-------
LEVEL
& TRANSMITTER
FLUSHING GATES
:^
FLUSHING GATE
OPEN
g^s^pg^^
^v^^-^^^---^^'-^--^
FLUSHING GATE
CLOSED
FLUSHING GATE
CLOSED
18- FLUSHING GATE INSTALLATION
39
-------
15 -6 x 17 -9" box. Since sedinentation problems had occurred and
since sludge, sand and gravel deposits were blocking a portion of the
Conner Gravity barrels where in-system storage was desirable, the con-
cept ^ of using ^ dry weather flow to flush out a sewer following storage
was included an the project. The plan was to attempt to dislodge the
existing deposits and evaluate whether this concept could be used to
Keep this and other multi-barrel sewers clean in the future.
i
During normal dry weather flow, two of the gates would be lowered and
the flow forced through the remaining barrel. Initially the flushing
of the east and west barrels would be for short periods in order to
avoid overloading the grit chamber and primary sedimentation tanks at
the Wastewater Irea-bnent Plant. As the deposits were removed, it was
planned to flush each barrel routinely for a four-hour period. For
storms, however, all the gates would be raised to insure that upstream
riooding did not occur. As the system was dewatered and the level up-
stream of the gates was less than seven feet, two of the gates would be
lowered and the flushing procedure initiated.
SYSTEM SURVEILLANCE ,
i
Among the 76 overflow points are i 45 float-controlled regulators. Prior
to the monitorojig system one of the routine maintenance programs was to
check each regulator and backwater gate once a week (10, 18). Thus if
the regulator malfunctioned, it v?as possible to overflow sanitary sewage
to the rivers without knowing it ;for a period of days. The advent of
the monitoring system now would give the DMWD a 2M~hour surveillance
capability. _ If a backwater gate |is open, an overflow weir is being .
topped, or if a gate does not close properly, a maintenance crew could
^r^8111 OUt to rectify "the situation immediately. Specifically., the
DMWD would be able to discover gates which are jammed open, regulator
malfunctions and sewers clogged with debris solely through the surveil-
lance aspect of the overflow monitoring portion of the system. It is
felt that these overflow eliminations would be particularly significant
because these would be overflows bf strictly sanitary sewage and there-
fore grossly polluted (10, 19, 22).
40
-------
SECTION VI
: MONITORING AND REMOTE CONTROL EQUIPMENT
In order to implement the conceptual mode of operation, a real-time,
sensor based, data acquisition system was installed to provide data on
interceptor, combined sewer and pump station wet well levels, rainfall
intensity and total, and the status of overflow points. In addition,
a supervisory control system, independent of the data acquisition system,
was installed to provide for remote control of pump stations and regulators
and to display pump status, regulator operation, wet well levels and
scanner alarms. .
Sensor data is telemetered to the Systems Control Center in the Water
Board Building, which is located in downtown Detroit. .The telemetered
signals are interfaced with a digital computer which processes the data
and supervises output to three data logging.typewriters. Pre-programmed
alarms are printed on a teletype unit which also may be used for pro-
gramming. Pump station wet well levels are interfaced with both the
computer system and the supervisory control system to provide a backup
in the' event of failure of either system. Table M- lists the^equipment
that was installed under this demonstration project. The rain gages,
level sensors, proximity sensors, electrode sensors, audio tone trans-
mitters and receivers, and the computer and output devices comprise the
monitoring system. The remaining equipment listed is part of the remote
control system. . '
RAIN GAGES
The fourteen tipping bucket rain gages were installed at the locations
shown on Figure 19. Because these gages require both power and communi-
cation links, and because ready access was to be maintained, the gages
were located on municipal buildings rather than private buildings. •
Wherever possible, DMWD buildings were used. Ideally the gages should
be placed at ground level but, because of the desire to keep the gages
away from various types of outside interference, it was decided to_
place the instruments on the top of large, flat roofs. A typical on- ,
stallation is pictured in Figure 20. It was felt that placing the gages
on a flat roof would not subject them to unnatural air currents and that
the inaccuracy caused by rainfall being blown into the gage would not-
occur. However, finding a public building with a large, flat roof in a
desired area became difficult. From an examination of Figure 19 it is
evident that the gages could have had a better distribution (20).
The gage consists of a funnel, the large end of which is open to the en-
vironment, which collects rainfall and drains into a triangular shaped
bucket. There are two of these buckets placed so that as one of the
buckets fill, the weight of the collected rainfall causes the bucket to
tip, emptying it and placing the other bucket in position. The rainfall
41
-------
TABLE 4
MQNITORING AND REMOTE CONTROL EQUIPMENT
Number
Supplier
M-
118
30
38
200
1
3
1
13
18
6
5
Tipping Bucket Rain Gage
Level Sensors with
Level Transmitter ,
Proximity Sensors
with Amplifier
Electrode Sensors '
with Amplifier •
Audio Tone Transmitters
and Receivers ;
PDP8 Digital Computer
(4K Core and 32K Disc)
with Operator's Console
Data Logging Typewriters
with 30-inch Platens
Teletype
Sets of Transmitters,
Receivers, Contact Scanners
and Related Equipment for
Remote Control of Pump
Stations and Regulators
Mounted in Three Control
Panels
Strip Chart Recorders
Limitorque Operators
Gates (3 roller, 2 sluice)
Belfort Instrument Company
Baltimore, Maryland
Bristol Company
Glen Cove, New York
Minneapolis-Honeywell
B/W Controller Corp.
Birmingham, Michigan
Quindar^Electronics, Inc.
Springfield, New Jersey
Bristol Company
I.B.M. Corporation
Teletype Corporation
Skokie, Illinois
Quindar Electronics
Bristol Company
Philadelphia Gear Corp.
Rodney Hunt
42
-------
I I
|i«i|
I nil!
?*3?f?T??^lJ!S!
KKKKKKKKKK SES
i m
si
K'S
2
2
§
*T
»
a
43
-------
spilled from the first bucket drains out of the gage, eliminating the
need for an observer.
The gages are equipped with merctiry switches which momentarily interrupt
a continuously telemetered signal each time the bucket tips. Each signal
interruption is interpreted as OlOl inch of rainfall by the computer.
Figure 21 is a picture of the tipping bucket rain gage mechanism. The •- •
mercury switch is in the foreground of the opening and the actual tipping
bucket as located directly behind the mercury switch.
I
Because snowfall is common during the winter months, some of the gages
had electrical heating tapes wourid around the outside of the funnel as
shewn in Figure 22. It was believed that the tapes would develop enough
heat to cause the snowfall to melt and yield an accurate precipitation
measurement.
Several tipping bucket gages were placed next to existing weighing type
rain gages. The weighing gages are part of the Southeastern Michigan
Ram Gage Network. The close proximity of the two types of gages has
allowed an evaluation of the data obtained from the tipping bucket Biases
to be made.
LEVEL SENSORS
A total of 118 level sensors were installed at various points in. the in-
terceptors, the trunk line sewers! ten feet in diameter or larger, certain
critical smaller upstream sewers and in the wet. wells of all pump stations
as shewn in Figure 23. Figure 24' shows the detail of a level sensor cell.
Acting as a pressure cell, it consists of a 2-inch ID polyvinyl chloride
tube approximately 11 inches long, . This particular cell size was chosen
xn order to keep the volume of air in the cell at least ten times the
volume of air in the 1/4-inch tubing. This ratio was believed sufficient
to prevent wastewater from entering the tubing and possibly causing
plugging. Figure 24 indicates the level of water in the cell under
various conditions.
The cell is slanted about 15° downstream to help prevent fouling with
debris. A continuous section of 1/4-inch OD nylon tube is attached to
the cell and runs to the pedestal1box as shown on the installation
details in Figure 25. The-tubing is protected by a 3/4-inch metallic
conduit which runs frcm the top of the manhole to the pedestal base. At
locations where manholes were located in a street, it was necessary to
cut a slot approximately three indhes deep in the pavement. After in-
stallation of the conduit and tubing the slot was filled with either hot
asphalt or epoxy concrete to provide for a minimum of traffic interference.
Figure 26 shows a typical level sensor pedestal and slot cut in the pave-
ment. Both 120 VAC electrical service and leased communication line's
enter the pedestal through underground conduits from service drops on
utility poles as shown in Figure 27, if underground service is not available.
44
-------
Figure 20 -
TYPICAL RAIN GAGE
INSTALLATION
Figure 21 -
TIPPING BUCKET
MECHANISM
Figure 22 - HEATING TAPE INSTALLATION
45
-------
M UK.£t WIIT
Figur* 23.TELEMETERING DATA AND
46
-------
SOUTH MACOUB
ECTION
^-TYPICAL OVERFLOW
STATUS SENSOR
rjTIUJ CONTROL CENTER
LEGEND .
MONITORED POINT ON COMBINED SEWER ©
MONITORED POINT ON INTERCEPTOR
MONITORED RAIN GAUGE R-|
MONITORED OVERFLOW POINT. B-34
DETAIL AT CONNER PUWP STATION
REMOTE CONTROL LOCATIONS
47
-------
PVC CAP SOCKET TYPE
WITH £ TAPPED HOLE-i
14 NYLON TUBING
NYLON FITTING
SEWER FLOW
ID. PVC PIPE
MAJOR STORM EVENT
DRY WEATHER FLOW
ANCHOR SENSOR TO
CONCRETE WITH
EXPANSION BOLTS
•-U21/*
TYPICAL COMBINED SEWEF
NOTE!
SENSORS INSTALLED ON THE
INTERCEPTORS ARE SUBMERGED
AT ALL TIMES.
BOTTOM
Figure 24- DETAIL OF LEVEL SENSOR CELL
48
-------
FILL SLOT
WITH HOT
ASPHALT
&. CONDUIT
NYLON TUBING
ASPHALT SURFACE
CONCRETE BASE
POWER DISCONNECT
SWITCH AND TELEPHONE
TERMINAL BOX
Section Q-Q
POWER&TELEPHONE
CIRCUITS IN CONDUIT
LEVEL INDICATOR
& TRANSMITTER
, BURIED CABLE-
(CARRYING POWER
& TONE SIGNAL)
\: NYLON
'4 TUBING
NOTE:
THE PADLOCKS ON ALL LEVEL
SENSOR EQUIPMENT CABINETS
AND POWER DISCONNECT SWITCH
BOXES CAN BE OPENED WITH ONE
MASTER KEY.
25- LEVEL SENSOR INSTALLATION
49
-------
Figure 26 -
LEVEL SENSOR
PEDESTAL
Figure-27 - SERVICE DROP
Figure 28 -
INSIDE OF
PEDESTAL CABINET
Figure 29 - LEVEL TRANSMISSION
EQUIPMENT
50
-------
Referring to Figure 28 and 29, as the wastewater level rises, air pressure
in the level cell increases. This pressure increase is converted to a
mechanical motion through the bellows, located in the upper right corner
of the cabinet. Mechanical linkage attached to the bellows moves^the .
level indicator arm to provide for on-site 'level readout. In addition,
this linkage positions an arm on a cam which rotates at 12 rpm. The arm
linkage and cam are designed so that contact is always made for the first
second of the cycle. In the next three seconds contact is made for a
tine proportional to depth; i.e., zero seconds represent a zero depth
and three seconds represents 40-foot depth (full scale). In the last
second of the cycle no contact is made. The cam and cam motor and arm
are located directly behind the level scale in the center of the cabinet.
A tone, signal is transmitted whenever the cam and arm are in contact.
OVERFLOW.STATUS SENSORS
In order to study the frequency and duration of combined sewer overflows
status sensors were installed at all major overflow points. Thf3® sensors
are designed to transmit an interrupt signal to the data acquisition
system which the computer interprets as an overflow. Sketches of each of
these overflow points are shown in Figure 30. For a general location of
each of these points, refer to Figure 23. It can be seen that at
slightly more than half of the overflow points, dams are used to divert
dry weather flow into the. interceptor and prevent river water intrusion
into the system. At the remaining locations, backwater gates prevent
river water intrusion. The two types of sensors installed to detect
overflows were proximity sensors and electrode sensor's. Wherever possible,
proximity sensors were "installed to detect backwater gate openings. At
locations with dams, electrode type sensors were installed.
Proximity Sensors .
Proximity sensors are devices which detect the presence of a ferrous
metal. The sensors are mounted within the sewer outfall at the'backwater
' gate. A typical installation is detailed in Figure 31. , For a timber
gate, a block of 'ferrous metal is attached to the gate with the sensor
directly facing the metal- block. The sensitivity of the sensor is ad-
justable, but the normal distance between the sensor and the metal block
is approximately 1/2-inch. A continuous signal is maintained when the
backwater gate is in the closed prosition. As the backwater gate opens,
the distance between the sensor and the metal block increases to a point
at which the sensor no longer emits a signal. This in turn causes a relay
to open at the equipment pedestal. When the delay opens the tone signal
is no longer transmitted to the data acquisition center and a gate opening
is logged.
Electrode Sensors
At dams, as shown on Figure 32, a two-element electrode probe is anchored
upstream of the dam. The electrodes are positioned so that there is a
51
-------
Figor. 30-MONIUORED
52
-------
OVERFLOW POINTS
*H* DIVERSION WIER —yfl
8ACKWATCR CATE_
OVERFLOW SENSOR -
53
-------
Figure 30-MON1TOREI9 OVER-
54
-------
L
FlOW POINTS ( CONT.)
LEGEND
*
TT
DIVERSION WICR yM
OVERFLOW SENSOR \^y
55
-------
OPEN
TIMBER BACKWATER GATE
-"-CLEARANCE AS REQUIRED
PROXIMITY SENSOR
CONDUIT FITTING
FILLED WITH DUCT
SEAL
SIGNAL WIRE
^'CONDUIT TO JUNCTION
BGK
SENSOR BLOCK ^
TONE SIGNAL-
TRANSMITTER
DETAIL-PROXIMITY SENSOR
INSTALLATION
PROXIMITY
BACKWATER
HIGH WATER LEVE
OPENING
TO
REGULATOR
IJOW WATER LEVEL
u
SECTION-BACKWATER GATE CHAMBER
Figure 31-PROXIMITY TYPE STATUS SENSOR
56
-------
z
Ul
CO
Q
O
^
i—
u
fl)
3
CD
57
-------
clear space between the active
probes. The probes are protected
by a 4-inch ID casing with one of
the two electrodes located about
1/4-inch above the dam crest. When
an overflow occurs, both of the
probes are submerged and the circuit
is completed. The completed circuit
activates a relay which causes the
continuous signal transmission to
the data acquisition center to stop.
During the time for which no signal
is received, an overflow is logged.
DIGITAL COMPUTER AND INTERFACE
A total of eight tone transmitter
and receiver cabinets were installed.
Although'each cabinet may contain up
to 100 receivers or transmitters, in
general the cabinets were not filled
to capacity. Wherever possible, the
. equipment was. grouped — first ac-
cording to function and second ac-
; cording to geographic location of
; the monitored points. Sluice gates
•and regulator controls are in one
cabinet, level sensors and status
sensors require two cabinets and the
remaining five cabinets are used for pump station controls. Figure 33
shews a typical cabinet. Each horizontal bank contains ten receivers
and, since up td ten signals wikii different frequencies may be multi-
plexed on one leased line, maximum effort was made to have each hori-
zontal bank served by one phone; line. •
The computer and interface is shown in Figure 34. The present system
is designed to accept 475 sensor inputs conposed of 250 level sensor
signals of both 5-second and 15rsecond duration; 200 overflow sensor
signals; and 25 rain gage signals. Since only 200 sensor signals are
presently being received, the system may be expanded with only software
modifications.
The system software provides for a number of different functions to be
performed by the computer. The, software scales the analog-level sensor
signals and adds a stored constant to the level sensor reading to account
for cell height above invert; tallies pulse counts from rain gages and
computes the intensity of rainfall over the preceding 5-minute interval
and total rainfall; stores overflow status changes for a given time
interval as a check against false signals; .increments the real time clock
Figure 33 -
TONE RECEIVER
CABINET
58
-------
Figure 34 - COMPUTER AND INTERFACE
59
-------
for the system; generates alarm printouts from programmed set points;
and initiates three different printout cycles, an hourly cycle during
dry _ weather flow, a 15-minute cycle when rainfall is detected, and a
5-minute cycle if certain key points rise to alarm levels.
I/O HARDWARE
The operator console, shown on the left of Figure 35, allows the
operators to interrupt the executive routine and demand or suppress a
data logger printout at any time. This equipment may also be used to
monitor any specified sensor point so that. observations may be made of
the instantaneous changes taking place between data logger printouts. .
Normally it displays the real time of the computer clock.,
The teletypewriter shown on the;right side of Figure 36 is used primarily
as an output device to display programmed alarms such as high or low wet
well^elevations, rainfall detection or communications failure from a
specific monitor. This unit .also serves as a keyboard/paper tape input
device. Diagnostic routines and system software are stored on paper
tape to facilitate reloading of ithe computer if necessary.
Output from the data acquisition is logged on three typewriters with
30-inch long platens as shown in Figure 37. Output is in numeric
form only with the numeral "0" used to represent gate opening or dam
overflow and the numeral "1" to I signify that no overflow is occurring
at a given point. Level sensor ,data is printed to the nearest 0.1 foot
and rainfall total and intensity to the nearest 0.01 inch and 0.0.1 inch/
hour respectively.
SUPERVISORY CONTROL SYSTEM
The three control panels installed under this' project were designed to
match the existing water distribution system control panels. The three
leftmost panels in Figure 38 are used for control of pump stations, re-
gulators and gates. Figure 39 is a more detailed view of the center
panel containing pump station controls and a partial view of the left
panel containing gate controls.-
The upper section of the control boards contain miniature, 31-day, trend
recorders which display wet well elevations independent of the data ac-
quisition system. Mimic busses of the pump stations are on the middle
section of -the panel. Select-push indicating control switches are used
to start^and stop pumps and gate operators, as well as to display status
and confirmation of changes. Immediately below the sluice gate switches
and also to the right of the pump station mimic busses are the alarm in-
dicating lights. ' , ! .
60
-------
Figure 35 - OPERATOR CONSOLE
Figure 36 - ALARM TELETYPE
Figure 37 - DATA LOGGERS
61
-------
Figure 38 - CONTROL PANELS
I I I I 'I TT
''
tTtTTt
Figure 39 - DETAIL OF CONTROL PANEL
62
-------
PUMP STATIONS
Seven pump stations containing a total of ~39 pumps were converted from
local operation to remote operation. The pumps, remotely controlled,
range iTsize from 7 cfs to 500 cfs. Modifications^ the existing ^
stations for remote operation required minor electrical work to tie in
the necessary transmitters, receivers, relays and scanners.
The eight 500 cfs storm pumps at the Conner Station were not set up for
remote control under this project. The storm pumps are over forty years
old and have raised ijipellers that require pruning. In October 1971,
after .the eighteen month project evaluation period had been completed,
one of the Conner Station storm pumps was modified for remote operation.
The control scheme used proved successful and the remaining seven pumps
are now remotely controlled. The pumps at the Wastewater Treatment Plant
are monitored at the System Control Center, but are not remotely con-
trolled. Coordination of the operation is done by telephone.
REGULATOR MODIFICATIONS , • . ' _
In order to selectively load the interceptor system and to provide for
in-system storage, four major regulators at various locations in the
system were modified to provide for remote operation.
Baby Creek Regulator
The Baby Creek Regulator, Figure 40, used two float-controlled shutter
eates to regulate flow into the Oakwood Interceptor. Hand-operated 72- .
inches x 48-inches sluice gates were located in the stop log chamber
ahead of each shutter gate. These gates were used for diverting flow
during maintenance of the shutter gates.
' Previous studies had indicated that during dry weather flow, only one of
the two regulators was needed. Further, during storm events, -tfie regula-
tors caused the interceptor to surcharge, thus backing up two suburban
connections. Secondly, the float-to-gate linkage_chains_frequently broke,
because of excessive wear due to surge phenomena in the interceptor.
For these reasons, both shutter gates were chained open; the westerly
sluice gate was closed and the easterly sluice gate is remotely controlled.
The existing hand-operated sluice gate mechanism required 496 revolutions
to fully opin the gate from a closed position, .The hand wheel was removed
and the existing shaft and gearing was coupled to a new gearing system
and electric motor drive, as'shown in Figure 41. Manual_override controls
are provided as well as a hand crank in case of power failure.
63
-------
L
3;3 3.3 3 3 3 ] 1
64
-------
Figure 41 - MOTOR OPERATOR AT BABY CREEK
Hubbell-Southfield Regulator
Modifications to the Hubbell-Southfield regulator, Figure 42, were
similar to those made at Baby Creek. In this case, however, sluice
gates had to be fabricated since there were no existing gates. The
two Brown and Brown float-controlled regulators were chained to the
full open position. As in all the installations, on-site manual over-
ride to the remote-control equipment is provided, as well as a hand
crank. .
The motor operator in Figure 43 is located above a grating in the stop
log chamber and supported by two wide flange beams. Because this in-
stallation was not in a separate chamber, as in the case of the Baby
Creek Regulator, and would be subjected to gases and moisture, the gate
stem was greased and then enclosed in a plexiglass tube.
Figure 44 shows the control equipment cabinets. The two larger _ cabinets
contain the control equipment for each gate and the smaller cabinet con-
tains level sensor equipment as discussed previously. Figure 45_is a
view of the inside of one of the control cabinets. The tone equipment
necessary for remote operation is mounted in an eleven-module rotating
rack. Tone 'modules are plugged into this rack and may be removed
easily if maintenance is required.' All other remote control installations
are similarily designed.
65
-------
LEVEL SENSOR •-==
EQUIPMENT CABINET fLJ
SUPERVISORY
CONTROL EQUIPMENT
CABINETS
NEW 30X60 SLUICE GATES
EXISTING FLOAT
OPERATED REGULATORS
[NOT IN USE]
PLAN VIEW OF
HUBBELL-SOUTHFIELD
DIVERSION AND REGULATION
FACILITIES
POWER CONDUIT
EXISTING FLOAT AND
REGULATOR CHAMBER
EXISTING UPSTREAM
STOP LOG CHAMBER
LIGHT SWITCH
MANUAL OVERRIDE
CONTROLS
NOT
EXISTING FLOAT OPERATED
REGULATOR CHAINED TO
THE FULL OPEN POSITION
STEM COVER
NEW MOTOR OPERATOR
TRANSMISSION
SHAFT
HUBBELL-
SOUTHFIELD
SEWER
EXISTING DOWNSTREAM
STOP LOG SLOTS
•-EXISTING STOP LOG SLOTS
-NEW SPECIAL CASTING
Pll. ^
NEW 30 X 60 SLUICE GATE
SHUTTER
GATE
TO INTERCEPTOR
SECTION A-A
Figur*42-MODIFICATIONS AT THE HUBBELL-SOUTHFIELD REGULATOR
66
-------
Figure 43 -
MOTOR OPERATOR
AT HUBBELL-
SOUTHFIELD
Figure 44 - CONTROL EQUIPMENT
CABINETS
Figure 45 - DETAIL OF CONTROL
EQUIPMENT
67
-------
Warren-Pierson Regulator
Prior to construction of the southerly section of the Northwest Inter-
ceptor, a^pump station was located at Warren and Pierson. With the
construction of the deep interceptor in 1958, the pumps were removed
and a float-controlled, hydraulic cylinder-operated gate was installed.
The section of the Northwest Interceptor above the Warren-Pierson Regu-
lator contains no regulators arid flow is diverted into the combined
sewer by dams. With the increased urbanization of the upper Rouge River
Basin, the river crested at higher elevations, topped the dams and
flooded the interceptor. The backwater gates ara now being installed at
outfalls north of the Warren-Pierson Regulator to alleviate 'this problem.
To implement the concept of "Selective overflowing", this regulator was
modified for remote control. The existing hydraulic cylinder was re-
conditioned and remote control equipment installed. Since failure of
the hydraulic cylinder at this location would cause raw sewage to over-
flow into the Rouge River and leaky seals on the cylinder would allow
the gate_to drift downward during dry weather flow, with the resultant
possibility of an overflow, a concrete counterweight was cast and con-
nected with stainless steel rope as shown in Figure 46. This is designed
to equalize the load on the hydraulic cylinder.
Conner Forebay Regulator
In 1962, the original toggle type regulators were replaced with float-
controlled, hydraulic cylinder-operated gates. Since these were relatively
new, it was decided to remotely control the cylinders rather than the
electric sluice gate operators installed in 1927. Figure 47 shows the
modifications made to the .regulator. The regulator controlled by the
48-inch knife gate is of sufficient capacity to allow dry weather flow
into the interceptor. Thus, by! remotely controlling all three gates, a
backup system is provided if a gate fails to operate.
CONNER FLUSHING GATES
A 10T-0" x 7'-0" roller gate was installed in each of the three barrels
of the Conner Gravity Sewer at the transition from a 12'-0" x 16'-6"
two-barrel box to a 15'-9" x 17?-6" three-barrel box. This transition
section is located approximately 8500 feet upstream of the Conner Fore-
bay Regulator. The gates are being used to test the concept of using
dry weather flow to flush combined- sewers.
As shown in Figure 48, a chamber was constructed on the top of the sewer.
The chamber size (and thus the gate height) was limited by the fact that
the top of^the sewer was about 11.5 feet below the street surface. Each
gate is raised and lowered by means of a cable drum hoist with a Limitorque
electric operator. Since debris and sludge deposits were found in the
sewer, it was believed that the'gates might jam due to uneven seating.
68
-------
SUPERVISORY
CONTROL
EQUIPMENT
CABINET
LEVEL SENSOR
EQUIPMENT
CABINET
•BULKHEAD
EXISTING
HYDRAULIC
CYLINDER
COUNTER
WEIGHT
FORMER
PUMP STATION
WET WELL
60 X72
SLUICE
GATE
6-3 NORTHWEST
INTERCEPTOR
wsttmm&K?
Figure 46- MODIFICATIONS AT THE
WARREN - PIERSON REGULATOR
69
-------
70
-------
: SUPERVISORY CONTROL ST£M FOR
EQUIPMENT FOR REMOTE EMERGENCY
OPERATION IS LOCATED ON THE MANUAL STREET
SURFACE IN A METAL CABINET OPERATION—y SURFACE—^
STD. REGULATOR ",;. .''•' '.'..
COVERS ON THE
^EQUIPMENT
ACCESS
LIMIT
SWITCH
LIGHT
SWITCH
MOTOR
OPERATOR
CABLE DRUM
1=
10'-OX 7-0
ROLLER GATE
(3 INSTALLED)
^-MANUAL
OVERRIDE
CONTROL
STAINLESS
STEEL
MASTERS
LEDGE FOR
MAINTENANCE
OF UPSTREAM
FACE OF GATE
Ps^yVyXNlvi A r
.• . -o . . .
"--••••—v-.^
3BBL® I5-9"X I7:6
COMBINED SEWER
GUIDE JAMB —
Figure 48-FLUSHING GATE EQUIPMENT
71
-------
h * Y?* f13 d?8i«9ed with *e c^les on each end of
n flcal linkage w*s installed so that the Limitoraue
^autonatically reverse gate, travel if one of the two cables slacken
ed prior to the preset gate travel limits. slacken-
Figure U9 shows the hoist and Limtorque operator' with the roller gate
°n Fure 50 ;hows ^ ^^ ove^ide oortoaf 2
-
.
naintenance crews are delayed in opening the gates
Figure 49 - MOTOR OPERATOR AT CONNER
Figure 50 - OVERRIDE CONTROLS
72
-------
SECTION VII
SAMPLING • ...
Efforts to optimize' in-system storage, first flush interception and
selective overflowing would have little value if the pollutional con-
centration of each sewer were unknown. Although studies related to_
rainfall analysis, runoff characteristics, flow in sewers and location
and duration of combined sewer overflows are valuable, they would not
be complete without sampling (21). Eor this reason, a combined sewer
sampling program was initiated. Both a manual grab sample program and
an automatic sampling program were set up.
GRAB SAMPLING PROGRAM . .:
A two-quart grab sample was taken hourly each weekday between^: 00 a.m..
and 3:00 p.m. at a point on each combined sewer outfall immediately
upstream of the regulator or diversion chamber. Due to the diversity
of outfall locations, approximately 6 months were required to obtain
samples from all of the outfalls. Two additional sampling circuits
were completed during the project period. Between 25 and 35 samples
were taken at every outfall for each circuit. The
-------
74-
-------
75
-------
76
-------
, 77
-------
78
-------
79
-------
80
-------
81
-------
82
-------
Approximately 200 grab samples were taken at each of twelve sample
points along the Rouge River and tributary branches (24). The results
of the analysis are summarized in Figure 60. It was hoped that this
sampling program would be of aid in determining which areas of the
Rouge River were nost heavily polluted. With this knowledge, steps
could then be taken to alleviate part of the problem by modifying exist-
ing dams or regulators. However, as can be seen in the averages shown,
there is no area in which there appears to be a significant increase in
pollutants. .
Following storm events, the concentrations of, the pollutants measured
did increase significantly. However, as in the case of the average
values, there were no trends toward specific areas being more polluted
than others.
AUTOMATIC SAMPLING PROGRAM . .
Those familiar with sampling of combined sewer overflows have probably
found that nature rarely cooperates with sampling crews. Combined _-
sewer overflows can be expected to occur any time the sampling crew is
not at the site. This prompted an automatic-sampling program to be
set up in conjunction with the grab sanple program.
The desired mode of operation of the sampler was to locate the^sample
line downstream of dams or regulators so that only actual combined sewer
overflows would be sampled. Because samples would be taken only during
actual overflows, a vacuum-type suction pump was selected. _This type
of pump can run continuously without damage to itself even if there is
no flow to be sampled. The Megator sampler selected for^the project
takes a continuous sample and cycles every 1/2 hour to fill a new sample
bottle. The sampler has 48 - 20-ounce sample bottles.
Because of the cyclic nature of the sampler and the type of pump selected,
not only could combined sewer overflow samples be obtained but also the
time of the start of.the overflow and the duration could be established
by noting which sample bottles were filled.
Since this was HMD's first experience with an automatic sampler for
combined sewer overflows, it was decided to locate the sampler on a
combined sewer upstream of the overflow for test purposes. The West
Fjid Relief Sewer was chosen since it carried a small amount of dry
weather flow (approximately three feet deep in a 14 feet box) with the
majority of its capacity used for storm relief. It was thought that
this sewer would be fairly representative of an outlet sample point
with regard to the pollutants it would contain during storm events.
Two other factors were considered in the site selection. Since the
pump was limited to about 18 feet of lift, it was necessary to locate
the pump within the manhole.- At the location selected, the manhole
83
-------
o -
10
co E:
CM
— •*
(O
M-
CM
(O
0)
CO
o —
CO
CM
(VJ
CM
CO
oo |£}
00
CO
—
CO *O
IT)
-------
was built after construction of the box sewer and an access hole was
drilled through the crown. A ledge existed in the manhole, and the_
pump.was located on this ledge. Secondly, the manhole was located in
a park which allowed easy access without traffic interference.
The sampler was located in a small shed which is shown in Figure 61.
The shed was set over an open manhole and chained to the manhole steps
to prevent vandalism. The sampler is serviced by a two-man crew and the
service van is equipped with a hoist to allow for lowering the sampler
into underground locations. The van and hoist are shown in Figure 62
and 63, respectively. The sampler may be operated with 120-volt A.C. or
12-volt B.C. At the location shown in Figure 61, two 12-volt automotive
batteries wired in parrel were used to provide the necessary amperage.
The sampler and batteries are shown in Figure 6M-. Since the batteries
had to be charged daily,. the sampling van was also equipped with a
battery charger.
The automatic sampling program was very disappointing. As others have
discovered, debris (paper, rags, plastics from disposable diapers, etc.)
in the combined sewer tend to wrap around the sampling head and cause
blockage (25). Several months were spent on testing different types of
protective devices for the sampling head. It was found that a #10 tin
can seemed to afford the best protection while at the same time allowing
a representative sample to be obtained. Although the plugging problem
was not eliminated, it was considerably improved.
Once the plugging problem was lessend •, a second major problem, arose. An
attempt was made to, study the variation in pollutional load in the com-
bined s'ewer during storm, events. Figures 65, 66 and 67 are plots of
suspended solids concentration versus time. Above each bar graph^is a
graph of rainfall versus time from a nearby rain gage. As is obvious
from, the graphs, the sampler did not obtain representative samples of
suspended solids. The widening variations in suspended solids was found
in almost every day's sampling. Although it would be reasonable to ex-
pect variations in suspended solids, the wide variations from 0 mg/1 up
to 1000 mg/1 occurring at random time periods indicate that samples were
not representative.
The pump was • of the flow inducer type, in which the sample travels through
a 3/8-inch plastic tube to a distributor arm in the sampler. The dis-
tributor arm traverses a series of holes arranged in an arc pattern and
connected to the various sampling bottles. The pump is adjustable from
approximately 1/3 gph to 4 gph, depending on sampling time desired. For
HMD's application, it was set at the lower pump rate to provide for a
24-hour sampling period. It appears that this lower flow rate resulted
in an accumulation of solids in the sample line and their release into
the sample bottle as'a "slug". Daily flushing and cleaning, of sample
lines did not improve this problem.
85
-------
Figure 61 - SAMPLING SHED AND
VAN
Figure 62 - SAMPLING VAN
Figure 63 - VAN AND HOIST,
Figure 64 - AUTOMATIC SAMPLER
86
-------
WEST END RELIEF SEWER
SATURDAY, AUG. 22,1970
RAINFALL
(inches)
SUSPENDED SOLIDS
(mg./l.)
87.
-------
WEST END RELIEF SEWER
MONDAY, SEPT 14, 1970
T
iq
O'
J
_
-------
1 1
o
RAINFAL
(inches
,
)L
WEST END RELIEF SEWER
THURSDAY, SEPT. 17, 1970
iA
1908 >
V
1
L_
H
"=r
-1— .
; /
h
i
!
I
1
1
'}
'- /
r i ii i ~~l
§ § § 8 8 § §
° » y> •* m (M -
SUSPENDED SOLIDS
fmg./l.)
't
CM
(M
~ N
-8
- °° (/>
!3
(0 U
-- tt
o
z
-S 3
Q.
uj"E <
.N 2 g in
p 2 0
r^ •* »»
5
_o ^
2
- s
•o
•
" 1
IL
-•*
-CM
0
89
-------
Based on EMWD's experience, thei concept used by the Megator sampler
would provide an ideal collecting device for combined sewer overflows,
since it would not only sample the overflow but also give information
as to time of overflow. However, the means to transfer the combined
sewer sample to the sampling bottles is entirely inadequate. It is
recommended^that the sample line be a minimum of 1-1/2 inch ID, the
pump be a minimum of 25 gpm capacity and that a primary grinder be
installed on the sample line. A flow-through type -system with a take-
off for the sampling device is recommended. To sample combined sewer
overflows, the system would have to be equipped with the necessary
sensors to detect overflows and;to start and stop pumping. It would
also require on-site power for the pump and coirminutor. Although this
system would have limitations, as to location, it is felt that if a truly
representative sample is desired a large flow must be maintained. With
a flow-through system, the sampler could be located away from 'the manhole.
This would be desirable since hydrogen sulfide could damage the equip-
ment. This type of damage did occur to the sampler tested and it was
necessary to rebuild the sampler.
• 90
-------
SECTION VIII , . •
MONITORING AND REMDTE CONTROL EQUIPMENT EVALUATION .
The Concept of system nonitoring and remote control with regard to com-
bined sewers is relatively new. Industrial processes and water supply
systems have used nonitoring and remote control systems for sometime.
The technology of systems control equipment has increased accordingly
in these fields.
Unlike the above systems in which the process flow variations can be
predicted and the constitutents of the process flows, in general, are
well defined, combined sewer flow is highly variable and the flow may
contain any number of chemicals,, solids and gases. Thus, many ,of the
system monitoring devices available are not suitable for combined
sewer systems (26). Because the Detroit sewer system is so large and
complex, other monitoring devices such as sonic level sensors, flumes,
acoustic flow meters, etc., are limited in application due to their
high cost. .
The purpose of this section is to evaluate the system monitoring and
remote control equipment installed by DMWD. It^is hoped that this evalua-
tion may aid others who are contemplating a similar system.
RAIN GAGES
The main consideration of the installation of telemetering rain gages
was the economics of the device to be used. Weighing gages did not
offer this economy. Because of the large area to be covered in the
rain gage network, it was not practical to install weighing gages that
must be emptied manually. Consideration was given to automatic emptying
of gages, but it appeared that this would be quite costly. The only
practical method would have been to remotely control the emptying pro-
cedure since on-site timing devices might empty the device while a
rain event was occurring. Thus, economics dictated that tipping bucket
rain gages be installed.
The rain gages were installed primarily to aid in operation of the system.
They provide the operators with data relative to location and intensity
of rainfall as well as the total rainfall. By knowing approximately where
it was raining and with what intensity, the operators could take the actions
necessary to minimize combined sewer overflows. To this end, they have
served their purpose well.
It was hoped that, in addition to operational data, the rainfall data
could be used for analytical studies relative to runoff characteristics
and system response. Before this type of study could be done, it was*
necessary to analyze the accuracy of the telemetered tipping bucket
rainfall data. For this reason, three of the tipping bucket gages were
91
-------
located next to weighing gages which are part of the Southeast Michigan
rain gage network.
Approximately 100 rainfall and snowfall events were used in the analysis
of the data. It was concluded that the rain gages are not accurate
enough for analysis purposes. As in the case of other investigators (27),
data transmission was found to .be the most significant source of error.
The major problem encountered is in the method of data transmission. As
was_discussed^in Section VI, a i continuous tone signal is broken momen-
tarily each time the bucket tips. Unlike weighing gages which can be
equipped to produce continuous | analog signals, tipping bucket gages
produce pulse type signals. The computer interprets each signal inter-
ruption as 0.01 inch of rain. There.is no on-site pulse counter. This
has led to two problems. If a pulse is received, there is no means
currently^available to check whether a valid pulse' occurred or whether
transmission line noise was creating erroneous pulses. Secondly, if
the transmission line is down, there is no means available to store
information locally until it can be transmitted. Thus, without some type
of on-site pulse counter, it is not possible to provide for transmission
checks. It is estimated that the cost of data transmission mil triple
if accuracy to +_ 0.01 inch is desired. This would require extensive
software modifications as well ias additional on-site data transmission
devices.
A second source of error in tipping bucket data is in the measurement
of snowfall. Seme gages were equipped with 60 watt• heaters. However,
field observations as well as the data indicated that these heaters were
not sufficient to melt accumulated snow. Currently, all tipping bucket
gages are being equipped with 500 watt heating tape to maintain 45°F
temperature.
It is believed that, with the installation of the heaters and more sophisti-
cated data transmission devices, the tipping bucket gages will provide
data which is sufficiently accurate to be used for analysis (28). •
LEVEL SENSOES
Because of the large number of level sensors installed, 118, and the
planned future expansion to about 250 level sensors, it was realized
that the, installation of bubbler type level sensors would be highly
impractical. Instead, air bellows type sensors were designed and in-
stalled. These_have proven to be reliable and almost maintenance-free.
With the exception of wet wells; and several sewers that can surcharge,
it has been found that the bellows unit could be changed frcm 0-40 feet
full scale to 0-20 • feet full scale. This increase in sensitivity would
be desirous but does not appear to be justified at this time.
92
-------
It was decided to purge the gages bi-yearly with compressed air as
routine maintenance precautions. Some gages, however, have been in
operation for over two years without any maintenance and continue to
perform satisfactorily. To date, very few of the sensors have been
plugged with either debris,.oils or grease. Other pressure cell in-
stallations in the area using stainless steel have been blocked with
grease. It appears that PVC pipe is much less subject to grease ac-
cumulation than other materials.
The original'plan for sensor installation did not call for the use of
conduit to enclose the 1/M-inch nylon tubing but it was included at _ the
Environmental Protection Agency request. This has proven to_be a wise
decision since it allows for the replacement of the tubing with a mini-
mum of .such problems as traffic disruption. There have been a few loca-
tions where the conduit has buckled out of the slot in the pavement due
to heavy traffic loads. To eliminate this problem in later installations,
the conduit has been cut and a sleeve, designed to allow for local ex-
pansion and contraction has been placed approximately halfway between
the manhole and the curb. This additional change has allowed the neces-
sary movement required to keep the conduit from buckling because of
traffic loads.
It has been found that once the l/M~inch ; nylon tubing has been cut or
if a leak should develop, an effective air tight splice is difficult to
achieve. One tube was cut with a pavement breaker and attempts at splicing
failed. The current procedure requires that the entire length of tubing
from the level cell to the valve inside the pedestal cabinet be replaced.
With the large number of sensors and diversity of location throughout the
city and suburbs, it was thought that vandalism and traffic accidents
would be a problem. Sensors have been in operation for over two years
and, to date, only two have been lost due to vandalism and one due to
an auto accident. This fact has eliminated many anticipated problems.
Periodically, sensors are recalibrated. Figure 28 shows the level in-
dicator in the pedestal. Note the embossed tape. It contains the
sensor number, sensor height above invert and the depth to invert.
The recalibration procedure first calls for the injection of compressed
air into the 1/M—inch nylon tube and the connected cell in order to clean
out any foreign material which may be in the level cell assembly. Next
the depth of flow is measured and the height at which the cell is located
above the invert is subtracted from this measurement. The tone signal
corresponding to this difference is what should be transmitted to System
Control; A call is placed to the System Control Center where a readout
is requested on the monitor. • Adjustments are made to the air bleedoff
valve and to the linkage until the reading at System Control is identical
to the measured reading. Calibration when needed has been found to be a
relatively simple procedure.
93
-------
A number of variables were analyzed to determine their effect on the
accuracy of ^ the level sensors. These included: variations in. barometric
pressure, air and wastewater temperature and wastewater density; solu-
ability of air in water; gases produced in the wastewater; and transmission
line noise.
The level sensors are designed to measure gage pressure, not absolute
pressure. Since both the equipment cabinet and sewer manholes are not
sealed, any difference in barometric pressure acting on;the components
of the measuring system is due to the difference in elevation between
the wastewater surface and the bellows. However, this difference in
elevation has a negligible effect on the level measurement.
Seasonal variations in the wastewater temperature and the resulting
change in density affect readings by less than +_ 0,2%. This was based
on the assumption that the temperature of the wastewater varies between
40°F and 80°F and that the specific weight of the wastewater is equiva-
lent to that of water for the temperature range shown.
The^effect of air temperature on the level sensing equipment was evaluated
during an extremely cold period,in December, 1968. A regulator on one
of the sewers was manually adjusted to cause wastewater level changes.
Temperatures as low as -6°F caused no measurable effect on the sensor
reading or_data telemetering. All of the level sensor cabinets are
equipped^with heating elements to maintain the temperature above 45°F.
In addition, the air bellows are designed to maintain linearity over
wide temperature ranges.
The diffusion of the air in the' cell into the wastewater, or the releasing
of gases into the system from the wastewater affects the volume of air
in the level cell-bellows system. In addition, the coating of the in-
side of the cell with waste material will also affect the air volume.
As the volume of air in the cell decreases, the wastewater must rise
higher in the cell. Assuming that, in the worst case, the air volume
is halved, and that the air in the system obeys the perfect gas law,
pv = RT, the resulting error in 'level measurement will be less than
-0.05 feet. Increases in air volume will cause positive errors in
level measurement of the same magnitude as decreases.
Transmission line noise has not been found to be a major problem at the
present time (29). Signals are .generated using a cam follower contact. '
Existing sensors are designed for 5-second cycle time for data trans-
mission. Due to wear and motor [maintenance, it-is planned to install
15-second cycle on all new sensor points. Currently, the system prints
out a real time reading based on the latest 5-second interval. All other
previous readings are ignored. The level in the sewer is printed out to
the nearest 0.1 foot which is equivalent to a 7.5 millisecond signal
duration. Several sensors were installed at a height above the sewer in-
vert such that they rarely transmit a signal to indicate rising levels.
A pre-programmed constant is added to the level sensor reading by the
94
-------
computer to account for height of sensor above the invert. With the ex- •
ception of major storms, the data printed has never been higher than this
constant. If line noise were a problem, one would expect a reading to
be affected occasionally. This has not been the case.
With the data available at this time, and considering the variables which
can affect level sensor measurements, it appears that the level sensor
data is accurate to +_ 2% in the worst possible combination of the above.
In general, accuracy to +_ 1% would be expected.
STATUS SENSOES
Two types of status sensors are used to determine when combined sewer
overflows into the Detroit and Rouge Rivers are occurring. Wherever
possible, backwater gate proximity sensors are used. In locations where
this Was not possible, probe type dam overflow sensors are used.
One problem common to both types of sensor has been at locations where
there are multiple gates-or dams. It has been found that in several
locations, only one of several backwater gates has opened, and the open
status was not recorded due to the fact that the monitored gate did not
open. In order to monitor all gate openings, sensors will be_installed
in series on multiple gate and/or dam locations. Although this^will .
hot allow the monitoring of each gate individually, it will indicate
overflows occurring at a given location.
Proximity Sensors
Proximity sensors have yielded excellent data'regarding backwater gate
status. To date, there have been negligible maintenance .problems even
though many of the devices have been submerged continuously.
'.Two problems have been encountered, however. Wave action has yielded
false readings. • This 'has been eliminated where tandem gates exist and
the sensors were remounted on the upstream gates'. Line noise has _ also
given false .readings. Computer software is designed to neglect signal
changes that occur in 15-second intervals or less. "This has reduced
the majority of false readings. It has also been found that the trans-
mitter must be periodically checked to insure that full signal strength
is maintained. As the signal weakens, line noise tends to affect signals
to a greater degree.
Electrode Sensors
Several problems have been encountered with the electrode probe in-
stallations. Due to the. geometry of the sensor as shown in Figure 32,
there has been a tendency for debris to accumulate in the casing and
cause erroneous overflow indications. Several sewers which contain a
high amount of grease have caused considerable maintenance problems, as
95
-------
the grease tended to harden within the protective casing. Rags, plastic
sheets and paper have also been jfound to create problems by wrapping
around the electrodes. Although; the protective conduit was designed to
allow flushing, it has been found ineffective in some instances. Ice
bridges and condensation have occasionally caused false readings.
From experience with the electrode installations, all future installa-
tions will call for separate protective conduits for each electrode.
Minimum spacing of electrodes should be at least four inches so that the
larger debris cannot cause false readings. In addition, electrodes
should be easily adjustable in height. At certain dams, flashboards
have been installed to raise dam height. It has proved difficult to
adjust the electrodes.
COMPUTER SYSTEM HARDWARE AND SOFTWARE
In its initial conception, the data acquisition system was to be used
as an aid in the remote operation of the pumping stations and to acquire
data relative to the response of the combined sewer system during storm
events.^ It was not intended to process data in any manner other than
to provide hard copy output on the data loggers and to annunciate pro-
grammed alarm conditions via the teletypewriter.
For process control applications, a backup computer is usually in-
stalled. However, because of the complexity and size of the Detroit
combined sewer system, it was decided that before any computer control
of the system could be implemented, system response to storm events
would have to be well documented in order to define the appropriate
mathematical model, yarious studies on predicting system response and
the related problems inherent to such mathematical models require large
computer systems (27, 30, 31). Also, since the sensor based data ac-
quisition system was experimental in nature, it was not deemed advisable
to control the sewer system until sensor reliability could be determined.
For the above^reasons, a small computer designed for data acquisition
purposes was installed. As a backup to the computer data acquisition
system, ^scanners have been placed at all remotely operated pump stations
and sluice gates. The scanners monitor various equipment as to whether
the equipment is properly operating at each location. In addition, wet
well levels and selected upstream level sensor data are displayed on the
control panels independent of the computer via strip chart trend recorders.
This backup system has proved invaluable since the computer has been down
several times since its initial installation. Although some information
was lost, pump station operations; were not affected to any great: extent.
As with most computer installations, system startup was not without pro-
blems. Hardware problems include?} failure of transistors, loss of memory
due to variations of line voltage1 and calibrating the telemetered signal •
to correspond with the data logged. Once these problems were solved, data
output became quite dependable.
96
-------
All programming for -the computer must be done in an octal based machine
language. This type of programming is unfamiliar to many prograirmers, and
has hindered development of new programs. Also, since there is no fore-
ground-background capability or backup computer, it is not possible to
test program changes prior to implementation on the on-line system. Thus,
only minor programming changes have been made.
DATA LOGGING /
There are several limitations to the data logging system that have
created difficulties in the analysis of data and in operations. The
alarm teletype is quiet noisy when in operation. This has proven to
be very useful since it not only alerts the operators to alarm condi-
tions but also prints out the source of the alarm. The data logging
typewriters are also quite noisy when printing. However, it has been
found that the operators tend to suppress printouts of data to reduce
noise levels. Normally, the data is suppressed if a rain cycle U5-
minute printout interval) is initiated due to afalse ram indication
signal. It has also been found that operators tend to suppress projit-
out of data on the 5-minute or 15-minute cycle following the passage
of storm. Although the data is no longer needed for operations, it
would be useful for analysis in determining optimal system dewatering
procedures. The installation of magnetic storage media capable of re-
cording data on a 5-minute cycle for up to 24- hours after a storm, would
eliminate this problem.
Level Sensor Data . .
Figure 68 is a portion of the output from Data Logger No. l^and also a'
portion of the index sheets which is used as a key to identify sensor
points. The output shown .represents various levels in the. sewer system
and on the Rouge River. Two sewer level sensors, A-5 and 20, were out
• of service and river sensor, A-40, was jafrmed with debris. This parti-
cular data set was selected in order to show the various print cycles.
The print cycle was initially'at 15-minute intervals. This indicates
-that rain was being recorded but no alarm levels at any point in the
sewer system had been reached. The print cycle shifted ta5 ramutes
for one data set, indicating alarm levels. It then shifted back to a
15-minute cycle. Note that hourly cycles are always logged. The rain
stopped at about 6:30 a.m. and the logging shifted back to the normal
hourly print cycle.
;
Although the output may look somewhat confusing, it has been found that
experienced'operators have little trouble in finding the appropriate
sensor output. The data could be nuch more easily read if only one -
line was used for each real time printout and each point was identified
oh the logger platen. However, the amount of data printed would not
allow a one-line printout in the space available for data loggers.
97
-------
*
a:
UJ
£
1U
a.
X
UJ
a
Z
A-15
A-t*
A-13
A-12
A-l
A-2
A-3
A-4
A-5
A-6
A-7
A-e
A-40/A-9
A-39/A-II
A-38/A-IO
20 / A-33
SEC
MIN
HOUR
YR
DAY
MO
t*^ t^ t-» l*»" as Is* |>
£3 eg ei PI e* of •?*
o o o o o o p
so "* so as eo i/J oo
t£5 UJ 1/5 in SO* SO OO
p o o o -o o . o
o o • o o o o o
o o o o o o o
eo eo m m oo es eo
3 '8 8 8 S S 8*
£* »-* 00 00 OO OS O
O O O O O O r-H
oo "i1 m oo •* o oo
00 OS OS os O — 1 -4
O O O O <-i -H — <
co r-~ oo co oo •* -H
O O O. -H t-l ^H ^H
c*4 oo r-i i— i o ^l* eo
55 ON O O -M — T -4
O O O O O O O
8 8 8 8 8 8 8
o os I-H eNi ^ oo m
•• . • * • • • •
so in so o
co •* in o -H eo o
m in m so so so r-T
p o o o o o o
£££££££
« • • • « . . .
Os ON os Os OS OS Os
ex CM
-------
For. certain key sensors on the interceptor and near pimp stations, em-
bossed tape has been placed on the data logger-platens to aid operators.
CRT displays and high speed printers are being investigated as additional
output aids to the operator. Preliminary studies have indicated that the
most useful form of display would be CRT's. Operators have indicated that
CRT output should include a schematic diagram of the sewer plan and level
sensor locations with pump station location, if applicable, together with
alphanumeric level display. It has also been suggested that multicolor
CRT's be used.. Rising levels, falling levels,' and alarm levels could be
indicated by various colors. If data logging or alarm printing is re-
placed by CRT's, there is still a necessity for audio alarms to insure
that operators are alerted.
Rain Gage Data
Figure 69 shows the rain gage data output for the same day and time. Two
lines are required for output. The first line shows the intensity over
the preceding 5 minutes and the second line shows the total rainfall.
The index sheet is shown above the data. One problem in the software is
evidenced by this data. Printout is either 5, 15 or 60 minutes apart
depending on the print cycle. Rainfall intensity, however, is based on
a 5-minute cycle. Thus, the intensity shown in the output is for the
preceding 5 minutes only. Referring to the data, it can be seen that at
tines- during a 15-minute cycle, the intensity is 0.00 while total rain-
fall has increased. At other times, the intensity indicates that total
rainfall should be higher or lower than that shown if calculations are
based on the 15-minute cycle shown. Although this has not created any
operational problems, it would be more convenient if intensity was based
on the print cycle time rather than the preceding five minutes.
Overflow Status Sensor Data
The overflow sensor data is shown in Figure 70. The period selected
corresponds to the real time shown for the level sensor and rain gage
data. As in the case of level sensor and rain gage data, an _ index
sheet is used to identify the various sensor points. A portion of this
index is shown above the data. The numeral "1" in the data indicates
that no overflow is occurring while the numeral "0" is used to indicate
overflows.
Status sensor printout follows the same cycle as level sensor and rain
gage sensor printout. This can be seen by noting the printouts at ap-
proximately 5:30, 5:45, 5:50, 6:00, 6:15, 6:30 and 7:00 which correspond ' -
in time to the level sensor and rain gage printout times. There is ap-
proximately a 2-minute interval between the start time of print of level
sensor data and the start time of print of overflow sensor data. Thus,
the cyclic times shown in this data lag the times shown for level
sensors. The rain gage real time lags the level sensor time by approx- ^
imately half this amount. The lag times are due to the print time required
99
-------
o:
Ul
QC
UJ
Q.
Q
Z
88888888
o* •* •* ob oi ejv
-H o o» o -t o
§" O O O O O O
o o o o o o
•* vO •* 01 •*• •* \O O
O» O O» -H O)l —ICOi-IOI
888
§0 •*
O 01
8 8 8
o
O
o
-------
-71-NOT IN USE
-69-FRISBEE
-67-7 Ml W
-65-MC NICHOLS
-63-PURITAN STA
-61-LYNDON
70-PEMBROKE E
68-7 Ml E
66-GLENHURST
64-PURITAN ~
62-FENKELL
60-SCHOOLCRAFT ^.^ PARKWAY
58-LAHSER -57-BURT (GLENDALE)
— 56-PLYMOUTH 55_w CH| w
—54-W CHI' E
—52-WARREN
— SO-BABY CREEK'
— 48-CARBON
45-DEARBORN
41-DRAGOON
39- JUNCTION
37-FERDINAND
35-MC KINSTRY
•33-SWAN
31 -24 TH
29-18 TH
27-12 TH
25-CABACIER
23-GR1SWOLD
21 -BATES
19 -HASTINGS
17 -RIOPELLE
I6A-ORLEANS
— 14-DU BOIS
-53-TIREMAN
-51 -HUBBELL-S'FIELD
-49-FLORA
-46-PULASKI
-44-CARY
-42-SCHRODER
-40-CAMPBELL
-38-MORRELL
-36- SUMM IT - CLARK
-34-SCOTTEN
-32-W GRAND
-30-21 ST
—28-VERMONT
-26-11 TH
— 24- I ST HAMILTON
-22-WOODWARD
— 2O-RANDOLPH
— I8-RIVARD
— 16B-ORLEANS
— 15 - ST AUBIN
— 13 -CHENE
— 12-JOS CAMPAU
— IO-L1EB
— 8 -HELEN
6-FISCHER
£ I 4 -NOT IN USE
2-NOT IN USE
-II-ADAIR
- 9-MT ELLIOT
- 7-IROQUOIS
- 5-MC CLELLAN
- 3-CONNER
- I-FOX CREEK
OVERFLOW
INDEX
Figure 70 -OVERFLOW STATUS SI-NSOR DATA
101
-------
for each data set. • In addition to the standard cycle times, status
sensor data is also printed whenever a change in status is recorded. An
overflow start and an overflow stop are indicated on the data selected
to illustrate this additional printout.
The major problem associated with the logging of overflow status sensor
data has been "gate chatter." Gate chatter is perhaps a misnomer, but
in this report it is used to describe a condition in which the data
logger output indicates a series of gate openings and closings within a
short period of tame. This can be seen in the data set where four over-
flow starts and stops occurred within approximately one minute . This is
believed to be due to river wave! action, sewer surges or transmission
line noise or combinations thereof. The result is that a large number
of data sets have been logged.
To reduce the number of printouts , computer software was modified to
check overflow status at the beginning and end of a 15-second tiire in-
terval. If the status change signal is received, a status change is
logged. Software is being modified to check status continuously over a
two-minute interval . If the status remains the same, it will b5 logged.
-. Sat® cotter is occurring, the total tine of chatter will be stored
internally and printed out at the end of the day. .This should -reduce
the number of data ^ sets logged, while at the same time provide a summary
or total chatter time. The surcmary can then be used to check for the
cause of the chatter.
SUPERVISORS CONTROL SYSTEM
Pump Stations
was well experienced in the 'remote operation of pump stations. The
water distribution system has been remotely controlled since 1962 and
to date no major problems have arisen in that system. Several of the
suburban communities which have sewage pump stations expressed interest
in remote operation of their own facilities but were hesitant to do so
Various concerns were expressed such as station fires and inpel]er or
shaft failures which may go undetected. One community investigated the
possibility of installing closed circuit television but found the cost
B^S*1^'. T1^S WaS mainly due to the len^th of coaxal cable required.
With the monitoring equipment available, it is basically a question of
economics as to the degree of sophistication in equipment and decisions
as to what parameters are most critical.
Mpnitored pump station parameters at the EMWD Systems Control Center in-
clude transmission line failure, power failure, control channel failure
supervisory equipment failure and station alarm. These conditions are '
displayed on the control panel as shown in Section VI, Figures 38 g 39
The station alarm may include bearing cooling water loss , high motor and
transformer temperature, excessive motor vibration, low voltage, loss of
102
-------
station ventilation, DC power supply loss, combustible gas Detection, and
high wet well level. Although this alarm system does not include all the
-potential sources of concern, it does include the more important parameters.
An additional aid to station monitoring has been the strip chart recorders.
As mentioned in Section VI, these are used to record wet well levels. They
are also used by the operators as. an indication of pump.operations. The
rate of rise or fall in wet well elevation gives an indication of in-
creasing or decreasing flow. The effect on the rate of change of level
when a pump is turned on or off Is also known to some extent. Thus the
operators, through experience, know if a pump is not operating correctly.-
This was evidenced at one pump station where the impeller was -badly.-worn..
Although this pump was'rated at the same capacity as other pumps at the
station, it did not produce a comparable rate of change of wet well level.
The maintenance crews were dispatched to the station and determined that
the impeller would have to be replaced.
Sluice Gate Operators
As in the case of pump stations, EMWD had some experience with remote
operation of gates and valves in the water system. Both hydraulic and
electric gate operators were used• in the wastewater system. The choice
of whether to install hydraulic or electric systems depended on avail-
able space and existing equipment.
In the initial conception of operational procedures, it was believed that
all gates would be fully open or fully closed. For this reason, no gate
position indicators were installed. Gate status, either "open", closed ,
or "running" is indicated on the select-push indicating control switches.
It has been found desirable to operate gates at partially open positions.
'Since the total time of travel from fully open to fully closed is known,
the operators are able to set the,gates at intermediate positions.
When the desired mode of operation is for partial gate opening, the elec-
tric operators are irnch better.. Since travel time is slower, more ac-
curate setting can be obtained. The main advantage of the electric
operators is that once a gate is positioned, it will not drift, tfydraulic
cylinders, on the other hand, do tend to drift. This problem was of
particular concern at the Warren-Pierson Regulator where drifting of the •
gate would cause dry weather flow to overflow into the Rouge River. For
the partial operation of hydraulically operated control equipment, gate
position indicators would be highly desirable. Unlike electric operators
which can be equipped with rheostats to give gate position, hydraulic
cylinders must be equipped with some type of mechanical linkage. _DMWD
is currently investigating position indicators of both the electrical
contact switch type and Metritape position transducers. The effects of
the sewer environment on these two positions indicating devices will be
of primary concern.
103
-------
All of the gear boxes on .the electric operators ae equipped with manual
operators that may be used to raise or lower gates in the event of power
loss or motor failure. The hydraulic systems do not lend themselves to
1TLf-tyPe of raanual control. Unless a counterweight mechanism was in-
stalled, a mobile crane would have to be used to lift some of the large
gates in the event of cylinder failure. In the event of power failure
hand pumps may be used to actuate the cylinder.
At the present time, hydraulic cylinders are used only to control regu-
lators . If a major failure was to occur, the wastewater would overflow
into_the receiving streams. Although this is not desirable, it does
provide a safety factor in that the system will not be affected to the
extent of flooding. For in-system storage, it is believed that hydraulic
cylinders may not provide a sufficient degree of safety to warrant their
use. ^Electric operators can be equipped in such a way that a single
individual can manually operate the gate in an emergency. This is'not
true of the hydraulic cylinders. , In locations where the electric opera-
tors may be subjected to an explosive environment or to submergence, it
would be better to either bear the cost of sumbersible, explosion proof
operators or to build a separate equipment vault for the operators
rattier than install the intrinsically safe hydraulic cylinder with their
poor emergency operability.
Conner Flushing Gates
The effectiveness of the flushing: gates will be discussed elsewhere in
this report. However, one mechanical problem which may have limited
their effectiveness was noted on a recent inspection. It was found that
the gates were not closing completely. This was allowing significant
flow to pass under the gate which, of course, would reduce the flow to
the barrel that was being flushed. It was a simple matter of adjusting
the limit switches on the electric operators to resolve this problem.
A second problem did not become apparent until after several months of
operations. The flushing gates are equipped with a control system de-
signed to recycle_the closing procedure if the gate should become jammed.
une o± the essential elements of the system was a spring which h=ld an
armature tight against the cable as shown in Figure 48. If the cable
becomes slack before the gate seats, the spring connected armature would
tnp a lojnit switch causing the gate to reverse its direction of travel
It was found that the spring was not holding the armature tight .against
the cable and thus tripping the switch. The steel spring originally
installed was replaced with a stainless steel spring. However, again
after several months, the device malfunctioned. The linkage system
was replaced with a proximity sensor switch. No problems were encountered
after this change.
104
-------
SECTION IX
OPERATIONS EVALUATION
Detroit has long been interested in pollution control. Its Wastewater
Treatment Plant, constructed in 1939, was one of the first large muni-
cipal plants in the country. Naturally, in the interest of pollution
control, the DMM3 was anxious to iboplement the monitoring and remote
control program as soon as possible. This, however, made analysis dif-
ficult. Rather than operate the wastewater collection system during
storm events' in an identical manner to the way it was operated before
the implementation of the monitoring and remote control program, the
DMWD. -ijmiediately embarked upon its storm capture and flow routing pro-
cedure. 'Although this was good for the environment, it did not provide
any base data which was needed to determine the improvement on reducing
storm runoff overflows due to the new system. As could be expected, a
three-month study to gain base data by operating the system by the old,
fixed point methods of storm pumping produced one of the driest three-
month periods in weather bureau history. Information from this test
period is minimal indeed.
OVERFLOW REDUCTION
However, analysis has been done on the .improvements accomplished due to
an increased understanding of system capabilities. The assumptions were
made that: 1) a certain amount, of time was necessary for the operators
to familiarize themselves with the system control concept of operations;
2) determining the full potential of the system was a tnal-and-error
process — a time-consuming procedure; and 3) several months of field
investigation were necessary to test data reliability. Therefore the
data collected from the first six months of. operations, along with the .
data from the three-month test period, was compared^to the data collected
during the last six months of.the demonstration period.
Data used in analysis comprised the number of points at which an overflow
occurred, the total number of gallons storm pumped to the river, the_dura-
tion of the rain and the Equivalent Uniform Depth (E.U.D.) of the rain as
determined from the. telemetered rain gage data. Because the rate, of runoff
is highly dependent upon the intensity of, the rainstorm and the total
amount of rainfall, it is essential that only similar rainfall occurrences
be compared. As expected, the number of such occurrences over this brief
period is not large. Table 5 is a summary of the pertinent_data collected
and analyzed for the cases which adhere to the above criteria. The word
"no"'in the row labeled "Was System Used to Capacity?" reflects operations
during the first six months of operation and the three month test period.
The word "yes" reflects operations during the last six months of the
demonstration period.
105
-------
a
H
CO
CO
CD
CO
LO
CM
CO
CD
co
CM
LO
o
CO
H
H
w
LO
a
I
CO
LO
LO
•
00
CD
CD
CD
00
M
LO
>H
CD
CM
00
H
H
u
!
r
o o
0)
PL,
ra
I
106
-------
CASE I; The 7/30/71 storm was more intense and of greater E.U.D. than
the 4/21/70 storm. This would suggest a higher amount of runoff. Al-
though there was no storm wastewater pumpage for the earlier storm (com-
pared to the 6.7 million gallons pumped for the later storm),' there were
13 points at which overflow occurred. With the Detroit relief points
ranging up to two barrels of 14'- 6" x 17'-6" dimensions, it is better to
have a short duration pumpage than a number of longer duration overflows.
Although the 6.7 million gallon pumpage appears large, it represents less
than a four-minute, pumpage at the Freud Pumping Station.
CASE II includes' the only storm during the three-month test period. During
this May 19, 1971, storm, the pump stations were activated strictly be-
cause of high wet well levels. System pumpdown, system storage or flow
routing were not used. A comparison of the storms themselves show that
the test storm was slightly greater in E.U.D. and of a 1/2 hour shorter
duration. This would not account for the large difference in overflow
parameters. By using the storm anticipation and storage techniques that
the DMWD devised, the number of overflows were reduced from 29 to 18 (38-5}
and the storm wastewater pumpage went from 75.3 million gallons to zero.
This is a significant demonstration of the effectiveness of system moni-
toring and remote control.
CASE III is another indication of improvements gained through experience.
Although the storms dropped an aljnost equal anount of rain, the latter
storm was of much shorter duration. Thus it would be expected that there
would be more runoff and hence more overflows and pumpage during the
latter storm. However, the number of overflows were reduced and the
storm wastewater pumpage was 53 million gallons less. Both storms
dropped a substantial amount of rainfall and, even though overflows were
reduced a significant amount, there was still more storm wastewater pump-
age and a greater number of overflows than is desired.
Another indication of the successful efforts using the monitoring and
remote control program has been in the amount of rainfall completely
captured within the sewer system. In the first six months of the pro- _
gram, the largest rainfall which the DMWD was able to completely contain
(no overflows and no storm wastewater pumpage) had an E.U.D. of 0.07
inches over Detroit's 137.89 square miles. In the last six months of the
demonstration period, the DMWD was able to contain a rainfall of 0.14
inches E.U.D.
STORM WASTEWATER PUMPAGE REDUCTION
Figure 71 is a graphical depiction of the improvement in storm waste-
water pumpage to the river due to the monitoring and remote control
program. In 1968 before the project was initiated, storm pumps were
started if the wet well level rose beyond a certain fixed point as
previously discussed. It was believed that pumping to the river was _
less detrimental than running the risk of flooding portions of the city.
107
-------
k\\\\\\\\\\\\\\\\\\V
ill
g
is
STORM PUMPAGE (MILLION GALLONS)
108
-------
By 1970, the storm wastewater pimping stations' mode of operation was
changed to conform to the monitoring and remote control concept of
operation.
All rainfall in the study period was segmented into 1/4-inch intervals.
The number of rainfall events in each interval was tallied along with
the amount of storm wastewater pumped to the river as a result of the
rainfall. The average amount of storm wastewater pumpage which was
likely to occur for a rainfall of a given size was determined. The total
rainfall was the only criteria; storm intensity, moisture content of the
soil, temperature and other factors affecting the amount of runoff likely
to occur were not included in the analysis. By examining Figure 71, it
is obvious that there was a substantial reduction in the amount of storm
wastewater pumpage resulting from the project. However, because total
rainfall was the only criteria for rainfall grouping, the exact amount of
reduction could not be determined. Note that some of the storm wastewater
pumpage volumes for E.U.D.'s greater than 1.25 inches are not shown due
to lack of data.
A frequency distribution for the rainfall in 1968 and 1970 along with the
number of pumpage events and the average amount of storm wastewater pumped
to the river for each rain appears in Table 6. It is important to note
that before monitoring, storm wastewater pumpage occurred for every rain
in excess of 0.50 inch. After monitoring, the activation of the storm [
wastewater pumping stations was no longer a necessity for a rainstorm
of this magnitude, and on one occasion the DMWD was able to sustain a
1.35-inch rainfall event without storm wastewater pumpage.
IN-SYSTEM STORAGE VOLUMES
A sufficient amount of data, has been assembled to calculate the minimum.
volume of storm wastewater which was stored in conjunction with the moni- •
toring and remote control program. The actual total volume of storm
wastewater stored would be difficult to determine due to the fact that
storage occurs many miles upstream of any regulator structure which is
set for storm wastewater capture-. Thus, only the major portions of the
sewers near the outfalls were used in the volume determinations.
The data used for volume determination was a combination of pump station
records, sluice gate operation records, level cell data and sewer hydro-
graph-analysis . Some locations of large potential storage volumes were
used only for large storms and consequently did not have a correspon-
dingly large amount of captured storm wastewater. This is evident at the'
Fraud Pump Station. Other locations are of relatively small volume but
are always used in the storm capture procedure and therefore have a cor-
respondingly high amount of retained storm wastewater.
Table . 7 lists the amount of storm wastewater captured at each storage
location during the 18-month demonstration period. The total volume of
109
-------
o
o
o
CM
O
O
O
CD
O
O
00.
(N
zl-
H
CO
CM
o
CD
1
o
o
LO
o
o
o
O
o
o
CO
CM
CM
O
CM
CD
CM
00
LO
CO
H
o
CM
'•
o
LO
o
LO
CM
110
-------
TABLE 7
RUNOFF CAPTURED AT EACH STORAGE LOCATION
LOCATION
Oakwood P.S.
Freud P.S.
Woodmere P.S.
Blue Hill P.S.
Conner- Gravity
Hubbell-Southfield
Baby Creek
TOTAL
MILLION GALLONS STORED
167.5
749.8
236.0
590.2
1077.0
318.6 ,
2551.6
5690.7
storm wastewater captured by using the storage capacity already in the
system amounted to 5.7 billion gallons. Assuming an average of 280 mg/1
of suspended solids and 145 mg/1 of BOD (32) entering the treatment
plant during days..on which rainfall, occurred, DMWD was able to prevent
13 million pounds of suspended solids and 7 million pounds of BOD from
pouring untreated into the Rouge and Detroit Rivers.
OPERATOR EVALUATION
The DMWD had been operating much of its potable water distribution, system
by remote control for several years prior to the advent of the remote
control and date acquisition program for the wastewater collection
system. • In order to strengthen the continuity of a total systems control
capability, it was decided to place the electrical hardware and its
associated ductwork with the existing electrical control equipment for
the water system. • The combining of the two systems would reduce ex-
penses and streamline the operation.
At the inception of the monitoring and remote control program the DMWD ,
had as its system operators at least one operator experienced with the
remote control of the water system and one operator transferred from
the remote wastewater pump stations. The operators experienced in re-
mote operation understood the ijiportance of trends and the interdependence
of the different segments of the system. This was a completely different
concept from the fixed point, high wet well level alarm which was pre-
viously used. The operators experienced in the pump station operation
aided the remote control system with their knowledge of station capabil-
ities and the immediate impact of storm wastewater pumping. Together
the operators supplemented each other's knowledge so that they were able
to control the wastewater collection system independently.
ill
-------
Physical Layout of Control Center
Figure 72 shows the physical layout of the DMWD Systems Control Center.
As the drawing indicates, the control panels are displayed along one-
third of the perimeter of the control room. With the operators' desks
situated near the focal point of the panel configuration, it is possible
for each of the operators to watch both systems. The alarm printer is
placed so^that it is accessible jto each operator and the operator's con-
sole is directly behind the operators. Next to the operator's console is
a two-way field communications r|adio used to coordinate field operations.
The data loggers are situated at a considerable distance from the operators.
This was a result of the physical proximity to the computer and the desire
to keep the control center uncluttered by numerous electrical ducts. Be-
cause the monitoring and remote control system is largely a data acquisition
system, for most instances it was believed that this arrangement was suffi-
cient.
The control panels consist of the control switches for the pump stations
and associated status lights and the strip chart trend recorders. Because
the outside panels are at the visual limit of the operators, if the systems
were to be expanded (expansion for both systems is expected) the opera-
tors will no longer be able to have full view of the whole system fron
one point. The result will be that one of the operators will be placed
at some distance from the alarm printer and the operator's console.
The dataloggers provide the hard copy of the rainfall and the level cell
data. With the data logger being away from the operators, there has been
less than full utilization of existing data. When storm events occur,
the operators are not always able to 'spend the time away from the control
panel needed to study the hard copy data. Thus, it happens that rainfall
may subside or a rising water level in the sewer may crest long before the
operators realize it. ,
Operational Effectiveness of Equipment
The M7D^ has met with varying degrees of success with the equipment in-
stalled in the Systems Control Center. This is not to be construed as
any endorsement or condemnation of any equipment manufacturer, but rather,
an evaluation of DMWD's ability to utilize the equipment's capabilities.
As stated above, system control is a dual operation concerned with potable
water distribution and wastewater collection system operation. The water
distribution system is strictly a remote control facility centered around
control_of reservoir levels and set point modulation. The wastewater
collection system, operation uses1 remote control capabilities over sanitary
and storm pimps, regulators and sluice gates. It also has extensive data
acquisition capabilities which utilizes a small computer. The computer
installation has limited storage: and processing ability but is pre-
programmed to activate a teletype printer under certain conditions. It is
this last quality which has been most useful for operations.
112
-------
Ul
8
Ul
UJ
.113
-------
Present programming now calls for certain alarms to be displayed on the
teletype. These alarms occur for such reasons as high or low wet well
elevations at a pump station, high level readings in certain sewers,
initial rainfall occurrences and comnunications failure. Previously, the
possibility existed in which certain conditions could occur without the
operator's full conscious knowledge of the event, such as the wet well of
a sanitary pump station approaching a low level. This type of gradual
change occurring over two or three hours can be easily overlooked. With
the alarm system, this problem has been virtually eliminated frcm the
control of the wastewater section. In comparison, detecting a gradual
change in the potable water distribution system without the computer sur-
veillance still can be a problem. Although the operators continue to
keep a careful watch on the wastewater system, they are reassured that
a vigilant backup system is present.
As it often happens, the advent of rain is synchronous with the advent
of increased system control operator activity. With the simultaneous
needs of two systems, the operators can be pressed for time.' Hence the
alarm system becomes increasingly valuable during storm events by reducing
the pressure placed on the operators. This often gives the operators the
time necessary to analyze incoming data and act in accordance with their
function as a system control specialist.
The strip chart trend recorders are the next most successful item in system
control. Designed as a continuous recording device, they indicate not
only levels, but also the rate of change in level of a particular point.
This has given the operators the lability to determine how fast conditions
are changing in the system. If the operator observes a rise in the wet
well elevation tapering off, he is inclined to withhold any storm pumping
or> storage reduction procedure for as long as possible, in the hope that
it will prove unnecessary. If, on the other hand, the trend recorder
indicates an extremely rapid rise in wet well elevation, the operator may
initiate storm pumping procedures somewhat earlier than normal, rather
than flood portions of the City.
In addition to the strip chart recorders which indicate rate of change in
levels, the data logger is also programmed to display trend alarms on the
teletype. However3 when the trend alarms were programmed into 1iie com-
puter, there was little available data which could be used to set these
alarm limits and the values selected were too high. Analysis of system
response to major storm events has indicated that open channel gravity
waves and pressure surges occur during intense storms. Although these
waves can be detected at the wet wells by noting rapid level increases on
the strip charts, the operators 'could use more lead time in order to
effectively anticipate these waves. Upstream level data has given some
lead time but in certain cases the operators have not detected these up-
stream trends. From the data which has been collected, it appears that a
rate of rise in level of more than 0.5 feet per minute should be considered
an alarm condition on the major sewers.
114
-------
Some difficulty- has been encountered with the sluice gate equipment and
controls. The control switches for the sluice gates do not have a
position indicator for any position other than fully open or fully closed.
It often happens that the operators throttle down a sluice gate but do
not fully close the gate. This information is not displayed on the panel,
and it would have been useful to the operator if it had been displayed.
The operators have also experienced some difficulty with the long travel
time of the sluice gates. Presently some of the gates have as much as
a seven-minute travel time. If the operator wishes to operate the
sluice gate at partial flow, he must watch and time the gate operation
for three or four minutes. This is a continuous three or four minutes
of the operator's time which is tied up each time one of these gates
are activated. The DMWD does not believe that this much time should be
devoted to one gate. Either sluice gates with shorter travel times or
position indicators with supervisory computer positioning are required.
Dual System Operation
As previously stated, the Systems Control Center is a dual operation
controlling both water and wastewater systems. Dual operation also
has its associated good and bad characteristics.. Because of the nature
of potable water use, the water distribution system is constantly changing
.to meet varying demands. The operators are on a constant vigil watching
the water system and are alert and responsive at all times. If the
operators were to operate only the wastewater collection system, periods
of dry weather would be quite boring indeed, for the system practically
runs itself during dry weather. In addition, with rainfall.occurring
only 5 per cent of the time, using the same operators for the water
system operation is good economics.
At the advent of storm events however, the simultaneous control of both
the water distribution system and the wastewater collection system con-
flict with good operations. As rain threatens, the demand for potable
water either drops or is expected to drop and, therefore, the pumping
rates in the water system must be reduced. At the same time the operation
of the wastewater collection system must be prepared to cope with ex-'
pected and/or actual runoff. Under present procedures one of the operators
controls the water system and the other operator controls the wastewater
collection system.
Because of the telemetered rain data, the DMWD has also become the best
source of information concerning instantaneous rainfall for the metro
area. With 55 communities draining into the Detroit system, rainfall
means a barrage of information requests. The information;requests are
handled by the operators as best they can, in addition to their primary
task of controlling the water and wastewater systems. This arrangement
does not, at times, leave much opportunity for the operators to examine
incoming data displayed on the data loggers situated at some distance
from the operators. Even though the operators do have some "leisure" time
. 115
-------
during dry weather, the operations required during rainfall events have.
the operators approaching their functional limit. This could be partially
alleviated if the incoming data was in a more accessible position to the
operators. It is believed that a maximum of information (within limits)
should be presented before the operators without the expenditure of effort
on the operator's part. It must be remembered that supplying the water
needs of 3.5 million people while trying to capture a maximum amount of
storm runoff can become somewhat difficult.
Traveling Operators
It is well known that any electrical or mechanical equipment can, at any
given time, be counted on to malfunction. This simple fact of :iTidustrial
prowess has caused the EMWD to introduce three separate, 24-hour-a-day,
traveling operator crews. They inspect each water and wastewater pumping
station daily and are on call for any system malfunction. Although
maintaining these crews is expensive, it is an excellent safety factor
which the DMWD feels its customers deserve.
CONNERS FLUSHING OPERATION
The visual inspection and sample analysis of the deposits in the Conner
Gravity Sewer indicated that dry weather .flow could be used to flush the
sewer. The fact that the center barrel carried all dry weather flew and
was relatively free of deposits further reinforced the flushing concept.
For these reasons, remotely operated roller gates were installed at the
transition section as shown in Figure 18. The electric operators and
controls were discussed in detail in previous sections of this report.
Shortly after the flushing gate operation began, it was found that the ,
deposits in the east barrel immediately downstream of the transition
section were of sufficient'depth to prevent flushing of this barrel with-
out the wastewater topping the other two gates when they were in the
lowered position. The mode of operation was then changed to concentrate
flushing on the center and west barrels. The east flushing gate was left
in the open position so that the turbulence created at the transition
would undermine the deposits and thereby reduce their depth to allow
flushing to be performed in this barrel. Figure 73 shows the deposits at
the transition section along the easterly wall in the east barrel prior
to installation of the flushing gates. This may be compared to Figure 74
which shows the effectiveness of this operation after about two years.
The large sludge bank on the left of Figure 73 was reduced to a depth of
about one foot. A channel about 1.5 feet deep has been cut through the
deposits at the location behind the shovel shown in Figure 75. Note the
outcropping of debris shown in Figure 75 to the right of the shovel handle.
This same formation can be seen at the top of the sludge bank in Figure 74.
The west barrel, in which most of the flushing operation was concentrated,
demonstrated that the concept of using dry weather flow to flush multiple
116
-------
Figure 74 - E. BBL. AFTER
FLUSHING
Figure 73 - E. BBL. SLUDGE BANK
Figure .75 - E. BBL.
BEFORE
FLUSHING
Figure 76 - W. BBL. AFTER FLUSHING
117
-------
barrel sewers was both practical and effective. The sludge deposits
which were almost the same depth as those shown in the east barrel were
reduced to a depth of about two feet. The material remaining was very
coarse, consisting mostly of brickbats, gravel and concrete. It was
quite difficult to walk on the material due to its coarseness. A sludge
bank about five feet deep and 15 feet long extending about four feet out
from, the east wall of the barrel was the only remaining noticeable deposit.
This can be seen in Figure 76.
As was mentioned previously, the limit switches on the motor operators
were found to be slightly out of adjustment. This prevented the gates
from closing fully and thereby reduced the effectiveness of the; flushing
operation. When the center and west gates were closed, the flow under
these gates was such that the flow through the east barrel was not suf-
ficient to flush the barrel effectively. At the time this concLition was
noted, about 5.5 feet of head was required to top the sludge deposits in
the east barrel. The center and west barrels required only about two
feet or less head and thus gate leakage was not a major problem when these
barrels were flushed. For this reason, it is difficult to evaluate the
east barrel flushing operation.
Any attempt to capture storm wastewater necessitates the installation of
some type of retention facility with the inherent problem of sedimentation
(11, 12). In these retention basins, the sludge deposits must be removed
by either manual cleaning or by some system of backflushing. Both of '
these cleaning methods result in high operating costs. The successful
flushing of the Conner Gravity Sewer has proved that dry weather flow
can be used to remove sludge deposits. It has also indicated that in-
system storage of storm wastewater and the flushing of the sewer -fol-
lowing storage is both economical and practical. However, it has been
EMWD's experience that large, heavy debris does enter the sewer1 system
and provisions for periodic manual removal of such items should, be pro- .
vided.
Based on the experience gained, it appears that in-system storage and
on-line retention facilities are the most desirable methods of capturing
storm wastewater. In a combined sewer system similar to the Detroit
system, where many of the sewers: are on relatively flat grades, it may
be possible to add additional barrels next to existing barrels in con-
junction with upstream control devices in order to implement storage.
Using concepts similar to the Conner flushing operation, it would be
possible to store large volumes of storm wastewater and to flush these
sewers following dewatering. In, addition, existing relief sewers could
be modified to provide for storage during small storms and allow for
flushing if controlled connections were provided. Multiple, large volume
connections between major sewers' could also be constructed to provide
additional retention. In both of the methods, dry weather flow could be
routed to provide for flushing.
118
-------
SURVELLIANCE . • .
The 24-hour surveillance characteristics of the monitoring and remote
control program have been particularly valuable. This feature has emerged
as a "second maintenance crew." If a backwater gate does not close after
a rainstorm, a maintenance crew is immediately dispatched, possibly days
before the regular maintenance program would have caught the backwater
gate malfunction. Regulator malfunctions also have been discovered in a
similar fashion.
One of the surprises resulting from the surveillance aspect occurred
through unusual circumstances. As often happens in the Detroit area,
a rainfall of an inch or more occurs during the winter when the ground
is frozen and covered with ice. As expected, the runoff under these
conditions is very high. During one of these storms, the system over-
flowed into the Rouge River at most of the available relief locations.
However, despite extremely high sewer levels, three of the backwater
gates did not open. Upon investigation, it has found that approximately
12-inches of ice had formed on the Rouge River in such a manner that
the backwater gates could not possible open. Although overflowing into
the Rouge River is not desirable, neither is flooding portions of the
city. A number of the backwater gates along the Rouge River were sub-
sequently redesigned — solely because of information gained from the
monitoring and remote control system. This design change is shown in
Figure 77. As can be seen, the backwater gates are being located atop
the diversion dams. During the winter months, the river level will be
below the crest of the dam and ice formation-will not hinder gate openings,
DEWATERING
The DMWD also relies heavily on the monitoring and remote control program
in dewatering operations. As previously mentioned, many of the suburbs
whose sanitary sewage empties into the DMWD system have constructed storm
water retention basins as a method of preventing overflows from their
drainage districts. After storm events, it is necessary to dewater these
basins into the Detroit System. By using the information provided by
the monitoring and remote control programs. Systems Control Center can
determine the capacity available in the various sewers. The Systems
Control Center operators have thus been able to modulate the amount of
stored runoff entering the system from the suburban retention basins so
as to create the least detrimental effects within the system.- Without
the level sensors, this was impossible to accomplish, and stored runoff
entered the system on an "as available" basis. Thus the dangers of
having a system running at full capacity should rainfall once again
occur during the dewatering procedure have been reduced.
The suburban communities have been most -cooperative in coordination of
their dewatering activities with the Systems Control Center. However,
this has placed an additional burden on the Systems Control Center
119
-------
120
-------
operators since all coordination mist be done by telephone. In addition
to the requests for rainfall data, the operators must also furnish in-
formation relative to available system capacity and suburban dewatering
priorities.
Several modifications to the Systems Control Center could be made which
would significantly improve dewatering operations while at the same
time alleviating some operational difficulties. One of these would be
to graphically display the major retention basins and those sewers which
service each basin along with the basin and sewer levels. The graphic
display would enable operators to visualize both basin and downstreams
levels and to use this information to establish dewatering priorities
based on the needs of each suburb.
Another modification would be to monitor levels upstream of suburban con-
nections to the Detroit system. It has been found that spot storms have
occurred in particular suburban districts while bypassing the majority
of the service area. If the storm bypassed existing rain gages, the only
indication of the occurrence of the storm to the operators at the Systems
Control Center was rising levels in the upper reaches of the sewers within
the City of Detroit. If operators had advance warning of these storm
events by monitoring levels in various suburban systems, they could better
prepare the Detroit system to receive this additional flow. In addition,
several of the metered suburban .connections are adjustable. They are
set for contracted flow amounts and are designed to shut off flow into the
Detroit system when levels in the Detroit system rise above certain fixed
points. If, in addition to monitoring levels upstream of the connections,
the Systems Control Center is given the capability to override the auto-
matic on-site controls, it will-be possible to accept additional flow
through these connections during spot storms.
A third modification would be to provide data links from the data ac-
quisition system to the suburban coirmunities. Instead of calling the
Systems Control Center operators for rainfall data, available system
capacity and dewatering priorities, this information could be sent directly
by the computer. Data transmission could be blocked so that each com-
munity would receive only that information which is pertinent to its system.
SYSTEM SHUTDOWN
Figure 78 is a copy of the actual system shutdown procedure used to faci-
litate construction at the Wastewater Treatment Plant. The first stop-
page of flow is initiated at the suburban communities which have storage
capability and are tributary to the DMWD treatment plant. This is done
at midnight at the locations furthest from the plant. Starting at 1:00
AM. DMWD starts to store in the upper reaches of its own system be-
ginning at the Bluehill pump station. The Puritan and Conner pump
stations are the next facilities to be taken off line. Notice that once
the 53 million gallons of storage at the Conner System is filled the
121
-------
He: Control Center Operations 3/22/71
22 Midnight (21-22) - Operators on duty to call the following:
Chapaton Basin
Milk River Basin
Bed Run Basin
Confirm shutdown of pumps bring flow into Detroit System.
1:OO A.M. - Shut down Bluefaill - begin storage
1:3O A.M. - Shut down Conner Sanitary Station - close Forebay and
store, (if level rises to fast, close flushing jjates
to effect storage upstream)
Shut down Puritan Station
Levels in Conner System will rise and overflow into
Freud and Algonquin.
2:00 A.M. - Shut down Fairview Station
2:30 A.M. - Close Hubbell-Southfield gates
3:OO A.M. - Shut down Oakwood Station
kiOQ AM. - Close Baby Creek gates
Shut down pumps at Wastewater Plant
5:30 A.M. - Begin purge of chlorine lines
We are attempting to store Tna-g-i-mum flow within the System without flooding or
overflowing. First point back on line will probably be the gates at Hubbell-
Southfield when Sensor B-*H reaches 6* or B-51 opens. Baby Creek should be
opened if Sensor B-5O signals open. If Forebay overflows, start a pump at
Fairview and then open Forebay to interceptor. System may be placed back on
1,1 ne as levels require or at 11 A.M. Call basins and bring flow back on line.
Check periodically with operators at Plant (VI. 2-1900) and see how work is
progressing after 8 A.M. and see if system can be placed on line earlier than
11 A.M. If sensors nearest the Plant indicate overflows call Plant and have
one pump or more as required turned on to start lowering interceptor. Elant
operators will turn on pumps at the Plant when necessary and notify Central
Control. If any station or gate is opened, notify Plant and keep operator in-
formed that pumping will soon be required. Try to key start-up of system with
pump start-up at Plant. Hold if possible until may-jimim number of pumps are on at
the Plant.
If rainfall occurs, notify Plant to go on line and place system back on line.
Figure 7S-CONTROL CENTER OPERATIONS 5-22-71
122
-------
flow is routed to the 20 million gallon Trued system. From 2:00 AM to
4:00 AM, additional control points are closed and pump stations shut
down until the pumps at the treatment plant can be taken off line. Be-
cause the work to be done is on equipment near the end of the treatment
process, 1-1/2 hours is given for.the wastewater to flow through 'the •
plant before construction begins.
The primary factor involved in the sequencing of the facilities shut
down is the time lag from the facility to the treatment plant. By '
shutting down a control point after most of the flow from the preceding
upstream control point has passed, the system will be able to store a
maximum volume. The information following the shutdown procedure suggests
action to be taken by the operators for certain anticipated developments.
It has been found that the lowest flow occurs early Sunday morning. When
system shutdown coincides with the Sunday morning flow period, it has
been, possible to create a no-flow condition at the wastewater plant for
up to six hours. During this time all flow is stored internal to the,
system and no overflows occurred.
The DMWD also employs the system shutdown techniques in order to inspect
certain sewers. The letter in Figure 79 illustrates the procedure used
in conjunction with the inspection of the Oakwood Interceptor. By re-
ferring to Figure 79 it can be seen that storage is being implemented
in the eastern portion of the city while the western portion of the city,
which is serviced by the Oakwood Interceptor, is being drawn down. This'
differs from the first pumpdown procedure shown in which the system was
uniformly drawn down. The uniform drawdown procedure was used to allow
the treatment plant to .go off line for as along as possible without
concern about high sewer levels upstream, whereas the second procedure
was intended to keep the interceptor level as low as possible but for
only a short period of time. Because the portion of the interceptor
to be inspected is directly upstream of the treatment plant and because
there are no screening devices between the inspection site and the pump
intake, it was necessary to take the pumps off-line before the inspection
team could enter the sewer. For this reason, the wet well .elevation had
to be lowered as much as possible as referenced in the Figure 79. Once
the inspection team entered the interceptor the system was completely
shut down and system control was limited solely to a monitoring function.
The preceding examples ,of system shutdown illustrate how the treatment
plant may be shut down for short periods without causing overflows. The
first example illustrated a total system pumpdown, while the second ex-
ample illustrated a partial pumpdown of the western part of the system.
Other sewer inspections have been made" by following similar shutdown
procedures. The major difference is that the pumps .at the Wastewater
Plant are not taken off line. As an example, the Fairview Interceptor
Lift Station was taken off line for over eight hours while an inspection
was made of' the 9-foot diameter discharge conduit. The monitoring and
remote, control capability has. provided additional safety to individuals
making these inspections while at the same time providing a continuous
check on overflow status to insure that sanitary wastewater overflows
do not occur.
123-
-------
Re: Sunday Inspection of Oakwood Interceptor
Control Center Operations-Feb. 20, 1972
03ie following outline will govern the operations at th£ System Control Canter and
Wastewater Plant Pumping Station on Sunday, February 20, 1972. These steps
are necessary to permit the physical inspection of the Oakwood Interceptor
from Carbon Avenue Into the Wastewater Plant site.
Operators on duty to call the following stations and confirm shutdown of pumps
bringing flow into Detroit:
3:00 A.M. - Chapaton Basin
- Milk River Basin
- Red Run Basin
- Call Wastewater Plant to start system pumpdown
, - Shut down Bluehill — begin storage
l4-:00 A.M. - Shut down Conner Sanitary Station
- Close Conner Forebay Regulator gates (if additional storage is
required, lower ail flushing gates and store upstream.)
- Shut down Puritan •
NOTE: Levels in Conner System will rise and overflow into the
Freud pumped system.
- Check Wastewater Plant — request additional pumps if wet well is
above elevation 72.0.
ty:30 A.M. - Shut down Fairview
- Check with Wastewater Plant — 1 Hr. to shutdown
Request additional pumps if wet well is .above elevation 71.0.
5:00 A.M. - Close Hubbell-Southfield gates — begin storage. Open gates at
first overflow indication (B-51) just enough to contain flow within
the system.
- Call Dearborn —- confirm shutdown
5:30 A.M. - Close Baby Creek gates
5:30 to
6:00 A.M.
6:00 A.M. -
Shut down pumps at Wastewater Plant when call is received from
System Control. Wet well to be at Elevation 69.0. Restart pumps
only after call is received from System Control.
Inspection team will enter the interceptor at Carbon Avenua and
exit at the Plant after checking the area where the PC-215 tunnel
crosses under interceptor.
- After all-clear is received from inspection team, start up system.
'Call suburbs to end storage. (All suburban areas have been requested
to hold a minimum of four hours•)
In the event of rain, Inspection shall be cancelled and rescheduled for,Sunday,
February 26, 1972. Call all suburbs in case of cancellation.
Figur. 7*-CONTROL CENTER OPERATIONS 2-20-72
124
-------
SECTION X
DATA UTILIZATION
DATA REDUCTION
As a result of the Monitoring and Remote Control Project, a voluminous
amount of data was collected. As previously discussed, the data output
is of hard copy form and the data sheets measuring 28-inches by 22-inches
with up to 70 different items of data per line. For a one-month period,
approximately 75 data sheets of sewer levels, rainfall and overflow
information is generated. The data on sewer levels was to be used for
the engineering purposes of studying the effects of rainfall on certain
sewer levels and determining the amount which the system is utilized.
It was believed that the information would not be of significant value to.
anyone not directly involved in the project. However, both the rainfall,
data and the overflow data could be of interest to various individuals.
Becausexthe raw data is difficult to understand and reproduce, the data
was firsts, reduced by'hand'into an acceptable form and them processed
by an off-line computer for further refinement. The results of one
month's data constitutes Appendix B of this report.
The first section of the monthly report consists of the data on com-
bined sewer overflows. Page one is a graphic summary for the month
shewing the days on which overflows occurred, the daily and monthly over-
flow totals, the total number of times an overflow occurred at a specific
point, the number of storm wastewater pumpages and the Equivalent Uniform
Depth of rainfall if any occurred for a particular day. The rest of the '
overflow report is an expanded listing of the date, the number of over-
flows for the day, type of sensor at the point of overflow, sewer location
and size, the duration of overflow at each location and additional pertinent
remarks concerning overflow information. 'On the last page of the overflow
section appears a brief summary for the month and a listing of any sensor
which was not in operation for any time during the month.
The second section of the monthly report is a display of the reduced data
gathered from the telemetered rain gages. The amount of rainfall at each
gage along with the produce of the rainfall amount times the Thiessen -•
Polygon weighing factor is listed for each day of the month. The summation
is listed and the area used in the E.U.D. computations. On certain occa-
sions, a particular gage may be experiencing conmunication difficulties
and the data is therefore not used in the E.U.D. computations. It may
also be desirable to limit attention to just one or two rain gate locations
in order to study spot rain cells. The last page lists the. rain gage
location and the corresponding area of the Thiessen Polygon in square miles.
It is believed that the monthly report supplies a sufficient amount'of in-
formation to interested individuals without being too overburdened with
details. As it was, there were approximately 220 pages generated in
125
-------
preparing the monthly report for'the 18 months of study. This was,
however, an improvement over the approximately 14-00 pages of 1iie over-
sized output. It is hoped that the single representative monthly report
supplied in Appendix B will adequately convey the information output.
The raw data is available for those interested in utilizing it for model
studies or other applications.
OVERFLOW INFORMATION
Although the monthly synopsis is useful in obtaining an overall view of
the system performance, engineering analysis demands a much closer examina-
tion of the 'many pages of hard copy data. From this data total rainfall,
storm intensities, runoff rates, total runoff, sewer hydrographs and over-
flow duration can be calculated; points of needed relief and the amount
of sewer utilization can be determined. All of this information is ^ of
significant importance when planning which methods of reducing combined
sewer overflows lend themselves to a particular wastewater collection
system.
Table 8 is a list of the various overflow points from, the wastewater
collection system into the Detroit and Rouge Rivers. From the hard copy
data sheets, the duration of each overflow was determined for each rain-
storm. The overflow durations for each point were summed and divided
by the duration of the study period to obtain the percentage of the time
for which overflows occurred.
Note that the Detroit system as a whole overflowed an average of 1.29% of
the time or 112.8 hours per year per outfall. This figure is a. direct
average of the numbers found'in the right-hand column labeled "% time •
overflow". Some of the overflow points were never used while others seemed
to overflow at the slightest amount of rainfall. Although there could be
no direct computation as to the volume of storm wastewater discharged at
the various outfalls, the information does indicate which portions of
the wastewater collection system are in the greatest need of overflow
abatement measures.
These overflow computations will allow the DMWD to plan future construc-
tion projects which will alleviate those sewers which present the greatest
overflow problem. These decisions would be based on the sewer size, the
duration of overflow and sampling.
UTILIZATION OF EXISTING FACILITIES '. .
As indicated, the most chronic overflow point is at the Conant-Mt. Elliott
regulator. Because the sewer is large, 16'-3" ID, and possesses a cor-
responding flow volume, it is believed that a control facility operated
from the system control center would result in a substantial reduction in
overflows and therefore be of significant value.
126
-------
TABLE 8
SUMMATION OF OVERFLOW DATA
SEN.
B-01
B-02
B-03
B-04
B-05
B-06
B-07
B-08
B-09
B-10
B-ll
B-12
B-13
B-14
B-15
B16A
B16B
B-17
B-18
B-19
B-20
B-21
B-22
B-23
B-24
B-25
B-26
B-27
B-28
B-29
B-30
B-31
B-32
B-33
B-34
B-35
B-36
B-37
B-38
B-39
B-40
TYPE
SEN.
P
P
E
P
E
E
E
E
P
P
E
E
E
E
E
E
P
E
P
P
E
P
P
E
E
E
E
E
P
E
P
E
E
E
P
E
P
P
P
SEWER
DESIGNATION
FOX CREEK
NOT USED
CONNER GRAVITY
NOT USED
MC CLELLAN-CADILLAC
FISCHER
IROQUOIS-V DK-CRANE
HELEN PLUS FOUR
MT ELLIOTT (MELDRUM)
CONANT-MT.ELT.(LEIB)
ADAIR
JOS CAMPAU
CHENE
DUBOIS
ST. AUBIN
ORLEANS
ORLEANS RELIEF
RIOPELLE
RIVARD
HASTINGS
RANDOLPH
BATES (BRUSH)
WOODWARD
GRISWOLD
FIRST-HAMILTON
CABACIER
ELEVENTH ST.
TWELFTH ST.
VERMONT
EIGHTEENTH
TWENTYFIRST
TWENTHFOURTH
WEST GRAND BLVD.
SWAIN
SCOTTEN
MC KINSTRY
CLARK
FERDINAND
MORRELL
JUNCTION(CALVARY)
CAMPBELL-MILT-JUNC
NO.
BBLS.
1
2
1
3
3
1
4
2
1
•3
1
3
1
1
1
2
1
1
1
3
1
2
2
1
2
1
2
1
1
1
1
1
2
2
3
2
1
1
3
SEWER
SIZE
15-6 X 18-0
17-6 X 22-11
6-3
10-6 X 11-6
6-0
9-0
5-0
11-0
5-0
6-8 X 8-8
4-0 X 5-3
5-0
3-6 X 4-8
3-0
5-0 X 7-0
3-6
5-3 X 6-0
4-0 X 3-6
8-0 X 5-0
8-6 X 6-4
8-0 / 9-0
7-0
10-0 X 10-6
4-8 X 3-8
5-0
5-2
4-3
7-0
4-6 X 6-0
8-0
3-0
3-0
4-8
4-6
8-8 X 7-6
4-6
9-0
13-0
6-6
% TIME
OVERFLOW
0.00
0.16
1.40
,4.29
0.34
2.54
6.12
7.66
2.70
0.00
1.15
1.08
0.21
0.00
0.25
0.44
0.51
0.46
0.18
4.69
1.83
4.00
2.73
0.37
0.22
0.28
0.45
0.00
0.02
0.91
0.01 ,
0.01
2.25
0.15
1.38
1.40
0.91
1.76
0.66
127
-------
TABLE 8 (Cont.)
SUMMATION OF OVERFLOW DATA
SEN.
B-41
B-42
B-43
B-44
B-45
B-46
B-47
B-48
B-49
B-50
B-51
B-52
B-53
B-54
B-55
B-56
B-57
B-58
B-59
B-60
B-61
B-62
B-63
B-64
B-65
B-66
B-67
B-68
B-69
B-70
B-71
TYPE
SEN.
E
E
E
E
P
P
P
P
P
P
P
P
P
E
E
P
P
E
E
E
E
P
E
P
P
E
E
E
E
SEWER
DESIGNATION
DRAGOON(LIVERNOIS)
SCHROEDER-WATERMAN
SOLVAY
GARY
DEARBORN (SLOAN)
PULASKI- PORTLAND
NOT USED
CARBON
FLORA
BABY CREEK
HUBBELL-SOUTHFIELD
WARREN
TIREMAN-JOY
W.CHICAGO-PLYMOUTH
W.CHI(W OF ROUGE)
PLYMOUTH (W OF ROUGE)
GLENDALE(S'CFT REL)
DOLSON(SCHOOLCRAFT)
WEST PARKWAY
SCHOOLCRAFT(WEST)
LYNDON
FENKELL
PURITAN STATION
PURITAN (E OF ROUGE)
MC NICHOLS
GLENHURST
7 MILE(W OF ROUGE)
7 MILE(E OF ROUGE)
FRISBEE
PEMBROKE
NOT USED
NO.
BBLS.
1
1
1
1
1
1
1
1
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
, 1
SEWER
SIZE
9-3 X 10-6
6-0 / 6-10
2-0
3-4 X 5-3
5-9
5-0 / 6-6
3-4 X 5-0
3-4 X 5-3
14-6 X 17-6
14-6 X 12-0
7-6
13-9
14-0
4-6
4-0
13-0
10-6
6-6
4-6
12-0
2-0
10-3
8-0
14-0 X 14-0
1-9 / 1-0
9-3 ,
8-6
8-3
13-0
% TIME
OVERFLOW
0,01
0.00
1.39
0.46
0.00
0.00
0.43
0.06
1,97
1.37
1,24
0,82
0.09
4.48
0.70
0.59
0.06
0.00
0.00
5.08
0.51
5.70
1.78
1.57
0.15
1.25
0.00
1.74
1.69
NOTE:
1. P = Proximity Sensor @ Backwater Gate
2. E = Electrode Probe @ Weir
3. Average % Time Overflow of System = 1.29
128
-------
Another possibility based on the overflow data is revealed from an examina-
tion of level sensor 41 on the Livemois Relief Sewer. The Livernois
Relief Sewer (Figure 80) was constructed in 1950 as a storm relief for
the upper portion of the Baby Creek district. Flew into this sewer was
designed to enter through high level, porthole connections from the
17'-9" x 13*-5" Wetherby arch. The sewer is 10'-6" in diameter and has
a length of 21,879 feet with a grade of only 0.05% for most of its length.
Built as a relief, it is below most sewers that it crosses and has very
few interconnections with any of the other sewers in the area. Overflow
data indicated that the Livemois Relief Sewer overflowed only 0.01% of
the time — far below the average for the system as a whole. Before the
monitoring and remote control program, it was not known how often or to
what extent the relief sewer was used. With the data obtained, it was
found that there was seldom any .storm flow in the sewer. Because of its
relative depth, its flat grade, its length and the few interconnections
with other parts of the wastewater collection system, it is believed
that the Ldvernois Relief Sewer could serve as a very successful retention
facility. If the present high level connections to the Wetherby sewer
were modified to a remotely controlled, low level connection, the ability
to route both.storm and dry weather flow to the sewer would be present.
Thus the DMWD would be able to route storm wastewater flow into the
relief sewer for storage and route dry weather flow to the sewer in order
to flush out sedimentation which occurred as a result of the storage.
Not only can this redesign add additional storage and flow routing capa-
bility at the Livernois Relief Sewer, but any relief sewer with high
level connections can be altered by the basic design of Figure 81 to add
storage and flow routing to many locations in the City. Other sionilar
locations are presently being investigated.
POSSIBLE ADDITIONS TO EXISTING SYSTEM
In Section IV, In-System Storage, it was mentioned that the storage in
the Hubbell-Southfield Sewer is limited by the depth of water at the
backwater gate. Normal river level at the backwater gate is approxi-
mately 1.5 feet as shown on Figure 82. Although the Hubbell-Southfield
Sewer is a lM-'-6!l x 12' double box, the amount of storage is only 3.5
million gallons before the 1.5 feet head at the backwater gate permits
an overflow to occur. However, as also previously discussed, during
periods of heavy runoff the Rouge River has risen to a level which
exerts 8 feet to 10 feet of head at the Hubbell-Southfield backwater
gate. Under these conditions an estimated 23 million gallons of storage
is available in the sewer.
From the data collected under the monitoring and remote control program,
it has been found that the Hubbell-Southfield Sewer overflowed 1.37% of
the time (Table 8). It has also been found from the level sensor data
that during the high runoff conditions when appreciable amounts of
storage have taken place, there have been no adverse effects upstream.
With this information it is evident that significant storage could be
129
-------
5
uj
cr
o
o
.
-10
U
oo
I
l!
O
o
a
Sfe
I
Of
O
Ul
GO
O
o
00
o>
130
-------
131
-------
Ul
UJ
CO
O
HUBBELL SECTION
^-HUBBELL-SOUTHFIELD SECTION
SOUTHFCLD SECTION
\
in
132
-------
achieved in this sewer during the lew river levels at the backwater gate.
The additional 20 million gallons of storage could be used to capture
storm wastewater from the less severe rainfalls. ,
The largest amount of storage could be achieved if some type of_level
control device was placed at the outfall. This would also provide a
maximum amount of flow control, since the regulator just upstream of the
outfall is already remotely controlled. However, it must be remembered
that overflows are of an emergency nature and if storage is to be con-
trolled at the outfall, it must be done by a "fail-safe" method. One
such method which displays promise is the inflatable rubber dam. This
dam can be designed so that the dam height is controlled from the Systems
Control Center to regulate storage. In addition, the dam can be provided
with independent control routines to insure overflow capability should
power or communications be interrupted or should high storage levels be
ignored by the control operators. This will provide the necessary form
of emergency backup along with providing controlled storage. Having the
ability to implement storage where previously an overflow would have
occurred should cut the overflow time well below the present 1.37% and
capture an additional 20 million gallons of storm wastewater. Presently
this type of installation is being considered to further augment the
existing monitoring and remote control program.
POSSIBLE ADDITIONAL INFORMATION DEVICE
Ocassionally, rain cells develop in the area which do not follow the
usual west-to-east weather pattern. These storms either- come across
Canada from the south moving in a northerly direction or from the north-
east moving in a southerly direction. These conditions were not ade-
quately foreseen when deploying the telemetering rain gages and therefore
no gages were placed in the area north of Detroit or in Canada to the
south. When these conditions occur, little, if any, warning is received
by Systems Control Center. In many instances these storm_cells have been
unusually severe. One recent storm dropped nearly three inches of rain
in a one-hour period as recorded on one of the weighing type gages, and
sections of the City were flooded as a result. At times the first indi-
cation of these rain cells at Systems Control Center has been rapidly
rising sewer levels... In these instances, it is impossible to prepare
the system to receive additional flow using the storm anticipation techni-
ques that have been developed.
It also happens that rain cells following the normal weather pattern
skirt the outlying telemetered rain gages and, although rain does not
fall on the City of Detroit, heavy rains occur in"the suburbs tributary
to the Detroit system. This condition has also added large_volumes of
storm wastewater to the system without any indication of rain recieved
at Systems Control Center.
133
-------
The storm anticipation technique employed by Detroit can only be effec-
tive if sufficient information is received by the system operators. Al-
though great quantities of information have been gathered through the
monitoring aspect of the program, it has been found that in some areas
the information is insufficient. Rainfall data collection can be, at
times, the area in which the most apparent lack of information occurs.
In Detroit, as in most portions of the country, the U.S. Weather Bureau
uses radar technology to yield information on current rainfall status
in the area. Information such as the intensity of the storm, direction
of travel and the form of precipitation (either rain or snow) is known.
The Weather Bureau has recently made this information available! in a form
which can be of great promise to the capture of storm wastewater.
Specifically, it is now possible to receive the information from the
Weather Bureau's radar scope by remote readout. Thus the system control
operators would have access to two independent sources of rainfall in-
formation. The telemetering rain gages can give intensities and totals
for specific points and the radar readout would yield the area of rain-
fall, direction and speed of travel and the intensity of the storm for
an area with a radius of 125 miles around Detroit. This data should
give the Systems Control Center Cperators enough information to negate
the possibility of being caught unaware of impending rain. Therefore
serious consideration is being given to the inclusion of the remote
readout from the U.S. Weather Bureau to the DMWD system.
ALTERNATIVES TO IN-SYSTEM STORAGE
The research effort into the various methods to reduce pollution from
combined sewer overflows appears to have taken several courses (33).
These alternatives may be divided into five major groups: sewer1 separa-
tion, combined sewer overflow treatment, surface drainage area control,
off-line storage and in-system storage. Each of the methods has certain
advantages and disadvantages and will be discussed briefly for comparison
purposes.
Sewer* Separation
Total sewer separation on a national scale has been estimated to cost
over seventy billion dollars (34). In Detroit it is estimated that it
would take 30-40 years to design and construct a completely separated
system at a cost of over two billion dollars (35). In addition, the
effectiveness of sewer separation in reducing.pollution of our water-
ways has been seriously questioned (3, 5, 36-38). It is evident that
urban runoff contains high amounts of solids, fertilizers, pesticide
residues, oil and organic pollutants from animal droppings, leaves,
litter and grass clippings, and even separation of the sewer system
would not eliminate the pollutional problem of urban runoff (39-M-2).
134
-------
Combined Sewer Overflow Treatment
Biological, physical and chemical treatment of combined sewer overflows
have been investigated (33, 43-49). All of these methods have limita-
tions which make them unacceptable for use at a majority of the Detroit
outfalls. Very few of the 76 overflow points in the Detroit system are
designed for less than 100 cfs relief with many approaching 1000 cfs and
the largest being 12,400 cfs. 'In addition, most of the outfalls are
located within the existing street rights-of-way and are bordered by
either private residences or industry. Thus, the amount of land required
for the installation of treatment devices is limited, and large amounts
of additional land would have to be purchased before any such method
could be integrated into the system. Therefore, if on-the-site treatment
is not a suitable solution, other methods designed to eliminate the
pollutional effects of combined sewer overflows must be examined.
Surface Drainage Area Control
The concept of surface drainage area control is'used to alter storm
runoff in such a manner so that the resulting influx of runoff would
have a minimal effect on combined sewer systems. This concept may
include surface housekeeping to prevent solids -from entering sewers (34);
capturing storm runoff from buildings to use for lawn sprinkling or in-
creased fire protection; zoning and site grading restrictions to increase
time of concentration and decreases the solids load of the runoff (50);
throttling of catch basin inlets to restrict inflow into the system;
utilizing porous pavement to reduce runoff (51); or employing any other
method to reduce pollutional inflow to the sewer.
The concept of surface drainage area'control must be considered a part
of urban runoff management. Although this concept will not provide for
a direct solution to the problem of combined sewer overflows, its appli-
cation will help alleviate the problem and control the tendency toward
the ever-increasing effect of runoff from urbanized area's.
Off-line Storage • .
In this context, off-line storage is used to connote the storage_of com-
bined sewer wastewater during wet weather periods in holding facilities
located at or near outfalls. These facilities may include rigid or
flexible tanks, deep tunnels or ponds, and they may or may not have
provisions for treatment of volumes in excess of their capacity. The
major problems encountered by tnese storage facilities are land costs
and availability, adverst aesthetic impacts, operating costs and
solids removal-(52).
As in the case of combined sewer overflow treatment, land availability
is one of the major limiting factors in the installation of retention
facilities. The only possible location in the Detroit area for retention
facilities are in parks along the Rouge River and in a few park areas near
135
-------
the Detroit River. However, many of these locations are wooded areas
and the possible .destruction of large segments of the parkland would not
be desirable.
In-System Storage
In-system storage has been defined as the utilization of excess capacity
in the trunk and interceptor sewers (52). Although this definition
describes the concept used in this project, it should be expanded to
include any storage facility which may be used to carry dry weather flow
but, which through some mechanical means, may also be used to store
storm wastewater flow. The Livernois Relief Sewer as discussed earlier,
would not be considered as an in-system storage facility until the modi-
fications mentioned were made.
ANALYSIS OF ALTERNATIVES . . . ' '
As in the case of others (53), separation of sewers was not considered
to be a practical approach to the solution of combined sewer overflows.
Not only is this alternative of tremendous cost, but there is no assurance
that all cross-connections, between sanitary and storm sewers will be
eliminated or that future illegal connections would not be made. In
addition, urban storm runoff is polluted and it would not seem logical
to allow this polluted runoff to enter the receiving waters. For these
reasons, separation of sewers into sanitary and storm is not justifiable.
Given the existing technology in the treatment of combined sewer over-
flows, it appears that a suitable method for treating large flows has not
been developed. In addition, maintenance requirements and sludge hand-
ling techniques are not suitable for the Detroit system with its rela-
tively large number of outfalls and diversity of location. It may be
argued that some treatment is better than no treatment. Although this
is a valid point, after analysis of the results of this project and
others, it' is believed that the monies involved could be better spent
in the capturing of storm flow and subsequent higher degree of treatment
at a wastewater treatment plant.
Surface drainage area control is dependent upon the support of the
general public and those in all levels of government. Zoning and building
code changes designed to alleviate either the volume or the pollutional
load of runoff must have general support. The need for neighborhood
surface housekeeping and judicious use of pesticides and fertilizers
must be impressed upon the general public. Engineers who are a part of
the DMWD speakers group have found that large segments -of the public ' .
are unaware as to individual and group action that may be taken to
reduce pollution although almost all are aware that a pollution problem
exists. It appears that, as public awareness of individual po3.1ution
control actions increase, this method of aiding in the reduction of
pollution from combined sewer overflows will play an increasingly impor-
tant role.
136
-------
Off-line retention basins have been considered for several locations in
the Detroit system where land is available. The economic analysis _ of .
this type of facility is based on the costs for four retention basins
which.have been constructed in the Detroit area. These basins 'are all
located near outfalls and range in size from 7 million gallons to 62
million gallons. One basin is gravity filled and dewatered while the
others require pumpage for dewatering. Figure 83 shows the estimated
cost for a retention facility that contains a skijnming weir overflow,
facilities for chlorination of excess volumes and hydraulic flushing
(54). The costs shown have been updated to 1971 using the Engineering
News Record Construction Cost Index. The first cost of the four basins
ranges between $0.20/gallon and $0.50/gallon based on 1971 costs.
The major problem with all retention facilities is the deposition of
solids. These solids may be removed mechanically or manually or may be
flushed back into the system either hydraulically or by mechanical re-
suspension. All of these'methods add to operation and maintenance costs.
In addition, combined sewers have been found to contain such items as
12 foot long timbers, 5 gallon pails, 6 inch globe valves, large chunks
of concrete, oxy-acetylene tanks, step ladders and miscellaneous debris
which would be difficult to remove from basins or screening devices with-
out some manual cleaning. Any on-site cleaning results in a solids waste
disposal problem, and these disposal costs must be included in the opera-
tion of the facility.
The demonstration project has resulted in a controlled storage volume of
approximately 140 million gallons at a 1971.cost of about $2.7 million
or about $0.02/gallon of storage. This figure cannot be compared with
that of retention basins since it was based on the use of existing
capacity in the system. In addition, the locations used for storage were
selected based on maximum storage at minimum cost. Additional in-system
storage locations have been investigated and the costs have increased to
about $0.06/gallon. It is expected.that, as the system is utilized more
fully, these costs for added increments of storage will continue to rise.
It is also evident that the capacity for in-system storage_does not exist
in every district in the system, and the problem, of capturing storm flow
will still exist. As a solution to this problem, DMO is considering the
construction of additional sewers to be used for storage and to provide
relief capacity during severe storms. Figure 8M- shows the estimated 1971
cost of sewers. Costs were derived from previous tunnel and open-cut
sewer and water main construction contracts (55). "In certain areas of
Detroit, open-cut construction could be used to add additional barrels
to existing sewers in an arrangement similar to the Conner Gravity Sewer.
In other areas, tunnel construction would be necessary. The cost/gallon
of the larger sewers is competitive with those costs shown for retention
basins. In addition, by using dry weather flow for flushing, the opera-
ting costs will be considerably less than for retention basins.
137
-------
cr
1 mv
o
o
z
o
— i 4Q-
_I
3,
r-
(0
O 30-
0
o
X
S
jr
S
/
jS
/
40 80 120 16
/
/
<
0 2C
X
>0 2fi
>0
RETENTION CAPACITY (MILLION GALLONS)
Figure 83-ESTIMATED COST OF RETENTION FACILITIES
138
-------
RETENTION CAPACITY
(MILLION GALLONS/ 10,000 LINEAL FEET)
11.5
I5.O
19.0
8 10 12 14
DIAMETER (FEET)
16
18
Figure 84 -ESTIMATED COST OF SEWERS
139
-------
An additional benefit to in-system is the flow routing capability. Many
of the sewers in the Detroit system are over 50 years old. The Detroit
River Interceptor was constructed in 1929 while other large sewers were
constructed around 1900. These sewers will need to be rebuilt eventually.
At the present time, if one of the major sewers fails, raw sewage may have
to be routed to the river. With multiple cross-connections similar to
water distribution systems, flow could be routed around failures or areas
of ^rebuilding without the necessity of dumping raw sewage into the re-
ceiving streams or the necessity of costly pumping of the wastewater
around the failure area.
Based on the experience obtained under this demonstration project, it is
believed that in-system storage can provide an economical solution to
the problem of pollution resulting from combined sewer overflows. Al-
though first costs for additional open-cut or tunnel sewer construction
may be greater than_retention facilities, it is believed that this approach
will be more economical on a long-term basis when operating 'and maintenance
costs are considered.
140
-------
SECTION XI . • ' .
ADDENDUM
Since the end of the evaluation period in August, 1971, the Detroit
Metro Water Departnent (DMWD) has continued the evaluation and expansion
of the original monitoring project. This section will serve_to update
some of the information contained in the report and to highlight work
which has been done since the original evaluation period.
IDNITOKENG EQUIPMENT MODIFICATIONS AND ADDITIONS
Rain Gages
The network of 14 tipping bucket rain gages has been expanded to a •• 25
• gage network. Several of the new gages were added within^the City of
Detroit to increase the existing network.density. In addition, gages
were installed in'sons of the westerly and northerly suburban cormunities
to aid in the operation of their systems. All gages are now equipped
with 500 watt heaters and frequency shift telemetry devices. It is
believed that both the accuracy of data transmission and .the calculation
of expected runoff will be enhanced by these additions and modifications.
Radar Remoting System
As a supplement to the rain gage data, a radar, remoting system to provide
instant weather information by telephone lines has been installed in the
Systems Control Center. The radar system displays the weather pattern
within a 125 mile radius of the Detroit Metro Airport.
Experience to date has shown that once a storm approaches within about
50 miles of Detroit, it is relatively certain that the storm will_affect
the wastewater collection system. Pumpdown procedures are now initialized
whenever a storm is between the 25 and 50 mile radii circles on the display.
Level Sensors
The level sensor network has been expanded and a total of 214 sensors are
now in place. All of the new sensors have a 15 second cycle time and_most
sensors have a 0-20 feet range. Thirty of the new sensors have been in-
stalled in various suburban sewer systems and retention_basins to facilitate
dewatering operations and coordination of system operations. Some of
these; levels are telemetered to both the DMWD Systems Control Center and
to suburban control centers.
Status Sensors
Additional status sensors have been installed at overflow points which have
multiple backwater gates or dams. Thus, if any one of the gates opens, an
overflow'will be recorded for that location. Overflow status sensors have
141
-------
also been installed on several suburban systems to provide data to
those communities on the frequency and duration of their overflows.
A total of 110 overflow points are now monitored.
COMPUTER SYSTEM
The existing data logger was replaced with a Control Data SC1700 computer
system. The central processor has 24K core and a 3 million word cartridge
disk subsystem. The mass storage operating system is stored on a.1.5
million word fixed disk and sensor data will be stored on removable
cartridge disks. These data disks will be taken to the Wastewater Treat-
ment Plant for processing of monitoring data utilizing the SC1700 systems
being installed there for process control.
Data logging at the Systems Control Center consists of hourly printouts
of level sensor readings, rainfall:totals, and hourly intensities. Over-
flow status changes and pump status changes will be logged in real time.
In addition, a daily summary of pump running hours, total rainfall, and
overflow signal malfunctions will be printed at 24-00 hours. Output format
of the logs has been changed to allow easier identification of sensor
points.
The operator interface to the system consists of a CRT and an alarm
printer. The wastewater collection system has been divided into 12 sub-
sections. A schematic plan view of the major sewers and level sensor
locations within a subsection are displayed at the top of each CRT page.
The bottom half of the page displays the level sensor data for the pre-
ceding three 5 minute intervals, three 15 minute intervals, and seven
60 minute intervals for a total of ,8 hours of data. The latest 5 minute
rainfall intensity and daily total rainfall for any rain gage located
within the subsection is also displayed on the appropriate page. The
system is designed to display a total of 30 pages. However, only 19
pages of sensor data are being utilized at the present time.
Alarms are logged on a Selectric Typewriter. All points godug into
alarm are logged in red. Returns from alarm or other system occurrences
are logged in black. In addition, all alarms must be acknowledged by
the operator using the CRT keyboard. Unacknowledged alarms are rsprinted
every 5 minutes until acknowledged.
SUPERVISORY CONTROL SYSTEM
The supervisory control system has been expanded with the addition of
four new control panels to the three existing panels installed under the
grant project. Remote control facilities including three wastewater
pump stations, four interceptor regulators, three fabridams, two in-system
storage gates, one flow routing gate, and one suburban connection have
been added to the supervisory control system. In addition, four suburban
142
-------
retention basins and 11 suburban pump stations are,displayed schematically
with basin levels and pump status indicated.
PUMP STATIONS .
Supervisory Control -
The eight 500 cfs storm pumps at the Conner Pump Station were remotely
controlled shortly after the end of the project. Since the pumps have
raised impellers, they must be prijned. In addition, an exciter is used
for each pair of pump motors and mist be running prior to pump starting.
A single push-indicating switch is used to sequence operations for each
pump. The switch is turned to the "start" position and pushed. This
causes the exciter to start if it is not already running. When the ex-
citer is running, the "running" light on the -switch is lit. After about
one minute the switch is depressed again to start the pump motors and
the automatic priming sequence. The "running" light turns off to confirm
that the priming sequence has started. Approximately 6 to 8 minutes later .
when the vacuum in the pump casing reaches 21" Hg, a limit switch shuts
the vacuum pump off and the "running" light is lit to.confirm that the
pump priming sequence has terminated.
Operators have indicated that the above procedure has led to some dif-
ficulties. The major problems are the time required to start the pumps
and the possible failure, of the pumps to prime correctly. As discussed
in Section IX, the operators are quite active during storm events. The
time lag associated with starting the storm pumps forces the operators
to anticipate the number of pumps required to prevent upstream flooding.
Furthermore, the operator must remember the status of each pump in the
starting sequence. Once an indication that the pump has started is re-
ceived, the operator must then observe the wet well level indicators to ^
determine if the pumps have been primed correctly and are actually pumping
wastewater.
Two other pump stations have been completed since the grant period ended
and both are remotely controlled. One station contains 3-M-O cfs pumps
• and one standby pump. The other station currently^has 2-150 cfs and
1-100 cfs pumps. The operation of these stations is the same as the
stations remotely controlled under the original grant project.
Pump Station Discharge Conduit Modifications
All flow from the easterly service area of the DMWD roust be pumped through
the Fairview Interceptor Lift Station. The service area of this station
is shown in Figure 3. Prior to monitoring, the station was shut down
during storm events and the Conner-and Freud storm pumps were used to pre-
vent most flow from reaching the station.
After several months of operation of the system utilizing monitoring and
remote control, it became apparent that the Fairview Station could be
143
-------
operated during storms without significantly affecting the system
downstream of the station. This was especially true,' for spot storms
which occurred over the easterly portion of the service area. In addi-
tion, over 100 million gallons of wastewater could.be stored in the
easterly service area during a storm event and it was noted that 3 to 1
days were required to dewater the system.
Although Fairview has a pumping capacity of 525 cfs, it was rarely
operated at more than 300 cfs prior to monitoring. Normal operation
ranged between 75 and 225 cfs. Previous reports had indicated that
surge problems could be anticipated at flows greater than 300 cfs.
Surge analysis studies indicated the magnitude of the problem. The
results of these studies were verified by utilizing level sensor data
and by field observation. The discharge conduit was inspected to
determine its condition utilizing system shutdown procedures outlined
in Section IX. Based on the level data and observations, various al-
ternatives were introduced into the surge model for analysis. The
results indicated that the most economical solution was to modify.the
discharge conduit as shewn in Figure 85. An electrode sensor was .
installed on the surge overflow wier and .the status is displayed on the
control panel in the Systems Control Center.
Although the station has been in operation for only a few months since
it was modified, sufficient information,is available to indicate that
the surge problem has been eliminated when the station is pumping at
its 525 cfs capacity. Level and surge overflow data may be used to
further rafine the model so that it may be used in other applications
if necessary.
The Oakwood Pump Station serves the district shewn in Figure 3. The
S'-O" discharge conduit as detailed in Figure 23 enters the system just
upstream of the Rouge River siphons. The conduit was originally designed
with a capacity equal to one sanitary pump. During storm events, the
sanitary wastewater pump was shut1 down and only storm pumps were utilized.
With the change in philosphy of system operation, the sanitary pump was
operated during storm events in order to capture as raach flow as possible.
Monitoring data indicated that the system should be -capable of accepting
the flow from both sanitary pumps' during storms. Pumping records were
analyzed and it was determined that, as an 'average, approximately 4-
million gallons per storm event could be pumped to the wastewater treat-
ment plant if both sanitary pumps;were operated.
To implement this mode of operation, the only modifications necessary were
the installation of weighting collars and gasketed manhole covers on two
manholes so that the discharge conduit could be used as a force main.
144
-------
tf>
tt _l_l
^ oo
°H??
_
1
.o
z
IN
*$•
d°
•JS^sss
] ] ]
^ f-
Vc<
- (&
h-
LjjQ
li - - —
i
£*?;*
i
i
]
pe?^
^
r
i//
<
00
o
k
D)
145
-------
REGULATOR MODIFICATIONS
Leib Street Regulator
The Leib Street regulators transfer flow from the Conant-Mt. Elliot com-
bined sewer, as shown in Figure 17 , to the Detroit River Interceptor.
At this point, two 11'-0" diameter barrels discharge into a regulator
chamber containing four No. 12 Brown and Brown float controlled regulators.
Overflow monitoring data indicated that this sewer had the highest, fre-
quency of overflows among the monitored points. Level sensor data in-
dicated that under certain conditions the regulators could have been kept
open longer to permit additional flow to be diverted to the interceptor.
Remotely operated hydraulic cylinders were installed on each regulator
gate. A typical installation is shown in Figure 86. Note that the
existing float mechanism was modified by the installation of a counter-
weight on each float. This serves to relieve the load on the cylinder
and also to open the regulator if the cylinder were to fail.
McClellan-Cadillac Regulator
The McClellan-Cadillac regulator serves the 6'-3" McClellan combined
sewer and the 5'-9" Cadillac combined sewer utilizing two 24-inch,
Type "C", McNulty regulators. The regulators operated as a function of
the depth of flow in both the McClellan and Cadillac sewers and the
Detroit River Interceptor. The regulators were set to be completely
closed when the elevation in the float chamber reached elevation 99.0
feet Detroit datum. The McClellan sewers' overflow weir is at elevation
98.0 feet and the Cadillac sewers' at elevation 98.5 feet.
The regulator chamber is located approximately 2000 feet downstream of
the Fairview Interceptor Lift Station. It was found that flow from the
interceptor will sometimes back up through the regulator chamber and top
the overflow weirs when the Fairview pump station is operated during
storm events.
The regulator has been modified for remote control as shown in Figure 87.
Since the regulator was built in 1927, it was decided to remove the existing
float mechanism and gate and install new sluice gates as shown. With the
exception of the telemetering equipment, all control equipment was located
within one of the existing float chambers. A sump pit, sump pump and de-
humidifier were installed in the chamber.
Dearborn Regulator
Due to the expansion at the Wastewater Treatment Plant, the Portland-
Harbaugh (Pulaski) Sewer was relocated to the Dearborn sewer. The outfall
and regulator chamber at Dearborn were replaced with larger capacity
facilities. 'As shown in Table 8, neither of the outfalls experienced
overflows. This was due to both the decrease in service area of-the
146
-------
STOP LOG SILL
(TYPICAL)
TELLTALE
VALVE CLOSED
(TYPICAL)
I DRY WEATHERU* -TORY WEATHER
FLOW F 3J FLOW
DETROIT RIVER
INTERCEPTOR
NEW FLASHBOARDS ON
EXISTING DIVERSION DAM
PLAN VIEW OF LEIB STREET DIVERSION
AND REGULATION FACILITIES
STREET SURFACE
SITE PLAN-LEIB STREET
REGULATOR FACILITES
NEW
HYDRAULIC
CYLINDER
EXISTING
FLOAT
LINKAGE
• SECTION (A)
( MODIFICATIONS TO EASTERLY 2 GATES SIMILIAR)
DETAIL OF HYDRAULIC CYLINDER
AND GATE POSITION INDICATOR
( ALL GATES SIMILAR )
Figure 86-MOD1FDCAT1ONS AT THE LEIB REGULATOR
147
-------
,•—NEW GATE POSITION
\ INDICATOR
N
STREET SURFACE
tXtSTING FLOAt
CHANWiH USED
TOW ELECTRIC
ANO HYt RAtT.lC
fKXJIPMENT
PLAN VIEW OF McCLELLAN-CADILLAC
DIVERSION AND REGULATION FACILITIES
NEW SUMP PUMP -
AND SUMP PIT
EXISTING FLOAT AND
REGULATOR CHAMBER
SECTION (A)
r EXISTING FLOAT LINKAGE WALL OPENING
\ FILLED WITH EXPANSIVE CONCRETE
STREET SURFACE
EXISTING FLAP GATES
CHAINED OPEN (TYPICAL)
EXISTING S' SLUICE GATE
REMOVED (TYPICAL)
NEW 24' X 24"
SLUICE GATE
EXISTING 24' REGULATOR, FLOAT
AND LINKAGE REMOVE.D
SECTION (B)
Figure 87- MODIFICATIONS AT THE McCLELLAN
CADILLAC REGULATOR
148
-------
sewe:ps due ±o plant expansion and to the fact that the float operated
regulators do not close because of drawdown in the interceptor.
For the above reasons, remote control facilities were. installed on the
regulator under the relocation work. It is anticipated that following
the first flush interception during a storm event, the regulator will
be closed to allow in-system storage. It will then be modulated to
prevent an overflow while.maximizing storage.
Evergreen-Farmington Regulator
The Evergreen-Farmington regulator and meter chamber is located at the
north end of the Southfield Sewer (Figure 17). The combined sewer
discharge from this suburban connection is regulated by a telltale into
the Southfield Sewer. In order to protect the Detroit system from ex-
cessive flows, the regulator closes during storm events and causes this
suburban flow to overflow into the Rouge River.
Monitoring data indicated that on many occasions, storms .would pass over
the northwest corner of Detroit. Because of the inflow into- the Detroit
system, the suburban regulator would close even though additional capacity
existed in the Detroit System further downstream. In an attempt to reduce
the overflow problem at this location, a remotely controlled override has
been installed on the regulator so that additional flow may be taken into'
the Detroit system when conditions permit. The overflow status is also
indicated on the control panel in the DMWD Systems Control Center so that
. the operators know if an overflow is occurring and may perhaps be able to
open the regulator gate to accept additional flow.
Hubbell-Southfield Fabridams
As discussed in Section X, the Hubbell-Southfield 11'-6" x 12'-0" double
box sewer had a storage potential of about 23 million gallons although
only about 3.5 million gallons was utilized. A 6r-3" fabridam was in-
stalled in each barrel of the sewer immediately upstream of the backwater
gates to fully utilize the sewer's storage potential. The installation
Is similar to the Minneapolis-St. Paul fabridams (27).
A condensate water drain line, an 8-inch diameter blow-off standpipe and
dam pressure telemetry devices as suggested by Minneapolis were installed.
In addition, the dam controls were designed for automatic on-site operation
with override controls at the Systems Control Center. The dams will
modulate automatically to maintain the level in the sewer at 2 -feet below
crown. If the wastewater level continues to rise to within 18 inches
below the crown, a second solenoid valve will open. At-12" below crown,
a spring loaded relief valve opens and the blow off standpipe is activated.
Since Minneapolis reported that surges tended to activate their blow-off,
the DMWD standpipes were equipped with submergence limiters ,to allow
adjustment of the blow-off standpipe water level should loss- of standpipe
water become'a problem.
149
-------
Livernois-Relief Sewer Modifications
The possible utilization of the Livernois Relief Sewer (Figure 80) for
in-system storage was discussed in Section X. Approximately 8 million
gallons of in-system storage was obtained by the installation of remote
control facilities at four locations on the sewer. These included
1-cross connection gate, 2-storage gates, 1-fabridam, and modifications
to the regulator. The details of these installations are shewn in
Figures 88 and 89.
Two electrically operated 78" x 78" sluice gates were installed at Warren
Avenue. Electric operators were selected for the same reasons as out-
lined in previous sections of this report. Emergency manual ovesrrides
and a plastic stem cover were installed. In addition, plastic gating
was installed on the motor support floor to provide some relief capacity
if both gates should fail and the emergency crew is delayed.
The farridam installation at Ranspatch Avenue utilizes the same -type of
control scheme as that described for the Hubbell-Southfield Fabridams.
DATA UTILIZATION . ' '
The utilization of the data obtained during the original study period was
described in Sections IX and X. In general, the DMWD has continued to
apply the data in a similar manner. Other uses of the data which have
been investigated or are currently being implemented will be briefly
discussed here.
Flow in Sewers and Overflow Volumes
Attempts to correlate sewer levels with flow and to calculate overflow
volumes have not been greatly successful. The locations selected for
the majority of level sensors were areas where operational information
was desired. In general, these locations are either upstream or down-
stream of major connections or relief ports; upstream of control facilities;
in pump station wet wells; or near transition sections. As a result,
flow calculations based on depth of flow in open channels have not yielded
reliable information. The major error source is backwater effect which
tends to yield much higher calculated flow values than would be expected.
In those locations where there is a relatively long reach of sewer upstream
and downstream of the level sensor so that "normal depth of flow" condi-
tions may be assumed, flow predictions based on depth of flow appear to
be reliable. However, in some of the older sewers, movement of the sewer
and/or construction procedures have changed assumed values for the slope
and as a result affected flow calculations. This was evidenced in cases
where upstream flows were higher than downstream flows when calculated
from level data. . .
150
-------
200O 0 200O
SCALE IN FEET
151
-------
152
-------
By utilizing the energy equation for gradually varied flow, backwater
effects may be considered in flew calculations. However, very few
level sensors are located in areas where a second boundary condition is
known or can be reasonably approximated. It was originally anticipated
that an "M curve" (56) could be generated to predict overflow volumes -
at outfalls having weirs. Iterative techniques assuming a flow over
the weir and solving for the upstream level were considered but not im-
plemented. Since flow through the regulator immediately upstream of the
dam was difficult to predict and since there were connections into most .
of the major sewers between the level sensor and the weir, flow deter-
minations would not have been accurate enough to justify the work in-
volved in their determination.
System Response Predictions
The DMWD has four major objectives in the utilization of the sewer system
monitoring data. These are: 1) to aid in.the operation of the system;
2) to predict and verify system response to storm events; 3) to establish
priorities for overflow abatement projects; and 4) to develop computer
control algorithms for the various remote control facilities. .
Because of the complexity of the Detroit sewer system with its multipli-
city of subdistrict and district interconnections, only three sewer
districts can be independently analyzed. As shown in Figure 3, the Blue-
hill and Oakwood Districts containing approximately 2200 and 1500 acres
respectively have all flow pumped. The Puritan District containing about
500 acres-has all sanitary flew pumped and has three gravity outfalls
each with a diversion weir and backwater gate. Additional level sensors
have been installed in these districts to provide system response data.
One of these new level sensors was installed in the Bluehill District
approximately 2000 feet downstream of an existing level sensor along a
reach.of sewer with no other inlet sewers. Although three levels are
necessary to solve the equations of motion if errors in the assumed pipe
friction, slope or level readings are to be accounted for (57), it is
believed that the flows predicted by only two levels will be sufficiently
accurate for system response predictions and control algorithm development
Acoustic Flow Meters '
As additional aids in predicting system response, acoustic flow meters
have been installed on two outfalls. One of these meters is located
on the Conant-Mt. Elliot sewer downstream of the Leib Street remotely
controlled regulators. The sewer has five level sensors at various points
upstream of the regulator and one level sensor at the regulator facility
as shown in Figures 17 and 23. In' addition, level sensors have been in-
stalled on the Detroit River Interceptor both upstream and downstream of
the'regulator facility. This information should be sufficient to verify
any sewer hydrographs which may be developed.
A second -acoustic flow meter has been installed on the West Chicago Sewer
Outfall. This outfall serves as a relief to both the Hubbell-Southfield
153
-------
Sewer and the Northwest Interceptor. There are 16 level sensors: on the
various sewers that affect overflow volumes in this outfall. Although
the sewer system is quite complex in this area with high level reliefs
and various interconnections , it is anticipated that the system response
in this area may be predicted and verified using the data.
Additional Storage Facilities
Currently, the DMWD is investigating both in-system and off-lines storage
in the Rouge River, Conner Creek and Fox Creek Districts as shown in
Figure 6. Preliminary studies in the Rouge River District have indicated
that at least two sewer systems are amenable to in-system storage. Modi-
fications to the outfalls in these system by the installation of fabri-
dams or other control devices are estimated to cost between $0.06 and
$0.13 per gallon of additional in-system storage capacity. Based on
rainfall and overflow duration data, the above modifications will reduce
overflows by 50% to 75% for.those sewers.
Off-line retention facilities are being investigated for the Conner
Creek and Fox Creek Districts. The existing 100 million gallons of in-
system storage will be augmented by off-line storage which could be in
excess of 100 million gallons. Sensor data could not be used in the
preliminary design of these facilities to predict the volume of overflow
which could be expected for a given storm.
However, the data is being utilized to determine operational procedures
for dewatering. As indicated previously, 3 to 4 days are required to
dewater the existing storage facilities. Normally, the Conner Gravity
System, Figure 7, is dewatered first followed by the Conner 'Pumped -
System and then the Freud System. • Level data in the Interceptor and
pump station records have indicated that the new facilities may be de-
watered concurrently with the Conner ]?umped and Freud Systems. Critical
to the dewatering operation is the sizing of the dewatering pumps'. With
the data available, the pump selection may be optimized to provide
maximum dewatering rates with a minimum affects on the system operation.
Flooding Complaints
The EMWD periodically receives complaints of flooded streets or backups
of wastewater into basements. In general, investigation of these com-
plaints indicates minor problems with either the lateral sewers or the
individual, householder's connection. Occasionally, however, several
complaints fron a sewer district are received. If the area is near one
of the larger monitored sewers, the level and rainfall data is analyzed
to determine if the problem is due to the system operation. Special
emphasis is placed on areas where remote control facilities are located.
To date, only one of these complaints could be traced to•operational ,
problems. In this case, surge in a sewer downstream of a pump station
caused air and mist to blow through the holes in a manhole cover. After
analysis of level data, inspection of the manhole and sewer, and inter-
views with the individual who reported the incident, it was concluded
154
-------
that the surge was not serious and that no change in operation was re-
quired. No other similar complaints have been received for the sewer
in question.
Summary
The•DMWD believes that the Sewer Monitoring and Remote Control Program
has'been an invaluable aid in the operation of the wastewater collection
system. During the evaluation period, the emphasis was placed on system
operation and maxinization of in-system storage. As the operation of
the system was refined, efforts have since been directed to utilization
of the monitoring data in the determination of overflow abatement programs,
analysis of system response, and design of additional facilities where
the data has indicated relief sewers are necessary.
The emphasis in the report has been placed on the equipment installed,
the reasons for the particular installation, and the evaluation of both
the equipment and the utilization of the equipment. It is intended that
this report may serve as a guideline for those contemplating either the
installation of new monitoring and control facilities or the modification
•or expansion of existing facilities.
155
-------
SECTION XII
._..,..:':. REFERENCES -.*..-.-..-
1. Anon, "Problems of Combined Sewer Facilities and Overflows - 1967,"
Federal Water Pollution Control Administration, Publ. No. WP 20-11,
Washington, B.C., December 1967.
2. Viessman, W., Jr., "Assessing the Quality of Urban Drainage,"
Public Works, 100, No. 10, October 1969.
3. Benzie, W. J., and Courchaine, R. J., "Discharge from Separate
Storm Sewers and Coiibined Sewers," Journal Water Pollution Control
Federation, 38, No. 3, March 1966. .
4-. PalJier, C. L., "The Pollutional Effects of Storm Water Overflows
from Combined Sewers," Sewage and Industrial Waste,22, No. 2,
February 1950.
5. Burm, R. J., "The Bacteriological Effect of Combined Sewer Over-
flows on the Detroit River," Journal Water Pollution Control
Federation,39, No. 3, March 1967.
6. Detroit Metro Water Department, "The 121st Annual Operating Report
for the Fiscal Year Ended June 30, 1973," Detroit, Michigan,
July 1973.
7. "Proceedings - Conference in the Matter of Pollution of the Navi-
gable Waters of the Detroit River and Lake Erie and their Tributaries
in the State of Michigan," U. S. Department of Health, Education and
Welfare, Washington, D. C., June 1965.
8. "Local Climatological Data - Detroit, Michigan," National Oceanic
and Atmospheric Administration, Environmental Data Service,, U. S.
Department of Commerce, Washington, D. C., June 1971. "
9. Strommen, N. D., "Urban Influences of Rainfall in the Detroit,
Michigan area, Unpublished Paper.
10. American Public Works Association, "Combined Sewer Regulator Over-
flow Facilities," Federal Water Quality Administration, Publication
No. 11022 DMV 7/70, Washington, D. C., July 1970.
U. Fair 9 Gordon M.; Geyer, John C.; and Okun, Daniel A., Water and
Waste Engineering, Vol. 1, Water Supply and Wastewatef Removal,
John Wiley & Sons, Incorporated, New York, New York (1958),,
12. Raths, C. H. and McCauley, "Deposition in a Sanitary Sewer,," Water
and Sewage Works,109, 1962.
156
-------
13. Ackers, P. and Harrison, A. J. M., "Attenuation of Flood Waves in
Part-Full Pipes," Proceedings, Institution of Civil Engineers,
London, 28, July 1964.
14, Martin, C. Samual and DeFazio, Frank G., "Open-Channel Surge Simu-
lation by Digital Computer," Journal of the Hydraulics Division,
American Society of Civil Engineers, 95, No. HY6, Nov. 1969.
15. Bryan, E. H., "Quality of Stormwater Drainage from Urban Land,"
Draft paper presented at 7th American Water Resources Conference,
Washington, D.C., October 1971.
16. DeFilippi, J. A., and Shih, C. S., "Characteristics of Separated
Storm and Combined Sewer Flows," Journal Water Pollution Control
Federation,43, p. 2033, No. 10, October 1971.
17. McPherson, M. B., "The Nature of Changes in Urban Water Sheds and
Their Importance in the Decades Ahead," Paper presented at the
Conference on "The Effects of Watershed Changes on Stream Flow"
University of Texas at Austin, October 1968.
18. American Public Works Association, "Combined Sewer Regulation and
Management - A Manual of Practice," Federal Water Quality Admini-
stration, Publication No. 11022 DMV 08/70, Washington, D. C.,
August 1970.
19. American Public Works Association, "Problems of Combined Sewer
Facilities and Overflows - 1967," Federal Water Quality Admini-
stration, Publ. No. WP-20-11, Washington, D. C., December 1967.
20. Linsley, Ray K., Jr., Kohler, Max A., and Paulhus, Joseph L. H.,
Hydrology for Engineers, McGraw-Hill Book Company, Incorporated
(1958).;
21. Riis-Carstensen, E., "Improving the Efficiency of Existing Inter-
ceptors," Sewage and Industrial Wastes,27, No. 10, October 1953.
22. Phillips, M. B., "Maintenance of Storm. Flow Regulators," Sewage
and Industrial Wastes, 31, No. 7, July 1959.
23. Black, H. H., "Procedures for Sampling .and Measuring Industrial
Waste," Sewage and Industrial Wastes, 24, No. 1, January 1952.
24. Haney, P. D., and Schmidt", J., "Representative Sampling and Analy-
tical Methods in Stream Studies," Sewage and Industrial Wastes, 30,
No. 6, June 1958. • '
25. Hayes, Seay, MaHern Architects-Engineers, "Engineering Investigation
of Sewer Overflow Problem", Federal Water Quality Administration,
Publ. No. 11024, BM3 05/70, Washington, D. C., May 1970.
157
-------
26. American Public Works Association, "Feasibility of Computer Control
of Wastewater Treatment," Environmental Protection Agency, Publ. No.
17090DOV 12/70, Washington, D. C., December 1970.
27. Metropolitan Sewer Board of St. Paul, Minnesota, "Dispatching System
for Control of Combined Sewer Losses", Environmental Protection
Agency, Publ. No. 11020 FAQ 03/71, Washington, D. C. March 1971.
28. Anon., "Data Communications: A Primer", Journal American Water Works •
Association, 63, No. 8, August 1971. '
29. Daneker, James R., "Problems of Telephone Communications in Telemetry"
Journal American Water Works Association, 62, No. 11, November 1970.
30. Harris, Garth S. "Development of a Computer Program to Route Runoff
in the MinneapoLLs-St. Paul Interceptor Sewers", St. Anthony Falls
Hydraulic Laboratory, Memorandum No. M 121, December 1968.
31. Municipality of Metropolitan Seattle, "Maximizing Storage in Combined
Sewer Systems", Environmental Protection Agency, Publ. No. 11022 ELK
12/71, Washington, D. C., December 1971.
32. Hubbell, Roth & Clark, Inc., Consulting Engineers, "Basis of Design
for Detroit Water Services Pollution Control Program Wastewater
Plant", Bloonfield Hills, Mich., February 1967.
33. Rosenkranz, W., "The Storm and Combined Sewer Demonstration Projects,"
Federal Water Pollution Control Administration Publication. No. DAST-
36, Washington, D. C., January 1970.
34. Heaney, J. P., and Sullivan, R. H., "Source Control'of Urban Water
Pollution," Journal Water Pollution Control Federation, 43, No. 4,
April 1971.
35. Brown, J. W., and Suhre, D. G., "Sewer Monitoring and Remote Control
Detroit," Preprint 1035, Presented at the A.S.C.E. Annual and
Environmental Meeting, Chicago, 111., October 1967.
36. Burn, R. J., Drawczyk, D. F., and Harlcw, G. L., "Chemical, and Phy-
sical Comparison of Combined and Separate Sewer Discharges", Journal
Water Pollution Control Federation, 40, No. 1, January 1968.
37. Burm, R. J., and Vaughn, R. D., "Bacteriological Comparison Between
Combined and Separate Sewer Discharges in Southeastern Michigan,"
Journal Water Pollution Control Federation, 38, No. 3, March 1966.
38. Burgess and Niple, Limited, Consulting Engineers, "Stream Pollution
and Abatement from. Combined Sewer Overflows - Bucyrus,- Ohio,"
Federal Water Quality Administration, Publication No. 11024 FKN 11/69,
Washington, D. C., November 1969.
158
-------
39. Cleveland, J. G., Reid, G. W., and Walters, P. R., "Storm Water Pol-
lution from Urban Land Activity," Paper presented at A.S.C.E. Annual
and Environmental Meeting, Chicago, Illinois, October 1969.
4-0. Spitzer, E. F., "Cities Play a Major Role in Eutrophication", Ameri-
can City, 83, No. 1 August 1967.
41. Weibel, S. R., et al., "Pesticides and Other Contaminants in,Rainfall
and Runoff", Journal American Water Works Association, 58, p. 1075,
July 1966.
42. Weibel, S. R., Anderson, R. J., and Woodward, R. L., "Urban Land
Runoff as a Factor in Stream Pollution", Journal Water Pollution
Control Federation, 36, No. 7, July 1964.
43. Erivirogenics Company, "In-Sewer Fixed Screening of Combined Sewer
Overflow", Environmental Protection Agency Water Quality Office,
Publication No. 11024 FKJ 10/70, Washington, D. C., October 1970.
44. Cornell, Rowland, Hayes and Merryfield, Consulting Engineers and
Planners, "Rotary Vibratory Fine Screening of Combined Sewer Over-
Flows", Federal Water Quality Administration, Publication No. 11023
FDD 03/70, Washington, D. C., March 1970.
45. Department of Public Works - Portland, Oregon, "Demonstration of
Rotary Screening for Combined Sewer Overflows", Environmental
Protection Agency - Research and Monitoring, Publication No. 11023
ED 07/71, Washington, D. C., July 1971.
46. Crane Company, "Microstraining and Disinfection of Combined Sewer
Overflows", Federal Water Quality Administration Publicaton No.
11023 EVO 6/70, Washington, D. C., October 1968.
47. Hercules, Incorporated, "Crazed Resin Filtration of Combined Sewer
Overflow", Federal Water Pollution Control Administration, Publi-
cation No. DAST-4, Washington, D. C., October 1968. ' • .
48. American Process Equipment Corporation, "Ultrasonic Filtration of
Combined Sewer Overflows", Environmental Protection Agency, Publi-
cation No. 11023DZF 6/70, Washington, D. C., June 1970.
49. Dow Chemical Company, "Chemical Treatment of.Combined Sewer Over-
flows", Environmental Protection Agency, Publication No. 11023
FDB 9/70, Washington, D. C., September 1970.
50. AVCO Economic Systems Corporation, "Storm Water Pollution From
Urban Land Activity", Federal Water Quality Administration,
Publication No. 11034 FKL 07/70, Washington4 D.C., July 1970.
159
-------
51. Thelen, E., et al., "Investigation of Porous Pavements fop Urban
Runoff Control", Environmental Protection Agency, Publication No.
11034- DUY 03/72, Washington, D. C., March 1972.
52. Field, R. and Struzeski, E. J.", Jr., "Management and Control of
Conibined Sewer Overflows", Journal Water Pollution Control Federa-
tion, HM-, No. 7, July 1972.
53. Harza Engineering Company and Bauer Engineering Company, "Chicago-
land Deep Tunnel System for Pollution and Flood Control - First
Construction Zone Definite Project Report", Metropolitan Sanitary
District of Greater Chicago, Chicago, Illinois, May 1968.
54* Hubbell, Roth & Clark, Incorporated, Consulting Engineers, "Study
and Report on Abatement of Pollution of the Red Run by the Twelve
Towns Relief Drains District", Bloonfield Hills, Michigan,
September 1969.
55. Detroit Metro Water Department, "Summary Report - Lake Shore Inter-
ceptor Study Gravity System - Pressure System, Oakland-Macoiib
Interceptor System", Detroit, Michigan, March 1971.
56. Daugherty, R. L. and Franzini, J. B., Fluid Mechanics With Engineer-
ing Applications McGraw-Hill Book Company, Incorporated (1965). ~
57. Liggett, J. A., "Mathematical Flow Determination In Open Channels",
Journal of the Engineering Mechanics Division, American Society of
Civil Engineers, 94, No. EM 4, August 1968.
160
-------
SECTION XIII
PUBLICATIONS ' . " ; " . ,:..:
1. ; Brown, J. W., and Suhre, D. G., "Sewer Monitoring and Remote Control
- Detroit", Paper presented at ASCE Annual and Environmental Meeting,
Chicago, Illinois, October 1969
2. Detroit Metro Water Department, "Detroit Sewer Monitoring and Remote
Control" Combined Sewer Overflow Abatement Technology, FvJQA Report
No. 11024-06/70, pp 219-220, June 1970
3. Remus, Gerald, "Storm-Water Retention Can Work. . .and Prevent the
Heavily Polluted 'First Flush' from Overflowing to Damage the Re-
ceiving River", Am. City, 85, No. 10, pp 68-69, October 1970
4. Suhre, D. G., "Cleaner Streams from Busier Sewers", Water and Sewage
Works, 117, pp R109-R112, November 1970
161
-------
backwater curve
backwater gate
biochemical oxygen
demand - (BOD)
capacity
combined sewer
combined wastewater
concentration time
CRT
cubic foot per
second (cfs)
dewater
drainage basin
SECTION XIV
GLOSSARY
The longitudinal shape of the water .
surface in a stream or open conduit
where such water surface is raised or
lowered above its normal level by a
natural or artificial constriction.
A gate installed at the end of a drain
or outlet pipe to prevent the backward
flow of water or wastewater. Generally
used on sewer outlets into streams to
prevent backward flow during times of
flood or high tide. Also called a tide
gate.
The quantity of oxygen used in the bio-
chemical oxidation of organic matter in
a specified time, at a specified tem-
perature, and under specified conditions.
The quantity that can be contained
exactly, or the rate of flow--that can
be carried exactly.
A sewer intended to receive both waste-
water and storm or surface water runoff.
A mixture of surface runoff and other
wastewater such,as domestic or indus-
trial wastewater.
The period of time required for storm
runoff to flow from the most: remote
point of catchment or drainage area to
the outlet or point under consideration.
Cathode ray tube. Television for visual
data presentation.
A unit of measure of the rate of liquid
flow past a given point equal to one
cubic foot in one second.
To drain or remove water or wastewater
from an enclosure.
The area served by a sewer system or
watercourse receiving storm and surface
water.
162
-------
dry weather flow
equivalent uniform
depth "(E.U rPTT
flashboard
gravity sewer
hardware
In-system storage
interceptor
interface
milligrams per
liter (mg/1)
million gallons
per day (MGD)
outfall
pollutional load
The flow of wastewater in a combined sewer
during dry weather. Such flow consists mainly
of wastewater, with no storm, wastewater
included.
The average amount of rainfall over an area
developed from the constituent rain gage
stations and their associated Thie.ssen Poly-
gons contained within the network of gaging ~
stations.
A temporary barrier, of relatively low height
and usually constructed of horizontal wooden
planks, placed along the crest of a dam to
prevent inflow into combined sewers due to
high river levels.
A sewer in which the wastewater runs on
descending gradients from source to outlet,
and where no pumping is required.
The physical equipment and devices which
comprise a computer or computer system
component.
Unfilled, enclosed volumes within a sewer
system capable of accepting and retaining
wastewater for a period of time.
A sewer that receives dry weather flow from
a number of tranverse sewers and additional
predetermined quantities of storm water and
conducts such waters to a point of treatment.
A common boundary between parts of a computer
system.
A unit of concentration of wastewater consti-
tuent. It is 0.001 gram in 1 liter of water.
A unit of measure of the rate of liquid flow
past a given point equal to one million
gallons in one day.
The point, location, or structure where
wastewater or drainage discharges from a sewer.
The quantity of material in a waste stream
that requires treatment or exerts an adverse
effect on the receiving stream.
163
-------
regulator
sanitary wastewater
sluice gate
software
sewer system
storm wastewater
Thiessen polygon
translatory wave
uncontrolled storage
wastewater
watershed
- A device for regulating the flow into
an interceptor from a combined sewer.
- Domestic wastewater with storm and
surface water excluded.
- A vertically sliding gate of any shape
used to control or shut off flow to a
sewer or in a sewer.
- The programs or instructions which con-
trol the hardware to perform some com-
puter operation.
- Collectively, all of the property
involved in the operation of a sewer
utility. It includes land, wastewater
lines and appurtenances, pumping sta-
tions, treatment works'and general pro-
perty. Also referred to as a sewerage
system or wastewater collection system.
- .That portion of liquid, resulting from
precipitation runoff, flowing in com-
bined sewers, during or after a period
of rainfall.
- A device for determining the zone
within which data taken at a rain gage
station are applicable in a ne1:work of
gaging stations.
- A moving or advancing wave or series of
waves that tend to overtake eacih other
and form a single larger wave. It is
'caused by any sudden change in condi-
tions of flow.
- Storage not controlled by any remotely
operated gates but depending entirely
on weir or river elevations.
- The spent water of a community. It may
be a combination of the liquid and
water-carried wastes from residences,
commercial buildings, industrial plants
and institutions together with any storm
water that may be present.
- The divide between drainage basins.
164
-------
SECTION XV
APPENDICES
NO.
TITLE
PAGE
A.
GRAB SAMPLING PROGRAM RESULTS
Table 1: Average of Daily Grab Samples,
June 68 - December 68
Table 2: Average of Daily Grab Samples,
January 69 - July 69
Table 3: Average of Daily Grab Samples,
August 69 - March 70
166
168
170
B.
MONTHLY OVERFLOW REPORT
Graphical Summary
Overflow Report
Equivalent Uniform Depth of Rainfall
172
173
177
165
-------
APPENDIX A
TABLE 1
Average of Daily Grab Samples - 1968
Sewer Location
Pembroke
Prisbee E.S.
Frisbee W.S.
7 Mile E.S.
7 Mile W.S.
McNichols
Puritan E.S.
Puritan W.S.
Fenkell-Lyndon
Schoolcraft
Glendale
Plymouth
W. Chicago
Joy Road
Tireman
Hubbell-Southfield
Dearborn (Miller Rd. )
Baby Creek
Flora
Carbon
Pulaski
Dearborn Avenue
Schroeder
Dragoon
Campbell
Morrell
Ferdinand
Sunmitt
Scotten
Susp. Sol.
mg/1
195
194
192
532
790
180
332
335,
423
180
482
592
1350
915
387
78
270
502
238
452
817
733
223
702
185 ;
166
146
327
195
BOD
mg/1
111
90
148
458
267
158
181
234
184
192
113
298
197
•202
149
43
144
227
162
109
181
222
97
203
107
125
111
183
95
Tot.P
ng/1
14.8
10.2
16.1
15.0
16.8
11.2
18.2
16.6
15.4
16.3
11.4
16.3
15.9
16.4
15i3
3.1
7.4
7.9
8.3
11.3
10.4
10.6
10.5
4.3
8.0
4.3
8.8
4.4
4.7
Phenols
ug/1
78
79
104
177
151
117 '
163
144
137
214
81
89
89
111
195
75
9700
2775
235
200
276
129
193
236
348
. 109 ;
144
238
145
Oil & Grease
mg/1
34
32
70
75
83
95
58
86
49
48 .
67
230
100
100
26
22
55
2775
1395
62
140
77
116
20
122
65
84
82
443
166
-------
• APPENDIX A
TABLE 1
(Continued)
Sewer Location
McKinstry
Swain
W. Gd. Blvd.
24th
21st
18th
12th
llth
3rd
First-Hamilton
Woodward
Brush-Bates
St. Antoine
Hastings
Rivard
Riopelle
Orleans
St. Aiibin
Dubois
Chene
Adair
Leib
Mt. Elliott
Helen
Iroquois
McClellan
Conner
Ashland'
Manistique
Fox Creek
Susp . Sol .
ng/1
203
69
• 117
149
498
648
387
92
239
140
155
342
53
180
342
142
1005
335
234
336
. 333
165
232
209
182
220
348
327
384
411
BOD
rag/1
92
21
88
217
270
390
208
116
174
186
38
221
120
90
201
68
723
299
305
233
233
109
208
170
108
148
84
156
146.
204
Tot.P
nE/1
5.4
4.7
5.1
6.9
7.5
7.0
4.4
5.9
4.9
2.7
2.9
3.6
3.1
7.1
5.4
4.1
12.7
8.6
5.9
8.8
8.0
5.8
6.6
9.5
6.9
6.3
2.7
6.1
8.8
6.7
Phenols
ug/1
183
227
60
78
113
125
98
58
111
39
110
136
81
147
115
98
87
76
220
185
223
190
166
164
156
169
278
220
613
232
Oils £ Grease
UK/1
177
11
25
104
858
99
116
14 ,
163
31
38
84
23
52
39
36
301
127
66
408
60
25
71
45
34
51
123
40
52
88
'167
-------
APPENDIX A
TABLE ?
Average of Daily Grab Samples - 1969_
Sewer Location
Pentoroke
Erisbee E.S.
Frisbee W.S.
7 Mile E.S.
7 Mile W.S.
McNichols
Puritan E.S.
Puritan W.S.
FerikeH-Lyndon
Schoolcraft
Glendale
Plymouth
W. Chicago
Joy Road
Tireman
Hubbell-Southfield
Dearborn (Miller Rd. )
Baby Creek
Flora
Carbon
Pulaski
Ifearborn Avenue
Schroeder
Dragoon
Campbell
Morrell
Ferdinand
Sumnitt
Scotten
Susp. Sol.
mg/i
149
105
117
172
70
115
85
253
411
302
399
200
163
147
170
78 •
402
446
294
309
453
115
239
625
205
197
213
163
204
BOD
mg/1
198
192
120
80
211
98
110
99
134
104
64
260
111
107
122
43
166
109
52
110
528
230
251
246
151
154
259
205
29
Tot.P
mg/1
13.3
12.9
9.6
11.4
11.4
6.8
11.5
10.6
13.6
15.0
12.5
17.3
9.0
13.5
15.4
3.1
8.3
7.6
6.8
8.5
7.3
9.3
7.5
8.0
7.8
5.9
8.2
3.1
8.8
Phenols
ug/1
130
222
180
212
326
163
212
218
239
286
159
247
239
148
149,
75
1250
963
257
214
99 •
249
269
142
165
181
109
152
125
Oil & Grease
mg/1
. 57
53
29
14
60
24
19
19
31
26
28
42
25
24
19
19
64
121
182
34
689
41
81
680
87
74
550
89
29
168
-------
APPENDIX A
TABLE 2
(Continued)
Sewer Location
McKinstry
Swain
W. Gd. Blvd.
24th
21st
18th
12th
'llth
3rd :
First-Hamilton
Woodward
Brush-Bates
St. Antoine
Hastings
Rivard
Riopelle
Orleans
St. Aubin
Dubois
Chene
Adair
Leib
Mt. Elliott
Helen
Iroquois
McClellan
Conner
Ashland
Manistique
Fox Creek
Susp. Sol.
rog/1
471
189
210
169
271
762
80
227 ,
422
184
413
225
119
218
142
• 224
1005
729
267
338
460
148
229
268
240
268
248
257
101
228
BOP
rag/1
213
84
128
86
210
328
24
139
152
76
103
206
49
163
124
169
730
424
350
233
226 •
109
157
157
199
198
84
195 .
Ill
146
Tot.P
rag/1
6.5
3.2
6.4
4.8
5.0
6.1
2.0
8.5
4.3
3.7
4. '6
3.1
3.5
5.4
5.2
8.2
11.9
10.5
6.7
8.9
7.0
5.8
7.6
11.5
8.1
9.1
3.7
5.8
4.0
4.8
Phenols
ug/1
199
329
191
179
131
188
81
49
107
18
122
115
73
206
120
243
70
112
265
185
236
173
179
105
96
95
322
146
130
228
Oils £ Grease
rag/1
188
23
44
91
162
132
13
38
157
45
65
33
14
57
21
133
240
73
81
409
91
25
44
51
55
35
43
27
22
26
169
-------
APPENDIX A
TABLE 3
Average of Daily Grab Samples - 1970
Sewer1 Location
Pembroke
Frisbee E.S.
Frisbee W.S.
7 Mile E.S.
7 Mile W.S.
McNichols
Puritan E.S.
Puritan W.S.
FenkeH-Lyndon
Schoolcraft
Glendale
Plymouth
W. Chicago
Joy Road
Tireman
Hubbell-Southfield
Dearborn .(Miller Rd. )
Baby Creek
Flora
Carbon
Pulaski
Dearborn Avenue
Schroeder
Dragoon
Campbell
Morrell
Ferdinand
Summitt
Scotten
Susp. Sol.
mg/1
169
278
307
746
5069
707
168
238
256
435
4294
1453
318
25,789
348
2384
342
152
217
1101
408
83
285
359
138
104
410 :
6115
72
BOD
mg/i
293
225
269
268
382
184
183
157
259
292
348
211
260
270
293
302
174
156
154
230
233
168
183
187
140
152
99
272
75
Tot.P
mg/1
13.7
•8.0
13.0
10.6
25.8
10.4
18.2
14.2
' 15.5
20.4
38.0
48.2
17.0
9.9
5.6
6.2
11.5
8.6
2.3
7.3
8.1
6.7
5.2
8.9
'17.6
, 3.0
Phenols
ug/1
108
78
156
88
155
234
99
230
120
218
60
155
.,
.-
— — —
460
1358
1500
160
490
98
68
100
279
134
224
170
320
195
Oil £ Grease
mg/1
48
296
48
32
113
112
177
53
51 '
63
176
184
116
543
81
253
179
553
65
407
185
58
112
81
99
55
136
428
25
170
-------
APPENDIX A
TABLE 3
(Continued)
Sewer Location
McKinstry
Swain
W. Gd. Blvd.
24th '
21st
18th
12th
llth
3rd :
First-Hamilton
Woodward
Brush-Bates
St. Antoine
Hastings
Rivard
Riopelle
Orleans
St. Aubin
Dubois
Chene
Adair
Leib
Mt. Elliott
Helen
Iroquois
McClellan
Conner
Ashland
Manistique
Fox Creek
Susp. Sol.
ng/1
465
53
162
181.
. 177
528
289
143
213
69
256
382
29
188
595
234
1006
607
322
367
382
399
. 241
133
109
156
103
BOD
rag/1
238
127
102
120
166
317
235
167
237 .
187
229
259
79
237
394
354
727
541
465
258
269
327
279
127
138
163
118
Tot.P
mg/1
7.2
1.5
4.1
4.5
4.5
. 5.1
5.7
8.3
4.0
4.6
4.3
4.6
4.0
4.0
4.6
8.4
8.8
6.9
6.1
15.7
10.8
13.9
9.8
7o4
9.8
8.6
8.0
Phenols
ug/1
189
• 360
278
442
117
250
215
185
131
124
410
269
161'
169
172
154
161
179
132
•258
199
370
269
141
214 •
605
146
Oils S Grease
mg/1
173
20
,36
25
33
81
160
31-
42
31
78 ,
72
30
82
68
98
229
94
73
142 ,
61
. — —
—
87
75
37
. —
. 33
25
62
171
-------
APPENDIX B
172
-------
APPENDIX B
• z
t~O
a.**
ui»-
OU
: |U
of. to
UI
l-_l
<0
oz
0:0
O
a:.
a:
a:
uj
O
a
3t Z
O O
ui'a:
> •=>
o a
CM o
O I-
«
o a:
ui
x »- ae
t- »-ui
o. «i-
z: < u
a: a:
o t- in
u. i-
_. ~ <
z z o
z> o
••*
> •- o
•-. «J O
^ z ui
a o v:
UJ ~HJ
< I UJ
or CD en
O Of. •£.
o ui
5 o •—
0> 3 • W *
»-« — o: >- •-«.
• ui o
o eg »- •
Oil- Z O
I — —
CD s: i
• to i-
Q o: z— •
UI UI UI
i-o z
_l XUI I- -I
_i u o z _i
ui o o ui
O CE *J *J
o —
z «n o -i z
»• Z QC -J '-'
< UJ < UJ <
Qj'lfl SO Qi
uio
aa.
o
H
ui m
_» o
ui a:
u ui
i-
z <
•- 3
^C 3k£
tK U
o cc
• z
o o
u
to
o
a.
ui
C£
o
a:
ui
UJ •—
«/» U)
• in
o -i
o z ta
o>
tc z
3 o
is »-.
^ Wf "IT
3 Z
kJ O
- z
a. z
>- ui
ui
1 to
u o o
o fit.
o
ui r-
i- I
< «-«
o o
ao
O
UJ
ao
CM
03
UI
o
o
tc
•
UI
CD
O
O
r^
I
o
0
f1^
1
o
I
to
o
o
^
1
o
1
ao
0
0
^^
1
0
1
00
o
o
^*
1
r-
o
1
CO
0
0
1
ao
0
1
03
O
O
f^B
1
C^
o
1
00
o
o
f~
1
o
*-»
1
oo
0
I I
.-« CM
I
99
O
I
ao
O
173
-------
APPENDIX B
00
* o
m oc
o a.
oc.
lO
• z
I-O
Q.—
Ult-
Q U
UJ
or «/>
UI
111
ce
<0
oz
(KO
CM
Ul
aa.
a: *
o H
a •
UJ N
a: •
3 M
o u
_l M
U- II
C£ M
UJ N
> B
O «
O
r-
o
13
uj at
Of
U) X
_j z
-I UJ
Ul -I
Z 3
o o
•O I
I i-
oc. t~- t-«
UJ UJ O O> oo -• _i or uj or
or t^-i^ _i o or
ui < or ui oc w u.
3 z < or or o
U) «-tU O UJ tD 3
m s Q -
I J X X >v | X | || »-
O^ m f^- in *i!^ ^^
• -f> H m r-9 CM CM CM fil
— O •
z a a _i v>
«-« 3 a: _i •-
x< • uj o a: t- > u,i
t- t- s o w uj >
«o.< a: _i
CO « UJ
u.
KZ
O O
I- U
•-•
zuj
UJ •
a. z
>- uj a. o. n. o. ui
i- «o
•o co IA o m
CA c*» -* in ift
I I I I I
uj
U)
I U)
a: 3
UJ O
> _i
o u.
UI
»-
o
•o
UJ
in
CO
o
r-
oo
o
UJQ.UJQ.UJO.Q.UJU]
U)
UJ
o
o
oc
O>-l>-ICM
-------
APPENDIX B
z
o
«-
I- O <
ur- u:
UJ O
_j o O
UJ CM Z
> m m o o o
o o co
ee.
t-
z
<
ee
o
•
o
eo
UJ
H- z
tat UJ
Q. Z
z: •-•
o
o
,
o o
z; i-
_
o <
t- O
M UJ
z: ^
UJ U
Z UJ
5 5
CO
u uj
—• a
o o
Z IK
« t-
UJ
u
00
o
U.UJ
Z Z
<
o:
o
o
UJ
o: <
ui CK
O O «•! O O "H O
UJO
oo.
o: m
kl UJ : I i i
u •— i*- •* «^
tf)
•o
m ru
«o CM
O ex
mo
I l
«A
UJ
<,
K
11.
O
I
\J
oe
o
o.
UJ
.£>£
O
IK
III
O
O
cc
a.
O
f-
o>
in
o
o
oo
zo:
•-•i-
ocz
oo
i-u
•~«
ziu
ot-
O — 1 «•* «"«
Z CO
cc
z
o
£K 1- >- UI
3 Z «CQ£
W •- 3
V)
UJ
O
UI •
0. Z
> LJ tJ O-
i- •»
• sT CM
z •* m
uj n i
oe 3
U] O
o u.
UI
i—
<
o
•-« •-«
UJ >•-
IS «)
13
O UI OL
CC 5^ 3t
or>
u. o: a.
O CD
s: Q
3 UJO
— Q. O
»-• 3
X ic
u <
« O
3
U1UJ
m o •
in ^* u^
II •
to CD a.
•*
•-•
UJ
o
3
o
cc
u.
o
3
»••
•«!
. X
U
•
3
IU
m
in
1
CD
r*
0
r-
I
o
CM
1
oo
0
f-«
UJ
o
3J
0
cc.
u.
o
3
*^
t-i
X
u
•
3 .
UJ
m
m
CD
r->
o
r-
1
t-i
CM
1
oo
O
•H
UI
o
3)
o
a:
u.
O
3
*^
*-4
X
U
•
3
IU
m
in
m
>-i
o
P»
I
CM
CM
I
oo
0
UI
O
o
3
UJ
m
m
o ^«
UJ
2:
i I- <
UJ
in O
O O
CD m
U
UJ
UJ
•
UJ
0
^~
t
o
CM
|
co
0
o
f-
1
t-l
CM
|
oo
O
o
1
CM
CM
oo
0
o
r-
1
c«v
CM
|
OO
o
o
r-
-*
CM
1
ao
O
o
r-
I
in
CM
1
O
0
r-
1
«o
CM
1
OO
0
o
f—
1
f»
01
1
ao
o
o
r—
1
oo
CM
1
ao
O
o
1
o*
CM
1
03
0
175
-------
APPENDIX B
• z
t-O
O.M
Wl-
OU
kl
CC.U)
Ul
III
te
3:0:
oz
a: o
i-u
UJ
ZZ
O
013
•a-
UJ
a.
iij,
u
z
•••
<
Of.
o
•
o
UJ
UJ O
QO.
xz
o o m
_1 «— CO
a:
uj of
> u
OQ
a:
ui ui o o o
3«M I I I
UJ •-••-« m •-«
Of
O
a.
UJ
ee
§
a.
of
UJ
o
• to
O _I€M
CO —
CD
at z —
3 o •
o •-•»—
3 OS I- _J
< UJ < Ul
to
UJ
o
a:
u
X
o
a:
Qu
«J
3
Of
•*> «o
cr> o»
o.
_
3 2k
u aeo
ZZ
o o
.
01 Z
O O
U)
O
••*
I-
<
u
o
r>zi-
oj'-'U
• 3
aoa:
in
13 O W
< MOUIZ
o: 0:00
(9UI OO'U
z uj
«JU
H D
Ct 0£
tf)W)
tf)
V> O
z o:
•••»-
a: z
oo
I--U
»•«
ZUJ
SUJ
wo:
»-
10 Q
>-Z
V) - UJIU
i- m-
o
•H
CO
ec
»4
U)
H- «
O »-
•<
o
Ul Ul
O* ^5
a:
<
at
ui
«o
a: 3e
in o «o
o u_
w>
1
z
o
111
o
o
ec
t-
u
UJ
U1
UJ
o
a at
ui
z o
o r-
Q u.
uio
ac
01-
o <
a
ui
U. 3
ce u
> o
o
U.
Of
o
O£ *-• UIUI<
30 ~ - -
a z
OU.
»-• o* *~i ui
i- ceo
< an- t-
ec a. oca
UJ CC O ZUJ
a. o x«-o
o «~» u>uio
A * 00
zee a: ec
O O
U«J
>->-
«
3: 2
X X
190
•>-«'•-«
XX
00
I- O O '•< O
o a. a. u. a.
z »-
O' O 3 O
Ul Z Z O Z
at
cc o «
O Z M.
UIUI
33
eta
oca:
ui ui
s§
a. a.
o o
ZUI I-
ui > 3.
«/> M O
o i
Z Z
•-O
i -co:
IAOUI
O O o
I O
CO OQ
I
O
00
o
o
r-
r-»
cm
09
o
176
-------
APPENDIX B
• z
1- O
Q. —
- UJ 1-
Q U
UJ
at m
UJ
t_ i
< o
3 at
t«
O Z
o: o
ui u
^r 7
rt
W M
•-« 1—
O 3
t— j
UJ O
o o.
0 O
Z O-
cc z
0 0
tr U
Z UJ
O t-
IUJ
UJ at
I/} Q
<
r-l
Ul
^i
Q.
i
<
Z
< —
at a
Si
UJ
x z
a z.
UJQ
f UNIFORM 0
[SSEN POLYG
Z UJ
.UJ X
j t-
^ ~*
^
i
UJ
o
I
(HI
o
CO
o
(S
I
o
r-
l
o
00
o
•
• v a
SATURDA
08-O1-7
•
i
H
N
•
H
It >
It <
H a
n a:
a u.
n
M
H
II
II
II
n >•
H <
II O
ii in
a at
II 3
II X
u t-
u
ji
u
u
n >-
u <
II Q
n m
n ui
II Z
n a
II Ul
II 3
n
n
u
H
H
. II >•
II <
II O
ii m
It Ul
II 3
II h-
II
II
H
tt
:i v
N <
II *"*
II Z
II O
ii x.
.!
n
n
H
II
II
II >•
II <
II 0
N Z
H 3
H in
u
II
tt
II
N
H
U
H
g
< M <
H r
X II
M (
H
H
« f
H <
at N
N I
M
U
< II
II '
X II
U
Q£ H
II
H
n
n
at n
it
M
H
< II
H
x n
it
at H
it
u
n
ii
at ii
u
u
u
< n
X II
it
at ii
H
n
. u
at n
u
u
u
< u
ii
X II
ii
at u
n
n
u
ii
at n
u
n
it
< H
U
X 11
H
at H
H
H
01
Qt tt
It
II
II
< II
II
X II
It
at u
H
II
II
H
at ii
u
u
UJ N
O N
< H
O N
tt
3 O f» « O r- •* f- J
3OOOOOCOOO tt
II ^14-
N • •
-iorcM<-
3 §O
m
-------
APPENDIX B
• z
UI I
ec.
« >- o
<«oooooooooo«
u < r- x ii
II O I
OOOOOOOOOO N
H O O M O II
ii a: - r- x «
H OOOOOOOOOO H
N <
II O O H O II
ii a r- a: H
MOOOOOOOOOOIIO-»IIOU
ii o r- ii o ii
< o
§
N OOOOOOOOOO H
II OOOOOOOOOO II
II OOOOOOOOOO II
III O
a a.
H >- O
< N OOOOOOOOOO II
ii «c r- x H
II O O O O O O O O O O II
ii a
O O H O II
u uj - o
< n oooooooooo u
H •« r- x »
II OOOOOOOOOO II
n oooooooooo ii o
H o o n on
ii o ii
UJ UJ
< <
O I-
a. z
ui O
n o
II UJ O
n 3 a: u
u 0000000,000 n
II OOOOOOOOOO II
II QOOpOOOOOO II
-4- II
-1 -I
o o
o a.
u.
M Z
Z UJ
3 W
U)
I— •-«
Z UJ
~ O
z a:
UJ OC,
U) O
H >- r~ x ii
N OOOOr-4OO-*OO II
H a
H U) O
H UI I
H 3 00
ii in o H
nooooooo--«oonirv-*iioii
H • • II • II
II <-l f- II O H
IIOOOOr-IOOaOOOH
HOOOOOOOOOOH
HOOOOOOOOOOH
O O O O O II
< H OOOOOOOOOO
H a N
< II I I
I I
I N
3: H
(9 H ac.ac.oc.ac.oc.oc.BC.oc.oc.oc. n oo n o H
a o: a M cr N
o o
z z
* z
Of.
U} 10 UJ
H a N
o oo
178
-------
APPENDIX B
»- o
a. -•
UJI-
UJ
: to
< o
s a:
O Z
a: O
in
•s. z
o
m
ui
ui o
aa.
: — o
' O f-
O 1
. I m
UJ I
X S. . eo
i- o
a. z
Ul O X
Q IS O
>- 15
Z. -J C
0:O K
O Q. X
U. I-
-•- z
z u» &
o i
i- ••• o
z uj o
Ul X I
J I— CO
«t — o
o
UJ
o o
z or
-•i-
or z
OO
I-U
s: c
mo
>- z
SATURDAY I
it
*
H >•
II <
II 0
n u.
n
n
a
H
M >-
U <
n O
n to
n or
H O
II X
H 1—
II
H
H
o
r- x
m or
l
00
0
or
o
r- x
•* or
I-H
1
CO
0
QC
O
r- x
1
tn or
•-•
l
CO
o
or
SSERKXKSSBSSS
WEDNESDAY 1
08-12-70 1
R R X A i
TUESDAY
MONDAY ]
o
IV X
i
•-i or
i
CO
o
or
0
l>- X
1
o or
r-l
1
CD
O
o
>- 1^ X
•f 1
a o or
Z 0
•O 00
0
or
UJ
o
II
II
II
II
II
II
II
U
N
n
n
*
it
ii
n
n
n
n
H
11
II
B
H
II
a
u
n
n
n
u
u
u
n
u
ii
u
H
11
n
n
N
II
II
H
II
N
9
U
II
It
n
H
M
H
M
H
II
II
II
H
II
n
H
n
ii
H
II
H
n
*
II
II
II
II
II
II
O
o
0
o
c
o
o
o
o
c
c
c
o
o
•0
o
c
c
c
o
z
o
o
o
o
o
o
o
o
o
o
c
c.
o
o
o
o
o
0
o
c
c
f-t
o
or
o o
o o
0 0
o o
0 0
o o
o o
c o
0 O
o o
o o
o o
o o
o c
o c
z z
0 0
0 O
o o
o en
o o>
c o
z z
c -a-
o o
0 C
C 0
o c
o o
o o
o c
O 0
o o
O 0
c o
o o
o o
O 0
o o
0 O
o.o
0 O
O 0
o o
o o
o o
O 0
o o
0 0
0 O
O 0
o o
o. o
o o
0 0
o o
o o
0 O
o o
z z
0 0
o c
o o
CO O>
•* CM
f> l«-
«-• o
IM m
O rH
0 0
o o
c o
0 0
0 O
0 C
O 0
o o
c o
o o
o o
O 0
O CA
c •-•
O 0
o *-•
c c
c o
o
o
o
0
o
o
o
0
o
o
o
o
o
o
0
z
o
o
0>
o
o
z
0
0
0
o
o
o
o
o
o
0
c
o
o
0
o
o
0
o
c
o
CM m -» *> o
o o o o o
11(11
or o; or tr or
o o
o o
o o
o o
O 0
o o
0 O
C 0
o o
0 0
o o
o o
o o
c o
0 0
z. z
o o
o o
o o
o o
o o
o c
z z
o o
o o
0 0
o c
0 C
o o
0 O
c o
c o
o c
o c;
o c
o c
o o
o o
0 O
o o
o o
O 0
o o
o o
c o
o o
0 O
0 O
o o
o* o
CM O
0 0
z
CM O
o o
0 O
o o
o o
o c
z z
o o
o o
6 o
o o
o c
o o
o o
o o
o o
o o
0 O
o o
o o
O 0
o o o o
o o
O 0
0 O
o o
C 0
o c
0 0
C 0
oo
O 0
o c
o o
«-« tMsn -it
•-• «•* •-••-•
1 1 1 0
e: or tr e:
o
o
o
o
o
o
oc
o
o
r-
-a*
•-•
o
r--
CM
•^
r- 0
o
0
o
o>
CM
0
.*
0 r-
•-4
'
•
o
o
o
o r^
•#
t~«
fl O
•-• •*
• •
O f-
•-•
3
o
or
CM
o
or
o
o
o
o
o
o
en
0
o
»
2
o
o
o
o
o
o
c
o
o
o
2
o
Ul U
o o
o. a.
»— ^~
v>
z
<
C
x
t-
a.
UJ
o
OL
O
U.
X »-
z
a: < ui
o o >
s: S' o
300
IO Ul UJ
o oo
or
-------
APPENDIX B
1- O
a. •-
UJI-
Ul
a: w
ui
t- _j
< o
*£.
o z
0:0 -s-
tu uj
•s.z. o
0 <
>-< 1—
0 3
ac-i
t- j
til O
o a
-J 5.:
_J
u.
z
o: o r>-
O 1
U. X C*
O t- ~ °
3
UI
_1
O O
ZE
ec. z
00
»«
Z fcl
Ol-
zr o
• z
X U
U K
t—
we
> z
<
>- o
a 1
a: CM oc
3 - O 1
a i i
O CK 1
cr CM i
3 1 1
X CD 1
1- O t
I
1
1
< 1
>- O 1
a i i
w o> o: i
UI <-l 1
Z 1 1
a co
UI O
<
o
< i
a «o tr
U) •-«
Ul 1
3 CO
h- O
' <
C
•< 1
Z *-t
0 1
S. CO
o
<
0
< 1
o « tr
z«-«
3 1
O «:
O
u
o
a
CM^r^ascMr^^iOcM-*-
ooo>ocncnin
(Mf-ICMCMCM«-ICMr-l«-ICM
OOOOOOOOOO
1 «
1 OO-< i
1 •HO«M-*r-l**«-i'*'-ICO 1
i cMO<-tr-tf>4>mcMt-nr*-*
OOOOOO<-tt-»-«>-l
teiiii.iiii
ec&eKaccccrcacc:
•o 4-
• •
o r—
m *
r*
t
r- •* i
• • 1
O.I-- 1
•* 1
r-l 1
1
1
1
1
.
1
0 •* 1
• • 1
o r- i
•a- i
ft i
i
i
i
i
i
i
CO O 1
• • i
•£• U> 1
j- m i
fH 1
i
i
i
i
i
i
O •*,!
• «'•
O t-- 1
•a- i
f i
i
0 •*
°s
•-I
o •«•
• *
o r-
•«•
ft
• •
^i m
oi
ec. tc.
CM
•
O
II
' X
*
O 1
• 1
O 1
1
H
u
1
1
1
1
H
O 1
• 1
O 1
1
1
1
1
1
1
1
1
to 1
• n
O 1
i
i
«
K
1
H
1
O
•
1 O
1
1
0
0
o
•
o
•
m
|
'
.
-
•
.
•
•
•
,
"
UJ UJ
<<
a. a.
,
i- t-
•£. 3513
3 3O
u> m ui
mam
ft
-------
APPENDIX B
l-O
Q. -•
Ull-
ou
u
oc. - 3
^ J O
a: O ac.
o a. x
o
OO
l-O
<
>- 0
Q 1
DC (MK
3 01
1- 1
< eo
- O
0 1
i?
I a
1- O
<
>- 0
O I
U>
UI O
•<
O
< 1
Q w> ac
•n
1- O
<
O
< 1
O •» £E
i?
SE a>
O
<
O
< i
o « a:
ZCii
3 1
«o a
o
•
ui
o
o
r-iooooooomo
COOOOOOOO^O
t-tOOOOOOOOO
Z. ZZZZZZ Z
cnooooooomo
f-IOOOOOOOOO
oooooooooo
oooooooooo
o o o o o p o o o o
oooooooooo
oooooooooo
oooooooooo
oooooooooo
'
oooooooooo
oooooooooo
oooooooooo
oooooooooo
oooooooooo
oooooooooo
oooooooooo
O O O O O'O O O O O
oooooooooo
oooooooooo
oooooooooo
oooooooooo
oooooooooo
oooooooooo
oooooooooo
oooooooooo
oooooooooo
oooooooooo
ooooooooom
OOOOOOOOO-H
oooooooooo
OOOOOOOOOi-l
oooooooooo
oooooooooo
.
o<-o«to>moo«
«-l.-f- •*
O f-
•*
a-l
r<«M
oo
ac a:
'
0
•
o
•
o
0
0
o
0
o
o
•
o
.
'
o
o
o-
o
o
o
«
o
t£
UI UI
o o
a. a
O O
»-t «M
UI UI
O O
UI UI
UI U
UJ <*>
u.
z
o
x,
a.
ui
o
IK
O
<
X
ec.
U-li. <
O O >
.
3 3O
u> at UI
O O O
a: a: a:
181
-------
APPENDIX B
II
11 < X
B Q
u ce or
H 3
HI- N
H < ' H
U - X It
II < H
II Q ££ If
II «-• II
ii a: ti
II U. II
H n
n a: ii
N II
II H
i
H
1
1
1
II
II
II
II
II
II
It
II
II
II
II
3 II
j- n
t n
•4 II
O II
1 II
0 II
3 II
II
C II
3 II
3 II
C u
C H
H
H
3 »
- U
H
3 U
i u
B
C II
U
K
U
II
II
II
11
H
B
B
II
n
B
II
U
n
u
B
H
B
It
H
II
11
H
II
B
II
H
B
II
>- II
< X II
o u
IO a: II
OZ II
X 11
1— II
cr ii
n
n
n
< n
>• u
< X II
a u
to a: ii
UJ II
z n
O II
UJ II
2 a: ii
ii
n
< n
it
> X II
< H
o. o: H
to H
UI H
o n
1— H
Q; H
II
H
II
< II
O II
>- r~ x ii
< I u
O •-« CC. II
Z CO H
0 1 H
S cc u
O H.
H
B
«
< H
O H
< 1 B
o o c: u
z to B
3 1 H
UJ OC B
C B
t*s K
H
U
H
Ul B
O B
O B
B
O O O O
0 C O O
o o o o
O O 0 O
o o o o
o o o o
O O CO O
CO O O O
o o co o
z
«> O CO C
C O *-< O
0000
•-« CM CO <}•
O 0 0 O
1 1 1 1
ir <£ o: a:
coo
COG
C C G
0 O O
o o o
O 0 0
o o o
O C CO
000
z
O •-« CM
G G O
o o c
O O <-l
1 1 1
cc.oc.cc.
coo
o c o
ceo
coo
o o o
oop
0 0 O
r- c o
o c o
z z
-* o o
o G o
G G C
CM CO -3-
«-4 v-f «-fl
1 1 1
OC.CC.CC.
B
B
B
II
B
B
B
B
II
B
B
B
R
B
B
II
a
11
II
H
II
II
H
u
II
II
B
II
II
II
II
II
u
II
II
II
II
II
II
11
n
II
II
II
II
II
II
II
II
II
II
n
M
u
H
B
H
B
B
u
B
B
II
B
II
u
u
II
B
B
tl
n
II
II
H
H
II
II
B
II
II
II
H
H
II
u
B
11
o o
o •*
0 f-
*
r-l
-4" CO
•* -<
• •
•4- r-
00
•HCM
oo
oc. oc
•
B
B
u
B
B
B
B
u
B
B
II
B
H
B
II
II
H
II
H
H
II
II
II
II
II
•II
II
II
II
II
II
II
II
II
11
II
II
II
II
II
u
11
II
II
II
It
II
II
u
II
B
B
B
H
B
B
B
B
B
B
II
B
B
B
B
B
B
N
B
B
B
B
B
B
B
B
II
B
B
B
B
B
B
II
II
II
B
B
B
B
0
C
o
m
o
o
CO
o
oc.
B
B
B
B
II
B
B
B
II
B
B
II
It
II
B
II
II
II
II
II
n
u
it
n
II
u
ii
n
u
II
M
II
II
II
II
II
II
II '
U
II UJ
u O
H <
n a.
n
ii i—
II co
II . <
II -1
•II
II Z
II
B *H
H
U UJ
U h-
B O
H Z
It
U UJ
n uj
B (0
n
B *
U
B
U
B
B
B
B
B
B
B
B
B
B
B
B
tt
B
It
H
II
B
B
B
B
H
B
B
B
B
B
B
UJ
a.
i—
to
z
O _li
_l
CM <
U.
UJ Z
0 <
Z IX.
UJ U.
Ul O
in
Z »-:
OJ
Ul
0
•5.
cc.
o
u.
•-•
z
«r o
X t-
z
o:
s: s: 3
0 3 C,
-------
APPENDIX B
o
t-»
I-
<
t- O
Q. •-•
Ul»-
QU
UJ
ae. m i-
Ul
I- _l 111
uimsui-jmi— i—Q
o a uj m zm -
o _j
Ul Z
u <
••*
> z
ce. —
ui
M O
Ul
u. o
o z
a
UJ
M
<
in
z
O
o
a.
«— o
UJ Ul "»
UIKUZOXXX
t-ZI-«—3l-l-l-t-
o
cc
o
jr z:
oo
1- U
o
0
1 1 1 1 1 1 1 1 1 1
ix.oLae.ce.ff.cf.tc.ct.ee.ac.
lO O
>- z
o o
z z
a a
UI UJ
a: a:
UJ UJ
o o
o
CK1K
Ul UJ
u. u.
o
a.
m
UJ
ae.
ae.
O
u
o
m
a
z
UJ
in
ce.
o
w
UJ V
—• Q
UJ UJ
o: >
Z IK
•- o
O Q
o ui
>• Q
-i a:
o o
Q_ ^
Ul
Z (£
Ul
in _i
m _J
ui u.
xz
u. a:
O
a: o
< i-
183
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-670/2-75-020
-------
-------
Cr CD
O 2
< CD
CD Co
8. o
Q. 3:
3 "J
en O
rn
X Q. CD
PS Q.
co c
rn ~ ~ °
D
;£• s
CD x
~ Q>
§ ct *
CD Q. O
?• -i a-
S, CD m
3 Co 3
tQ CO 4Q
Q. --*
CD O)'
C ~
m >"
£) ^
^ r:
r^!
O -ri
Tl
O -o >
H 5! CD
0)
m
rn
(fl
T^
r «i
rn
m
0)
0)
03
O rr:
01
ro
C/J
O rn
w_ CD
O m 2,
5' <. 3D
Ct) 3
-— «> o
§1
3. rr
DO g. g
m _ ~!
o m
CD O
oo o £•
O 3
§ o
i m
z
o
o o
z
-------