVERFLOW QUALITY VARIATION
316
-------
0°.
COoo"
LU
cn
o
o
Q
o
Q
Q
o
CD
OOo'
ED
o
Q
a
o
o
Q
8.00 16.00
SEQUENCE
8.00 16.00
SEQUENCE
8.00 16.00
SEQUENCE
'.00 8.00 16.00
SEQUENCE
o
o
CD
Q
O
o
o
Oj
cn.
o
>0
X
Q_
(D
cng
HflNFORD
8.00 16.00
SEQUENCE
8.00 16.00
SEQUENCE
.00 8.00 16.00
SEQUENCE
8.00 16.00
SEQUENCE
FIGURE 93
HANPORD OVERFLOW QUALITY VARIATION
317
-------
CO-
UJ
in
o
o
"tl.OO 4.00 8.00
SEQUENCE
o
o
=!•=;•
o
Q_
O
O
o
=»•
COo'
o
O
o
o
COco'
o
o
5.00 9.00
SEQUENCE
°1.00 5.00 9.00
SEQUENCE
10 5.00 9.00
SEQUENCE
„ o
0°.
o.
=f.
IT. 00 4.00 8.00
SEQUENCE
o
o
o
CD_
a.-
CD
Q
o
"b.OO 4.00 8.00
SEQUENCE
o
o_
CO
o
o_
IT)
CD
cng
4.00 8.00
SEQUENCE
4.00 8.00
SEQUENCE
DENNY LRKE UNION
FIGURE 94
DENNY LAKE UNION OVERFLOW QUALITY VARIATION
318
-------
—'2
cn
o
o
°1.00 9.00 17.00
SEQUENCE
o
o
o
o
o
rr
COo'
o
o
o
o
o
o
10.00 20.00
SEQUENCE
10.00 20.00
SEQUENCE
". If) 10.00 20.00
SEQUENCE
og
Jl'.00 9.00 17.00
SEQUENCE
o
o
Qm
O
CD
o
o
°1.00 9.00 17.00
SEQUENCE
o_
o
cn
J1.00 9.00 17.00
SEQUENCE
._ o
*o
o_
o
CO
cn§
..oo
MICHIGRN
9.00 17.00
SEQUENCE
FIGURE 95
MICHIGAN OVERFLOW QUALITY VARIATION
319
-------
The first flush effect appears to be represented best in
Hanford Street regulator (Figure 93) for ammonia nitrogen/ total
phosphate/ and BOD. Nitrite-nitrate nitrpgeri, solids and COD
peaks lag by about one hour, with a steady increase of the nitrite-
nitrate with time. The Denny Way-Lake Union Regulator (Figure 94)
has ammonia-nitrogen, total phosphate/ solids, BOD and COD with
highest values toward the end of the sampling sequence similar to
Connecticut Street.
At Michigan Street Regulator (Figure 95), the high initial
peak occurs for ammonia nitrogen, BOD and to a lesser extent, for
COD. The nitrite-nitrate nitrogen and total phosphate peak is
near the end of the sequence while the solids have a major peak
after 1.5 hours.
Comparing parameter profiles between stations, some generali-
zations can be made. Stations 14, 23 and 31 have similar sequential
profiles for the solids and COD concentrations. Stations 23 and
36 have some similarities for BOD, P04 and nitrite-nitrate nitrogen,
while Stations 26 and 36 have similar responses to ammonia nitro-
gen loading.
Sequential sampling data from the five regulators is typical
of the large amount of information collected during this study
(Appendix M). The "first flush effect" is not obvious from this
data. The automatic samplers were programmed to begin operating
as the outfall gate opened. Overflows generally occur well after
a storm has begun and water levels throughout the collection
system have built up to their limits. This means that, at most
stations, the heavy concentration of solids and other materials
has probably already been diverted to the interceptor before the
overflow begins and sampling takes place. At a few stations,
like Hanford, Norfolk and to a lesser extent, Michigan, the drain-
age basin is spread out so the first flush from some portion of
the basin would frequently be seen during an overflow.
Factors which might play a role in influencing arrival time
of peak concentrations include drainage system configuration (see
Figure 96), pipe slope and size, degree of separation, distances
traveled, and land use of the service area. When the average
sequential sample plots for each water quality parameter are over-
laid on one another (Figure 97), distinctly different three-dimen-
sional profiles become apparent for each regulator station. Each
profile is much like a station "fingerprint" which usually repeats
itself for different storm types.
320
-------
FIGURE 96
REGULATOR DRAINAGE BASINS
321
-------
\A/VW\A
VOLATILE SOLIDS
SUSPENDED SOLIDS
SETTLEABLE SOLIDS
COD
BOD
PHOSPHATE
NITRATE- N
AMMONIA -N
HANFORD
CONNECTICUT
VOLATILE SOLIDS
SUSPENDED SOLIDS
SETTLEABLE SOLIDS
COD
BOD
PHOSPHATE
NITRATE - N
AMMONIA-N
MICHIGAN ST
CHELAN
FIGURE 97
DISTINCT OVERFLOW PROFILES
322
-------
Regulator water quality .response to, variations in rain inten-
sity was. anticipated hut the variations. :in the response between
regulators was quite dramatic, Table. 34 shows maxtoium concentra*-
tions' of various- water quality parameters; against average rain ,
interis-ity. Eighth. Avenue South, regulator seldom overflowed so
limited data is available. For those few overflows which did occur,
all water quality parameters- peaked during high average rain inten-
sities (0.09 r 0.10 inch/hrl. Most parameters at Harbor! Avenue and
Brandon Street peaked at relatively low1 rain intensities of 0,02
to 0.03 inch per hour. The storage capacity of Harbor Avenue and
Brandon Street regulators are 0.6 to 5.14 hours and dry-weather
flows, 0.02 to 0.3 million gallons, respectively Csee Table 3). If
Brandon Street is compared to Michigan Street, which has similar
storage time but twice the storage capacity, peak values usually
occurred at higher storm rates for solids but not for the nutrient
parameters measured. The nutrient relationship cannot be entirely
explained at this time. Both stations are served by a very long
stagnant overflow line where septic waste would be unaffected by
light rains. The lower solids peaks at both stations may be due
to settling of materials and biological activity in the overflow
lines.
The different response of each regulator to various rain in-
tensities reaffirms that these stations are unique in their
responses and impact on the environment. Present automatic con-
trol strategy is directed to minimze overflow volumes and, in-
directly, loadings; but also considers receiving water capacities
and standards. Future control strategies may need to consider
the intensity and timing of peak overflow loading as a discharge
limitation. If the objective is to minimize loading of a specific
parameter, overflow priorities of individual stations should be
assigned according to their optimum or lowest concentration res-
ponse to rainfall intensity. If suspended solids were to be con-
sidered the critical parameter, then the sequence of overflowing
might be as follows: Station 30, 34, 14, 13, 37, 31, 26, 25, 36,
23, 20 and 28 respectively. If BOD loading were the critical
factor, the sequence would be: Station 36, 37, 13, 14, 30, 31,
20, 25, 26, 23 and 38.
323
-------
s a
§ -3
324
-------
I
System Overflow Loading to Receiving Waters
Difficulties were encountered in total loading calculation
because of the previously reported variations in loading with
time and rain intensity and because of some missing data resulting
from sampling or recording equipment problems. Therefore/ the
total loading and probably loading reduction resulting from CATAD
operation must be assessed from a total summation of available data
on each parameter. By treating the extreme values generated before
CATAD was fully operational as conservative estimates of loading
without control, then differences noted in the remaining 80-90
percent of the data should reflect reduced loading with control.
This approach to assessing nitrite-nitrate nitrogen loading
in Figure 98 shows the initial large increases appearing by 0.2
inch of rain and then the rapid decline with dilution around 0.82
inch. At the extreme of 2.72 inches of rain, a slight increase
with both controlled and uncontrolled loads coincides as the system
approaches its hydraulic limits and surface runoff is at its maxi-
mum flushing capability. The apparent reduction in peak nitrite-
nitrate nitrogen loading is 78 percent with CATAD control. Figure
99 shows a probable reduction of 84 percent in peak loading of
ammonia-nitrogen with CATAD control. Figure 100 shows an apparent
reduction of at least an 81 percent in peak phosphate load. The
background COD, data fits a well defined curve which is 88 percent
higher than the CATAD controlled data in Figure 101. Normalized
loading values for the suspended, settleable and volatile solids
are combined in Figure 102. The apparent peak solids load reduc-
tion is around 85 percent. If the total loading, as reflected by
the area under the curve, is computed, reduction in COD loading
between 0.01 and .2.72 inches of rainfall is 80 percent. In the
case of nitrite-nitrate nitrogen, loading reduction for the same
range of rainfall would be 70 percent based on the area under each
curve in Figure 99.
Combining the peak loading data from all stations into Table
35 shifts the total system reaction to control strategies and the
peak loads to a higher intensity of rainfall. Thus it takes larger
storms occurring at a less frequent time interval to produce a
similar loading peak; another encouraging indicator of CATAD per-
formance. The next pages of this section will examine total
system reaction to storms concluding with an evaluation of system
performance.
325
-------
o
o
saNnod
326
-------
o
o
1
V\
-§s a
I 8
saNnod
VINOWWV
327
-------
SQNnOd
HlVHdSOHd
328
-------
SdNClOd '9NIQVOT 000
329
-------
SQNnOd dNVSnOHl 9NIQVO1 SdllOS
330
-------
Table. 35. TOTAL RAIN IN INCHES FOR
ESTIMATED: PEAR LOADING ALL STATIONS
BEFORE CATAD CONTROL
NH3 - 0.30
N03 - 0.55
COD - 0.50
PO4 - 0.45
Combined
Solid 0.60
Analysis
AFTER CATAD CONTROL
0.60
0.60
0.65
0.50
0.85
Average: 0.48
0.64
331
-------
OVERFLOW ANALYSES
The Delta.
The most direct evaluation of the effectiveness of the "total
systems management* concept as- applied to combined sewer control
facilities such as- CATADf is- system-wide comparison of overflow
and rainfall data through the' various control phases. The over-
flow data includes volume, number of overflows for the system
during any storm, and the' number of storms during which overflows
occurred. Rainfall data includes total rainfall in inches per
storm, maximum rainfall intensity in inches per hour, average rain-
fall intensity and storm type.
Computer programs condensed the data from separate overflow
and rain gaging stations as discussed earlier in this section. The
condensed data was merged into individual records for each storm,
then analyzed by a table-generating program described later in this
section. Computer plots presented these comparisons. Data volume
and the number of different analyses desired suggested this form
of automated data processing using the same computer which monitors
and controls the sewage collection system. A total of 514 storm
events was recorded during the three and one-half year demonstration
grant period. In 341 of these storm events, a total of 762 separate
overflows was recorded. The average of 2.23 overflows per storm
is considered low because of the five to ten percent data loss
caused by mechanical failures of recording and sampling equipment.
However, more than 60 percent of the recorded storms resulted in
only one overflow, or no overflows at all; so the average is greatly
influenced by multiple overflows during a few heavy storms.
The drainage study areas shown in Figure 96 and itemized in
Appendix A total approximately 13,000 acres. Overflow and loading
data for those overflows where nearly all data was available are
listed in Appendix L.
Overflow Volumes
Figure 103 is a computer plot of every overflow event during
the grant study except for the 5 percent of storms which fall out-
side the range of the axis of the graph and for those small over-
flows during storms under 0.05 inch of rainfall. The obvious scatter
effect indicates the difficulty in correlating system overflows
because of the many variables which influence the nature of the
storm and station reaction. The computer generated regression line
evaluates every overflow, whether it is plotted on the graph or not;
so the line represents the best estimate of overflow expected from
the entire system for any given rainfall. Note that the vast
majority of overflows above the regression line occurred during
the local control period.
332
-------
0
o
o
r~
o
o
20
00
-In
Ll_or»
ac
LU
>
00
o
o
tM
0
O
»
^4
0
o
z
LQCfl
SURE
RUT 0
X
+ X
, J_ *
L CQNTR
RVISORY
MflTIC C
f
X
V^
J*&*
QL = 2
CQNTRQ
QNTROL
2
X
X
s
s
+ ^s*
.JZT.
*l ^3
•«JB^r-
X
+
= Y
S
3J
S X S
2
J^^ *
+
1**+ +* +
i*^ *
s
s
i^^
s
+
, X
s XZY r
x ^ %
s
s
X
^
s
X
s x
+
X
X
+
X
X
Hi. bo o'iio 0*20 0.30 o.uo o.so u.bo u. /u
TOTflL RflIN, IN.
FIGURE 103
OVERFLOWS DURING ALL CONTROL PHASES COMBINED
333
-------
Figure 104 shows similar computer regression lines drawn
through overflow events occurring for each of the three control
modes. ^The vertical axis changes on' each plot providing early
indication of the reduced system overflow volume for advanced
control techniques. The individual plots of Figure 104 are merged
onto Figure 105 for comparisons of each mode on the same vertical
scale. In addition to the three control modes previously plotted
and combined on this figure, an additional line has been added to
show the improvement of local control, meaning mechanically driven
gate regulators, compared to "uncontrolled" static-type regulators
which have side spill weirs or orifice plate devices. Information ,
from this figure is used to project overflow amounts and cost savings
which may be expected by cities implementing one or more of these
control phases. Other agencies considering some form or combination
of these control modes can.use this figure to anticipate overflow
volume from similar drainage areas and storm volumes.
Indication of total system performance is obtained from Figure
105 by drawing a vertical line at a chosen rainfall amount and
comparing the overflow volume indicated by the regression lines.
For example, at 0.30 inch total rainfall, the estimated overflow
volumes from Figure 105 are shown on Table 36 together with a
percentage of reduction over less complex control modes.
Table 36. ESTIMATED OVERFLOWS AND REDUCTIONS FOR A SYSTEM
Control Mode
Static Regulator
Local Control
Supervisory Control
Automatic Control
Overflow, MGC
15.38
8.98
4.91
.49
Percent Overflow Reduction
vs. "Local" vs. "Static"
43.3
94.5
41.6
68.1
96.8
aFrom Figure 105 at 0.30 in. rain.
These performance estimates should be considered with some
degree of caution when comparing results among cities because of
different operating bases upon which overflow reduction calculations
are made. Automatic control reductions above 90 percent are based
on limited data for the summer period of 1973. These values should
not be considered firm, but rather an indication of excellent system
reaction to the unusually light rainfalls experienced in the Seattle
area for the early part of 1973. The 60 percent overflow reduction
during the supervisory control period is an expectable lower limit,
and a final performance level will probably fall somewhere between
334
-------
UJo
_J3-"
LOCflL CONTROL
-1/20/70-U/1/72-
.00 0.
20 0.30 U.llO 0
TOTHL RflIN, IN.
SO 0.
60 0.70
SUPERVISORY CONTROL
- 1/1/72 - H/l/73
0.20 0.30 O.UO
TQTflL RRIN, IN.
0.60
O
o
o
to"
o
in
1
3
,tn
j
3
§
o
in
o
o
o
=t
flUTO
U/l
HflTIC C
/73 - 7
^
3NTROL
Y
Y
^-
Y
Y
Y
Y
•^
^-
1
•^
"ao ' .10 0.20 .30 .10 0.50 0.60 0.70
TOTHL RflIN, IN.
FIGURE 104
OVERFLOWS FOR EACH CONTROL MODE
335
-------
o
o
o
o
o
to
DRAINAGE AREA
13,120 Acres
o
o
»
o,
CD W
UJo
LOCAL
STATIC
CONTROL
REGULATOR
.
32
O
LOCAL CONTROL
DYNAMIC REGULATOR
00
Jo
SUPERVISORY
CONTROL
AUTOMATIC
CONTROL
) 00
0 10
0.20 0 30 0
TOTflL RfllN. IN,
0 50
0.60
0 70
FIGURE 105
REGRESSION LINES FOR EACH CONTROL MODE
336
-------
these two extremes as more rainfall data is gathered and.additional
routines are added to the automatic control program.
Tabular Overflow Analyses
Standard statistical analyses merely repeated the obvious
conclusion that there is no good correlation between overflow
volumes and measured storm characteristics? therefore, a program
was written to segment the data into storm ranges and analyze
average overflows for each of these storm periods. Storm volumes
and the number of overflows during storms were analyzed in this
manner, and the program was made flexible enough to not only seg-
ment the data into ranges of total rain, maximum intensity or average
intensity, but also to separate the data according to its control
mode, storm pattern, or time period. A sample output table from
this program is included as Table M-l in Appendix M. Percent over-
flow reductions are computed for the different control modes versus
local control for each selected storm range; then summarized as
an overall percent reduction or a weighted percent reduction.
Weighting is performed by multiplying the percent reduction for a
given range by the number of storm events measured and averaged
for the higher level control mode.
Table 37 shows how the program reacts in computing percent
reductions by varying the rainfall range over which the computa-
tions are made. Note the influence of the weighting factors. In
most cases, the weighting slightly improves the percentage re-
duction because the greater number of observations are usually in
the lower rainfall ranges where percent reductions are slightly
better. This is not universally true as demonstrated by the storm
interval of 0.05 inch where, under combined automatic and super-
visory control, weighting decreases the performance from a minus
1.3 to a minus 48.4 percent reduction. The final column in Table
37 summarizes the percent reductions calculated under all rainfall
ranges. Since the range beginning at 0.1 inch is quite close to
that summary, this total rainfall range was selected for compari-
sons made in this section.
A negative percent reduction is equivalent to reporting a
percent increase. These increases are nearly all in the "number
of overflows" column, reflecting the gate control performed during
the advanced control modes. By selectively overflowing and shaving
off certain peak overflows at many regulators, the system signifi-
cantly decreases overflow volume to the receiving water. The former
local control procedure allowed a particular regulator gate to
remain open for a long period of time, recording only one overflow.
However, that overflow was very large in volume.
337
-------
co
J*
s
c4
j^i
o
EH
i
H
pj
* •
(-1
^
w
s
o
&
CM
O
CO
H
to
>H
1
iaj
*i
U
1
P4
CO
0)
I
c
•H
4J
O
P
•O
O
n<
o
rH
S
O
J
51
u
CJ
r
?
a
L
+
o
4-
tt
re
o
c
.
in
>
u
H
•4J
M
2
•P
3
fil
U)
**
>l
O
W
r|
i)
a
A
13
(1
4J
C
(U
J
d
a
rc
M
a
^
tJ
0
+j
c
(C
3-
rt
53
•^
r_
rH
;
o
.
i
3
L
irt
*^
rH
rH
Hi
O
O
0)
T>
3
0)
Cr>
RJ
M
)
CO -cf
CO CO
1
rH
O CO
rH rH
VO
|
VO CO
vo vo
en i-H
CO
VO CO
•«!• in
en rH
in o
r- r~
ro in
1
rH
0 rH
VO rH
CN •*
Overflow
Volume
Number
Overflows
in
o
0
vo o
r- 10
en
vo vo
in M1
vo
o r^
r* co
vo
P- fO
^* vo
en ,n
CO CN
o o
in CN
en
CO rH
•* in
co ro
1 .
Overflow
Volume
Number
Overflows
O
O
ro ro
co H
m
VO CN
in rH
in rH
en in
CO
rH -ST
VO .rH
en CN
r^ CN
VO CN
1
in
in «*
ro o
rH ro
l
Overflow
Volume
Number
Overflows
in
CN
*
o
en H
CN VO
CO
ro
vo ro
VO CN
in
CN CN
• •
CN 0
en CN
CN
ro en
^ ro
en
••d* in
CO O
VO rH
1
CO
O CN
r- H
rH ro
Overflow
Volume
Number
Overflows
m
o
CTV rH
*3* ro
t- rH
CN
co en
VO rH
1 1
1 f
1 1
rH
in in
in vo
en ro
»* CN
• •
*^)* rH
in rH
1
CN
en in
1
Overflow
Volume
Number
Overflows
o
rH
1
rH
O O
CO M"
t-> rH
1 1
rH
VO O
vo en
en ro
1 l
1 i
1 1
rH
ro o
en rH
in rH
Overflow
Volume
Number
Overflows
- — i
o
CN
f^ CO
t^ rH
1
m H
m r-
VO
co m
* •
m CN
en ro
ro H
in CN
en CN
OO CN
in in
in CN
1
ro CN
-3< CN
ro ro
1
Overflow
Volume
Number
Overflows
i
P (H
CO ftf
rH
2
rt
o
rj
'S
a
.(0
01 •
(UO)
CT>rl
SI
o
rHO
rH
nJO
tH-P
•rig
no
+J
MH 0)
o
0)
M £
0} O
3 W
(U
represents
le had at 1
s o
3 e
a
338
-------
Total system summer performance is compared with annual per-
formance in Table 38, In this table, the effect of limited data
becomes quite apparent. Intense, short duration storms on two
occasions during the automatic and supervisory control periods
were compared with a single long duration storm during the local
control period. This single comparison had a very negative effect
on the overall summertime performance. Since the annual performance
is considerably higher, supervisory and automatic control"appear
most effective in reducing overflows during the winter rainy season
when local control proved ineffective in minimizing total system
overflows. The evidence of some improvement in receiving water
conditions during summer months does indicate some question re-
mains regarding the limited summer overflow data. Further summer
data and comparisons are required to evaluate total system summer
performance.
The relationship between storm patterns and system performance
for various control modes was analyzed. The results are presented
in Table 39. The best automatic control performance was during
the localized type of storm, as expected. The least effective
performance was for uniform area-wide storms where fewer alter-
native control actions are possible. However, supervisory control
was more effective for uniform storms by reducing the large number
of local control overflows during this, the most common storm type
recorded during the grant period.
Analyzing total rainfall alone ignores the fact rainfall
duration plays an important role in total system reaction to a
storm. Table 40 indicates how the system reacts to various storm
characteristics. Lower overall performance for maximum rate is
discounted because maximum rain may fall for very long or short
durations. Therefore, maximum rainfall rates are not expected to
truly represent actual performance. The average rainfall rate
considers not only total rainfall, but also the duration over
which the rain fell. As shown by the table, both supervisory and
automatic control performance for average rainfall rates have been
outstanding when compared to local control. Since earlier reports
have relied upon total rainfall as the performance criteria, this
method is used later in this section to measure CATAD total system
performance for various control modes.
System Overflow Loading Analyses
The tabular analysis program was next applied to the eight
different loading parameters recorded during the course of the
grant study. Loading percentage reductions are presented in Table
41. Because of the limited amount of loading data available for
the automatic control mode, supervisory and automatic control periods
were combined and considered as advanced control measures. The over-
all loading reductions for advance control vary from 57.7 percent
reduction in ammonia loadinq to 75.7 percent reduction in COD loading?
339
-------
•s
ao
in
CN CN
fl tp fc
fd nJ Q)
•Cj M ^
> 0)
M id ^J
w ft
0) • 0)
-) G to
••-^
H CN
340
-------
CO
0)
rH
JO
id
EH
3 0)
in +J
•H fi
> tn
-1-rH
U 0)
VO
•
f-^
r^
3 ;
rH
+ rH
id
O Vi
JJ
rt!0
rH
•J 0)
0 -P
tr>
• -H
VI (U
o
H tH
-P rH
id id
0
3 O
<
id
o ro
O 0)
1-3 -P
. tn
M -H
> 0)
Vl —
O rH
M rH
•H id
> VI
Vi (U
0) >
D O
CO
Vl
0)
4J
•P
id
At
O
-p
M
VD
«
in
VD
o
r-!
en
^
•
^*
en
rH
oo
m
oo
•
^t*
ro
5
0
rH 0)
i| [ c2
Vl 3
0) rH
> 0
0 >
13
Q) G
O4-H
E^ £j
o
rH O
rH
*
o rq oj
• • •
vo r^** o^
in vo
1
VO CO rH
• . • •
«* in vo
VD rH
r*^ ^* ro
oo ro o
ro oo r-
1
ro in in
« • •
VO VO CM
rH 00 VO
1
VO O rH
VD rH OO
CM ro vo
1 '
rH VO f~
• • •
in CO rH
ro ro ro
1 1
Ul U)
S & &
O 0 O
Vl rH rH (U Vl rH
fll *U1 U 1 C fi\ tl 4
»Q Vi J-i 3 *Q J-i
g Q) (U rH g 0)
3 > > 0 3 >
S O O > 2 O
0)
•rs
•H
^*
gl
id
O (U
VH vi
-H ril
C
D
CTi CM
• •
nj* ^t1
rH CM
1
in r~
• *
CT\ VO
rH
1
CO CM
in »*
cy> CM
1
en r-
• •
in oo
en
I
ro cr>
VO ^1*
VO ^3*
1 '
^
• •
o^ o^
in in
I i
(0
^ &
O O
rH Q) Vl rH
0 3 >
0 > S3 0
tn
Vl C
O -H
rH
rH Q)
id >
-P id
C Vi
O EH
fa
in oo
* *
^ O"»
^3*
r- -*
* •
LO r*"*
in CM
CM ro
r- t~-
en CM
CM ro
• •
r*^- r~^
en CM
en oo
O rH
ro
i
^. ^.
• •
CM O
rH CM
1
tn
^ ^
O O
rH (U Vl rH
M-1 g 0 3 >
O > !3 O
•d
d)
Vl N
O -H
1 — (
•P id
(U U
X O
U iJ
0
341
-------
XI
3
13
O
C
•H
CJ
1-1
id
0
s
o
•H
4J
Id
B
O
0 S
H
1-1
u
3 O
a
H K
O
•rt
B e
a $
3
u rt
a
H
a, a
u
o
to
0
w
•H
a
a.
o
CO
T3
a)
+>
£
t
1
^
r-
n
s
D
a
c
g
^
j.
Q
O
T:
a
4-
t
a
rH
(G
a
o
a) id
tn'j»
id M
j-i a)
c
N
H 4J
d C *H
SJ H *H
£
•rt
II
10
u
•H
4J
to
•H
O -8
+> O
w id
M
id
JC
o
10
•
r-
*0
in
V0
o
r^
m
r-
•^r
CTl
00
•
o
tn
CO
m
1
rH O
^ i
O •-<
> 0
O >
0
V0
r-
cc
n
ro
10
rH
•
O
CN
1
rH
m
>
(U il-t
Is
so
*0
rH
O
O
01
H
0
o
rH
rH OT
rH flj 0)
•P C O
O-H C
HgH
•
in
o
vo
^p
rH
10
cn
^p
vo
en
•
o
n
I
r-
m
i
§
»« e
M 3
0)rH
o
fV
o\
)
r^-
o
r-
0%
rH
CO
10
tn
g
rH 0)
0 H
0
cr
CO
r-i
M
O*.
en
^p
ro
r-
to
•
o
rH
O
§
JH rH
i§!
!S O
H
O
O!
rH
O
rH
O
O
0) ?
&> d) o
to -p m
h n5 \
C
•5; H
342
-------
Table 41. REDUCTION IN LOADING PARAMETERS
Loading
Parameter
NH3-N
N03 N
P04
Settleable
Solids
Suspended
Solids
Volatile
Solids
BOD
COD
Avejrage of
All
Loadings
Percent Loading Reduction
Supervisory
Overall
15.5
59.6
35.8
39.0
36.5
42.0
28.8
55.5
39.1
f vs. Local Automatic vs. Local Super. & Auto.
, Weighted
0.0
38.2
25.8
10.0
9.7
23.2
28.3
42.6
22.2
Overall
63.4
67.6
72.9
78.5
70.6
Weighted
63.4
67.6
72.9
78.5
70.6
Overall
57.7
79.8
67.9
65.9
65.0
68.3
64.4
75.7
68.0
Weighted
50.0
69.1
62.9
51.7
52.0
59.2
64.2
69.4
59.8
343
-------
for an overa.ll average of 68,1 percent for all loading parameters,
The reasons for the different reduction percentages are -unclearf
but may be due to the overflow volume reducing performance of
individual regulators having high 'concentrations of particular
parameters,
The loading data has- a scattered effect much like the volume
data plotted in Figure 103. Loading trends, however, became more
obvious after applying a moving-average technique and plotting the
data with the computer driven digital plotter. Figures 106 and 107
show loading trends for the different parameters as related to in-
creasing total rainfall. The moving average technique clips off
peaks and shifts the data location so that these figures should
not be used to indicate any actual overflow data point, but rather
to indicate trends. Under local control, there appears to be a
lag in the buildup of loading discharged to 'the receiving water as
rainfall volume increases until approximately 0.2 inch of rainfall.
Then the buildup becomes steady and reaches a peak at about 0.5
inch. After the 0.5 inch peak is reached, some cases show a de-
cline, but others do not show an obvious trend.
These figures can be interpreted to mean that a low rainfall
intensity the pollutants analyzed are not appreciably dislodged '
from their source. Increasing rains then loosen these materials
and transport them through the system to interception and over-
flow points until, for a storm in excess of 0.5 inch, there is a
washed-out effect, when very little more material can be washed
from the storm drainage system into the more constant pollutant
quantities in the sanitary sewer sources of the combined sewer
system. Table 42 compares the average peaks on these moving-average
diagrams to give another indication of the improvement with advanced
controls. Percent reductions were not calculated because the tabular
data is considered more indicative of performance. However, Table
42 eliminates confusion caused by the differences in vertical scale
(Figures 106, 107) which the computer plotter automatically assigns.
344
-------
"•a
I Q HQ ff BQ Q
TOTRL RRIN. INCHES
•»' M t'.n «'.w O'.N I'.M r.i
T8TBL llfllN. INCHES
1
a
LGCflL CONTROL
FIGURE 106
LOADING TRENDS DURING LOCAL CONTROL
345
-------
•-S-
s
.M .
TOTflL BHIN. INCHES
•Via
OIfll'"
TOIfl'HIN.
Sit* t'.n
TOTAL (WIN
."'mcHES0'**
8
.*
« o.so o.u
TOTRL HfllN. INCHES
8.10 o.n
ft
*T9T(lL'15«IN.0"l"cHES
o-w .. ••« «•«
SUPERVISORY CONTROL
FIGURE 107
LOADING TRENDS DURING ADVANCED CONTROL
346
-------
Table 42. COMPARISON OF LOADING PEAKS
Parameter
Peak Value in Pounds
Local
Control
Advanced
Control
Settleable Solids
Chemical Oxygen
Demand
Phosphate
Biochemical Oxygen
Demand
Nitrate
Volatile Suspended
Solids
Ammonia
Suspended Solids
400
50,000
38
9,000
12
18,000
70
50,000
360
24,000
30
8,000
5
10,000
60
27,000
aAfter smoothing by moving average technique
347
-------
SECTION XI
COSTS AND BENEFITS
INTRODUCTION
The previous section discussed water quality improvements
resulting from CATAD automatic sewage collection system control.
Residual effects on the receiving water also result from the many
sewer improvements made between 1962 and 1971. These area sewer
improvements are itemized in Table 43.
Table 43. METRO COMBINED SEWER AREA IMPROVEMENTS
Areaa Total Cost (xlOOO)
Elliott Bay Interceptor
West Point Treatment Plant
West Point Outfall
North Interceptor
West Duwamish Interceptor
Alaskan Way Interceptor
$20,957
13,816
1,229
1,594
3,482
3,176
$44,2540
a "Interceptor" may include facilities such
as pump or regulator stations.
b Total cost is 37 percent of 1960-1970 construction plan.
The interim report (12) listed many water quality improve-
ments of the total construction program. Beginning in 1967, the
CATAD project added improvements costing approximately five per-
cent of the combined sewer work. In addition to the Metro work,
in 1967 the City of Seattle began a $78 million partial sewer
separation project which affected approximately 25 percent of the
combined sewers within the CATAD study area.
PROJECT COSTS
The CATAD project was largely stimulated by a $1.4 million
Federal demonstration grant from the Environmental Protection
Agency — 53.6 percent of the total project cost. A breakdown of
eligible costs for this study is shown in Table 44.
350
-------
Table 44. ELIGIBLE PROJECT COSTS
Pro j ect
Lake City Tunnel Regulator
Denny Way Regulator
CATAD
Regulator Controller Mods
Automatic Samplers
Interface at Remote Stations
Equipment
Report Productions
Contract Engineering Total
Legal
Subtotal
TOTAL
$ 143,700
502,000
864,800
36,500
40,500
145,700
20,400
$ 42,300
57,100
433,700*
4,300
4,500
293,700
1,800
20,000
$1,754,600 $857,300
* Includes $356,200 for engineering
and $77,500 for programming.
$ 185,900
559,100
1,298,500
41,800
45,000
439,400
22,200
20,000
$2,611,900
2,200
$1,754,600 $857,300 $2,614,100
Final project costs can be compared to the initial cost esti-
mates detailed in Table 43. The most important changes from the
initial estimates involved reallocating funds for regulators and
applying contingency and other monies for programming and engineer-
ing work.
Additional construction grant funds totaling $536,860 were
applied to regulator stations constructed or rebuilt during the
three-year span of this contract. Telemetry and control equip-
ment allowing the CATAD system to interface with remote stations
was considerably less expensive installed in new stations as com-
pared to modifying existing stations with the same equipment.
The incremental hardware cost for remote control on newly con-
structed or rebuilt stations is slightly over $13,000 versus
$25,000 to modify existing stations. Table 44 shows that engineer-
ing and programming costs for the CATAD project, at 32.8 percent,
are unusually high in comparison to many construction jobs. How-
ever, this was a research and development project, and such costs
are expected in developing new and unusual concepts. The table
also shows the high engineering cost for interfacing at remote
stations; nearly twice the actual contract cost. Metro assigned
consultants to research and record the existing controls at all
remote stations. Precise design specifications were complicated
by the basic control differences at each of the sixteen remote
sites. Identical control components and functions at each station
would have reduced interfacing costs. Similar interfacing equip-
ment designed into a new station would accrue a fraction of the
engineering costs reported above. ,
351
-------
Any agency considering construction of a CATAD-type total
management system must consider the cost of remote control regu-
lators at every major control point in the collection system.
Reduced overflow volume resulting from installation of dynamic
mechanically controlled gate regulators is shown in Figure 105.
Table 45 shows all costs associated with regulating stations,
telemetry system, and all control and computer equipment acquired
at the various consoles. On this basis, total project costs were
$5.2 million versus $2.6 million eligible project costs previously
listed.
Table 45. TOTAL SYSTEM COSTS3-
Project
Computer & Interfacing
West Duwamish Regulators
Hanford Street Regulator
Lander Street Regulator
Connecticut Street Regulator
King Street Regulator
Denny Way Regulator
Lake City Regulator^
Grand Total
Construction
Including
State Tax
$1,573
597
362
321
201
120
502
144
$3,820
Engineering
& Technical
Services
78
56
49
31
19
57
42
$1,443
Total
Cost
$2,684
675
418
370
232
139
559
186
$5,263
a Multiply all costs by 1000
k Includes $356,000 programming
c Added to project, not directly on main interceptor
The largest single CATAD system contract was for computer
hardware, telemetry and control equipment located at the central
station and at the remote stations. Table 46 itemizes these
hardware costs which a municipality may expect to pay for a
centralized computer control system. Adding computer monitor-
ing and control to each station in the CATAD system cost $15,000
for station modifications plus $10,000 for a telemetry control
unit. Improvements in electronic component miniaturization would
greatly reduce the purchase costs for similar equipment today.
352
-------
Table 46. HARDWARE COST DETAILS
Item
Process Control Computer
Air Conditioning
(Cooling tower system excluded)
Input/Output Terminal
Water Quality Interface
Peripheral Equipment
West Point Satellite Terminal
Telemetry Control Units
w/commanding - 28 @ 9,214 =
w/o commanding - 8 §9,170=
Central Operator's Console
Renton Satellite Interface
Spare Parts
Hardware Programming
TOTAL
Cost Cl'9'6'8. dollars)
$ 165,870
5,613
109.711
1,964
122,038
41,234
257,992
73,360
182,165
16,257
36,967
148,397
$1,161,568
Construction cost for a new regulator station with motor-
driven gates and local controls has averaged approximately
$200,000. Other cities have installed less complicated balloon-
type regulator gate structures at significantly lower cost. How-
ever, Metro's regulator structures are above ground, provide easy
access for maintenance and operation, and often enhance the sur-
rounding areas (see Figures 6 and 7).
MAINTENANCE AND OPERATING COSTS
Remote station and central station operating and maintenance
costs are shown in Table 47. Regulator station maintenance and
operation totaled $202,000 annually or $13,500 per regulator.
An unusual central station maintenance and operation cost item
is the $7,300 annual building rental cost. This cost item has
been reported in few, if any, other combined sewer overflow con-
trol project reports. As noted, capital amortization costs may
be added to maintenance and operation expenses for comparisons
between projects. CATAD system central control facilities alone
($2,614 million) would cost $225,000 per year at an amortization
figure of 6 percent for twenty years. Adding remote regulator
stations increases the amortization of $5.26 million to an annual
cost of $452,600.
353
-------
Table 47. OPERATING - MAINTENANCE COSTS
Cost Item
Remote Station
Operation
Maintenance
Subtotal/ Remote Costs
Central Station
Operation
Maintenance
Labor &
Fringes
$ 87,960
35,920
$123,880
$ 37,020
19,200
Utilities
$62,080
$62,080
$ 3,860
Other
$ 1,360
14,880
$16,240
$ 7,320a
Total
$151,400
50,800
$202,200
$ 48,200
19,200
Subtotal, Central Costs $ 56',220 $ 3,860 $ 7,320 $ 67,400
TOTAL, all phases $180,100 $65,940 $23^,560 $269,600b
a About 75 percent of this cost is building rental.
k Amortization of capital: $2.6 million @ 6 percent for
20 years would add $224,000/year.
With capital and operating costs now itemized, some compari-
sons can be made between the CATAD system and other alternative
solutions to the combined sewer overflow problem. The volume of
combined sewer overflow captured annually can be determined from
Figure 105 and compared to overflow amounts for the various con-
trol systems with average annual rainfall amount shown on the
figure. Known costs, drainage area and annual overflow volume
estimates provide the basis for Table 48.
Table 48. AVERAGE OVERFLOW VOLUMES FOR CONTROL MODES
MG per Storm x 150 Avg. Storms Annual Reduction
Avg. Overflow Year Percent
12'. 36
7.68
4.20
0.41
1850
1150
630
62
38
66
97
Control
Phase
Static Reg.
Dynamic Gate Reg.
Supervisory Cont.
Automatic Contro]
The same techniques were applied to compare other overflow
problem solutions.
354
-------
Table 49 shows that the CATAD project compares well with other
methods of combined sewer overflow control. Comparisons should be
viewed with some caution because most of the other, projects are
designed to handle essentially all overflow volumes. (Refer to
"Percent Overflow Captured" column.) The low per-acre costs of
centralized control projects in Minneapolis and Seattle would
undoubtedly be considerably greater if required to achieve the
capture percentage of the other listed solutions. The Minnea-
polis system seems least expensive; however, cost data is diffi-
cult to derive from the Minneapolis report (10). Minneapolis'
supervisory control system is expectedly less costly than Seattle's
automatic control system. Cost per thousand gallons of overflow
captured and treated is also less for the centralized control pro-
jects. High rate filtration and rotary screening have a very low
cost because of the very high estimated combined sewage capture
volume. For an area 1/50 of the CATAD project, rotary screening
estimates annual overflow capture equal to one third of the esti-
mated overflow amount from the entire 13/000 acres.
FEASIBILITY OF CENTRALIZED CONTROL
Municipalities considering centralized control systems should
realize that many research and development costs might not be
encountered because future projects can profit from the experi-
ence of Seattle and other cities which have adopted some level
of total systems management. In addition, costs of many of the
electronic components utilized in this system are rapidly drop'-
pingwhile labor, utility and other costs are increasing. A
detailed feasibility study must take all these factors into account,
Some comparisons can be made between the CATAD total systems
management concept and sewer separation as a solution to the com-
bined sewer overflow problem. In a combined sewer system, the
heavy load of solids and pollutants from street drainage, and
other materials settled into the sewer lines during low flow
periods are often washed from the trunk line to the interceptor
line for treatment until the combined flow exceeds sewer capa-
city. In a separate system, these materials are transmitted
directly from the storm drain to the receiving water. If con-
trol storage is sufficient, no storm drainage pollutants would
escape from the combined sewer system. The interim report (12)
estimated a 50 percent decrease in BOD loading for a partially
separated area versus a combined sewer area.
355
-------
O
CO
H
O
p
§
H
EH
O
CO
l^">
s
en
CD
H
•8
EH
rd
CD
Jj
O
•P
CQ
Q
O
> CD
O H
3
* -P
•P ft
o m
CM O
•
•P C w
W W rH
O ,c« fd
CJ EH O
O
O
O
rH
1
1
|
"%
o
o
*g*
,
1
1
I
I
i
Id
4J
•H
ft
rd
O
Cn
rd
CD
u
rtj
CO
vo
in
«.
CO
tH
1
1
£
A
O
•H
CD
"V^
•g
1
JgJ
O
•H
4J
-P
£*
CD
g
p
rd to
0) rH
t i __J
EH n3
pT|
*§ cd
CD
CD ft
tji pt
td-H
0 0
CO
^_^
O
CM
O
O
r— 1
1
1
1
1
|
I
O
0
vo
».
vo
o
0
o
CM
CO
CD
rd
0
-P
CO
rH
CD
fl 0
fi tn
EH O
•H
Q) CJ
CD
Q
,_
CO
•S
en
en
-a*
o
•
rH
,_,
CT»
H
VO
en
co
o
i_{
«,
rH
O
a\
en
co
4J
CD
1
(d
CD
£-1
K>1
^
fd
g
CM
4J
rH-H
fd O
O H
-H -P
g CD
0) Q
XJ
CJ
^^
en
3
1
I
I
CO
vo
•
CM
rH
^
co
CO
o
o
co
o
VD
rH
V
VD
in
rH
tQ
C
£H
CD
rd
Q
•P
CO
H
CD
•P
fd
^
M
CD
£
D
,_^
CN
CM
VO
CT*
VO
CO
•
in
CO
CO
en
CM
•H
o
CM
co
•»
CO
o
in
CM
£j
o
•H
-p
nj
H
•P
iH
•H
fe CD
rH
CD -P
-P -P
PH CD
1 CO
•H
W
^.
o
m
VO
en
CM
CM
•
CO
CO
CM
•tf
t~-
in
CO
rH
CNJ
o
in
CM
&1
p*
•H
a
0)
a)
U CD
CO rH
-p
M fd
fd CD
•P CQ
M
^^
H
in
in
O
CO
o
CO
[~-
CO
1
1
1
1
1
1
CO
in
0
o
0
*.,
o
CO
rH
O
-P
CJ
O
CJ CQ
•H
£ O
"rH Oi
jfj (Q
O CD
•P C
ft-H
to S
•H
Q
^
rH
^-'
CO
CO
CM
CM
in
en
en
•H
O
CM
rH
•b
co
rH
to
H
0
4J
H O
3 H
O-P
s
O CD
13 -P
•p
Q rd
rtj CD
EH CO
rj
i
i
CM
f^
in
in
H
CM
H
O
^J<
0
CM
rH
*,
co
H
to
^1
O
4J
(d
H
Fr»
f?
p3
*s^
g:
Q
5j
^
£J
1
1
•P
to
CD
CD
•P
•H
dP
VO
CQ
H
CO
CD
O
CM •
H
>i >i Q)
H H -P
Cl Q) flj
0 > &
O fi
CQ U H
C5 CD O
O H -P
CO CQ
•H rH
H fd O
(d 4-> -P
P| -r-1
g ft CD
O d m
H C!
(d n) to
-P CD
to to H
•P 0
to = a
O rtj tP
O= H
• • •
fd X> 0
356
-------
Table 47 was derived, from Figure 105 assuming an average storm
rate of 0.225 inch per storm and an average of 150 .storm events
per year. Being conservative and accepting temporarily only a
combined supervisory and automatic annual overflow estimate of
(630 + 62)/2 or 350, the table shows that supervisory control
replacing existing dynamic regulator gates provided an annual
overflow capture of 800 million gallons per year (1150 minus 350).
Since the entire 13,120 acres has an annual average overflow of
1150 million gallons in the dynamic regulator gate phase, an
equivalent separation area to achieve the 800 million gallons
removed by the CATAD system would require:
13,120 acres x 800 MG 7 1150 MG = 9110 acres
Next, a basic acreage cost for complete separation must be
established. A 1967 cost of $2,171 per acre for partial separa-
tion has previously been reported (54), for cities of 500,000
population or greater. Recent actual costs for partial separa-
tion in the Seattle area are $4,300 per acres. An escalation
factor can not be calculated:
$4,300 r $2,171 = 1.98
Applying this factor to the complete separation cost from
the same reference yields:
$13,568 x 1.98 = $26,874/acre
The estimate of $26,870 per acre for complete sewer separa-
tion is considered reasonable because it is well below the maxi-
mum reported value of $55,800 per acre (54) for partial and com-
plete separation.
Using these cost estimates of $4,300 and $26,870 per acre
for partial and complete separation, respectively, the equivalent
costs for either separation level is:
9110 acres x $4,300 = $39,173,000 (partial)
9110 acres x $26,870 = $244,786,000 (complete)
Obviously much less than $245 million, spent for automatic
model programming and additional storage tanks would provide an
improved total systems management project able to meet or exceed
any reasonable overflow reduction goal. CATAD-type centralized
control systems compare even more.favorably since sewer separa-
tion projects often result in overall loading reduction between
50 and 75 percent. CATAD cannot claim all overflow reductions
which have been observed in the three-year study. Partial separa-
tion of about one-fourth of the study area has also contributed
357
-------
to the near 80 percent overall reduction in combined sewer over-
flows. Metro will evaluate additional 'computer, optimization pro-
grams , more sewer separation/ storage tanks, and. other alternatives
which may be implemented to obtain even greater future efficiencies,
BENEFITS
An on-line computerized monitoring system has a very large
capacity for continuously collecting/ processing and storing valu-
able data about collection system facilities and the reaction of
this system or its components to storms and alternative control
strategies. Centralized monitoring of many facilities results
in a timely response to and analysis and correction of collection
system problems. Large investments in transportable test eguip-
.ment/ and related costs resulting from the dispersed activities
are eliminated.
Apart from the cost advantages already discussed, installing
a CATAD-type centralized control system realizes some benefits
which are intangible or difficult to quantify. Some of these
benefits are:
1. Routine data handling capacity
2. Time-shared services
3. Flexibility
4. Increased system capacity
5. Improved plant operation
6. Receiving water improvement
7. Public relations effect
A well designed control system often includes spare data
processing capacity which can be applied to solving common busi-
ness or scientific problems. The available computer time might
also be sold, if allowed, or traded with outside agencies or
businesses to reduce total central system operating costs. Many
service agencies, like Metro, perform several community functions.
There are obvious advantages in controlling more than one func-
tion with the same centralized system.
A computer system is easily modified to meet changing popu-
lation growth patterns, public attitudes or governmental policy.
This inherent flexibility is also valuable because the system
can respond to new discharge restrictions developed by govern-
mental agencies. Computer data files are readily available to
be scanned and analyzed to assist in design and control changes
needed to meet new regulations. The system can be converted to
optimize control of the modified equipment or strategies.
358
-------
The centralized computer system can increase Metro effective
dry weather capacity by relieving normal daily peaks in sewer
flows.. More 'uniform flow might stabilize some sensitive sewage
treatment processes such as biological secondary sewage treatment/
thereby improving overall plant efficiency.
Centralized system control provides a proven alternative to
sewer separation as a solution for combined sewer overflow pro-
blems. If still required, costs of a more expensive solution
such as sewer separation may be stretched out over a longer per-
iod of time, and possibly reduced in scale.
"Receiving water improvements accomplished by overflow reduc-
tion using a centralized control system may have both short and
long-terra effects. Higher recreational or commercial water use
can increase water body productivity. Greater fish catches or
improved tourist trade brought about by cleaner waters may repay
much of the investment in the computerized system.
Finally, there are benefits to be gained in public and govern-
mental relations from the control facility's appearance and per-
formance. The growing emphasis upon environmental impact and
community involvement make it very important for a water pollu-
tion control agency to consider this intangible benefit of a
computerized control facility.
SUMMARY
The CATAD system has met or exceeded nearly all its proposed
objectives. The going has sometimes been rough? however, the
assets gained and the performance of the system far outweigh
development difficulties. Other municipalities should seriously
consider centralized sewer system control as one of the most
reasonable solutions to combined sewer overflow problems. In
the Seattle area, a highly automated centralized control system,
aided by sewer separation of a portion of the drainage area, has
reduced combined overflows by nearly 80 percent. Other communi-
ties may find centralized control combined with other overflow
reduction techniques will provide performance records equal to
or greater than Seattle's. The information developed and avail-
able from the CATAD project and other control systems will help
water pollution control administrators decide on the type and
degree of control required for their communities.
359
-------
SECTION XII
REFERENCES
1.
3.
4.
7.
9.
10.
American Public Works Association, Public Works Computer
Applications, Special Report No. 38, Chicago, Illinois (1970).
McPherson, M. B.f Feasibility of the Metropolitan Water Intelli-
gence System Concept,Technical Memorandum No.15, American
Society of Civil Engineers, New York, New York (1971).
Tolle, W. A. "Management Information Systems," Journal of
American Waterworks Association, Volume 63, No. 11, Page
688-691 (1971).
Murdock, George B., A Systems Approach to Water Quality Manage-
ment , Preprint No. 22 for the National Symposium on Data and
Instrumentation for Water Quality Management, University of
Wisconsin, Madison, Wisconsin, July (1970).
American Society of Civil Engineers, Urban Water Resources
Research, First year report to the office of Water Resources
Research, U. S. Department of Interior, New York, New York,
pages 35-38/ Appendix G-J (1968).
Gordon, Geoffrey, System Simulation, Prentice-Hall, Inc.,
Englewood Cliffs/ New Jersey (1969) .
TEMPS Research, Environmental Management Information System,
a preliminary report to the River Basin Coordinating Committee,
Seattle, Washington (1973).
American Public Works Association, Combined Sewer Regulator
Overflow Facilities, final report on research project No. 68-1,
Contract 14-12-456, Federal Water Quality Administration, U. S.
Department of the Interior (1970).
Brown, John W. and Suhre, Darrel, Sewer Monitoring and Remote
Control - Detroit, Preprint 1035, American Society of Civil
Engineers environmental meeting, Chicago, Illinois (1969).
Minneapolis-St. Paul Sanitary District, Dispatching System for
Control of Combined Sewer Losses, Demonstration Grant No.
11020FAQ, Environmental Protection Agency Water Quality Office
(1971).
360
-------
11. Grigg, Neal S., Labadie, John, Smith, George L., Hill, Duane W.
and Bradford, Bruce, Metropolitan Water Intelligence Systems
Completion Report - Phase II, Grant Agreement No. 14-31-0001-
3685, U. S. Department of the Interior Office of Water Resources
Research (1973).
12. Municipality of Metropolitan Seattle, Maximizing Storage in
Combined SewerSystems, Project No. 11022ELK. Contract No.
13-Wash-l, Office of Research and Monitoring Environmental
Protection Agency (1971).
13. Federal Register, Volume 31, No. 94, May 14 (1966).
14. Municipality of Metropolitan Seattle, Post-Construction Studies
Evaluation, Report submitted to Federal Water Quality Adminis-
tration, Seattle, Washington (1967).
15. Gibbs, Charles V. and. Alexander, Stuart M., "CATAD System Con-
trol for Regulation of Combined Sewage Flows," Water .and Wastes
Engineering, Volume 6, No. 8, pages 46-49 (1969JT
16. Metropolitan Engineers, Sewage Disposal Project Contract No.
68-1 for Computer Augmented Treatment and Disposal System,
Municipality of Metropolitan Seattle, Seattle, Washington
(1968.).
17. "National Pollutant Discharge Elimination System," Federal
Register, Volume 38, No. 75, April (1973).
18. Condon, F. J. "Treatment of Urban Runoff," APWA Reporter,
Chicago, Illinois, March (1973).
19. Comptroller General of the United States, Need to .Control Dis-
dharges from Sewers Carrying Both Sewage and Storm Runof£~,
Report No. B-166506, a report to the Congress, March 28D-973).
20. Liebenow, Wilbur R., and Bieging, James K. , Storage and Treat-
ment of Combined Sewer Overflows, Report to the Office of
Research and Monitoring Environmental Protection Agency Project
No. 11023FIY, Contract EPA-R2-72-070 (1972).
21. Ministry of Housing and Local Government, Technical Committee
of Storm Overflows and the Disposal of Storm: Sewage, Her
Majesty's Stationery Office (1970).
22. Carl R. Rohrer Associates, Inc., Underwater Storage of Combined
Sewer Overflows, report for the Environmental Protection Agency
Program No. 11022ECD, Contract No. 14-12-143 (1971).
361
-------
23. Nelpar/ an American-Standard Co., Combined Sewer Temporary
Underwater Storage Facility, for the Federal Water Quality
Administration Department of the Interior/ Program No. 11022DPP,
Contract No. 14-12-133 (1970).
24. Gibbs, Charles V., Alexander, S. M. , and Leiser, C. ..P.,
"System for Regulation of -Combined Sewage Flows," Journal
of the Sanitary Engineering Division, American Society of
Civil Engineers, Volume 98, No. SA6, December (1972).
25. Eagleson, P. S., "Unit Hydrograph Characteristics for Sewered
Areas," Journal of the Hydraulics Division, American Society
of Civil Engineers, Volume 88, No. HY2, procedure paper 3069,
pages 1-25, March (1962).
26. Logger, John A., Pyatt, Edwin E. , and Shubinski, Robert P.,
Stormwater Management Model, Volume' I-IV, Environmental Pro-
tection Agency Report No. 14-12-501, Reference No. 11024DOC
July (1971).
27. Chow, Ven Te, Open Channel Hydraulics, McGraw-Hill, New York,
New York, pages 265-268 (1959).
28. The Boeing Company, Electromagnetic Compatibility Study, Report
No. MET216, Task Order No. 1, Seattle, Washington, June (1973).
29. Schlicke, H. N. and Struger, O. J., "Getting Noise Immunity
and Industrial Controls," IEEE Spectrum, Volume 10, No. 6,
page 3, June (1973).
30. Boaen, V., "Design Logic Circuits for High Noise Immunity,"
IEEE Spectrum, Volume 10, No. 1, pages 53-59, January (1972).
31. Allen, C. M. and Taylor, E. A., "Salt Velocity Method of Water
Measurement," American Society of Mechanical Engineers Journal,
Volume 46, No. 1, pages 13-51 (1924).
32. Tucker, L. S. and Hill, D. H., Social and Political Feasibility
of Automated Urban Sewer Systems, Technical Report No. 5 from
the Metropolitan Water Intelligence System Project, Colorado
State University, Fort Collins, Colorado (1972).
33. Metropolitan Engineers, Predesign Report on Second Stage Con-
struction of Comprehensive Sewerage Plan, for the Municipality
of Metropolitan Seattle, Seattle, Washington (1970).
34. Standard Methods for the' Examination 'of Water and' Waste Water -
13th Edition, American Public Health Association, New York(1971)
362
-------
35. Strickland, J. D. H., and Parsons,. T. R., A Manual of Sea
Water Analysis, Bulletin No. 125, page 61, 1st Edition,
Fisheries Research Board of Canada (1960).
36. Strickland, J. D, H., and Parsons, T. R., A Practical Handbook
of Seawater Analysis (1968).
37. King, H. W. , Handbook of Hydraulics, 4th Edition, McGraw-Hill
Book Co., Inc. (1954).
38. Zadorojny, C., Saxton, S., Finger, R., "Spectrophotqmetric
Determination of Ammonia," Journal Water Pollution Control
Federation, Volume 45, No. 5, pages 905-912, Washington, D.C.,
May (1973) .
39. Farris, G. D., Dalseg, R., and Machno, P. S., Freeway Runoff
from the 1-90 Corridor, Municipality of Metropolitan Seattle,
Seattle, Washington (1973).
40. "Soxhlet Extraction Method," Method 209A, page 409, Standard
Methods for the Examination of Water and Wastewater, 13th
Edition.American Public Health Association, New York (1971).
41. Gibbs, C. V., "Receiving Water Monitoring," Water Works and
Wastes Engineering, Volume 1, No. 9, pages 52-55(1964).
42. Gibbs, C. V., and Isaac, G. W., "Seattle Metro's Duwamish
Estuary Water Quality Program," Journal Water Pollution Con-
trol Federation, Volume 40, pages 385-394 (1968).
43. U. S. Department of Interior - Federal Water Pollution Control
Administration, Problems of Combined Sewer Facilities and Over-
flows - 1967, Water Pollution Control Research Series No.
WP-20-11, Washington, D. C., December (1967).
44. Department of Ecology, Implementation of and Enforcement Plans
for Water Quality Regulations, State of Washington, Olympia,
Washington, pages 7-10, September (1970).
45. Water Quality Criteria, Report of the National Technical Advis-
ory Committee to the Secretary of the Interior, Federal Water
Pollution Control Administration, Washington, D. C., April
(1968).
46. McKee, J. E. and Wolff, H. W./ Water Quality Criteria, State
Water Quality Control Resources Agency of California, Second
Edition, publication No. 3-A, Sacramento, California (1963).
363
-------
47.
48
49
50.
51.
American Public Works Association, Problems of Combined Sewer
Facilities and Overflows - 1967, U. S. Department of Interior
Federal Water Pollution Control Administration, Pollution Con-
trol Research Series No. WP-20-11, Washington, D. C., December
(1967).
Kielzer, Victor A., Bauer, William T. , and Dalton, Frank E.,
"The Chicago Area Deep Tunnel Project," •Journal of the Water
Pollution Control Federation, Volume 41, No. 3, Pages 515-534,
April (1969).
Dow Chemical Company, Chemical Treatment 'o'f Combined Sewer
Overflows, Water Quality Office of the Environmental Protec-
tion Agency, Project No. 11023FDB, Washington, D. C., Sept-
ember (1970).
Nebolsine, Ross, Harvey, Patrick J. and Fan, Chi-Yuan,' Hi-Rate
Filtration of Combined Sewer Overflows, Office of Research
and Monitoring, Environmental Protection Agency, Project No.
11023EYI, Washington, D. C. (1972).
Cornell, Rowland, Hayes and Merryfield, Rotary Laboratory Fine
Screening of Combined Sewer Overflows, Federal Water Quality
Administration, Project No. 11023FDD, Washington, D. C., March
(1970).
364
-------
SECTION XIII
LIST OF INVENTIONS AND PUBLICATIONS
Gibbs, Charles V. and Alexander, Stuart M. , "CATAD System Control
for Regulation of Combined Sewage Flows," Water and Wastes Engineer-
ing/ Volume 6, No. 8, pages 46-49 (1969).
Municipality of Metropolitan Seattle, Maximizing Storage in Com-
bined Sewer Systems/ Project No. 11022ELK.Contract No. 13-Wash-l,
Office of Research and Monitoring Environmental Protection Agency
(1971).
Gibbs, Charles V., Alexander, S. M./ and Leiser, C. P./ "System for
Regulation of Combined Sewage Flows/" Journal of the Sanitary
Engineering Division/ American Society of Civil Engineers, Volume
98, No. SA6, December (1972).
Mallory, T. W. and Leiser, C. P., Control- of Combined Sewer Over-
flow Events, A paper prepared for the 1973 International Public
Works and Equipment Congress, Denver, Colorado, September 15-20
(1973).
365
-------
SECTION XIV
GLOSSARY OF TERMS AND ABBREVIATIONS
access time - time required by computer to locate a word in core
memory and transfer the word to a register.
ACU - Auxiliary Control Unit: cabinet containing electronics,
relays, etc. located between TCU and station local control system.
address - name or number which identifies a particular storage
location in the computer.
analog - the representation of a numerical quantity by some physi-
cal variable; e.g., translation, rotation, voltage or resistance.
as-built - engineering drawing completed after a facility is con-
structed, shows original design plus revisions made during con-
struction.
ASCII code - American Standard Code for Information Interchange;
a special eight channel paper tape code developed to facilitate
data transmission between machines manufactured by different
companies.
autoanalyzer - copywritten term referring to equipment which auto-
mates chemical tests on samples. After initial setup of equipment
and sampling unit, no further human effort is needed other than
interpreting a strip chart.
background - a low priority, unprotected processing area in a com-
puter where batches of programs are compiled, tested and run with-
out affecting other protected control and processing areas.
BCD - Binary Coded Decimal, a computer coding system.
benthal sediments - deposits on the bottom of a stream, often
containing decaying material which removes oxygen from the stream.
binary - the representation of a numerical quantity by use of the
two digits 0 and 1.
bit - abbreviation of binary digit;
computer storage.
smallest possible unit of
BOD - biochemical oxygen demand, a standard test used in assessing
wastewater strength. The, quantity of oxygen used in the biologi-
cal-chemical oxidation of organic- matter in a specified time under
standard conditions.
366
-------
buffer - a device which compensates for speed differences between
two machines, permitting them to operate together.
bug - a mistake or malfunction in the. design of a program or com-
puter.
byte - a sequence of adjacent Binary Digits operated as a unit.
CATAD - Computer Augmented Treatment and Disposal system
CCU - CATAD control unit: cabinet same as ACU but located at
pumping stations.
cfs - cubic feet per second, unit of quantity of liquid flow.
checkpoint - a point in time where computer processing is stopped,
all machine variables, registers and background area copied to a
rapid magnetic storage device so that a large foreground program
can temporarily use all or a part of what is normally called back-
ground area.
COD - chemical, oxygen demand, a standard rapid test qf the strength
of~~a wastewater. The COD determination provides a measure of the
oxygen equivalent of that portion of the organic matter in the
sample that is susceptible to oxidation by a strong chemical oxident.
core - computer storage area where binary data is represented by
the direction of the magnetic field in each unit of an array of
tiny donut-shaped rings (generally).
Coliform - bacteria found only in intestines of mammals, there-
fore used as an indication of the pollution level of a sample or
body of water.
composite - placing samples from a continuously changing stream
into one common storage point to test for an average value.
CPU - central processing unit: that portion of computer excluding
input-output and external storage units, where arithmetic, logical,
storage and control functions are centered.
CRT - Cathode Ray Tube: television, for visual data presentation.
DC - direct current, a form of electrical power.
debug - the process of eliminating a mistake or malfunction in the
design of a program or computer.
demersal fish - the resident fish species for a particular water
body.
367
-------
DO - dissolved oxygen, the amount of gaseous oxygen dissolved
into a liquid sample.
digital - the use of discrete numbers to a given base to represent
all ^ quantities in a problem or calculation. Most often all infor-
mation is stored, transmitted or processed by a dual state condi-
tion; e.g., on-off, open-closed, true-false.
- dry weather flow: normal sewer flow from domestic and indus-
trial sources only.
dynamic regulator - a regulator which automatically makes adjust-
ments in control settings by responding to water levels in the com-
bined sewer or interceptor.
executive - a program, often supplied with a computer, which con-
trols loading and relocation of all software (much unknown to the
programmer) required to execute a job entered by a programmer.
fecal coliform, fecal streptococcus - bacteria associated with
humans, refinements to coliform tests to eliminate animal sources
when assessing the pollution level of a water sample.
first flush - heavy load of material, previously settled in sewers,
which is washed along by the initial flow resulting from a storm.
fps - feet per second, a measure of velocity.
force main - a pressure pipe joining the pump outlet at a wastewater
pumping station with a point of gravity flow.
FORTRAN - FORmula TRANslation; a programming system developed by
IBM which converts mathematical statements into computer language.
gpad - gallons per acre per day, a measure of infiltration or leak-
age between a pipeline and its surroundings.
hard copy - a printed copy of machine output.
hardware - the physical equipment and devices which comprise a com-
puter or computer system component.
head - distance in feet from a free liquid surface to some reference
point (which may be a different liquid surface)'.
hydrograph - a graphical representation of liquid flow versus time
with time as the horizontal axis.
interceptor - sewer which receives water from various traverse
sewers and carries water to a point for treatment and disposal.
368
-------
interface - a common boundary between parts of a computer system.
interrupt - a special signal which temporarily halts the normal
operation of some computer job for the purpose of accomplishing
another more important short task, after which normal operation
resumes.
loading - the dry weight, in pounds, of some material that is being
added to a process or disposed of to a receiving water.
logic - the science of combining electronic components in order to
define the interactions of signals in an automatic data processing
system.
mathematical model - the characterization of a process or concept
in terms of mathematics, which allows the simple manipulation of
variables in an equation to determine how the process would act in
different situations.
mg/1 - milligrams per liter, or the concentration of some chemical
in a liquid. If a letter appears after "mg" it represents the
chemical symbol; e.g., N for nitrogen, P for phosphorous.
mgd - millions gallons per day, a common term for quantity of
wastewater flow.
modem - device which converts between computer recognized signals
and tones transmitted over telemetry lines.
NH4, NC>3 - chemical shorthand for ammonia and nitrate in solution.
nutrient - is a major ingredient of a water or wastewater sample;
e.g., ammonia nitrogen, nitrate-nitrite nitrogen, phosphate-
phosphorous, or silica which may serve as growth media for micro-
organisms, phytoplankton or zooplankton.
off-line - system and equipment under human operator control, not
CPU control.
OGC - outfall gate controller: cabinet containing controllers and
recorders relating to the outfall gate at a mechanical regulator
station.
on-line - system and equipment under continuous automatic CPU con-
trol.
orifice plate regulator - a machined hole in a metal plate (usually
horizontal) allows a flow only slightly in excess of trunk dry
weather flow to enter the interceptor.
369
-------
overflow - undesirable emergency relief of sewer system by direct
transmission to receiving water generally without treatment.
outfall - pipeline which carries raw or treated sewage from the
collection system or treatment plant to the receiving body of water.
parity - a term relating to whether a given word or character (group
of digits) is even or odd. A single bit denoting the parity of a
word is usually attached to that word so that quick tests can be
made as to the validity of the data.
partial separation - removal of some portion of all the elements
of storm drainage into a combined sewer; e.g./ streets and parking
areas only, leaving roof and foundation drainage to enter the com-
bined sewer.
peripheral - specialized machines connected electrically to the com-
puter for converting between binary and other data forms; e.g.,
cards, tapes, typed pages.
pH - a measure of the degree of acidity or alkalinity of a solution.
PO4 - chemical shorthand for phosphate in solution.
priority - degree of importance assigned to some computer task.
psig - pound per square inch, gage: a measure of the pressure of
a fluid or gas based on atmospheric pressure.
purging - the act of reversing flow or using high pressure liquid
or gas to clear a pipeline of a plug of solids.
RAS - Random Access Storage: a technique in which the computer can
find one bit of data as quickly as any other, regardless of its
specific location in storage.
rational method - a means of computing storm drainage flow rates by
use of the formula A=ciA; where £ is a coefficient describing the
physical drainage area, _i is the rainfall intensity and A is the
area.
real-time control - control of a system by using computers and
timing such that the speed of response to the input information is
fast enough to effectively influence the performance of that system.
register - device consisting of miniature electronic components
including transistors, where a specific number of bits are stored
and operated upon.
370
-------
regulator - structure which, controls amount of sewage entering an
interceptor by storing in a trunk line or diverting some portion
of flow to an outfall.
retroreflective - technique of sensing periodic light reflections
from a constant light source bouncing off a mirror attached to a
rotating device.
routine - a sequence of instructions which perform a specific func-
tion within a larger program.
routing - storing, regulating, diverting or otherwise controlling
the peak flows of wastewater through a collection system according
to some prearranged plan.
rule curve. - a curve which relates storage and time for a given
reservoir under different control conditions.
sag curve - a curve which describes the gradual drop to some mini-
mum value followed by an increase to normal levels of the dissolved
oxygen in a flowing stream following the addition of some oxygen
demanding material such as sewage.
salt wedge - a wedge shaped volume of salt water beneath a body of
fresh water where a river and ocean meet in an estuary area.
secchi disk - a disk, painted in four quadrants of alternating
black and white, which is lowered by rope into a body of water.
The measured depth at which the disk is no longer visible from
the surface is a subjective measure of relative transparency.
sequential - samples taken at known time intervals and stored in
separate labeled containers.
side weir - a regulator which is essentially a long slot cut into
the side of a sewer. Normal dry weather flow continues through the
sewer while the increased depth during a storm will allow excessive
flows to exit through the slot to some alternate point as an over-
flow.
scan - the collection and storage of data from all points at all
stations in system by computer.
scan time - time set by operator which establishes the interval *
from the beginning of one scan to the next.
simulation - representation of physical systems and phenomena by
computers, models and other equipment.
371
-------
siphon - a u-shaped pipe used to carry wastewater under some obstacle
such as a stream or another, pipe.
sluice gate - a vertically sliding gate of any shape used to control
or shut off flow in a sewer or other channel.
software - the programs or instructions which often control the
hardware to perform some computer operation or extend the capa-
bilities of the system.
static regulator - a regulator which is either fixed or can be
adjusted only by manual actions of a human operator.
storage - unfilled/ enclosed volumes within or connected to a sewer
system capable of accepting and retaining wastewater for a period
of time.
subroutine - a compact set of instructions to perform some repeti-
tive task quickly and return to a main program.
table - a collection of data in a form suitable for ready reference
by a computer.
telemetry - data transmission over long distances via telephone or
telegraph lines by electro-magnetic means.
TCU - Telemetry Control Unit: interchangeable electronic cabinets,
convert between telemetry and station control signals (from ACU or
CCD) .
time sharing - use of a computer or device for two or more purposes
during the same overall time interval, done by interspersing com-
ponent actions in time.
total systems management - as applied to combined sewer systems,
means using both the transporting and retention capacity of a
sewer system to limit overflows.
trunk - large sewer which receives wastewater from tributary branch
sewers serving generally one drainage area.
trawl - a fish catching technique using nets pulled along a river
bottom (net sometimes also reaches to the surface) by boat.
up-dating - the art of bringing information up to the correct time
or value.
wastewater - the spent water of a community, including domestic,
industrial and commercial water-carried solids plus storm or other
water sources. Replacing the term sewage.
372
-------
word - a set of 16 or more bits stored and transferred as a unit
by the computer.
write - to transfer information, usually from main storage to an
output device.
373
-------
SECTION XV
APPENDICES
A. Remote Station Data 376
Table A-l: Regulator Station Construction Costs 377
Table A-2: Pumping Station Construction Costs 379
Table A-3: Basin Area - Hydrograph Data 381
B. Data Monitoring at Stations 382
Table B-l: Data Point or Contact Monitored 383
Table B-2: CATAD Interface Function Definitions 388
Table B-3: Classification of Remote Terminals 395
C. Station Equipment Calibration Data 396
D. Executive System Priority Assignments 413
E. Force Main Calibration Curves 419
Figure E-l: Kenmore Pump Station, Pump #3 420
Figure E-2: Kenmore Pump Station, Combined 421
Figure E-3: Matthews Park Pump Station 422
Figure E-4: Interbay Pump Station, East 423
Figure E-5: Interbay Pump Station, Combined 424
Figure E-6: Interbay Pump Station, West 425
F. Regulator Sampler Experience 426
G. Control System Description 432
Table G-l: Regulator Station Control Schematic 434
Table G-2: Pumping Station Control Schematic 436
H. Pump Station Force Main Calibrations 439
Table H-l: Force Main Calibration Equipment 440
Figure H-l: Force Main Calibration Setup 442
Figure H-2: Salinity-Time Graph 443
Figure H-3: Kenmore Force Main Calibration 444
Table H-2: Station Discharge Equations 446
Table H-3: Darcy Equation Summary 447
Figure H-4: Darcy "F" Values 448
Figure H-5: Force Main Pressure Transmitter Installation 449
Table H-4: Force Main Calibration Equipment: 450
Components and Characteristics .
Figure H-6: Typical Force Main Calibration Mask 451
I. Dexter Regulator Simulations 452
Figure 1-1: Flow Chart of Dexter Avenue Regulator 453
Simulation Model
J. Inventory Program Sample Output 456
Table J-l: Computer Division Manuals Inventory 457
Table J-2: Computer Division Spare Parts Inventory . . . 458
374
-------
K. Rainfall Studies Program Sample Output 459
L. Urban Storm Drainage 464
Table L-l: Physical-Chemical and Oils Analyses 465
Table L-2: Nutrient .and Bacterial Analyses . 466
Table 1-3: Heavy Metals and Solids Analyses . 467
M. System Performance Program 468
Table M-l: Rainfall vs. Overflow 469
Table M-2: Rainfall vs. Loading 472
375
-------
APPENDIX A
REMOTE STATION DATA
Table A-l: Regulator Station Construction Costs
Table A-2: Pumping Station Construction Costs
Table A-3: Basin Area - Hydrograph Data
376
-------
id
EH
H
-P a
o g
id o
n u
-P
O M
U >H
C M
O Vl
-H id
-P H
O rH
S fl
-p 1
V) -P
c: 01
0 0
u o
0) -P
O Q)
id d)
CO
o
M i-l
o id
O 3
rH tT
h CO
k -P
-P-rH
0 3
O VI
rH -H
o) u
EH
^
id
•P
4J
o
•H
rH
rH
a
CO
o
CM
CM
^^
CO
C
O
-d
§
n
t- rH
vo r-
cn
rfl rtj
M X«
HI +J
0) -H
Q H
00 rH
vo r»
o^ o^
o o
o o
o o
O CO
rH rf
rH CO
O O
en oo
rH ^3*
VO VO
CO "3*
rH rH
+J -P
4J -P
O O
•H -H
rH H
rH rH
H W
•o -d
O 0
in in
c a
s x
o o
0 CO
in in
^^
CM
•~*
rH CM
=»= =»=
'd *d
M k
o o
IM M-l
C C
id id
K n
P* fH
VO ["*
en en
rH rH
0 0
o o
•q. o
in i>
f~ rH
rH
0 0
O CM
CM CM
rH VO
-* r-
H
^
id
x:
w -P
•H -P
6 O
id *H
S rH
3 rH
Q W
& 23
o o
O 00
•» in
4J
0)
o
4J
M CO
O
M C
id -H
ffi W
en rH
vo P-.
en en
rH rH
0 0
o o
T O
p- in
CO O
rH CO
0 0
CO O
r-~ oo
•3- VO
m m
CM rH
^i
id
t^ 03
3 -P
M 1 *
EH O
•H
Xi rH
•P rH
OB
•
K a
o o
rH *4*
r» in
.*^
CM
•~^
rH
0)
C
C
3
EH
£*i
.p
•H Vl
O 0
•d
^ id
CM CO
P~ p^
en en
ffl Q
0 0
o o
t~- r~
vo en
co co
rH rH
0
CM
CM
CM
CTt P*
£*1 ^1
id id
« ffl
4J -P
+J -P
0 0
-H -H
rH rH
rH rH
P3 H
CO CO
O 0
•^ VD
CM CM
*-^ *— *
CO CM
•—* *~^
C
« M
CT> rH
•H o
x: t-i
O VI
•H O
s a
p>
vo
en
rH
O
0
I~
VO
in
o
co
rH
rH
CO
X!
Ul
•H
g
id
3
Q
ES
0
in
^i
C
id
tr>
•H
X!
O
•H
a
•
s
01
C
O
•H
-P
id
o
o
(U
rH
O.
•H
rH
3
S
•P
id
Q)
^l
id
01
4J
C
(U
c
o
a.
g
o
o
c
o
•rH
V
id
4J
0)
VM
•rH
*
01
-P
•rH
C
3
.
•O H
3 C
rH 0)
0 >
a (D
•H
•d
01 a
0) 4J
o i
<4-l XJ
0)
•d a)
3 c
rH O
0
G -P
•rH Hj
-P 0}
O G
C O
•rH
01 -P
CD id
O 4->
rd 01
•P X!
01 4->
o o
CJ (H
• •
XI 0
377
-------
•O
o
U
01
I
g
co
XI
id
0)
0) -H
> G
•H Id
O Cj
a o
0)
s
ui
ui
id
a
D CQ
Ul
0)
O rH
c id
LJ tj I
I -P
0) g
N O
•H
CO
0) M
4J O
id -P
o id
0)
VH
co
C
ui id
o IH
o
H
I tH
0) -P
•P a
a) a)
id n
•rl Q)
Q -P
Mrl
0)
CO rH
rH
tr> n3
C >M
•rl 4J
4J 3
O O
d)
£ b
o o
•r)
rH
3
rl O O
>i O O
A
•w ,
•P -H
i-H i-H
3 3
M 0 M O O O
>i O >i O O O
A £>
•H •»
rH t-l
3 3
H M O 0
>i >i Q Q
A
o
vo
n
oooo
OOCTIITI
CO "3<
rH 00
^
CS
co «* o vo vo o
••* CM m CT> M CM
o
vo
co vo- i id o>
O I I M
O
in
rH
co
0) 0)
* G
1)
0)
M
4J
V
3
>i
-P
(d
CN
O
X CO CN
t-
X
G
•rl
G
0)
a
O
H_j
0
o
•H
Ul
G
01
1
Q
O
•H
A
0)
•P
tn
T3
•H
0)
XI
Cn
G
•rl
d
(U
O
o
Ul
(U
m
(U
rl
0)
N
•rl
W
•
^|
id
rH
3
I
O
0)
)H
rH
rH
id
2*1
rH
id
3
4J
O
id
0)
id
Ul
4J
id
O
.
Ul
fe
O
fH
4-1
C
•H
3
04
a
0
U
M
o
-------
CM
4
(1)
^^
X!
>d
EH
V)
z
o
H
1
co
o
z
M
D
PJ
rH
-P a
o e
a c
k u
1 1
G •
0 rl
O ><
C 01
O IH
••H ro
•P rH
O H
3 0
•P I
01 -P
C Bl
O O
o u
0) -P
U fl)
id
P! PL.
CO
0)
14 V-1
o id
O 3
H O
Cu CO
£"l
1 1 1 1
4J-H
a> 3
£ O
0) M
H -rl
g MH
0 (U
o as
c
o
•H
-P
Primary
legment Loca
W
^1
•p
•H .
•-H O
•r^ J3
r\
9
fa
i
01 -P
H -H
CO
IH IH
(U
M -H
Bellevue-Me
Montlake Un
O O
vo o
rH rH
0)
0
id
rH
PH
3 IH
> -H
01 O
i-l >
rH rH
CD 0)
CQ CO
en 1
4->
•rl
(0
»H
>i >i
S -G
4-1 o
t3
01 G
M id
O rH
G VH
(U -H
X X
r^ o
VO t~
en en
o o
o o
o en
ro CM
on rH
*•
ro
O 0
0 0
o r-
•• CM
o
CM
•w in
VO CM
CM ro
.
D]
H
J^
0)
o
North Trunk
Bellevue-Me:
O O
CM i-l
r-- CM
T3
G
id
^ rH
IH 01
id H
u 01
£ 0
(U H
SI 01
•p a
rd •
a 3
CM in
r- vo
en o^
O
o o
0 0
in o
r- o
H 0
f~ in
*
ro
o o
o o
H 0
•-I •>
O
ro
CM in
CM en
i-H CM
01
t) 4->
IH G
O id
IH rH
Rainier.-Han
Treatment P
o o
"* ro
•H CM
(1)
3
§£LI
•
> EH
W
rl
a) c
•H O
G 4J
•H C!
R) 0)
in in
vo vo
on en
o o
o o
00 CM
on vo
f- rH
rH CM
O O
^- en
H
in in
ro i-H
ro ro
t t
01 01
H M
^1 ^|
(1) 0)
O 0
Bellevue-Me
Bellevue-Me:
0 O
CM in
CM rH
73
G
id
rH
01
M C
i
0)
W CO
rH f~
r~ vo
on en
id
o o
o o
o o
in CM
oo vo
CO CM
o b
o oo
H 0
CM CM
co 1-1
"31 CM
CM '
>1
4J
•H
01
0)
•rl
Montlake Un
W. Duwamish
O 0
rH "3"
rH -31
•
m
23 ^i
« id
0) S
< rH
id
5 -H
•H M
4-> id
•H
EH IS
in vo
vo vo
on on
rH rH
O
O O
O O
cn o
^ji o
in o
H O
^p
0 0
•* o
on oo
in
CO
^* ro
in oo
CO CM
0)
td 4->
«-( C
4J rd
01 rH
Issaguah Ea
Treatment P!
0 o
-a1 vo
rH VO
i-H
•
CU
EH
CO
4J
a G
o -n
4J 0
3 &
XI 4^
rH 01
•rl 0)
SB S
co
1
rH
cn
H
T3
0)
rH
0)
73
o
g
0)
n
CD
o
4-1 •
4J
• 01 C
01 (3 Id
- O rH
0-rl &
*»• 4J
en id 4J
rH 4J C
01 0)
c e
•H *43
•O id
•a a> o)
0) rH rl
4JrH 4J
o id
3 4-i U
H 01 CJi
4J c id
01 -rl S
C 0)
O 4J W •
O 0) 0)
>i G i-H
01 -H XI
(1> 4J J3 Id
•H O 4J rH
4-> C -rl -rl
-rl S Id
rH 0} >
•H ft 01 Id
u G a)
Id 3-rl 4J
O 0)-rl W
G O tn
a id G
r-J Ml t] 1 .^J
•-I O S
id tr* id
T3 rH CU
o) id g **
-P > 3 Q)
id a u
6 w id
•rl 3 UH &
4J rH O 01
01 a
0) 0) Vl
4J 3 O
01 01 rH O
0) O Id rH
T3 0 > 4J 4J
•*H nj nJ fl3
e e e
4-> -H -H -H
W 4-1 4J 4J
O W 01 01
CJ H W H
• • • •
Jd £1 O H3
379
-------
7
C
o
u
»•••» CO
o
{N H
1 EH
« g
fl) W
E-4 H
1
n
0)
ui u
in C
a H
i«4 •
CQ id
•H
Q
Ou-P
e m
3 0)
•a «.
0) t!
rd m
PS K
a
E
1
•P 4J
•H id
O £
Id'rt
Id rH
O D
Cn
C
•rt rH
g -H
3 -P
C
rH H
id
-P
O
EH
fj
Q) 0)
rH 4J
rH Id
id E
4J-H
(0 4->
C H
H 3
(fl
D)
« »r
irH
a
it
c
•r
4->
n
4J
CO
vo
*3* VO
CN rO C2J
CN
o o in
o ro r~
en CN H
00 O 0
ro in o
rH rH LO
rH
CO O O
ro in o
rH rH in
rH
ro ro ro
ro ro ro
0)
o
(d
rH
Pi
CU 43
3 (4 Ul
> -H -rt
0) O B
rH > Id
•-H rH S
0) Q) 3
CQ M Q
^D "d"
ro CN
TT 0
VD O
H en
o
•» 1
o ro
o *r
CN
ro 1
id
CN CN
rH
id
•H
Oi 0
M G
flj -H
S P-4
* •
H H
CO O CM < <
U> t~ \ \
2 S
o o o o o
o en en CN CN
co rH ro CN ro
rH rH
O 0
I IOOI
CO r-l
r-l
O CN O O O
ID r- 0 rH CN
CO rH H
rH
1 1 M •» 1
id id id
r>j CN ro ro ro
tn
43
Cn
•H
•a a)
rH C B
CO O >i
•rt to id id CD
H-| H ,Q -P M
x: cu to -rt o
JJ -O 0) C I
id c 4-> id S
0) Q) rt 3 0
K K H ra «
"^J*
CN
hit ^*
4 A °°
CO CN *3* CN ^D ^3* O VO rij "T
^» CN t^* en CN ro ro *\ CN CB»
4CN
t-HOOrH rHHi-HCNHCNrrtrO
rH in •*
ooco ooo^-orooo
cNror» cncNcoooo^cocN
i-icn cnrHCNrHcN ro
i-H ^*
cn^^rovDro^rororo^-
rororororororororool^1
H
, •
rrt *& i2* p|
Id Q) Id • id EH
,14 rH 3 • rH M
Id H 0) • M C < rH
p. > EH 0> id 4J
M < >V4^:43CCrt
rrJOTO) CQtt)O4J-rtO'rt
C^OM OOCUCn4JO
M 4-> -rt C 01 -rt rH Ul
MSZPlP
O
rt
Ul
i-H
a.
1
43
rQ
cu
4J
id
•rt
Ul
H
380
-------
APPENDIX A
TABLE A-3
BASIN AREA - HYDROGRAPH DATA
STATION
AREA
TOTAL
ST. MILES
SEPARATED
MAX. FLOW
of s
PEAK TIME
Min.
West Michigan
8th Avenue So.
Harbor
Chelan
Norfolk
Michigan
Brandon
Hanford
King
Connecticut
Lander
Denny-Local
Denny-Lake Union
TOTAL
0.334
0.512
0.692
1.775
5.30
1.172
0.266
6.84
0.394
1.290
0.783
0.380
1.866
20.508
-
-
-
-
4.70
-
-
2.84
-
-
-
-
—
7.50
446
780
842
1640
1420
1324
436
5630
593
1570
1080
575
2087
21
18
23
28
29
24
18
55
20
22
20
18
23
381
-------
APPENDIX B
DATA MONITORED AT STATIONS
Table B-l: Data Point or Contact Monitored
Table B-2: CATAD Interface Function Definitions
Table B-3: Classification of Remote Terminals
382
-------
rH
I
PQ
CD
1
p
H.Q
O
EH O
M M
O
S to
0)
EH 'd
O O
< o
EH
6 O
U M-l
O I
ca
EH
^1 t*~\
°-3
CL, fa
EH
EH Q)
iNai
A3T IH3
A31 INI
SOd SO
aj
aows
ajas
avs
'a/ifz
9Na
Z~Z9dV
z-iaav
aoio
z-aas
z-ais
z-aHi
i am
iaH3
asdd
aoao
as30
avio
aovo
xnvoo
xnvoo
xnvsso
xnvssa
xnvoa
xnvoa
a93
avi
aov
ajo
t-aio
z-aoa
z-aoa
0 80A1
Regulator Name
and Gatetsl in
Usa°
_
-
t-
>o
in
•»
f\
u
-
^
h
n
lo
10
n
n
S
o
in
•a-
r-»
(M
-
X
X
X
X
X
X
z-yoo
z-yoo
X
X
X
X
X
X
X
X
X
X
X
•n
Denny Way - Local
Outfal 1 /Regulator
X
X
X
X
X
X
X
X
X
X
X
x
X
X
X
X
X
X
X
X X
1*
OJ
> t-
L. ID
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
*
IHanford St. No, 1
Regu 1 a tor
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
•*
£
c
si
5?
1
383
-------
O 0)
H W
H 3
O to
*3
88
£H
B O
^ o
Q)
O
CM
0)
(!)
A3! ZH3
AJI 3011
SOd IdlS
A3! smi
A3! IH3
A3! INI
SQd SO
SOd Sa
aj
HOWS
aja3
BVS
HAfrZ
SSI
Z-ISdV
WTO
£-sn«
Z-MHS
j-sns
2-am
i-atu.
ZilH3
iaH3
badj
aoao
asso
uvio
B3VO
xnvoo
xnvoo
xnvsso
xnvssa
xnvoa
xnvcw
H33
avi
aav
a jo
£-alo
i-wa
z-aoa
. ,ui
Regulator Nano
and QatoCs) In
Uso°
—
o
o*
CO
r-
yj
in
^
t*\
(M
-
ft
K
r
in
s?
n
*|
ft*
¥
?
Pj
rl
P]
&
pi
n
fM
M
C*
CD
r-
^
".'
*r
Kt
—
O
Ov
CO
r»
King Street
Outfal] /Regulator
—
(A
™
P
"
a
s.
+•
•o £
8-2
i a
Is
fc°
~1
"o o
c £
21
Ix
(0 A
384
-------
C.Q
1-1
1
ca
1 1 TORE
used)
o w
£ <1>
no
EH O
U O
EH M
|2< O
81"
CN
ffl 1
^^ PQ
Q) EH (U
« II
EH Pq EH
< 0)
EH Q)
*< CO
C —
£Sd
ZSd
ZJdWJ
MdWJ
Sid
A3! 9IS dd
A31 MM
«M H3
9Sd
SSd
PSd
ssia
dd
l-8aWO
1-iaWO
l-9aWO
i-sawo
1-fraWO
i-£awo
1-ZaWO
1-iaHO
i-aAfrz
i-ajad
i-ajMM
9Na
2-l9dV
i-asM
£-aia
z-aas
2-ais
Z*aiMM
I'SaSd
i-^sasd
i-ata
i-iasdj
t-aasa
i*avis
i^aoai
i-aA8t
i-aiMM
PgbTld
i-sand
l-WJtld
(•"sand
i -zand
i-iand
i-aD3
i-avi
i-aov
iftasd
i"£asd
i-zasd
i-iasd
m SOA1
Pumping Station
Name3
~
o
01
CO
r~
vo
in
•*
Kl
N
-
O
in
n
n
n
K
?\
n
n
—
n
Oi
™
fc>
K
K
y
R
(N
S
N
M
CM
—
9
iil
tu
r--
^
^
•*
rn
i^
—
o
0*
co
1^
-------
•
Q
£V» »
O 'O
EH Q)
H M
8
S w
Q)
H O O
t < O
pq EH
Q) O O
H CJ «H
£J
flJ At o*j
H O 1
E-i
K 0)
M H
O .Q
ft r3
<1
EH O
«3MJd
l-rusdj
(•*2M^dJ
1-IMSdJ
1-UAC2I
I'CKMOd
1-ZlOMId
1-ltDAOd
I-BACO
1*9 DMd
1-itlMd
l->4MOd
i-£a«w
l-£^AOd
l-IHAOd
I-SHS3
1-2 HS3
1-1BSS
1-21 bro
[-111*0
a •dAi
Ptmptng Station
HIM11
S)
:«
5?
irf
*p
^
!n
iK
In
K
!n
^,
In
K
ft
TT
T
£
TT
T
**r
?
X
X
X
X
X
X
X
X
X
X
N
1
£
e
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
K>
Mattfwws Park
•o
t
V)
te«
t^
° H
sl
•fc "
s
•5.|
I, O
m in
a,?
SS
H
ro-a
386
-------
r-)
I
CQ
O
1
EH
O
EH
H
W
CO
U
< H
EH O
O
U
-sawo
-ZaNO
-iawo
-aA«
i-ajad
1-adMM
9Nd
Z-l9dV
1-aSM
t-aia
z-aas
Z-H1S
Z-anMM
i-gasd
i-sasd
i-aH3
i-iasdj
i-aasa
i-avis
l-a09l
i-aASC
i-aiHR
i-9and
irsaru
i-tand
i-cand
i-zand
i-iand
i-as3
i-avi
i-aov
i-frasd
i-tasd
i-zasd
i-iasd
„ SOAI
c
O
(0
?
5ro
11
0
o»
CO
-
u>
in
*r
m
CM
_
o
ij?
"?
r-
«o
^
n
S
tM
—
r
o>
s
0
f?
h*
\o
rl
^i
in
p|
o<
CO
1-
iu
111
T
—
—
-~
*-»
Oi
CO
r-
«3
in
•*
1*1
O4
-
X
X
X
X
X
x
X
X
x
X
X
X
x
x
X
X
X
X
X
X
X
X
X
x
X
X
X
X
-.
Belvotr Place
X
X
X
X
x
X
xx
x
X
X
x
X
x
X
X
X
X
X
X
X
x
X
X
x
X
X
x
X
X
X
X
X
-
•z.
o
5
1
X
x
X
x
X
X
WHd
X
X
• X
X
x
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
x
X
X
X
X
X
-
1
«
c
i
X
X
X
X
X
X
X
X
X
X
X
x
X
x
X
X
X
X
X
X
X
X
X
X
X
-
I
g
E1
i
LJ
X
X
X
X
X
X
x
X
X
X
x
X
• x
X
X
X
X
X
X
X
X
X
- X
X
X
X
X
X
X
X
X
0)
c
(Z
LJJ
x
X
X
x
x
X
X
X
X
X
X
X
X
x
X
X
X
X
X
in
1
o
to
1
X
X
X
x
x
.
°l
o u
g"
— t)
Z S
.» =
387
-------
CO
§
H
EH
M
23
rH
PM
Q
S
O
H
EH
o
jZj
Pn
8
<
S
p£
pQ
M
1
g
*
04
n
£)
r-(
•3
B
CO
^
B
•^
•E
tt
PH
5S
o
H
a
s
Q
e
o
2
fn
1
i
*™*.
0
o
EH
a
•rl
m
V
M
•a
§
a
p
o
td
•P
c
o
u
•—*
1
CO
§
u
EE
p«
a
2
g
o
o
H
(1)
0 •
id a>
MH .p
rl Id
i
rl
tn-P
G'H
•rl 3
in o
O rl
rH-rl
o o
1
01
-p
id
tn
M
o
•p
0) id
H O P rH
S-p id 3
id tn en
•P en a)
C JH rl
0 H OrH
u o -P id u
•p id u o
•P Id rH 0 U G
•rl IB M 01 Id
Q "aSI
a) e 03 p4 o
4J O O g O
fa Qjr-l O
o o u u
rl
o
rH
3
en
0 o o hi
« KK V
J
rH
o
M
-p
G
o
u
u
s
•rl
Q
O
•p
id
o
rH
id
4J
3
o
«
o
a) P
p id
id en
Cn
rH
rH rH
rH Id
Id MH
MH P
•P 3
8°
01
G in
o> o
OirH
0 U
o o
o o
•
0)
p
rd
tn
rH
rH
id
H_j
•p
l^
o
MH
w
•H
p
•H
•p
(U
0
fll
Ul
a
in
id
M •
a> ai
co er
ro
0
• p
p in
•G
•H (U
O W
P (U
(U rl
in o
G
(U-rl
en
•H -a
PS id
ID
" -p
in id
G Cn
o
•rl H
id id
P MH
CO P
3
rH O
rH
id
•rl
in
rH
d)
a)
rH
(U
en
•rl
id
• • •
in in
G rH
o o
•rl H
id G
•P O
CO U
&1 Q^l
G g
•rl 3
3 o
n. jj
1
4J
G
•H
O
a,
•P
0)
CO
-jj
en G
G-H
-------
•d
§
o
ca
§
H
EH
H
§
H
EH
1
•a
EH
§
H
EH
H
Q
£3
C
H
EH
PH
PC
PC
iH
§
CO
PM
0
M
O
d
01
15.
•H
O
4J
d
o
o
4J
O
0)~
M
•H
a
ft
£~
P
O<
Q
CO
•P •
M ca
id ft
4J O
CO -P
CO
ft
3 i*
ft 3
ft
ca
cu ca
ca d
O o)
•H ft
O O
4J 4J
O O
id (d
•p -p
d d
0 0
o o
•o ti
S3
S 3
0 S
0 0
d d
g O
0) ft 01
-P rH
Id 0)
tn-P H H
id O 0)
H S714J >
O ft 0)
+J rH 01 rH
Id rH O
rH Id H ^
3 1-J 0) d
O14J -P 3
0) 3 d M
Ptj O H EH
J,
> a)
co ca at i-3
0 0 iJ
ft ft M
OS O H EH
rH
0)
^
0)
•H
rH
id
•a
•H
EH
.£,
3
H
Q
H
EH
d
0
•H
-P
•H
CO
O
.p
d
•r-l
O
ft
•p
0)
cn
ca
O
ft
EH
Cn
EH
CO
rH
01
0)
H
•a
M
ffl
N
aJ
tr{
rj
losio
a
H
>
01
a
-p
a
0)
3
co
id
0)
g
H
0)
0)
rH
H
id
3
-P
1
rH
01
0)
•H
rH
rH
1
•P
S
K,
0)
rH
0)
rH
O
4J
(1
0
o
Q.
§
p,
0
4J
rH
id
d
tn
•H
CO
9
EH
d
•
rH
01
^
0)
iH
H
id
d
tn
•H
co
g
3
^
tn
O
ft
ft
3
PM
tn
-H
CO >
0)
PM
0)
^|
3
CO
ca
0)
M
ft
d
•H
0)
0
M
O
fa
M
0)
•P
id
M
M
0)
4J
0)
o
rH
fa
fa
^
•a
QJ
01
ft
ca
ft
3
CM
cn
a<
389
-------
4
i
f~*»
•O
—
•P
C
o
CJ
W
O
H
EH
H
Z
M
b
W
Q
Z
O
H
EH
tj
Z
O
pC4
W
9
^
1
H
§
EH
"
•
C*4
I
CQ
0)
i
EH
1
1*
0
0
2
pc
3
§
H
EH
H
Z
M
pL4
U
Q
H
EH
g
1
Q
§
W
z
0
H
EH
O
Z
ED
&
P?
g
H
Z
S
UJ
•HE*
3
**
J
rtj
*— *
CO
g
rf
EH
W
H
H
H
rH
id
01 -P
•P G
O 01
IS U
.p
G g
O O
U M
44
. iQ
01 2
Pi IS
IS g
K S
J4 1H
0 O
(S
m -P
a
4J -H
IH 0)
o u
PI m
(U OH
PS •
•a o 01
G -P
iS 0) 3
g tn a
CJ O U
<
, •
CO
^1
(d
r- 4
0)
J_j
•a
§
g
o
o
•p
o
0)
M
•rl
•a
0)
4J
IS
Cn
H
O
•P
is
i-H
3
Cn
0)
K
^•^ ^-^ ^^
^i
>i IS
>1 iSrH
>1 Id rH 01
IS rH 01 M
rH 0) M
0) fc 01 >1
fc G 01 Id
a>l 0) O rH
Pi 6 (Q £lj rH Q)
is a n a) o) rH
rH S -P -P IS
0) S O rH IS IS O
H O rH id tn Cn O
rH rH O rH
arH o o n IH
g OIHrH O O M
a *d o) id id -p
•P G *P 1— 1 rH 3
"0 U) O O 3 3 a
id ^4 O g Cn Cn g
0) *H 0) 0) 0) 0) O
j-3 pt| CO P* Pi Pi O
««««««
3J2w§8«u
*
0}
S-
(3
,_ {
0)
V4
t)
id
g
S
O
0
•p
0
4)
^H
•H
•O
0)
4J
id
Cn
rH
rH.
id
MH
•P
^^ ^-% ^^
>i id
Id rH
rH 01
01 H
^4
0) >i
G 01 id
arH o) ~
O 0 V4 V4
0)
0) 0) rH S
•P -P IS O
Cn Cn O
rH +J
rH rH G
(0 nj jj o
m m o a.
3 3 (1) fl)
O O P5 CO
« BJ « «
Soolw
0)
5
o
•p
43
Cn
3
O
3
G
•rl
IS
43
0
irt
pj
id
o
0
01 4-1
>i G
•-To
o) a
M '1-^
0)
"O ui
G
id o>
g 43 •
g -P 0)
O -P
u G id
•rl Cn
•P
G 'O rH
•H 01 rH
O -P IS
a is >H
•P O 4J
01 O 3
"£~~
01
•HrH
id is
o
•P rH
G 1
•H Q)
O I *
a o
•p g
01 Q)
CO OH
« «
s«
01 0) -P
•P 01 -H
0 0) 3
IS 43 O
_|J Jj t|
G -rl
OHO
0 O
H
M C O
01 O iH
43-H 4J
•P -P G
O 0 O
G U
01 <4H rH
G n)
O rH O
•H IS O
•p arH
O 'H
G 0 T3
3 G G
MH -H IS
•p a 01
0 -P
id o) o
•P43 IS
G -P -P
O G
0 01 O
3 0
O B "d
IS G
43 IS
•P oi I
M -P O
O-H O
a a
01 0 D
•H EH
01 U
43 0)
•PrH 43
O 4J
O M
•P -P G
GOO)
DOS
•H 4J
~ -P rH 01 ~
•H IS 43
TJ O
T3 O 01
IS rH >1
is
G 01 rH
H 43 0)
•P M
— G Cn
01 -H G
>1 -rl
IS 01 01
rH 5 O
oi M a
H 01 n
01 0)
oi G
a >1-rl
3 i?
O rH 01
M 01 IS
Cn ^4
Ul
rH 01 -H
rH Ul
IS 0) 01
43 >i
M -P id
O rH
1H G 0)
" *" O V4 ~
.,
EH
0
z
01
!H
O
•rt
G
O
4J
G
0)
e
a
•H
3
O1
Cn
G
•H
IS
01
£
ffl
4J
•rl
O
a
0)
U]
G
•H
V
Pj
•*^
^J
o
IS
0)
G .
0) Cn
Cn G
•rl
01 -P
U] IS
rH V4
3 01
a a
o
G
0) 01
43-H
01 -rl
0) 3
w o
rH M
3-rl
a o
•P rH
O O
Id V4
•P -P
G G
o o
U 0
IS
rH
01
^1
id
•rl
rH
•rl
rj
id
o
4J
id
0)
G
i
g rH
5 G
0 O
G 01
•H S
3
!> 4J
•-H -P
id id
^ «5
•h.^
!H
01 43
> 0
O -P
01 -rl
Ul S
O 01
0 -P
•rl
G g
OI-H
43 rH
01 43
01
01 13
O 0)
rH 4J
o id
o
4J -H
o -a
n) G
4J -H
O U]
y id
&
r-H
n
13
0)
01
o
rH
o
01
rH
id
^1
01
o
Ul
Ul
o
M
u
K
u
0
o
390
-------
•a
o
•P M
id a)
0) -P
01 !H
3
0)
&rH '01 "
O C-rl M
5
M
id
.p
tn
•H
0)
•8-
0)
H
o
•rl
+J
•H
0]
O
ft
g
+J
0)
ft
O
H -H
C
>1
rH
$
0)
H
I
O
xi
o
•p
•a
0)
•H-H
M O
0)
C (H
a, o)
0) rl
•H id
•rl
O M
O
4->
•H
01
a
c
•H
-P
Id
O
•H
•a
•> «—'
rH
-H
0)
-P
o
•H
+J
to
O
a.
-P
•H
01
§
o
ca
s
o
EH
H
S
H
t,
H
Q
§
H
EH
O
2!
&<
M
%
%
u
S
H
0)
rH
•3
C
£•
•a
§
O
CO
"Z
o
H
EH
3
b
P?
g
H
S
g
ALA
njra
O. O
id id
iw tn
C
0)
(«
01
C
O
5
01
&
(tfidS -P (D-d 0) 0 (D-H H7l
-PO4
-HO
a -d
SB
-H
01
!'
o
G
0)
fl)
>
I
01
O
o
o
O!
0) G
wo
tn
So)
tn
«
i
o
&
ai
^3
o
•H
-p
id
M
0)
M
0)
-d
fl)
>
>l
-— M Oi
a
Jd OVrl
o c 01
oi>,a> 01
idtntn
H
o
4J
3
no)
4J
-pid
-H&>
00
i! 0] 01 rH O
rH 0-rl >,-rl >, 0) >1 Ol -P
rHjJrOidnSidoinjCid
Id-H rH rH rH -H rH
0)
H
s
o
0)
>i
M
= 0)
rH 01 0)
id 10 tji
o -P id
O 01 oi
I 4S rH
0) O rH
4-> -P id
O -H MH
0)
H
O)
a
0
$
id
rH
tl ^J
O-H
•P O
M -P
0) 0] IH >i
•P -P id
Id— O, rH
cji c o a)
o o M
O JJ rH —
4J Id >irH CO
id M id a) co
rH 0) rH S S
A K
05
D
X X
X
8
88
8
8
8
8
3S
391
-------
I
g
H
H
S
O
M
Cu
U
O
S
H
1
CJ
m
«B
RE
•
I
CO
o
H
fr4
CJ
w
CO
ations
F
id
rH
<
I
o
Cn
G
•H
-P
H
(U
ft
10
to
O
W
(D
e
•H
b1
o
g
H id O
IH
•H CO CO
o
I
H
S
, 4J
O
O
H
•H
3
5
to
>1
to
o
ft
ojo
COJ^H
i 0)
G
• O'
§3
•H id
•H -P rH 0)
M G 0) H
ft O (D ft
8801
o i O.G O O
O id
to •
co M —~
O O H
H +J
o 3 -H
§-p
o
O
0)
G H
0) -H
Cn as
4J
cobocDOfljocoo'iJE
•P o> -P H a)
oo
tn
G
•p
id
M
o>
O G
•H
o id
•p to
O rH
id
O -P
10 O
to at
o -p o
H"S
G 4J
O G to
(U id
to E
i
&
>i
ft-O
•H C
•P
S
r9
u
h
$23
rt! ffl
S
O
o
O
J,
>d
«;
N
(d
o
o
ft
p.,
U
CO
m
o w
B3
w
«
•p
o
o
rH >,
0)>ild
-HdJld
idtn
'H+J
0) -P g
-P Cd 3
a^H
ftoi-p
S G to
OOJG
UCJJH
Ofjta
o
H
•p
o
I
o
+J
id
(d>Hidftto
o -P
H M rH H H >, G i
rH 0) rH 4J rH Id 0) I
4J ft 4J 0 4J H IH P
a a a-na
O O O WO
rtj
OH
U
o
W
o
oo
392
-------
^^
•d
c
0
u
CO
0
H
EH
H
H
W
Q
£5
O
H
EH
U
&4
W
U
(2
w
EH
H
g
EH
g
•
CN
1
0>
rH
id
EH
CO
K
s
PS
§
H
EH
H
z;
H
fa
H
Q
O
H
EH
REVIA'
ffl
M
<
-O
_
pH
O
o
CO
o
M
EH
U
Z
D
^-^
os
g
H
O
l£
s
3
rij
M
<
• co
D
EH
CO
M
M
H
.
g
0
O
N
ft
3
ft
-P
C
-H
Cn
•H
•a
o
o
m
m
o
en
C
£j
•H
• tn •
<
C
O CQ
U
0) G
l 0
rH (d
0)
o
•0-rl
0 -P
33
M-l CQ
g M
0 0
O -P
(_{
ft 3
g tn
3 01
OS «
05 ft
OH «
M
(U
N -P
0) Id rH
> S H
•H o) -P N
a) rH s c o
• O Id -H MH
ft 0) 0) 4J O
3 N M 0) ft • N
S -P MO)
'O N * *P '' Q) 4) f
M Q) *rl rH *rl (H Wl ^ O
•rH J4 (B fl) Cj "H O ft
id o > 3 M ft
•H O» ffl TJO rH
0) ftC Htn 0) MH rHO
O -H G A ON
•iH 4-1 4 * rgj .|-| Q J^ J_j ^J
> ON tnft id 0) -PC
N Cj id *r) g QJ * S CO
O -P tC3 NCO OO
CO CO CO ft O ft O
•H -~ CO -rl O
id >i id co id i i -a
(tf ^ i~ I M rc3 rO •
OlrHO) NO) 0)0) >CO
N 0) N -P N > ft> > in ft
3 N3 ^3 0) O -a- oooig
CO CO COHCN<*rH3
CO N CO 4J ra 42 ft
o) o)O OO-P
CGC Nraco *-" tn C
3 3 3 4JO)-P-P +J-P +JW4JC-P+JO
UH OO OOgO-HO-HOOO
id *>H idid*ididO)id id ft id id
§iO^o+JooiB
CO CO CO ONUU-POOcOCJEHOcoUUg
1 1 1 1 1 1 1 1 1 1 1 1
•p >1
1 O M M N N
< N -P >i N
EH>lQ)CO rH Id NNO)
rtJidgC O H >iO)0)&
CJrHOJ'H N 0) fd^^O
0) 4J>,N>irHOOft
cnc>ic >i cid id o) a ft
OOldOlBOr-HrHi-HN rH
-HN-HrH-H rH OO) 0)0) i-H rHO
>iidSidMid N N 0) O-N N-P
ld4JO4J -P 0)N Hg O -P -PC
rHUlftCO!-l01>iN +J-H 0) rH C CO
0) Old-HCOIdJ^-PMHOOO
N N rH N -P N rH id id tn c/1 *O O
OOOldOO) gtn-rl^rH O
M+>M4JN-PNO) C .CCO rH 0 Ol
0) id -P id o) id u g *ri o) i i ld
SrHCrHCrH^l-H 0)4-> XN S 13 "O
O3O30)3-ri> +JN CO) >
fttnocritntnidN raid 3-P -P > >m
o) o) 0) o) >i-P ri id o) •* co
O WH <; SEn BBco S ^)* OOCM
PS B PS co coco EH|S S CN ^3-rH
393
-------
-P
^_^
•a
4J
c
o
o
-* w
w 2
o «s
H U
EH 3
H
Z
M
W
Q
Z
0
M
g
Cjj
CM
W
U
§
"
Z
H
9
EH
f"t
•
CM
1
m
o
43 Z
RS 0
EH H
EH
H
Z
H
tii
W
Q
Z
O
H
EH
a
1
CO
§
CO
Z
o
H
EH
H
1
Q
EH
Z
S
Bl
H
1
•
H
13
tt) 0)
tt) ft 4
(3 43 -H
H •
4-) W O 4J
O 13 i
• C • 4J O H H r
•PO D*0)OO-P
3
U
T!
3)
H
0
-1
D
/]
H
•H -ri O rH M
0 Di i-l (d -P -H
tf = c w o ty a -H a) o
O t~^ *•"! >~l M C *H ^ O
fj] n] E (ti 4-) nj to £5 ^
CJO M C C flj 3 M-i
O • «U -H O O fi H
O ^H W 4-J € O -P ,f^ -rH (U
43s M fi C! -P
•P i^T3(UrOtt)tt)SC
4
ft
w
1)
:n
T)
1343 3 §§DO (3
••H-Ptt) - ktnUM-
tt) M tn H -P O tt)
tn era >i>ta)a)id airaH
4-> -H (1) S-l 1-1 4J rt IdO
C 4-» M -H rH S O -1
o 43 a) tn to id id
O O rH tt) tt) J-4 £ *!-t tt) rH
4-) O W U O O 1-1 fO -1
H
TJ
MH
fl rH
o id ;
<4-l
P 13
U-H
[J Id
J )H
id's 43 p O O 43 ft43 O O >H
BB tn 0 S W S EH CO 4J K! 00
1 1 1 1 1 1 1 1
A
U
-p
•H 43
SO H
W 4J -P tt) -P
•H -H rH -H
M S C H rH C
O 01 D tt) O O
4J -P -P H (H
O MrH-rl(3t 13 13 -P tn >i
o rHkooa ko)
hi rHIdOOOi-H -P tn
1 tt) -H O rH tt) Id
gS rH Q M id E tn
•H r< o .(3 >H 0) C
*d ttJ^r^Ofd!^ tt) Id
O ^i^CO^f^O ^ OH
W CO
CO COODOOtCJ DC3
iJ SOOOOO OZ
id
394
-------
1
H
? I
« 1
W P4
a
&
CO
i-l
o
4J
8-
U
0)
4-1
e
M
Jat
§
M
H
_j
r^
??
Q
0)
*J
id
VI
cd
(X
o>
CO
1-4
rH 0)
E
H
§
z
0)
9
Vt
o
CO
o>
2
o
CO
CO
e
9
Vi
H
to
01
^1
a
*y
jj
Vi
4J
CO
O "-I
0 0
Vi Vi
•H 4.
•o e
6 o
H 0
ll
« c
I-l
o
4J
§
U
4J
0)
Vi
4
X
X
X
1
1!
i
1
X
«
X
X
X
{
(
i
1
M
X
X
en
i
c
i
06
X
X
X
X
X
*i
(
c
i
i
06
X
X
(
«d
i
i
BJ
r-4
O
8* iJ
3 a
PL. O
U
SO)
M
« qs
Vi Vi
4J O
CO 4J
O.CO
e
01
0> i-l M
0) Vi P
Vi £) G
.c 3
H to a*
id
VI «
•3 o I
co 01 £
u
•H
Vi M «
1st
u
1-1
o> vi co
01 4J t
0)
£
§
4J
U
§
X
X
X
-
e
c
•r
P
X
CN
e
c
•r
A
\
X
X
X
en
tx
c
•^
1
04
X
X
X
sf
M
c
•r*
i
l
*
X
m
c
Op
i
0
395
-------
APPENDIX C
STATION EQUIPMENT CALIBRATION DATA
Regulator Instrument Calibration Sheet
Retroreflective Photo Sensor Schematic
396
-------
APPENDIX C
941
3 9J
-------
INSTRUMENT ADJUSTMENT
Connecticut Street Regulator
Trunk D/P
Span 144" H2<3
Elevate 0 to 2.66 VDC
Span 132" = 10.00 VDC, 5A(-),5B(1)
Tide D/P
Span 144" H00
£,
Elevate 0 to 3.00 VDC
Span 126.0" = 10.00 VDC, 5G(-),5H(+)
Interceptor D/P - Span 120" H2O
0 = 2.00 VDC
Span 120" = 10.00 VDC, 2E(-),2F(+)
398
-------
CALCULATION SHEET
METROPOLITAN ENGINEERS
Connecticut £
Station
Street Regulator Contract No. . 6S-J.
Dwg. Ref. Eb
Function Tide Level Xmitter Tag. No. LX 2G3A
Tnstrument Tvoe Electronic D/P Mfr. F&p
location MCP
INSTRUMENT DATR
Range Limits 200" H.
T.nad impedance
Model No. 13D2495RB
Serial No. 6911A0869J11
FIELD DATA
30 Datum Ref. MSL = 100.00 Ft.
Finished Grade Ref .El.
output Sianal 4"20 ma Location of Perm. Ref .
Ma x . Ove r rancre
Air Supply
Invert El. of Sewer 9S.50
voltage Supply 43vdc Bubbler Tube El.
Mountina Ra°k Transmitter Zero Setting
Calibrated Accuracy
XSpan "
Remarks
5% 93.50 El. Ref.
Transmitter span
' . Setting 12 Units fl-
or 9S-5 to llc--5 Ft.E!
Signal Proportion:
0 to Ft.= 3-15 psi
or to Ft.= 10-50 ma
or 0 to 12 Ft.= 4-20 ma
Line Pressure psi.
Scale
DATE , , Br
2/11/72 JoF.Lynct
1JOB NO. TITLE SHT. NO.
C200 Connecticut St. Regulator 1
Tide Level Xmitter
399
-------
CALCULATION SHEET
METROPOLITAN ENGINEERS'
SEATTLE, WASHINGTON
February 2.5, 1572
StationCo-inecticut Street Regulator
Function Tide Level Siyual Converter
Contract No. _.65-l
Dwg. Ref. E6
Instrument Type Electronic I/I
location
Tag No.
Mfr.
I/I 26?A
P&P
MCP
Model No. 50EK1CGO
Serial No. 69llA066yJ24
INSTRUMENT DATA
FIELD DATA
Input Signal
Set Points
4-20 -ai
LX2O3A
Signal Prom _
To LR207 & SP 209A
Output Signal a— ?.Q ma
Set Point Dial Accuracy
Input Impedance 130 -a-
Mounting Rack
Input Span Setting:
4 to
Output Span Setting,:
4 to
2O
20 ma
Voltage Supply
Air Supply
1 -XI — Tra.-
Remarks Fixed Calibration
ATE
2/11/74
DY
JOB NO.
caoo
TITLE
Connecticut Street
Tide Level Signal Converter
SHT. NO.
2
400
-------
CALCULATION SHEET
METROPOLITAN ENGINEERS
SEATTLE. WASHINGTON DflTE. February 28, 1972
__
Station Connecticut St. Regulator
Function Recorder
Contract No.
Dwg. Ref.
E6
Instrument Type Electronic
Location MCP
Tag No.
Mfr.
LR 207
F&P
Model No.51-4202BL02-BL02
Serial HQ.6911A3713J3
INSTRUMENT DATA
Input Signal 4-20 ma
FIELD DATA
Signal From
Input Impedance 250-/2.
Pens & Chart 3 pen 1 chart
Mounting
Plush
Red - LX 202A
Green - LX 203 A
Blue - R/I 206 B
Scale Length
and/or Arc
None
and/or Marking
Accuracy 5%
Chart Drive
Response
1" /hr.
120 ac
Voltage Supply
Remarks
D/2T/ll/72
BY
J.F.Lynch
JOB
Connecticut St. Regulator
Recorder
SHT.JIO.
M.QI
-------
CALCULATION SHEET
METROPOLITAN ENGINEERS
SEATTLE, WASHIN8TON DATE: February 28, 1972
Station Connecticut
Contract No. 69-1
Street Regulator Dwg. Ref . Efo
Function Hi9h Signal Selector Tag No. sp 209 A
Instrument Tvce Electronic Mfr. F&P
Lnnafc-inn MCP Model No. 55ES3512PA
Serial No. 6911AO869J20
INSTRUMENT DATA
Input Signal 4~20
FIELD DATA
ma signal Prom LX203A, SP209B
Input Impedance 250 -A- To SP209G
Output Siqnal 4-20
ma Set Point
Setting Hi Signal
Load Impedance 0-750 -S^
Mode Adjustment:
Proportional
Reset
Rate
Auto-Ma nua 1
P.rop. Setting
Reset Setting
Rate Setting
Contact: 1st 'I;.
L'-.W
2nd r-\
Output Meter 0-100%
Set Point - Man.. Yei
Isolation j input/ou
Mounting Rad
Low
3 Auto ^
3rd J'i
tput No
Lew
c
Available Contacts (No. of & Hi or Low)
None
Voltage Supply 120-vac
Remarks
DATE BY
2/11/72 J. P. Lynch
JOB NO. TITLE . SHT. NO.
„- Connecticut St. Regulator 4
C200 High Signal Selector
INDICATING CONTROLLER
402
-------
CALCULATION SHEET
METROPOLITAN ENGINEERS
SEATTLE. WASHINGTON mTE. February 28, 1972
Station Connecticut Street 'Regulator
Function Set Point Signal Converter
Instrument Type Electronic I/I
Location MCP
Contract No.
Dwg. Ref.
69-1
E-6
Tag No. I/I 209B
Mfr. F&P
Model No. 50EK1000
Serial No. 6911A0869J29
INSTRUMENT DATA
Input Signal 4-20 ma
Set Points
FIELD DATA
Signal From
To
SP209B
SP209A
Output Signal 4-20 ma
Input Span Setting:
Set Point Dial Accuracy
Input Impedance
Mounting
to 20 ma
Rack
Output Span Setting:
4 to
20 ma
Voltage Supply
Air Supply
120-vac
Remarks Fixed Calibration
DATE
2/11/72
BY
J.F.L.
JOB NO.
C200
TITLE
Connecticut St. Reg.
Set Point Signal Converter
SHT. NO.
5
TRANSDUCER & CONVERTERS
403
-------
CALCULATION SHEET
METROPOLITAN ENGINEERS
SEATTLE, WASHINGTON
February 28, 1572
_. . . Connecticut Street Regulator
Station v
Function Electronic Trip
Instrument Type Electronic Alarm
Location MCP
Contract No. fii-1
Dwg. Ref. E6
Tag No. SP 209 c
~ Mf r . Foxboro
'Model No. 63U-BT-OEJR-P _
Serial No. 223466^
INSTRUMENT DATA
FIELD DATA
Input Signal 4-20 ma Siqnal Prom SP 209 B
Input Impedance 2!?C -'*•
To SP 209 B
Output Siqnal Contact Set Point Red 53%
Load Impedance
Mode _Adjustment:
Proportional
Reset
Setting Green... 6u%
P-rop.
Reset
Rate
Setting
Se tting
Setting
Contact: 1st Hi See Remarks
Rate
Auto-Manual No
Output Meter No
Set Point - Man.. Yes
Auto Ho
Isolation i input/output
Mounting Rack
Lew
2nd !'•:.
Low
3rd Pi
Lew
Available Contacts (No. of & Hi or Low)
1 hi
1 lo
Voltage Supply 120 ac
Remarks Contacts drive motorized set point in SP 209 B through pulse
generating circuit.
DATE BY
2/11/72 J.F. Lynch
JOB NO. TITLE . . . . . SHT. NO.
ryr}rt Connecticut St. Regulator 5
Electronic Trip
INDICATING CONTROLLER
404
-------
CALCULATION SHEET
METROPOLITAN ENGINEERS
DATE: February 28. 1972
SEATTLE,WASHINGTON
Stationconnecticut S1
Contra c t No . 69-1
u. Keyul^Loi. j^^^ Ref> E6 •
Function Set Poillt Controller Taq No> gp 2Q9 B
Instrument Type Electronic Mfr. F&P
Location MCP
Model No. 53EC3111BB2
Serial No. 6911AO869J23
INSTRUMENT DATA
FIELD DATA
Input Sianal See Remarks Siqnal From sp 209 -C
Input impedance *one To SP209C, TCU & SP209A
Output Siqnal 4.-30 ma Set Point 60%
Setting
Load Impedance 0-750 -"•
Mode Adjustment:
proportiona 1
Reset
Rate
Au to-Ma nua 1 VP s
Output Meter °~1
P-rop. Setting
Reset Setting
Rate Setting
Contact: 1st 'I.i.
liT-W
2nd r-f.
00%
Low
Set Point - Man.. *es Auto *es
Isolation j input/ou
Mounting Ra
3rd Hi
tput yes
' : Lew
ck
Available Contacts (No. of & Hi or Low)
None _____
Voltage Supply 120-ac
Remarks Output signal proportional to motorized set point driven
•by SP209C or remotely by computer
DATE BY
2/11/72 J.F. Lynch
JOB NO. TITLE SHI, NO.
C200 , Connecticut St. Regulator 7
Set Point Controller
INDICATING CONTROLLER
405
-------
CALCULATION SHEET
METROPOLITAN ENGINEERS
SEATTLE, WASHINGTON DflTE. February 28, 1*72
Station Connecticut
Contract No. 6S-1
St. Reg. • Dwq. Ref . E6
Function Outfall Gate Controller Taq No. SP 209 D
Instrument Type Electronic Mfr. F&P
Location MCP Model No. 55EL3111BL2-BBE
Serial No. 6911A0869J30
INSTRUMENT DATA
FIELD DATA
,. „. , 4-20 ma , „ SP209A.LX202A
Input Signal •*-*•" "la Signal From '
Input Impedance 250
-fl- To Outfall Gate
Output Signal Contact Set Point
Load Impedance
Mode Adjustment:
Setting Auto from SP 209A
Prop. Setting Not used
Reset Setting Not Used
proportional 2-50%
Reset °-
Rate
Rate Setting
1UO%
Contact: 1st '11 Gate Open
Lew Gate Close
Au to-Manua 1 ye s
2nd i:.-.
Output Meter yes 0-100%-
Low
Set Point - Man., yes Auto yes
Isolation j input/ou
Mounting Rack.
3rd ?!i
tput yes
Lew
Available Contacts (No. of & Hi or Low)
Hi 1
Lo 1
Voltage Supply 120 ac
Remarks Only the contact output is used.
No function
DAT^2/ll/72 uXF. Lynch
for 0-100% scale, prop, or reset.
JOB NO. TITLE _ . „. „ SHT. NO.
C200 Connecticut St. Reg. 3
Outfall Gate Controller
INDICATING CONTROLLER
406
-------
CALCULATION SHEET
METROPOLITAN ENGINEERS
SEATTLE, WASHINGTON mTE: FebynaT.v ,
Station Connecticut St. I
Contract No. $9-1
ie F&P
Location MCP
Model No. 13D2495RB
Serial No. 6911M86W10
INSTRUMENT DATA
Ranae Limits 200" H?0
Load Impedance
Output Signal 4-20 ma
Max. Overrancre
Air Supply
Voltaqe Supply 43 vdc
Mountinq Rack
Calibrated Accuracy 5%
r-f . -,- ..
1-.- . - ' >
XSP3M
Remarks
FIELD DATA
Datum Ref. MSL = 100.00 Ft.
Finished Grade Ref .El.
Location of Perm. Ref .
Invert El. of Sewer 99.50
Bubbler Tube El. 100.00
Transmitter Zero Setting
99.00 El. Ref.
Transmitter span
Setting J-^ Units ".
or 99.00 to 111.00 Ft.E!
Signal Proportion:
0 to Ft.= 3-15 psi
or to Ft.= 10-50 ma
or 99.00 to lll.QQFt.=i= 4-20 ma
Line Pressure , psi.
Sea le
DATE BY 008 N
2/11/72 J.F. Lynch C20C
D. TITLE SHT. NO.
} Connecticut St- Reg. 9
Trunk Level Xinitter
TRANSMITTERS
407
-------
CALCULATION SHEET
METROPOLITAN ENGINEERS
SEATTLE, WASHINGTON mTE. February 28. 1972
Station Connecticut St. Reg..
Function Outfall Gate Position
Instrument Type Electronic R/I
location MCP
INSTRUMENT DATA
Incut Sianal 0-1000 -A
Set Points
Output Sianal 4_?n na
Set Point Dial Accuracy
Input Impedance
Mountina Side
Voltaae Supply 120 vac
Air Supply
Remarks Note: 4 ma equals 0
20 ma equals 1C
Gate 84 x 60, 3
DATE BY JOB NO.
2/11/72 J.P. Lynch C200 C
Contract No. §9-1
Dwg. Ref . E6
Tag No. R/I 206B
Mf r . Acromaq
Model No. 813-BX-U
Serial No. ADO 609
FIELD DATA
Sianal From P°S. X 206 A
To TCU & LR 207
Input Span Setting:
0 to 1000-/2-
Output Span Setting.:
4 to 20 ma
per cent open.
0 per cent open.
nvert 99.50
FITLE SHT. NO.
lonnecticut St. Regulator , n
utfall Gate Position iu
408
-------
CALCULATION SHEET
METROPOLITAN ENGINEERS
SEATTLE, WASHINGTON mTE. February 28,. 1972
StationConnecticut St. Regi
Contract No. 69-1
xlaLui • uy,^^ Ref> Efi
x
Function Interceptor Level Xmitter Taq. No. LX 2°1 A
Instrument Type Electronic D/P Mfr. F&P
Location MHW 10-137
Model No.13D2495RB
6911A0869J9
Serial No.
INSTRUMENT DATA
Rancre Limits 200" If 0
^
Load Impedance
Output Signal 4-20 ma
Max. Overrange
Air Supply
Voltage Supply 48 vdc
Mounting Rac]c
Ca libra ted Accuracy 5%
ItSnati
Remarks
FIELD DATA
Datum Ref. MSL = 100.00
Finished Grade Ref .El.
Location of Perm. Ref .
Ft.
Invert El. of Sewer 94.41
Bubbler Tube El. 34.90
•Transmitter Zero Setting
34.90 El. Ref.
Transmitter span
Setting ±u Units
or 94.90 to 104.
Signal Proportion:
0 to Ft.= :
or to Ft.= 1(
or 0 to 10 Ft.== <•
Line Pressure
Scale
ft.
90 Ft.E!
5-15 psi
)-50 ma
1-20 ma
psi.
DATE BY JOB N
2/11/72 JoF.Lynch C20
0. TITLE
0 :onriecticut St. Regulator
: Interceptor Level Xinitter
SHT. NO.
11
409
-------
CALCULATION SHEET
METROPOLITAN ENGINEERS
SEATTLE, WASHINGTON DATE:
February 23, 1.. 72
Contract No. 69-1
Station Connecticut Street Regulator Dwg. Ref. E~6
Function Reg. Gate Controller Tag No. LC201C
Instrument Type Electronic Alarm Mfr. Foxboro
location "CP
Model No. 63U-BT-OJER-F
Serial No. 2232664
INSTRUMENT DATA
FIELD DATA
Input Signal 4-20 ma Signal From LX 201 A
input impedance 250 -^ To Gate Control
Red... 62%
Output Siqnal Contact Set Point aroe.n ___*,p,°/.
Load Impedance
Mode Adjustment:
Proportional
Reset
Rate
Auto-Manual No
Output Meter No
Setting
Prop. Setting
Reset Setting
Rate Setting
Contact: 1st 'I;. Gate o -sen
L.rw Gate close
2nd r-\
Itsa
Set Point - Man.. Yes Auto No
Isolation j input/ou
Mounting Ra
3rd I'i '
tput Yes
Lew
ck
Available Contacts (No. of & Hi or Low)
M — — L_
Lo 1
Voltage Supply 120-ac
Remarks
DATE BY
2/11/72 J. F. Lynch
JOB NO. TITLE . SHT. NO.
C200 Connecticut St. Regulator j_2
Recr, Gate Controller
INDICATING CONTROLLER
410
-------
CALCULATION SHEET
METROPOLITAN ENGINEERS
DATE:
SEATTLE, WASHINGTON
February 28, 1972
Station Connecticut St. Regulator
_ .. Reg. Gate Position
Function a
Contract No. 69-1
Dwg. Ref.
E6
instrument Type Electronic R/I
Location
MCP
INSTRUMENT DATA
Input Signal
Set Points
0-1000 ~CL
Output Signa1 4-20 ma
Set Point Dial Accuracy
Input Impedance
Mounting Side
Voltage Supply
Air Supply
120 ac
R/I 205B
Tag No.
Mfr. Acromag
Model No. 813-BX-U
ADO - 608
Serial No.
FIELD DATA
Signal From
POS-X 205 A
To
TCU
Input Span Setting:
0 to 1000-^
Output Span Setting.:
4 to 20 ma
Remarks Note: 4 ma equals 0 per cent open
20 ma equals 100 per cent open
Gate 36 x 36, Invert 100.03
DATE
2/11/72
BY
J.F.Lynch
JOS NO.
C200
TITLE
Connecticut St. Regulator
SHT. NO.
13
411
-------
OUTPUT
TO COUNTER
(3VOLTS)
250 Mo
(SENSITIVITY)
A INDICATES VOLTAGE TEST POINT
NOTE: LIGHT EMITTING DIODE, D3, IS A "BITE"
IT INDICATES ALL CONDITIONS FOR ACCURATE OUTPUT ARE CORRECT
DIODE WILL INTERMITTENTLY GLOW RED WHEN:
a. Batteries are correctly connected & sensor is turned on.
b. Sensor is viewing a reflective (WHITE) target on rotating shaft
with black background.
c. Optical alignment and distance are correct.
d. Output voltage level is between 2.5 volts & 3 volts.
e. "Bl" battery voltage level is between 1.6 volts & 3 volts.
f. "B2" battery voltage level is between 5.6 volts & 9 volts.
g. That there is no failure of components within photo sensor assembly.
BATTERY LIFE:
"Bl" battery discharge continuously to
"B2" battery discharge continuously to
11 Bl" battery discharge 4hr/day to
"B2" battery discharge 4hr/day to
1.6 volts 70 hrs.
5.6 volts 86 hrs.
1.6 volts 90 hrs.
5.6 volts 120 hrs.
FIGURE C-l
RETROREFLECTIVE PHOTOSENSOR SCHEMATIC
412
-------
APPENDIX D
APPLICATIONS PROGRAMS
EXECUTIVE SYSTEM PRIORITY ASSIGNMENT
Level Sublevel
0
0
Name
STNDRIVE
1 CMNHAND
2 CMNDSTRT
3 WTRWRITE
4 FIREMESG
Function
Scan environment protection system
Bid clock dependent programs
Blink console lights
Initiate station contact changes
Initiate station demand and command
scans
Update station data file with new
scan data
Process station command scans in order
to check command progress and com-
plete or abort operations
Check and initiate station commands
Moves each record of Water Quality
data to the Water Quality data file
Outputs alarm messages when the envir-
onmental protection system detects
trouble
5-15
UNUSED
PROCREGL
PROCPUMP
Convert and alarm regulator analog
data
Check and alarm regulator status data
Convert and alarm pump station analog
station
Check and alarm pump station status
data
2-15
UNUSED
413
-------
Level S lib level Name
Function
SMTHREGL
SMTHPUMP
Smooth regulator analog data
Add calculated flows and new alarms
determined by Level 8 to the sta-
tion data file
Smooth pump station analog data
Add calculated flows and new alarms
determined by Level 8 to the sta-
tion data file
Check and alarm pump operation
2-15
UNUSED
ALRMREGL
3
3
2
3-15
ALRMPUMP
ALRMLOCL
UNUSED
Calculate and alarm regulator station
rates
Add alarm messages to the alarm table
Output alarm table overflow messages
Freeze station data for hourly log
Freeze station data for demand log
Calculate and alarm pump station rates
Add alarm messages to the alarm table
Output alarm table overflow messages
Freeze station data for hourly log
Freeze station data for demand log
Initiates Station local console command
4
4
4
4
1
2
3
4
CONSFORM Output the station data format to the
appropriate CRT screen
CONSUPDT Update console CRT and lights with
new station data
CONSENTR Initiate console commands when ENTER
button pushed
CONSCMND Check and log proposed console commands
CONSTATN Process new station push button re-
quests from console
CONSYSTM Process new system push button re-
quests from console
414
-------
Level Sublevel Name
4 6 CONSGMNT
4 7 CONSKEYS
4 8 CONSTEST
Function
Process new segment push button re-
quests from console
Process new digital entry module
push button requests from console
Output test message events and logging
printers
4
9-15
UNUSED
RTNRPOLL
RTNRLOCL
RTNRALRM
RTNRDISP
7-15
RTNRALRV
RTNRTEST
RTNRDATA
UNUSED
Poll Renton Remote Terminal periodically
Bid for proper sublevel if command
entered by Renton operator
Send time of day as computer running
indication
Issue commands to put station or sys-
tem requested by Renton operator
into local control
Display and print at Renton Remote
Terminal all alarm messages
Display and optionally print at Renton
Remote Terminal the operator re-
quested station's data and alarms
and status
Display and optionally print at Renton
Remote Terminal the alarms currently
active in the operator requested
station or system
Send and receive messages to test the
operation at the Renton Remote Ter-
minal
Accept data entered by Renton operator
and save it on a specified data file
415
-------
Bevel Sublevel Name
0 • WPTRPOLL
7-15
WPTRLOCL
WPTRALRM
WPTRDISP
WPTRALRV
WPTREST
WPTRDATA
UNUSED
Function
Poll West Point Remote Terminal periodi-
cally
Bid for proper sublevel if command
entered by West Point operator
Send time of day as computer running
indication
Issue commands to put station or system
requested by West Point operator into
local control
Display and print at West Point Remote
Terminal all alarm messages
Display and optionally print at West
Point Remote Terminal the operator
requested station's data and alarms
and status
Display and optionally print at West
Point Remote Terminal the alarms
currently active in the operator
requested station or system
Send and receive messages to test the
operation at the West Point Remote
Terminal
Accept data entered by West Point
operator and save it on a specified
data file
LOGRALRM Print alarm and normal messages on
events printer
LOGRHOUR Print an hourly log of the station
data on the logging printer
LOGRHADL Print an hourly storm log on the logg-
ing printer if there has been any
rainfall or overflows
LOGRDMND Print a log on the logging printer of
the station or system data requested
by the CATAD console operator
416
-------
Level Sublevel Name
7 4 LOGRDADL
7-15
LOGRALRV
LOGRWQAL
UNUSED
Function
Print a log on the logging printer of
the current storm log data for the
station or system requested by the
CATAD console operator
.Print a summary of the current alarms
at the station or system requested
by the CATAD console operator
Print an hourly log on the logging
printer of the data collected by
the Water Quality Monitoring System
8
8
8
2
3
4
FLOWREGL Calculate Regulator flow, interceptor
flow/ overflow
Accumulate overflow volumes
Calculate station maximum flow
Bid Level 8 sublevels 2, 4,6
FLOWPUMP Calculate pump discharge, force main
flow and bypass flow
Accumulate bypass volume
Calculate station maximum flow
Bid Level 8 sublevels 3, 5, 7
STORREGL Calculate trunk inflow, storage rate
STORPUMP Calculate inflow, storage rate
AUTOREGL Calculate safe diverted flow and set-
point. In storage mode, calculate
optimal diverted flow from rule
curve
Bid Level 8 sublevel 6
AUTOPUMP Calculate maximum safe discharge or
maximum safe setpoint
Bid Level 8 sublevel 7
SUPRREGL Calculate necessary regulator position
to give flow requested by operator
or AUTOREGL
417
-------
Level Sublevel Name
8 7 SUPRPUMP
8 8 RAINUPDT
Function
Calculate safe setpoint to give station
discharge required by AUTOPUMP or
operator
Saves rainfall data from Level 3 on
rain data file on RAD
8
9-15
UNUSED
9
9
1-14
15
BGNDCKPT
UNUSED
BGNDREST
Checkpoint background before funning
other Level 9 sublevels
Restart background after completion
of all other Level 9 sublevels
418
-------
APPENDIX E
FORCE MAIN CALIBRATION CURVES
Figure E-l:
Figure E-2:
Figure E-3:
Figure E-4:
Figure E-5:
Figure E-6:
Kenmore Pump Station/ Pump #3
Kenmore Pump Station/ Combined
Matthews Park Pump Station
Interbay Pump Station/ East
Interbay Pump Station/ Combined
Interbay Pump Station, West
419
-------
APPENDIX E
o
IM
o
•(
to
o
to
in
o
CO
o
-------
APPENDIX E
o
CM
o
a1
to
o
CO
o
CO
CD CM
He*;
O
a-
CM
O
CO
o
CO
KENMORE COMBINED
12/71
<2-0. 166xXs
0.10 0.80 1.20 .1.60 2.00
TOTflL HEflD-STflTIC HERD (PSD
2.40
FIGURE E-2
421
-------
APPENDIX E
o
to.
o
o
o
(O.
o
o
o
a1.
o
o
o
cv.
o
o
o
u.
o
o
o
o.
<0
o
o
o.
o
o
o.
CVJ
o
o
MRTTHEWS PflRK
00 0.80 1.60 2. HO 3.20
HEflD-STflTIC (PSD
li.OO
U.80
FIGURE E-3
422
-------
APPENDIX E
o
o
o
o
o
ID
o
o
o
o
o
o
o
C_J
o
o
o
a
o
OJ
INTERBflY-EflST
189xXs
TToo a'.oo s'.oo 7T
HEflD-STflTIC (PSD
5.00
6.00
FIGURE E-4
423
-------
APPENDIX E
o
o
o
co
o
o
CD.
o
3".
o
o
o
w.
o
o
o
o
o.
CO
o
o
o.
a1
o
o
o.
CM
INTERBflT-COMBINED
Y=-«-25. 80 lxX'-2. 803xX2+0. 17
1.00 2.00 3.00 li.OO
HEflD-STflTIC (PSD
FIGURE E-5
424
5.00
6.00
-------
APPENDIX E
INTERBflT-WEST
Y=+25. 667xX'-2. 771xX2+0. 165xX9
°b.oo
I'.OO 2.00 3.00 4". 00 5.00
TOTflL HEflD-STflTIC HERD (PSD
FIGURE E-6
425
6.00
-------
APPENDIX F
REGULATOR-SAMPLER EXPERIENCE
426
-------
REGULATOR SAMPLER EXPERIENCE
Allocation of manhours to the study varied both during differ-
ent seasons and also according to the individuals involved.
During the periods of frequent overflows, the water quality
technician assigned to CATAD worked as follows: sample collec-
tion, 40 percent; lab analysis, 55 percent; data handling, 5
percent. Conflicting duties at these times were kept to a
minimum. When rainfall became infrequent and the number of over-
flows low, time was spent in the following way: field work, 10
percent; lab work, 5-10 percent; data handling, 50 percent;
other Metro water quality studies, 30-35 percent. Other lab
technicians involved contributed 6-8 percent of their time for
sampling during the summer periods. The instrument technicians
from the maintenance division used 4.5 percent of their time on
CATAD related duties. In the winter this involved 9-14 percent
of his time repairing samplers. Maintenance crews assigned to
routine maintenance and inspection work spent 1 percent of their
time on CATAD.
During the sampling study, a great deal of data was gathered
and considerable experience garnered. Some of this data later
proved to be unsuitable for use in the evaluation of the project.
Failure of the samplers proved to be the major cause of invalid
data collection. All the samplers had problems which were often
unique to the individual sampling device. The following is a
breakdown of major problems: Line plugged by debris or smashed
by closing gates, 36 percent; pump breakdown, 21 percent; pro-
grammer malfunction, 21 percent; refrigeration failure, 15 per-
cent; fuse blowing, 10 percent; sampler counter malfunction, 5
percent.
The most vexing problem experienced was recorder pen failure,
particularly during critical overflow periods. Ninety-five per-
cent of recorder problems reported dealt with this problem. Fail-
ure of any one of three level recording pens prevented reliable
computation of flow information. The loss of flow information
prevented calculation of loading figures and caused efforts in
sample collection and analysis to be of limited value. Continu-
ous checking proved to be the only way to keep this problem at a
reasonable level. Transferring coordinates from strip charts to
computer cards was excessively time consuming. Loss of flow data,
either through pen failure or misplacement of charts, has been
the biggest limiting factor in evaluation of sampling data.
427
-------
Continuous checking and maintenance appears to be the only
remedy to the sampler and recorder problems. The priority in
work scheduling was not sufficiently high enough to prevent the
downtime experienced in spite of instrument technicians avail-
able. If practical/ a central location for logging the data
would have simplified both surveillance and maintenance.
During the course of this study, few major problems with
lab analysis became evident. Unfortunately, the Technicon Auto-
Analyzer was not of great value in sample analysis. To be really
effective, a multi-channel installation would have been needed.
Experience showed that it was faster to run most of the tests
manually than to set up the single-channel Auto Analyzer for
each test. It did prove valuable to the lab in general. Much
needed space and equipment was liberated when the Analyzer per-
formed one of the routine quality analysis, the test for nitrate
nitrogen.
The biggest problem encountered with the sequential samplers
was in the method of timing sample collection. The programmer
disks often were not turning at the correct rate because linkup
was provided by friction between a groove and rubber 0-ring.
Slippage could have been avoided entirely had the disk been gear
driven. The method of changing sample bottles was time consuming
and sometimes hazardous. The brittle glass bottles occasionally
broke off at the neck in the sampler, leaving jagged debris in
the confined area of the refrigerator. Coating the bottle threads
with silicone stopcock grease partly alleviated this problem and
also improved sealing. The composite sampler mechanism was more
dependable than the, sequential sampling units. A minor incon-
venience, peculiar to this type of sampler, was the necessity of
prompt sample pickups. In the event of two storms close together,
sample mixing of two overflows could occur.
The samplers were a prototype and suffered from lack of re-
finement in design and workmanship. For example, the pumps used
(being oil-less) were not designed to operate in a damp environ-
ment. Due to design shortcomings, one of the main starting
switches would initiate the sampling cycle after power failures
or^emergency generator tests. The refrigeration units would not
maintain a constant temperature because of control by an external
thermostat. Because the units were manufactured in Europe, parts
were difficult to obtain. After the initial problems were corrected,
the samplers functioned fairly reliably, considering the condi-
tions under which they were used.
428
-------
The large number of samples that needed immediate attention
caused congestion at certain lab equipment used by other projects;
the spectrophotometer in particular. This problem was solved
when another spectrophotometer was purchased.
Problems also occurred with data handling. Routine logging
of data was put off during sample collection and laboratory analy-
sis and became harder to bring up to date later. Although_the
heavy work load necessitated postponement of some computation
and logging, more current updating of the data would have been
preferred. This would have assisted in pinpointing possible
deficiencies. A system of monthly reports to the supervisor in
the form of a computer printout and storage of data on easily
accessible tapes would have done much to eliminate the problem.
The more detailed evaluation could still have been reserved for
slack sampling periods. When undertaking a sampling program of
this type/ there are three points to keep in mind:
1. All those involved in such a study should have an under-
standing of the aims and priorities of his job and the
project as a whole.
2. Consistency in the laboratory and field sampling must .
be carefully watched and maintained by all participants.
3. Any changes in techniques or methods should be correlated
directly to the previous procedures before any change-
over is approved.
During the program, technicians and supervisors should have
a definite idea of the progress of the data gathering and how it
will relate to the stated aims of the study. While equipment
problems can be minimized with adequate experience and backup,
one must continually work on project coordination and direction
of the principals involved.
429
-------
ROUTINE MONITOR CLEANING PROCEDURE
Winter - once per week
Rest of Year — twice per week
1. Check float and hose for obvious problems.
2. Upon entering monitor station, check pump and meters, analyzer
dials and recorder for failures.
3.
5.
8.
a.
b.
If pump is found off or plugged (amperage O.K. but no
flow), call Renton Maintenance and report it (BA6-8294).
If recorder, analyzer or probe (except DO) problem found,
notify West Point Instrument Lab (Dick Trask, AT4-6336).
Clean probes, flow cells, reservoir-to-cell tubes, and note
on check sheet. Clean reservoirs if necessary; once per month
is probably sufficient.
4. Make recorder check and note conditions on check sheet.
Draw dissolved oxygen sample(s) from spigot(s), fix with dry
reagents provided and titrate. If necessary, adjust value
for conductivity. Compare the titrated value with the
analyzer reading and adjust analyzer if necessary (full-
scale set screw). Record any adjustment made as well as
other DO check information on appropriate part of check
sheet. If no adjustment ("trim") is available, or if
analyzer is off by more than 0.5 mg/1, the DO probe probably
needs to be replaced.
Temperature, pH and conductivity calibration checks will be
made on a monthly basis and recorded on check sheet. Other
than minor adjustments (+ or - a few units) should be
handled by the Instrument Technician.
Record on check sheet all information about cleaning done,
analyzer or probe changes, supplies needed, telemetry checks,
and date and time.
Telemeter check for each station once per week. Request
zero and span adjustments (from Disk Trask) when drift is
sufficiently large.
430
-------
10.
Record briefly in log book and on recorder chart the time of
visit/ adjustments made/ any problem/ and initials.
Upon leaving, be sure recorder is operating correctly/ thermo-
stat is set at 60°/ lights are out and door is locked.
431
-------
APPENDIX G
CONTROL SYSTEM DESCRIPTION
432
-------
APPENDIX G
CONTROL SYSTEM DESCRIPTION
DETAILED REGULATOR AND OUTFALL CONTROL SYSTEM
The following explanation of regulator and outfall gate con-
trols refers to Figure G-l. The regulator control system provides
a mi11iampere-transmitted signal reflecting the measured (bubbled)
interceptor level for local indication and for transmission to the
TCU A/D converter. The pneumatic bubbler level signal is also
applied to the bellows of the regulator gate controller instrument
which has manually adjusted high and low setpoint contacts. Raise
and lower contactors of the gate motor starter are controlled by
these contact closures when the central computer has released com-
mand of the regulator, by de-energizing the computer-local (CLR)
relay. When the central computer command picks up the CLR relay,
only direct contact closure of the regulator open (RO) relay or the
regulator close (RC) relay in the TCU can cause gate movement. The
same system controls the outfall gate in the four stations which
do not have usable trunk sewer storage.
The principal control elements of the electronic outfall gate
control system are the manual controller and the referenced set-
point unit. The manual controller is provided with a step-driven
setpoint which has a range of about 1,000 steps. The reference
setpoint electronic monitor unit has high and low contacts and an
adjustable setpoint. The monitor unit compares the electronic
output signal of the manual controller with its own setpoint. If
the two values differ by more than the preset instrument dead
band, then either the high or low contacts of the monitor will
close, completing the circuit through one pulse generator unit to
the manual controller setpoint. The setpoint moves until the
manual controller output and the monitor's setpoint agree, pro-
viding a constant trunk level system setpoint, which is affected
only by a tidal level within six inches of the storage level. In
this case, the high signal selector chooses the tidal level and
raises the setpoint of the gate controller instrument.
Under central computer command of the outfall gate, relay CLR
picks up, removing pulse source 1 and the reference setpoint monitor
from the system. Pulse source 2 (which may be set at a different
rate than pulse source 1) is connected to the manual controller
circuitry. A central computer gate operation command picks up the
outfall open (OO) relay or the outfall close (OC) relay in the TCU,
causing the manual controller setpoint to move at the preset pulsing
rate. Manual controller output feeds directly through I/I isolator
and high selector to the gate controller. In case of signal loss,
the manual controller setpoint and the reference setpoint may dis-
agree. The manual controller setpoint moves to balance the condi-
433
-------
CD
W
CD
H
PH
O
H
EH
O
CO
EH
O
co
CD
-------
tion, opening or closing the gate (if required). Two pulse gen-
erators were provided due to the design requirement that system
response speed for computer command be greater than the speed of
ramping back to fixed setpoint. Ramping speed is determined set
by monitoring after installation.
PUMPING STATION CONTROLS
Although the various pumping station control systems differed
greatly in detail and principle of control, level measurement using
a bubbler in the wet well was common to all stations.
Facility analysis and a study of the wide control variations
indicated that the simplest and most straight-forward central
control procedure would leave the individual control systems in
tact and insert a substitute signal to replace the actual level
measurement. Under this system, a computer link failure could
leave a station with an abnormally high wet well level for some
particular inflow rate; resulting in additional pumping unit being
unnecessarily brought on the line. To prevent this, a ramping
system similar to that described for control of the outfall gates
was designed (Figure G-2). This system gradually lowers the con-
trol point to meet the true momentary wet well level. It produces
a gradual pump speed change rather than an instantaneous reaction
in case computer control is lost.
DETAILED PUMPING STATION CONTROL SYSTEM
Pumping stations designated for central computer system control
were provided with ramping instrumentation systems. Stations not
designated for central control were provided with monitoring instru-
mentation only.
Most small pumping .stations were provided with direc-i- operating
level transmitter bellows. The transmitter was usually a Foxboro
large case contact controller with an auxiliary electronic trans-
mitter to provide the analog signal for pumping speed control. It
was necessary to add instrumentation to this system, but it was
deemed impractical to work with the low air pressures applied to
the contact controller bellows directly from the bubbler reaction
system. A pneumatic differential pressure transmitter measures
bubbler reaction pressure and converts it to the normal 3-15 psig
control air pressure. Local control feeds this pressure through
a solenoid valve (SV-1) to the original electric contact control
instrument (modified by addition of a 3-15 psi range receiver
bellows).
The basic local control system was unchanged in principle but
made subject to remote computer control. The electric contact con-
troller proportional level signal is applied to the pump speed
435
-------
r M *d
ii |fi P .
II ill *i 1 1
* t * si
i * i lyfiH
i i *ja
j L" u ,
1 J
,W ;
0.1^
•S^S^
i
h, , ,_f A ) w fSJi •-
I H- J >l yl^O ^
* * 2
1/1
t s i^ o: <*
5« « ^mH K ;. _, SS
HE i & i ' ii i
sli «! } J hsf ss c=:7nj
. ... HK jf-p wfnS ->p i , , H
,-*-, . . . . ?JF MF . . Qi
f- *- ± 1 -. -I Jl L._i ,_ J -*"1 V^r-. ^ (^1 \ ' '' *
T T L_tL_J ' T. sr j Hr ^ vy 1
-, i s ' H
5 « r 1 Ul
. . 1 • -ii I <
3j- 3»- .o M ' JL yiy
g'r ri" J 1 > 3 (pal
.,, i ill. , ..._- T
ll ill — i
u u 1 tf g '
IT « a il 1
^Z -. P5r
jjg
1 0
D!
Sla
O
H
EH
H
O
CO
0
CM p^
CD &
FIGURE
NG STATION CO
H
PH
i — .
s
436
-------
control system as the measured variable. Level changes result in
pump speed changes so that pumping capacity always matches the
station inflow.
A pneumatic-to-current converter (P/I) and a pressure switch
were tapped off the pneumatic output of the wet well level trans-
mitter. The pressure switch provides a safety measure to drop
out computer control (revert to local control) by de-energizing
solenoid valve SV-1, in case extremely high wet well level con-
ditions.
The output of the P/I converter provides the wet well level
for analog signal CATAD TCU monitoring and is also fed to an
electronic monitoring unit. This unit provides the permissive
contacts that control setpoint movement on a motorized setpoint
manual control station identical to that in the outfall gate con-
trol system. A pulse generator provides a continuous train of
contact closures at a preset rate to effect setpoint change at the
desired rate. The manual controller output is fed to the electronic
monitor unit for reference and to a current-to-pneumatic (I/P) con-
verter for control purposes. CATAD monitors the value of this analog
signal via the TCU. The pneumatic output of the I/P converter is
fed to the NC tap of solenoid valve SV-1 which is always energized
after CATAD has been given control of the station by the wet well
selector switch (WWSS).
The electronic monitor unit differs from the outfall gate con-
trols in that it is a pure differential unit with no fixed or manually
adjusted setpoint. If the two input signals are balanced, within a
small dead band, both output contacts are open. If one signal is
higher, a contact pair closes; and if the other signal is higher,
the second contact pair closes. Any differential between the manual
controller output and the true wet well level causes the manual con-
troller setpoint to move at a rate determined by the pulse generator
and in a direction to produce corrective action. When the signals
are balanced, the manual controller responds to the wet well signal
and moves with it.
Under computer control the TCU CLR relay is picked up and the
level raise or lower signal is given by command contact closures.
The controller setpoint moves up or down in response to the pulse
generator output. Output current change directly affects the level
contact controller and thence the pump speed. The wet well level
may or may not change depending upon conditions, but the level signal
from the bubbler/transmitter system is continually fed to the differ-
ential monitor and the monitor contacts are available for control if
the computer link is lost.
If the system is locked out by high-level actuation of the
pressure switch, the full pneumatic control signal is applied
437
-------
directly to the contact controller and whatever additional pumping
capacity is programmed to match this level is brought in to use.
Most stations provide a further stage of backup control derived
directly from an independent float switch in the wet well which
activates the standby pumping unit. Most stations have one pump
in standby service which is not programmed to operate directly in
the lead-follow system and is available for such backup operation.
438
-------
APPENDIX H
PUMP STATION FORCE MAIN CALIBRATIONS
Table H-l:
Figure H-l:
Figure H-2;
Figure H-3:
Table H-2:
Table H-3:
Figure H-4:
Figure H-5:
Table H-4:
Force Main Calibration Equipment
Force Main Calibration Setup
Salinity-Time Graph
Kenmore Force Main Calibration
Station Discharge Equations
Darcy Equation Summary
Darcy "F" Values
Force Main Pressure Transmitter Installation
Force Main Calibration Equipment:
Components and Characteristics
Figure H-6: Typical Force Main Calibration Mask
439
-------
APPENDIX H
PUMP STATION FORCE MAIN CALIBRATIONS
Field calibration instrumentation is listed in Table H-l.
Table H-l. FORCE MAIN CALIBRATION EQUIPMENT
Component
Recorder
Pressure Transducer
Pressure Gauges
1 each of 30, 60
and 100 psi range.
Rotometer
Salt Injector
Salinity Probe
Salinity Meter
Characteristics
Esterline Angus, 2-channel, Model
1102E, complete with AZAS Range 0-1
mv and 0-200 v d-c.
Rosemont Model 1101/ Range 0-100 psig,
Accuracy 0.15% of full scale. Output
0-5 v d-c.
Ametek/U.S. Gauge, Model 1972, 4%"
diameter, 316 ss bourdon tube fur-
nished with certified calibration
chart.
Differential Pressure, Fischer and
Porter Model No. 10A3135, 3.3 GPH,
200 psig max.
Metro maintenance constructed
Metro maintenance constructed
Low cost multimeter
The pressure transducer has an accuracy of 0.15 percent over its
total range and provides an output of 0 to 5 volts d-c proportional
to the range of the instrument. The recorder can spread any por-
tion of the 0 to 5- volt signal into the full range of the chart.
In this application, a 20-psi input range spread across the full
10-inch span of the chart provides resolution of the least marked
graduation of 0.2 psi and approximations of 0.02 psi.
The rotoraeter provides a small (l.Sgph) fresh water flow to
continually flush the pressure-sensing equipment. Two salt injectors
were job constructed; one of two-inch steel pipe for low flow stations,
and the second of four- inch steel pipe for large flow stations.
Outlet connections consist of a 1-inch steel pipe which is threaded
into either the pump suction or discharge piping, and a quarter-turn
ball valve for controlling salt release. A pressure fitting in the
salt injector cap permits injection of station air to blow salt out
440
-------
of the tube. The salinity probe is a woodon pole with two copper
strips located on the base which are exposed to the sewage. Wiring
leads up the probe to the salinity jaeter, a low cost millimeter
connected in series with a 6-volt d-c battery.
A typical calibration setup is shown in Figure H-l. The salt
injector is preferably connected to the pump suction, but can be
connected to the pump discharge when testing shows that salt injec-
tion into the suction causes pump air binding, a frequent occurance
on. smaller pumps. Salt injection into the pump suction is prefer-
able because less air pressure is required to discharge the salt.
The pressure transducer and associated equipment are connected to
the force main pressure tap. The recorder is set up in the motor
control room so the pump station operator can watch the events
taking place and increase, or decrease the pressure. The salinity
probe and indicator are located at the end of the force main where
it discharges into the interceptor. Stations within the Municipal-
ity's systems are ail automatic and generally do not require the
services of a manned operation. However, for this testing, all
pumping was controlled by a pump station operator assiigned by the
Municipality to the force main calibration crew. Pump control
rooms are generally located above grade while most of the force
main pressure taps and motors are located in the structure below
grade. To begin a calibration run, the station is shut down, all
pumping shut off, and the wet well levels allowed to build to the
highest safe levels which can be maintained in the system in order
to provide sufficient liquid for the calibration run. The force
main static pressure is checked against the as-built drawings to
verify the relative difference in elevation between the force main
discharge and the pressure tap. When sufficient storage is obtain-
ed, the pumps are started, and the pressure brought up to a constant
level. The recorder located next to the operator in the control
room provides the operator with the visual indication of trends of
.increasing or decreasing pressure. Pressure is maintained at a
constant level throughout a given run. Background miHiamperage is
checked at the salinity probe. When all values'are in order, the
cock on the salt injector (previously charged with salt and com-
pressed air) is opened. The salt is blown into the pump and forced
into the force main. Salt wedge travel time in the force main is
monitored with two stop watches which are started together and
checked for time correspondence approximately every hour. Time is
maintained to the nearest second. A typical salinity increase is
noted in Figure H-r2 which shows the salt wedge passing the salinity
probe at the end of the force main. Radio communication is main-
tained with personnel at the probe station. A typical station force
main calibration takes approximately three days. Salt injection is
approximately one pound for each million gallons per day of flow.
441
-------
UJ
CO
•z.
o
a: -i
CD <
IJ -
< O-
O >-
UJ
O
cc
o
u.
H
W
P^
1=3
CD
H
Pn
EH
W
en
o
M
EH
PQ
H
W
o
-------
40-
30-
eo
OL
20-
10
TIME OF FLOW
FOR CALCULATION —
10
TIME, (minutes)
FIGURE H-2
SALINITY - TIME GRAPH
A major modification of the standard method of calibration
force mains was made at the Kenmore pumping station. At this
station, the time of flow in the force mains was too short to
effectively measure the salt wedge transit time, so the calibration
method was modified as shown in Figure H-3A. Air bottles, rotome-
ters, and U-tube manometers were set up to measure the depth of
flow in the sewer downstream from the pumping station. The method
of sensing force main pressure was unchanged. The major modifica-
tion used the second recorder pen to indicate the passing of the
salt wedge. Three manholes downstream of the station were utilized
— one at the discharge of the force main; a second approximately
105 feet downstream; a third approximately 730 feet downstream.
The probes were connected in series with a battery and wires strung
from each probe back to the recorder. Milliampere flow in the line
was then connected to the second recorder pen. Steady state flow
volume was measured by the U-tube manometers and an air-bubbling
system installed in each manhole. The pump was brought up to a
stable pressure, the sewer line watched, and when both manometers
were leveled out and a steady state of flow was read in the sewer,
salt was injected at the upper manhole. Salt wedge passage at the
first manhole began the time and the salt wedge passage at the
second manhole ended the time. The time for each run was kept on
the chart paper itself. A typical graph is shown in Figure H-3B.
443
-------
3
U.
u.
o
OJ
p
0
*
z
^
ALCULATI
a:
o
ii_
r
x
<
<
J
N
/
^
\
1
D C
M
H
O
H
=,
Q.I-
-------
Comp u t a t i on s
The basic calculation of the steady state of flow Q, in cubic
feet persecond, was determined by dividing the total volume in the
force main between the point of salt injection and the salinity
probe by the time required for the salt wedge to pass. Volume was
determined from the as-built records of the total lengths of force
mains and the inside diameter for the particular pipe material used.
A small computer program determined the centroid of the passing salt
wedge in relationship to time and the increased salinity as shown
on the millimeter (Figure H-2).
The force main calibration test data for each station was
tested to determine relationships between discharged pressure,
static pressure, and discharge rate. An attempt was made to fit
the data with a Hazen and Williams or a Darcy-Weisbach type equa-
tion of the form:
where:
Q = K (Pd - Ps)n
Q - Rate of Discharge
Pd - Discharge Pressure
PS - Static pressure
k,n - Constants
It was found that only the Interbay and South Mercer Island
pumping station data followed the formula. Data from the other
pump stations deviated radically. The inability to fit the data
from the most pumping stations is probably due to the low level of
turbulence encountered in these tests. The data indicates that the
Reynolds number prevailing in the test area was in the transition
range between laminar and full turbulent flow. The friction coef-
ficient in this range is a function of discharge, making a Hazen
and Williams type representation invalid. Even at the Interbay
pumping station, the coefficient "n" in the above formula found by
the least-square analysis of the data did not match the values
associated with the Hazen and Williams formulas. Further analysis
reveals that the data from all the force mains examined can be
represented by the power series of the form:
Q = A0 •<-
Pd + A2 Pd2 + A3 Pd3 +
An Pd
n
Where the "A's" are constants, it was found that the first four
terms of this expression are sufficient for the representation
(i.e., a third order polynomial). Note that the static pressure
may be ignored. The formulation actually used in the computer
445
-------
programming conditioned the input level pressure to indicate only
the pressure rise above static and dropping the AQ value in the actual
formula. The determination of the pump station discharge rate by
the third order polynomial in the CATAD system involved a table-
lookup of the polynomial coefficients for each pump station. The
coefficients were determined by the least-square polynomial regress-
ion of discharge on pressure differential. The data from the two
force mains at the Sweyolocken pumping station were combined so that
only one set of coefficients need be used for that station. Like-
wise, the data for the two force mains at the Interbay pumping sta-
tion were combined. Calibration of the Kenmore pumping station also
allowed the use of just one formula for all three force mains. Cal-
culation results are shown in Table H-2.
TABLE H-2
Station Discharge Equations
Station
Matthews Park
30th N. E.
Kennore
Intaxbay
Sweyolocken
South Mercer Isl.
North Mercer Xsl.
Juanita Heights
Ballovuc
KirkIand
Discharge Equation
Q -
34.140X-5.15 3x2+0.412x3
10.053X+0.500x2-0.680x3
S.334x-0.664x2-0.166x3
25.BOlx-2.803x2+0.171x3
3.131x-0.244x2+0.008x3
1.101x-0.084x2+0.003x3
3.4 43X-0.542x2+0.034x3
2.730X-0.326x2+0.019X3
2.840X-0.260x2+0.010x3
1.352X-0.133x2+0.005x3
Static Pres.
psi
26.3
13.2
\t
10.58
10.90
25.15
26.15
4,50
38.15
51.10
26.85
57.70
Max . Pres .
psi (Test)
31.55
16.85
12.85
16.9
40.1
19.1
45.0
57.30
35.00
68.30
Max. Q
(Test) cfs
90.275
17.66
6.843
90.817
17.953
6.630
9.185
8.897
11.167
5.688
x « difference between total head and static head in psi
Q » discharge in cfs
Each station was tested to determine the relationship between
the head loss and the discharge rate by the Darcy equation of the
form:
h = f L V2
where:
h - Energy loss due to friction total
L - Pipeline length in feet
D - Pipeline diameter in feet
446
-------
V - Velocity in feet per second
g - Gravitational acceleration in feet
per second
f.- Friction loss coefficient
The friction loss coefficient "f" was developed for each
station. A plot of the composite is shown in. Figure H-4, and the
formula for the individual friction factors for each station are
listed in Table H-3.
TABLE H-3
DARCY EQUATION SUMMARY
EL/D
1736.54
1880.1
2862.9
464.9
3179.5
1792.67
306.71
987.75
2034.5
133.36
Station
Juanita Heights
Bellevue
Kirkland
Matthews Park
South Mercer Island
North Mercer Island
Kenmore
Interbay
Sweyolocken
30th N. E.
f Value
0.01737+0.04557-V2
0.0199+0.06121-V2
0.01929-0.00747-V2
0.0134+0.009 32-V2
0.0198+0.01764-V2
0.02123+0.02079-V2
0.01219+0.09192-V2
0.01713+0.13315-V2
0.01393+0.03457-V2
0.02746+0.36418-V2
The friction factor "f" was developed to provide a means of
Comparing the relative force main friction losses in each pumping
station. In most stations with a sum of L greater than 1000, the
D
friction factor is approximately f=0.013. Exceptions are the Matthews
Park and Sweyolocken pumping stations. The high friction factor
coefficient for the 30th N. E. pumping station is explained by the
station's age. This station's force main was installed in the early
1930's, and deterioration of the main is quite obvious. Using the
Darcy equation in future calibration test runs, the indicated
deterioration of the Municipality's force mains can be checked. It
447
-------
8
a
H
*« JH <
u K >
fw »
25 S
O. H
PM O
8
01 •0
BO'O WO 90'0 M'O HTO 80'0 20'0 U'O 00'
-------
was felt that the large number of factors considered in the Darcy
equation rendered it unsuitable for use in the computer.
PERMANENT EQUIPMENT SELECTION
A typical force main pressure transmitter installation is shown
in Figure H-5. The major components are listed in Table H-4.
LOCAL INDICATOR
TCU-
FORCE
MAIN-
PRESSURE
TRANSMITTER-
CATAD CENTRAL
DIAPHRAM SEAL
POWER
PURGE WATER
ROTAMETER
FIGURE H-5
FORCE MAIN
PRESSURE TRANSMITTER
INSTALLATION
449
-------
TABLE H-4
Component
f ••—^™-n~ . __ _.._.._H^M^»
Pressure Transmitter
Characteristics
Fischer and Porter Model 50
EP107 w/diaphragm seal
Local Indicator
Differential Pressure
Rotometer
1%, 250 degree scale, switch-
board indicator, 4-20 ma equals
0-100%
3.3 GPH
Fischer and Porter Model 10A3135,
200 psig max.
The pressure transmitter is a standard industrial unit with a
diaphragm seal which converts the transmitter's calibrated range
to a 4-20 milliamp signal which is fed to the telemetry control'
unit in the station for transmission to the CATAD central office.
A local indicator on the pumping station control panel provides
visual indication of the force main pressure. Pressures and cal-
ibration ranges vary so much across the system that a standard 0
to 100 percent indicator was used. A mask showing relative flow
was developed for each transmitter ^in the system. A sample
mask is shown in Figure H-6.
FLOW MGD
9 10
12
PRESSURE TRANSMITTER
SIGNAL GAGE
FIGURE H-6
TYPICAL FORCE MAIN CALIBRATION MASK
450
-------
Each transmitter is water purged to maintain a clear opening to the
force main and to prevent sewage solids from plugging the differen-
tial pressure rotometer sensing line. In many stations, booster
pumps were required to bring the local system water pressure up to
sufficiently overcome the maximum pressures which could be obtained
in the force mains. Calibration adjustments of the individual trans-
mitters in each station were made by modifying the test pressure
transducer used in the force main^calibration test runs. Metro's
electrical department modified output of the test pressure transducer
so that the pressure readings can be read- directly on a standard
digital voltmeter. Calibration was obtained to the nearest one-tenth
psi.
451
-------
APPENDIX I
DEXTER REGULATOR SIMULATIONS
The simulation model is based on the classic inflow-outflow-
storage equation represented incrementally as follows:
S2 - Si =
+ 12
t -
o2
where I
O
S--
t
inflow
outflow
storage
time increment
Subscripts refer to the present and preceeding time incre-
ments. Storage and outflow (82 and O2) are both unknown diptat-
ing a nested iterative solution for their values. The general
solution procedure is indicated in Figure 1-1.
Boundary conditions for solving the equation are:
1. Storage upstream of the regulator gate is approximated
by the volume below a horizontal plane extending from
the water surface at the station upstream to the inter-
section with the normal inflow depth or the Broad Street
siphon, whichever gives the shorter length. This pro-
vides the condition for calculating water surface eleva-
tion upstream of the regulator station in any time incre-
ment when storage is known.
2. Regulator station outflow is computed from orifice equa-
tions or broad crested weir equations as appropriate.
Station outflow is a function of water surface eleva-
tions upstream and downstream of the gate openings.
3. Water surface elevation downstream of the regulator sta-
tion is a function of outflow. It is computed from the
normal depth of flow over the overflow weir as applicable.
The interrelation between outflow and downstream water
surface elevation requires an iterative solution for
these values when upstream water surface elevations and
gate openings are known.
452
-------
PLOW CHART OF DEXTER AVENUE REGULATOR SIMULATION MODEL
jirtr
Read Initial. Values
Inflow, II; Outflow, 01
Water Surface Elevations
Set Point Conditions
Geometric Parameters
Time Increment, DT
Inflow Hydrograph Parameters
Initialize Solution Parameters
Compute Storage, SI
Compute Regulator Gate Opening
02TEMP. = 01
Etc.
Increment Inflow and Time Step
Compute Storage, S2 from Storage
Equation
52 = ((II + 121/2. - (01+02TEMP)/2.)*DT + SI
Compute Upstream Water Surface
Elevation Corresponding to S2
Compute Outflow, 02, and Downstream
Water Surface Elevation Given Upstream
Water Surface and Gate Openings
02TEMP = 02
Yes
Output Solution
Determine Mode of Operation
From Computed Water Surface
Elevations
Initiate Appropriate Gate
Movements .
Begin Next Time Step
SI = S2
01 = 02
02TEMP = 02
FIGURE 1-1
453
FLOW CHART OF DEXTER AVENUE
REGULATOR SIMULATION MODEL
-------
The general solution procedure in any time increment assumes
an outflow rate and computes the corresponding storage and upstream
water surface elevation using known values of the other parameters.
Based on upstream water surface elevation and the gate openings,
corresponding new values of outflow and downstream water surface
elevations are computed. The new outflow rate is then compared
with the assumed rate and a new iteration is begun if the two
values do not match within a specified percentage. When a solu-
tion is reached for a given time increment, parameter values are
output and appropriate gate movement initiated via a cascading
series of FORTRAN IF statements depending on the computed water
surface elevations and the setpoint conditions. The solution as
a function of time may be plotted on a CALCOMP plotter.
Explicit solution stability criteria have not been deter-
mined, but is is known that a time increment of 3-sec or less is
required. Although the simulation model has not been verified
by comparison with field data, the simulated regulator station
behavior is as expected and the results appear reasonable. The
assumed triangular inflow hydrograph used in the simulation studies
was defined by a rate of charge of inflow of 0.015 cfs/sec and a
peak flow rate of 200 cfs. These are believed to exceed corres-
ponding values in the real situation so that the computer simu-
lation is conservative.
Setpoint conditions for Mode 3 operation were proposed after
initial experience with the simulation model. These were modi-
fied on installation to provide a larger elevation difference
between the regulator gate open and close position and to over-
lap the neutral positions. Computer simulation using the in-
stalled setpoint conditions revealed that this situation would
result in Mode 3 upstream water surface fluctuation wide enough
to cause inadvertent bypass gate 'tripping. This points out the
need for computer simulation of station behavior under given
setpoint conditions prior to installation. In general, the model
revealed the following relationships:
1. Fluctuations in water surface elevation and storage
upstream of the regulator gate, and in overflow rate
downstream of the gate in Mode 3 operation are directly
proportional to the dead band width between the gate
open and close positions; and to the distance between
the extremes of the dead band and the respective neu-
tral positions. The speed of regulator gate motion may
also have an effect. This factor was not examined.
2. Mode 3 neutral positions overlapping with the gate clos-
ing neutral position at a greater elevation than the
gate opening neutral position cause excessively large
upstream water surface elevation fluctuations. This
454
-------
condition necessitates a lower elevation of the regula-
tor gate open position to protect against inadvertent
bypass gate tripping. Neutral positions should be set
no more than one-third of the dead band width away from
the respective gate open and close positions.
Further experience with the simulation model led to the
following recommended Mode 3 setpoint conditions.
Setpoint
Shift from Mode 2 to Mode 3
Regulator gate open position
Opening neutral position
Closing neutral position
Regulator gate close position
Bypass gate trip
Other setpoints were not modified.
Recommended
Elevations
44.0
44.0
43.9
43.7
43.6
44.6
The simulation model also determined the effect of the time
required for the bypass gate to open fully on the maximum subse-
quent overflow rate to Lake Union. Reducing bypass gate speed
spreads the resulting overflow hydrograph time, reducing the
peak overflow rate and surge in the outfall. The original sta-
tion design specified a 30-second opening period, which produces
a very sharp overflow hydrograph and an extreme peak overflow
rate.
Simulation studies revealed that a bypass gate opening
period in excess of 10 minutes reduces the subsequent peak over-r
flow rate to a value comparable to normal Mode 3 regulator opera-
tion. It was recommended that bypass gate speed be reduced to
6 inches per minute, providing an opening period of 11 minutes.
Maximum simulated upstream water surface elevations with this
gate speed ranged from 145.0 when the regulator gate was allowed
to function normally to 146.2 when it was fixed in a nearly closed
position to simulate gate malfunction. For reference, the regula-
tor station gate room floor is at elevation 146.7.
455
-------
APPENDIX J
INVENTORY PROGRAM SAMPLE OUTPUT
'"i
Table J-l: Computer Division Manuals Inventory
Table J-2: Computer Division Spare Parts Inventory
456
-------
iH
1
1-3
^
1-3
§
£••
»-
»- o:
Z UJ
< o
3 z «r
o o to
> o
IT 1- U
1- »-
Z UJ
Ul < O
<9 3 Ul
•< o or
a
i
_J (O
< or
3 i)-
z z •
< 05 : •x
u- U
•> cn
•» o
—i
,3 D . _l
Z •- U!
•* 3 Q •
i: a o xi
ui v .•»
t-
t
•
v
i
i
t
i
i
i
i
i
r
B
D
t
t
1
1
*
t
1
1
1
t
1
1
•
,
1
,
1
r
t
I
1
i
i
i
y
1
1
0
in
o
p)
*c
m
cu
o
IT"
r
cn
Ul
a:
Ul
CD
1—
z
*o
u
a.
x
_i
•^
•*
!^.
_j
)
C5
u:
z
(C
u
a_
\
' )
^'
*
rv
o
c
•*
m
*
tr\
o
0
CD
^
•_4
f~\
O.
X
1
o'
*•
(S^
_)
_J
to
at
JtT
CT
CJ
a.
\
i
o
^
i^
o
o
«4
o
0
t
^
^3
Z
•4
5:
u.
Ul
or
, u.
t-
z
o:
Q_
1^)
~2i
*_4
;
O
1..
£K
U.'
U.
u.
:E
IE
Ul
f—
^
or
a
LjJ
Z
•— >
_J
]j
c}-
^»
a
o
o
V-(
o
0
cr
_J
3
^
Z
U.
Ul
a:
or
Ul
^
^.
a.
Ul
z
*— 1
_J
o
U.'
a:
LL.*
u.
u.
!tl
r,.
UJ
t—
'j*
or
a
LiJ
z
•— <
i
*
4-
fv
o
o
•
in
in
o
a-
f_
to
U!
or
Ul
i—
z
t-*
X
a.
Ul
»^
_j
Q.
XI
n
x
CVl
^
21
ei
cn
3-
o
a-
UJ
t—
z
c:
a
LiJ
z.
•— *
i
^
*
rx
0
or
»-
z
c
UJ
ca
r
a:
cn
1C
CL
CD
<
Q
CK
Ul
1-
jl
or
a
Ul
•—I
_J
^.
^
^.
o
o
*o
^
•rt
o
jr.
•
cn
»_
f_
C/3
(D
Z
CD
<
o
Q-.
UJ
»—
s
JJ,
a
Ul
z
»•«
1
o
^>
)/x
o
o
•*
m
*
o
rs
0
a.
t
cn
u
i-
cn
0
TC
tn
•-Ll
a
x
_i
o
*
(^
or
Ul
t—
z
a:
a.
Ul
7
»— t
1
O
^
fx.
O
^. •
*H
O
o
o>
z
JL
U.
Lt
a.
x
i -
o
*
^
Lt
Ul
1—
^
a;
a
ill
z
•-M
i
^
4>
pv
o
to
o o
,_!
oo
m
o
Q*
H
cn
Ul
or
Ul
2J
(T
z •
33 -I
U (D
f >
JQ
^ T
n ui
u
rr •<
i- <
•g •>-
CJ ^
tc
Z!
cn
>
Ul
o
•-*
•3
O O
o> o
rv oo
»
IV 0 0 O
•^ ^^ ^
O O O O
jj-
m
u u. •
* * »< •*
iC «O ^ Ri
O> CT\ O -r-1
o o o o
O> C\ Iv O>
or
.
or
z
_1 _l CD
< 4
33 or
< < u.
s x:
z
u. u. <
or cy
CD z:
- >jj
3 a
•*t "^ ^
r s: CD
CD r-3 3 <
co tn cj o
3 3 :3 3
a. a a. a
u u u u
a; cv tx, ex
< -t < ~* •— •
co cn cn cn
0 O 0 O
rj -3 a o
00 OG 00 00
0
o
(TV
CO
03
O
o*.
—J
^{
or
U)
z
a.
ce.
Ul
a.
o:
o
u.
z
^
z:
5T
'S
Q.
CD
<
O
71
a
u
CVj
-*
a
»— <
cn
o
o
oo
O
o
G\
pj
00
O
0s
I
3
a
u
(U
•<
cc
»— 1
cn
o
o
UO
0
o
•&
PI
1
(XI
0
o
o
»v
7T
^
or
cs
(D
rr
a
cs
•*
a
^
a
u
CVj
<
cs
»-4
CO
c
o
30
0
o
J^
p>
f
OH
0*4
Al
o
s:
^
or
0
1C
or
a
CD
<
rs
^3
a
u
f\!
<
cs
»^
cn
o
00
457
-------
cuo>-i*-injoo«-iooooo euoooo oo •-< ni ••< '•-« *•>
Ok
UlfU
PI X.
Z> • I
(3
D »-.
nr o
T ui
in
t-
nr
<
a
U. O
o n.
y* .«i
^ ffl (T
*- U <
— Q.
-J cn
Q.
O
3
IT
i
ar
^D UJ
W ^ '1
nil- !•}
I- - z s- in i- or oc
7 Ul--r3 ,Z-»-l-
•— co f- a < u> in
ui ui x
z z en a. «u
(3 n "D KI 3C 7
•< •< cc n: •-• o
y s: .-n of i- a.
UJ UI
"O
_
rr y. tx
a. ui « ui
— > 7 K >
or uj —Q. LU
»- _i a. in j
UJ
a.
< >-
o i- x
00 UI ID
y -J tr
< ui ui
_l(- X
T O O O O O
U O O O O O
PO
7-tn
UJ U
n
*-• i
o
*-i
is.
or
ar ui
o ui—i--«njm* ^-
OOOIOOOO7
»
-------
APPENDIX K
RAINFALL STUDIES PROGRAM
SAMPLE OUTPUT
RAIN ST9R1 STUDIES FR8M ll/ 2/69 T9 7/ 1/73
INCREMENTS- TBTAL* 0*050 MAX.RATE* 0.010 AVS.RATE* 0.010
MINIMUMS • TOTAL* 0*010 MAX. RATE- 0.010 AVG.RATEw 0.010
SEASBN • FR8M H9NTH* 1 T8 MBNTH 12
3AQES • FR8M SAGE 1 TB GAQE 6
T8TAL
RAIN RAMGE NUMBER 8 1? STBRMS IN R A N 3 E
FRBM T8
0*001- 0*050
0.051- 0*100
0*101- 0*150
0.151- 0.200
0*201- 0.250
0.251» 0*300
0*301- 0*350
0*351- 0*400
0*401- 0*450
0*451- 0*500
0*501- 0*550
0*551- 0*600
0.601- 0*653
0*651- 0*700
0*701- 0*750
0*751- 0*800
0*801- 0*850
0*851- 0*900
0.901- 0*950
0*951- 1*000
1*001- 1*050
1.051- 1.100
1.101- 1*150
1*151- 1*200
1*201- 1*250
1*251- 1*300
1.301- 1*350
1*351- 1*400
1*401- 1.450
1*451- 1.500
1.501- 1*550
1*551- 1.600
1*601- 1.650
1.651- 1.700
1.701- 1.750
1*751* 1*800
1.801- 1*850
1*851* 1*900
1.901- 1.950
1.951- 2.000
2.001* 2.050
2*051- 2*iOO
GAGE 1
189
58
49
39
21
26
13
13
5
9
7
6
4
4
6
4
1
0
0
1
0
1
1
0
0 '
0
0
2
1
0
0
0
0
0
0
0
0
0
0
0
1
0
GAGE 2
154
74
41
43
30
21
16
11
13
15
7
5
8
5
6
7
3
2
3
0
0
0
1
2
0
1
0
1
3
0
0
0
0
1
1
0
0
0
0
0
0
0
3AGE 3
140
88
41
32
30
25
23
16
8
12
5
6
11
8
7
2
6
3
2
0
2
0
1
3
0
2
0
1
1
1
0
0
0
0
0
0
0
0
1
0
1
1
SAGE 4
96
54
35
28
31
22
14
9
7
8
7
6
4
5
3
3
3
6
4
1
1
0
1
0
0
0
3
0
0
1
0
0
1
0
1
0
0
0
0
0
1
0
3AGE 5
88
69
36
25
28
22
17
11
9
4
5
8
5
7
6
2
6
0
2
2
3
2
0
1
1
1
1
1
0
0
1
1
1
0
0
0
0
0
0
0
0
0
GA3E 6
30
19
11
8
9
10
3
1
3
0
1
1
2
1
1
1
1
0
0
0
0
1
0
0
1
0
0
0
0
0
1
0
0
0
0
0
1
0
0
0
0
0
A T
459
-------
2*101- 2*150
2*151* 2*200
2.201* 2*250
2.2S1- 2*300
2*301* 2*350
2.351* 2*400
2.401* 2.450
2*451- 2.500
2*501* 2*550
2*551* 2*600
2*601* 2*650
2*651- 2*700
T8UL ST8RMS
TSTAu RAINJ
MAXIMUM
RATE RANQE
rRBM T8
0.001- 0*010
0.011« 0.020
0.021- 0*030
0.031- 0*040
0.041* 0*050
0*051* 0*060
0*061* 0*070
0*071* 0*080
0*081* 0*090
0*091* 0*100
0.101- 0.110
0*111- 0*120
0*121- 0*130
0*131* 0*140
0*141- 0*150
0*151* 0*160
0*161* 0*170
0.171* 0*180
0*181- 0*190
0*191* 0*200
0.201- 0*210
0*211* 0*220
0*221* 0*230
0*231- 0*240
0*241- 0*250
0.251- 0.260
0.261* 0.270
0.271* 0*280
0.281- 0.290
0*291- 0*300
0*301- 0*310
0.311- 0*320
0*321* 0*330
0*331- 0*340
0
0
0
0
0
0
0
0
0
0
0
0
455
81*49
N J M B
GA3E 1
134
45
25
41
36
30
52
10
16
11
9
10
6
7
3
2
4
4
2
1
1
0
0
1
0
2
0
0
1
0
1
1
0
• o
0
0
0
0
0
0
0
0
1
0
0
0
475
. 107.92
E R 8 F1
GAGE 2
103
49
38
47
32
32
46
14
19
19
13
11
4
6
10
7
7
2
1
3
2
3
2
2
0
1
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
480
117.68
S T 8
GAGE 3
95
39
43
40
39
33
42
17
12
25
16
14
5
7
11
9
6
2
2
5
2
2
4
1
3
2
2
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
354
89.60
RMS i
GAQt 4
55
38
36
38
19
21
37
20
20
9
8
7
9
5
7
4
4
3
2
1
4
1
1
1
1
1
0
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
367
98.39
M R A
SAQE 5
57
34
40
33
23
31
43
17
15
13
10
8
7
4
7
3
3
3
2
0
2
1
1
1
1
2
0
0
0
1
3
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
107
26*85
N 3 E A
3A3E 6
17
9
12
9
9
13
16
2
2
3
1
1
1
0
2
1
0
1
1
0
2
0
1
0
1
0
0
1
0
1
0
0
0
0
460
-------
0*341- 0«350
0*351- 0.363
0.361- 0»37D
0*371- 0.380
0.331- 0.390
0.391- 0.400
0.401- 0.410
0.411« 0*420
0*421- 0.430
0>431« 0.440
0.441- 0*450
0*451* 0*460
0.461- 0*470
AVERAGE
RATE RAM3E
FRBM T8
0*001- 0*010
0*011- 0*020
0.021- 0*030
0.031- 0*040
0*041- 0*050
0.051- 0*060
0*061- 0*070
0*071- 0*080
0.081- 0*090
0*091- 0*100
0«101« 0*110
0*111- 0*120
0*121- 0*130
0*131- 0*140
0*141- 0*150
0*151- 0*160
0*161- 0*170
0*171- 0*180
0*181- 0*190
0*191- 0*200
0
0
0
0
0
0
0
0
0
0
0
0
0
M U M B
QA3E 1
166
88
78
49
25
, 17
19
Z
4
3
2
1
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
E R 9 F
GAGE 2
138
111
76
53
37
25
21
5
5
0
2
1
0
0
0
1
0
0
0
0
0
0
6
0
0
0
0
0
0
0
0
0
1
s T e
3AGE 3
126
112
83
55
29
30
23
9
1
3
4
1
1
1
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
1
0
0
RMS I
GAGE 4
89
86
63
43
25
19
22
1
2
1
1
0
0
0
0
1
1
0
0
0
0
1
0
0
0
0
0
1
0
0
0
0
0
Ni R A
QA3E 5
84
95
77
41
24
14
22
1
1
2
4
0
0
1
1
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
N 3 E
GAGE 6
23
25
28
9
7
1
7
1
2
1
2
0
0
1
0
0
0
0
0
0
A T
461
-------
F8R THE A3BVE PERI8D -
AREA-WIDE ST9RMS, T0TAL «
TRAVELLING 9R FR8NTAL* TBTAL «
PBCKET 8R L8CALIZED* T8TAL 5 182
183
TBTAL
RAIN RAMGE
FRS^I T8
0*001- 0*050
0*051- 0.100
0.101- 0*150
0.151- 0*200
0.201- 0*250
0.251- 0*300
0*301- 0*350
0.351- 0*400
0*401- 0*450
0*451- 0*500
0*501- 0*550
0*551- 0*600
0*601- 0*650
0.651- 0*700
0.701- 0*750
0.751- 0*800
0*801- 0*850
0*851- 0*900
0*901- 0*950
0.951- 1.000
1*001- 1.050
1.051- 1.100
1*101- 1*150
1.151- 1.200
1*201- 1.250
1.251- 1.300
1.301- 1*350
1.351- 1.400
1.401- 1.450
1.451- 1.500
1.501- 1.550
1.551- 1*600
1*601- 1*650
1.651- 1*700
1*701- 1.750
1.751- 1.800
1*801- 1*850
1*851- 1*900
1*901- 1*950
1*951- 2*000
2*001- 2.050
2.051- 2.100
2*101- 2*150
2*151- 2*200
2*201- 2.250
ST8RM PATTERN
AREA-rflOE
38
23
25
24
9
13
10
5
3
5
6
2
4
4
3
2
2
0
0
0
0
1
0
1
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
DISTRIBUT18N
FR8NTAL
53
20
11
9
7
12
3
7
1
4
2
2
2
0
1
. 2
0
0
0
1
0
0
1
1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
L3CALIZED
30
14
14
10
8
2
2
0
1
2
0
2
0
0
3
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
462
-------
2»2ii»
2t301-
2«351-
2»401»
2.501«
2i551»
2.601-
2*6S1«
ENID
2*300
2*350
2*400
2*450
2*500
2*S50
2*600
2*650
2*700
0
Q
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
463
-------
APPENDIX L
URBAN STORM DRAINAGE
Table L-l: Physical-Chemical and Oils Analyses
Table L-2: Nutrient and Bacterial Analyses
Table L-3: Heavy Metals and Solids Analyses
464
-------
m
o
X Q
M
Q S
w o
&4 EH
co
H
CO
I
CO
M
O
0
O
H
CO
At
Q)
rH
•a
*Q
O
0
§D
D!H
H nj
CO
i-q
3 H
1°
Se5
q X
S W
*
Q
§
O
•ie1
•51
0
CO
*
0
o
*
8
PQ
*
Q
Sri
£°
H
EH
1
STATION
rH co r- CM CM co r-
tH rH »* VO in -si1
F- rH o r~ o o o
CM co n vo «* oo n
«* CM CM
ro o o r- oo H o.
rH rH H CM CM "* IT)
rH
in o o o ro oo o
rH O O O CO O CO
rO 00 fO CM rH iH rH
• •
^ co
ro CM
o o o t-- r~ r~ CM
»* CO CO OO 00 CO «*
rH H rH
CO CT> rH CM CO rH
in vo o o <* ro
0 H rH 00 O P- rH
iH CM
oocM'tfr^invorHo
W rHrHrHCnOOrHrH
rH rH rH rH rH rH rH
in*3«*3<'3O^CT^CT>OOrH
OOOOOrHrHiH
VO ^O VO VO VD VO VO
co co ro ro ro co ro
o o o o o o o
ro ro ro co co co co
r» r» r^ r* r~ r» r^
0
iH CM co •>* in vo r~
1
rH
g
r^ in CM CM r- vo CM
rH in r~ vo vo •* "*
vo in CM in *# rH **
co ID oo cr> oo oo r~-
•<* m •«* CM in «* in
in in r* CM vo r»- ro
rH
o o o o o o in
in -^r in rH CM o o
^f CO CM CM rH rH rH
• •
00 0
f* co
in o o o o er> oo
o »* iH vo in o
rH iH CM rH rH rH
t» vo vo vo vo oo Ht
CO VO O O O CM CM
rH OO O O rH OO >!jt
CM H iH
cMor^oo1"* cotnoo
OOO OOO
rH rH rH iH rH rH
vocor-»cooooiHoo
r-» r^ r« r^-
0
O 0
iH CM ro -^j" in vo tH
I
•<*
U
en
oooooooo o
ooinrHinco'a
COroCM^frHiHiH'* rH
oooooooo o.
rHrHr>-cr>oincMco CM
vomooooincM'a'in in •
oor-ooooo p
rHCMcnc^corovoo CM
rHrH iHCMrHrHCM CM
ooooomoo o
CMCMrHVOOVOCnin O
"*«3*COCMCMrHrHCM CO
• • •
OO CM O
CM CM CM
Oooooooo o
o^Dr^oot~-ovoro o
coonin^cMrorovo vp
vot~vovovovor-r^ r^-
* * * * * * * * '
-
^000000° o
vovor~rHi~-inin^f oo
HrHrHCOCMCOinH VO
inooooinoinroinm
vovot^vooocTioovor^r^
oinoincMCMCMoooo
incMrOCMrHrHCMCMCOin
ininomoinoinoin
•^Jo^c«cr>oo
OOOOOOOOrHrH
vovovovovovovovovo
rorororocorocororo
ooooooooo
co co co co ro ro co ro co
r^r^r»r^r~t^-r^r-»r~
O
rHcMco-^invor^ooCTi
I
i-~
Q
m
r>
§
•H
4J
(1)
O
fl
O
u
465
-------
CM
0)
H
ft
a
2
0
P4
H
8
V->
<
(
1
^
i
rH
£
u
<2J
§
§
H
H
2
H
s
o
o
H
s
W
3
CQ
H
<£>
1
J
i
EH
FECAL
STREP
W
fr
i
o
o
K
&
§
« -*
Q O
&*
•^
i
CM
2
ft
co
3
•
9
I
1
g
PTATION
i
O O O O 0 0 0
in ro o in CT> *# o
H co in i* Is- vo vo
o o o o o o o
o o o o o o o
CN VO CF> •«* CO CO O
CM ro in vo r~ r^ r~ t^- r>-
O
rH CM oo •<* m vo t*-
rH
CM CM VO in CM O CM
H r- H
o m in in o in in
vo «* oo a\ vo oo cr>
CM rH
in co iH CM o o o
O 0 O O
H Is* o\ oo in vo «*
rH 0 0 0 rH rH 0
in o in in o in o
H vo PO r^ vo -^ ^
CM CO <***•># -sj* 'd'
o o o o o o o
CO VO O VO CM OO O
O CM CO CM CM rH rH
r~ CM o o o «* r~
CM CT> •* co rH r~ vo
•H rH rH
vo vo 10 vo VD 10 vo
co co co co co co co
o o o o o o o
co ro co co co co co
r> r~ r^ r~ t-» r^ r>
0
rH CM CO •<* in VO O
1 o
«* H
u
CD
o oooo o ooo
o oooo o ooo
o oooo o ooo
o ooooo or^o
CM ocooin CM mcooo
c~-co H
K ^
CM CM
OOOOO OOOO
OOOOO OOOO
ooooo oooo
O QSD OOO (DO OOO
O >D O 0 O >O OOO
CM OD *# CM CM oo in o o
CnOVOCOCO OVOCMCM
H iH
COCM COCOH (T> OCOO
VOt^- COCMrH VO COCOO
H rH
OO O-<*O O OOO
-*VD t~-^fCM CM COCOO
•*in CMrHrH rH iHrH'*
OO 000 0 OOO
•>*in r^o\c~- in inm^j*
mo moo o mmo
OrH iHrHrH rH rHrHiH
OO OOO 0 OOO
O"* VOOOO «# VOOOCO
f>VD ^CMCM rH CO^CO
oo ooo o ooo
oo oor~ oo ooo
Ho cncMoo -jrvocri
CO CO iH rH H rH rH
VDVD VOVOVD VD VOVOVO
coco cococo co cococo
oo ooo o ooo
CO CO CO CO CO CO CO CO CO
r^r- r^r-r>. r- r^r^r^
• ;
O
HCM co"*in vo r~oocn
r*
Q
m
o
466
-------
co
H
CO
CQ
Q
H
r3
O
CQ
CQ
i-3
IS
co
XI
(0
EH
rH
in
ta
^
O
H
rl
EH
&
M
0
a
0
o
dr
Q
rl
CO
W
•
03
-P
0)
CQ
C
t<3
(0
»-l<
•O
o
^
u
tn
W
0)
ta
XI
Ai
3
H
I
STATION
o o o o o o o
in ^r in I-H o co co
rH CM CM CM in rH rH
rj< vo oo o o vo oo
rH CO ^ r^ 00
OO rH VO rH O O CM
rH o o o in rH o\
CM <* rH
CM VO
rH rH rH CM ^P CO CO
t-~ rH rH •* CO CO CM
O O O O O O O
vo vo vo vo vo vo vo
rH rH rH rH rH rH rH
co co co co co co co
0 0 O 0 O 0 0
co co co co co co co
r~ r~ r- r- r^ r~ r^
o
rH eg co -^ in vo t-~
I
i-H
^
CN O O O O O O
in <* CT\ in co vo vo
CO CO CN CM rH rH rH
00 OO O O 00 ** ••*
CM rH 'ST CM CM
rH
rH rH 0 0 O rH rH
o o •«* «* r-« o o
CM CM CO
o o o o o o o
in co oo vo o in t-»
H CO -* "* •=* CM CM
in in in in in in in
o o o o o o o
•«*•<* rH ^P •* r- -*
O O rH O O O O
O O O O O O O
rH rH rH rH rH rH H
O O O O O O O
CM CM CM CM CM CM CM
O O O O O O O
O O O O O O O
O O O O O O O
r-~ in o o o in vo
in i*- CM vo ^ r-- ^
rH rH rH
o o o CM r- in r~-
rH co oo c» r- co CM
o vo co in m m
CM rH rH O H O 0
vo vo vo vo vo vo vo
rH rH rH rH rH rH rH
co co co co co co co
o o o o o o o
co co co co co co co
r-~ r~ r~ r~ r~- r-- r>
O
O 0
rH CM CO «* in VO rH
1
"*
o
03
OOCMCMOOOOO
co'*cocMCMor--'a<
CO CM rH rH rH rH rH
OOOOOOCMOO
OinCMVOt-rHr>-VDinoocM
OOOOOOOrHO
ooooooooo
ooooooooo
OrHinoooooo
cx5vor~-cor-coooo\
CM CM rH CM 3 rH
ooor^'3'inincMin
rHrHCMinr-^COCMrH
oocMcor^vovoo\r~
cM^covocomininvo
•=J- rH
VOVOVOVQVOVOVQVQVO
rHrHrHrHrHrHrHrHrH
cocooococococococo
OOOOOCDOOO
cocococococorococo
r^r--r--r-r--r~r~-r^t~~
O
rHCMco^invor-ooeri
I
r-
Q
ffl
U
467
-------
APPENDIX M
SYSTEM PERFORMANCE PROGRAM
Table M-l: Rainfall vs. Overflow
Table M-2: Rainfall vs. Loading
468
-------
H
Q
£^
W
1
PROGRAM
W
U
CM
o
<2
w
A<
s
••M
EH
CO
X
to
O
•
CO
i
£2
H
<:
a
•
i
*
^»*
(U
.H
•3
EH
. ST8RMS '
Ul
O_
0)
Ul
_l
CO
•<
fc-
{/)
I
»-
_l
Q?
»-
Z
U
_I
u
_J
r
I
Ul
tin
ce.
ffj
u.
Ul
Ul
•
(D
7
in
u
CD
CD
7
i
*
33
1-
a
r>
. »-
c/>
_j
A
K
~ "
or ui
I j^i
•« or
l-U
«-(
in
«~t
o in
o ro
O (VI
* o
•"»
vO *~*
(vi rv
•
OS -«
»—
O
O
r+
O
in
o rv
Q !XI
o o
.442938 69,
.027749 li
<»» -t
in
*-• (VI
en in
OO .H
rv o
J •
< S.O
»- >
OB . <
»-
0
o
(VI
o
o en
a rv
o en
O (U
(VI O
in -<
NO NO
— 1 (Tl
o u>
rv rv
* rv
NO
m
tv rv
NO OJ
>4
_l •
09 •<
fr-
O
O
*
o
in
*H
00
0 0
O 00
.536056 72i
.035737 4i
in •:•>
«-< »-i
NO -*
NOO
_1 •
•< to
•- >
ex
»-
o
o
>n
•
o
in
«-*
o rv
O no
.923625 103i
.461575 6,
-i in
JO (VI
* 0
in in
oj in
00 O
_i •
•< (i
t- •>
09-4
1-
o
o
NO
o
00
*-!
O *•<
O •rt
O NO
v-l
on
o m
-a- m
rv o
* m
«-< vVI
OJ O
ID nt
d\ rv
OJ 0
rt
_j •
< ti
i— >•
»-
o
o
00
o
00
o in
o ru
O •-!
.613000 49i
.326625 61
00 (VI
in PI
-t OO
O 00
rt 00
rv o
_i •
"C CO
t- >
*-
o
o
0
«_!
W IO *H
o o o o o o
00 O 0 00
o in o AJ o o
0^ -4" «H ^o r**1 r^
OJ sQ O^ OQ -H
* oi tn m o o
^ (*) w^ cQ fv, fv.
CM «-* -
o o o
in o o
(vi in o
» • »
•«
»-
O
0
o
•
in
Jl
J![
y
^_i
•
OL
cn
*
*
Ul
^j-
or:
Of
_l
T"
03
OC
7 ai
— •»-
OC Ul
y ,
-IU4
-«or
>-c»
«1
O rt
D rv
0 OJ
NO a
«-t
in o
rt in
0 O
oo in
m o
OJ O
•*• •*
Q\ &
in o
ru o
_i •
•C CO
>- >
OD 4
^-
o
o
«*
•
o
469
-------
M
»—
III
o o
o o
o o
o o
o o
o o
o o
o o
oo
o o
o o
oo
o o
o o
o o
O p
o o
00
o o
o a
00
o o
o o
o o
o o
o o
o o
o o
o o
o o
o o
-DO
o o
o o oo oo
oo oo oo oo oo oo oo
>
ID
o o
o o
o o
00
o o
op
o o
o o
o o
00
o o
oo
00
00
oo
O O 00 00
OOOCOT
ooncoa
00
o m
o ci
o m
r>m
o to
m m
4- ni
«>4
OO
o o
oo
o o
0*4
•r*Q
ocvj
(VI
oro
om
o ro
o -n
0 00
mm
* ra
«-(
100000
183333
r-» -i
o in
0 0
o ?»
r*
o o
o o
0 0
0 O
o *o
CO ~*
3 ftl
v4
800000
763000
m »
«
o a
o o
0 0
o o
O 3
mm
o o
OOOC08
OOOC08
T-l *-l
O O
o o
o o
00
o o
00
o o
w
n
a.
a
-<
ID
o o
o o
o o
o o
•p o
o o
o o
a o
o o
o o
a o
o o
o o
o o
o o
o o
o o
o o
o a
o o
o o
o o
o o
oo
o o
oo
oo
oo
oo
o o oo o o
3 O
o -<
in OA ^< r^
r- (\j r^m
••> O T-4 O
ao * * U) * -o
cu o mo -* o
o ~o
eo in
o -o oo
00 «-t »-* »H
oou mco
no oo
xO xO O O
-o -a o o
•f
ac.
oo ao oo oo oo
oo oo oo oo oo,
OO 00 OO 00 OO
00 00 00 00 OO
I- >
-
^ >
a> •*
< o
i->
IE •<
O
O
CM
o
o
TO
O
O
*
o
o
tn
o
o
o
o
eo
o o
o in
o cu
o
o
in
o
o
o
o
o
o
o «-«
-* (U
r in
* or
I.I Ir
n
« ^ 01
o
o
o
o
fu
O
O
m
o
o
o
o
in
470
-------
aoocu ooiv n-n o
iii o o ui iv oo uiiv •&
i-incu f-en c«> i-rvo
x • • x - •> i • •
13 o o en o o us o o
UJ
_
•«ro en
nr • -
UJ O O
> t
ID
_
•< en «-i
nr • «•
III O O
_
_jin •*
•< ^ o
or • *
a. o o
>
o o
00
o o
o o
(Tl O
•o in
01 4>
|v O
o o
a o
o o
o o
o o
o o
o o
o o
o o
o o
o a
o o
oo
o o
o o
O O
o o
00
oo
-o -o
50 00
o o o o
o o oo
o o o o
o o
00
o o
o o
o o
00
o o
o o
~o -o
o o
m rv
* -o
o o
•oa\
a\ cu
M>CX1
_i • •
< o o
u »
o o
o o
o o
• »
o o
o o
b o
• •
o o
in m
• •
00
*
00 >C
• •!
o o
to o
j- m
\o en
in --i
» oo
to ir>
m on
°a cu
tO OO
• • »
o o
I
o a\
* .-I
00 CU
_l O O
<
u
ID CU 00
_i rv
OB t
O
O
< a
I- >
03 •«
o
o
oo
_
•< a
»- >
« •«
o
o
o
•
« <* CB-4
o
in
cu
o
o
in
•< C5
»- >
e •<
o
o
o
o
o
o
«-* *-l CU
471
-------
u»
>
ui
•
<
w
,&
H
p
O
t-4
M
•j
§
**J
{X|
J2J
H
«
CM
t
SJ
GJ
H
•8
z
rn
z
UI
o.
1*
(/I
UI
_l
m
t-
cn
X
_i
iy
»—
2-
U
_l
<
U
03
_J
•
f
UI
r>
n
IE
oa
ta
a.
t
"i
a
-t
03
_J
z
<
»
»-
Q.
•-
cn
<
_i
j-
»
or
7 Ul
a ui
JUJ
•< or
0 O
00
oo
oo
o o
00 (VI
CO (XI
OTO
in m
•HO
00
1C •
_i •
•<(•>
»->
05 •<
H-
O
0
v-«
0
o
oai
0 CU
O (XI
a m
0(VI
* o
**
om
«-i *-t
m o
_j •
•«< C3
l_ ^
o o
•< ra
t_ ^
m <
I-
0
o
^
•
o
o o o m
o o am
o o o m
o o o m
o o o m
* (VI J3 -t
iu 4- rx m
01 Cl
O 00 O 00
co in o in
o * r^ in
O O ^J O
«H
_l • J •
«r e> •< o
t- > >- >
C ^ OD ^
»- 1-
o o
o o
in «
• •
o o
O (XI
o m
o
OB •<
>-
O
a
•a
«
o
00
a o
o o
O 0
om
mm
m o
0 rt
«H
O(XI
(XI 00
0000
000
_l •
•< ta
>->
ID •<
t-
o
o
o
«
v-l
OOOCS8'
ooocov
in m
o m
(VI O
-* «-•
*C (*3
t— >
02 •«
»—
0
m
(XI
•
vi
'
O O
a o
0 0
o o
o .n
m ^
o in
in r*
a\ *
(XI ^
_» •
< (D
f— >
03 <
»-
0
O
in
•
«-i
o o
o o
O 0
0 0
o in
oa oa
-O (XI
«-«
o m
(vi o
oo oo
O «-t
V*
_l •
< a
H" >
in •»
t~
o
O
O
•
in
•
_l
i-
^
u
•>
a.
to
t
•
ui
ri
IT. 7
*~ ~
cr a:
in
U -I
<
o
a.
»
r
'3
^
a
•<
0
.j
2
<
(K
f-
O.
»-
(/)
•<
_l
j-
m
n
h
f-
UJ
UI
it:
•103000
•103000
O 0
o o
ci m
0 0
o o
.j •
< 13
fr*" >
(15 -4
1-
o
o
v-l
•
o
472
-------
O
m
m -o
Ol Ix
Olx
*•* IX
rx oO
m'
«-• T-l
Jt *
«H «-<
•o at
ix rx
oo rx
in «->
n> rvi
oin
•a <-t
-* «•«
oo m
CIO
(VIO
•000
in o
oo
* m
171719
723906
•O oo
m *•<
000000
000000
oo
0 0
o o
o o
00
oo
oo
oo
m (vi
* m
-0 00
rx o
* IX
l1--t
00 -*
00
ao
oo
00
o o
00
0 O
(3
T
I
33
u.
a
«•« oo *•« 0")
*-» oo in •-*
« rx n» oo
«i ** m o
(vi o * »<
oo oo
*o rx ' -*• »^ o o
•& m ^ ao a o
iVt 00 *Q VQ O O
.» o o •<> o 10
o^ »^ in *^ DQ 01
00 O O O O
tn ni
o 01
ni •« * in
«•« »< mm ••«
in tn o\ rx
m cu oo *
* * -H m
oo mo
|x
tn •«
•< 13 < (3
•
•< (3
f- >
ID t
03 •«
f-
• Ul
m •<
*• ••
_i •
< n
in -*
• •
j »
< '.3
•• •
_l •
< (9
O
O
m
o
o
o
o
o
o
in
\
o
o
o
o
00
o
o
o
O
in
m
O O
O O
in o
O
O
o
Cf
it
O
o
o
o
(VI
o
o
o
o
*
o
o
in
os
o
m
i: —»
< ^
at ae ui
473
-------
OOJO OOO OOO
uiooo ui o o ui o o
i-roo i—o o »-oo
OJ O O OOO
ui
••t
Ul
Ul
_J%OO _l O O _J 00 O
_io\ o _ioo _«mo
Or • » Or » » Or**
Ul O O Ul O O UiOO
ID
03
O O
o o
o o
00
00
oo
o o
oo
oo
vO O
CO O
OV O
00
00
00
to o
vC O
m o
01 o
r^ o
oo o
o o
oo
oo
a\ o
m o
m o
00 O
vC O
m o
oo
oo
oo
* o
oo o
>o o
-4 O
o> o
»-~ o
o o
o o
oo
in o
01 o
oo o
o o
f> O
o o
o o
o o
o o
o o
o o
o o
o o
a o
o o
o o
a o
o o
o o
o o
o o
o o
o o
oo
o o o o o o
OO 00
vo a
OJ O
O 3
OO
o o
o o
o o
o o
o o
o o
00
o o
o o
o o
00
00
O O O O
oo
o o o o
oo oo oo oo oo oo
oo oo oo oo oo oo
oo oa oo oo oo oo
o o
o o
oo
00
m o
m o
» o
cvi o
oo o
o o
-I • •
O
OO
oo
oo
oo
oo
oo
•o o
•i O
•
** o
* o
in o
03 «< O
_J 00 O
•a o
O O O O 00 O O
O O
>-
Of
001 O
en o o
•-•m o
00 >• O O
an
_»oo
o o
en oo
c
t- •-! O
r> in o
_ _
< tO «3 < O
I- > t- > »- >
B •+ 03 •+ •) •<
o
o
vO
o
o
CO
o
o
o
o -<
uo a
«J •< Z
ocas
W_l
o.
U _l • O .
-> mi —> en f
ui u ui i> •<
or i :> or 3 > or
u. gi _i ff> rt
K _i t- u OD
z1 or • 3*1 o* • or
ui »o uj •> in n
u e? u 09 2
or or o
Ul Ul ?
a. a. ui
474
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-670/2-74-022
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
5. REPORT DATE
July 1974; Issuing Date
COMPUTER MANAGEMENT OF A COMBINED SEWER SYSTEM
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Curtis P. Leiser
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Municipality of Metropolitan Seattle
410 West Harrison Street
Seattle, Washington 98119
10. PROGRAM ELEMENT NO.
1BB034/ROAP 21-ASZ/TASK 37
11. CONTRACT/GRANT NO.
11022-ELK
12. SPONSORING AGENCY NAME AND ADDRESS
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
OR&D Final - 12/66 to 11/73
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
Supplement to "Maximizing Storage in Combined Sewer Systems," GPO No. EPZ.10:11022-
KLK-12/71; NTIS No. PB-209 861
16. ABSTRACT At the conclusion of a ten-year construction program which affected much of
Seattle's combined sewer system, a computer-controlled "total systems management"
complex was proposed, funded and constructed. Computer augmented treatment and dis-
posal (CATAD) takes advantage of storage in the sewers to limit overflows, and selects
overflow points based on water quality data.
Since the control system began operating in 1971, receiving water quality, espe**
cially dissolved oxygen and coliform levels, has shown significant improvement; over-
flow volume has decreased by 50 to 60 percent during supervisory control and in
excess of 90 percent during three months of limited automatic control. Eight
pollution loading parameters were measured and found to be 68 percent less than
before advanced control techniques.
Capital costs totaled $2.6 million for the control system at the 36 remote sta-
tions, or $5.3 million, including construction of 15 gate-driven regulator stations.
Annual maintenance and operation costs totaled $270,000. The system is worth roughly
$40 to $245 million in equivalent sewer separation.
Work continues on a fully automatic optimizing model to add predicative capability
to program decisions so the system could maintain an 80 percent overflow reduction.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
*Computer programs, *Flow regulators, *Flow'
control, *0verflows, *Corabined sewers, Oper-
ating costs, *Remote control, *Remote sensing,
*Computer systems hardware, Flow measurement,
Mathematical models, Surface water runoff,
Runoff, Water pollution, Waste water, Waste
treatment, Sampling, Storage tanks, Tele-
metry, Monitors, Rainfall intensity, Flow
seaparation, Construction, Maintenance
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
*Wastewater storage, *Water pollu-
tion control, *Computer models,
*Computer control, *In-sewer stor-
age, Storm runoff, Water pollution
effects, Water pollution sources,
•Wastewater treatment, Seattle,
CATAD, Total systems management,
Systems control
9B
13B
18. DISTRIBUTION STATEMENT
19. SECURITY CLASS (ThisReport)'
UNCLASSIFIED
21. NO. OF PAGES
487
RELEASE TO PUBLIC
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
475
•&U.S. GOVERNMENT PRINTING OFFICE: 197*1-757-58V533't Region No. 5-M
-------
------- |