WATER POLLUTION CONTROL RESEARCH SERIES • 11023 FDD 03/70
Rotary Vibratory Fine Screening
of
Combined Sewer Overflows
DEPARTMENT OF THE INTERIOR • FEDERAL WATER QUALITY ADMINISTRATION
-------
WATER POLLUTION CONTROL
The Water Pollution Control Research Reports describe the results and
progress in the control and abatement of pollution of our Nation's
waters. They provide a central source of information on the research,
development and demonstration activities of the Federal Water Quality
Administration, Department of the Interior, through in-house research
and grants and contracts with Federal, State, and local agencies,
research institutions, and industrial organizations.
Triplicate tear-out abstract cards are placed inside the back cover to
facilitate information retrieval. Space is provided on the card for
the user's accession number and for additional keywords.
Inquiries pertaining to Water Pollution Control Research Reports should
be directed to the Head, Project Reports System, Room 1108, Planning
and Resources Office, Office of Research and Development, Department
of the Interior, Federal Water Quality Administration, Washington, D.C.
2021*2.
Previously issued reports on the Storm and Combined Sewer Pollution
Control Progrram:
WP-20-11 Problems of Combined Sewer Facilities and Overflows -
1967.
WP-20-15 Water Pollution Aspects of Urban Runoff.
WP-20-16" Strainer/Filter Treatment of Combined Sewer Overflows.
WP-20-17 Dissolved Air Flotation Treatment of Combined SPWAT
Overflows.
WP-20-18 Improved Sealants for Infiltration Control.
WP-20-21 Selected Urban Storm Water Runoff Abstracts.
WP-20-22 Polymers for Sewer Flow Control.
ORD-U Combined Sewer Separation Using Pressure Sewers.
DAST-U Crazed Resin Filtration of Combined Sewer Overflows.
DAST-5 Rotary Vibratory Fine Screening of Combined Sewer
Overflows.
DAST-6 Storm Water Problems and Control in Sanitary Sewers,
Oakland and Berkeley, California.
DAST-9 Sewer Infiltration Reduction by Zone Pumping.
DAST-13 Design of a Combined Sewer Fluidic Regulator.
DAST-25 Rapid-Flow Filter for Sewer Overflows.
DAST-29 Control of Pollution by Underwater Storage.
DAST-32 Stream Pollution and -Abatement from Combined Sewer
Overflows - Bucyrus, Ohio.
DAST-36 Storm and Combined Sewer Demonstration Projects -
January 1970.
DAST-37 Combined Sewer Overflow Seminar Papers.
-------
ROTARY VIBRATORY FINE SCREENING
OF
COMBINED SEWER OVERFLOWS
Primary Treatment of Storm Water Overflow
from Combined Sewers by High-Rate,
Fine-Mesh Screens
FEDERAL WATER QUALITY ADMINISTRATION
DEPARTMENT OF THE INTERIOR
CONTRACT 14-12-128
by
Cornell, Howland, Hayes and Merryfield
Consulting Engineers and Planners
Corvallis, Oregon 97330
Research and Development Program No. 11023 FDD
March 1970
-------
FWPCA Review Notice
This report has been reviewed by the Federal Water
Pollution Control Administration and approved for
publication. Approval does not signify that the contents
necessarily reflect the views and policies of the Federal
Water Pollution Control Administration.
n
-------
ABSTRACT
The objective of this study was to determine the feasibility, effectiveness, and economics
of employing high-rate, fine-mesh screening for primary treatment of storm water
overflow from combined sewer systems.
The final form of the screening unit stands 63 inches high and has an outside diameter of
80 inches. The unit is fed by an 8-inch pipe carrying 1700 gpm (122 gal/min/ft2) which is
distributed to a 60-inch diameter rotating (60 rpm) stainless steel collar screen having 14
square feet of available screen area and a 165 mesh (105 micron opening, 47.1 percent
open area). The screen is backwashed at the rate of 0.235 gallons of backwash water per
1000 gallons of applied sewage.
Based on final performance tests run on dry-weather sewage, the unit is capable of 99
percent removal of floatable and settleable solids, 34 percent removal of total suspended
solids and 27 percent removal of COD. The screened effluent is typically 92 percent of
the influent flow.
On the basis of a scale-up design of a 25 mgd screening facility, the estimated cost of
treatment is 22 cents/1000 gallons. No finite cost comparisons were made with other
treatment methods; however, when compared to conventional primary sedimentation, the
selection of a screening facility as a treatment method is dependent on the value and
availability of land, the design capacity of the treatment facility, the character of rainfall
and runoff, and the available means of disinfection. It was observed that the proposed
screening facility required 1/10 to 1/20 the land required by a conventional primary
treatment plant.
This report was submitted in fulfillment of Contract No. 14-12-128 between the Federal
Water Pollution Control Administration and Cornell, Howland, Hayes and Merryfield.
-------
TABLE OF CONTENTS
SECTION TITLE PAGE
ABSTRACT Hi
TABLES v
FIGURES vi
1 CONCLUSIONS 1
2 RECOMMENDATIONS 4
3 INTRODUCTION 6
National Importance of Storm Water Overflows 6
Location Situation 7
Objectives 7
4 DEMONSTRATION PROCEDURE 8
Site Description . 8
Pilot Plant Operation 8
Sampling Program 15
Experimental Design and Data Reduction 17
5 INVESTIGATIONS 21
Characterization of Combined Sewage 21
Treatment Capabilities of Screening Unit 21
6 DISCUSSION OF RESULTS 25
Characterization of Combined Sewage 25
Treatment Capabilities of Screening Unit 26
7 PRELIMINARY DESIGN OF A SCREENING FACILITY 35
Design Criteria 35
Presentation of Proposed Screening Facility 37
Estimated Construction Cost of Seattle Facility 44
Estimated Annual Operation and Maintenance Costs 44
Estimated Total Annual Cost 44
Discussion of Feasibility 45
8 ACKNOWLEDGMENTS 47
9 GLOSSARY OF TERMS 48
10 APPENDIXES ' 51
Screen Specifications 52
Detergent Specifications 53
Experimental Data Presentation and Interpretation 54
-------
TABLES
NUMBER TITLE PAGE
1 Summary of Characterization of Combined Sewage,
Sullivan Gulch Pump Station, Portland, Oregon 22
2 Range and Level of Variables Tested 23
3 Comparison of Storm-Caused Combined Flow and Dry-Weather
Flow 27
4 Summary of Extended Test 33
5 Estimated Annual Operation and Maintenance'Costs 44
6 Estimated Total Annual Cost 45
Cl Summary of Responses, Experiment B, Original Form 57
C2 Summary of Main Effects and Interactions, Experiment B,
Original Form 57
C3 Summary of Data, Experiment C, Modification 1 61
C4 Summary of Responses, Experiment D, Modification 2 64
C5 Summary of Main Effect and Interactions, Experiment D,
Modification 2 64
-------
FIGURES
NUMBER TITLE PAGE
1 Experimental Riot Plant, Sullivan Gulch Pump Station,
Portland, Oregon 9
2 Daily Flows at Sullivan Gulch Pump Station,
Portland, Oregon -1969 10
3 Monthly Rainfall, Portland, Oregon -1969 11
4 General Layout, Sullivan Gulch Pump Station and Screening
Facility, Portland, Oregon 12
5 Original Screening Unit 14
6 Combined Sewage Overflow Screening Sampling Program 16
7 Experimental Design and Data Reduction 18
8 Development of a High-Rate, Fine-Mesh Screening Unit 24
9 Experimental Design and Observations of Final Experiment 28
10 Hydraulic Capacity of Screening Unit 30
11 Typical Screen Failures 31
12 Imhoff Cone Comparison of Flow Streams 34
13 Amount and Rate of Storm-Caused Combined Overflow at King Street
Regulator Drainage Basin, Seattle, Washington 36
14 Proposed Screening Facility, Perspective 38
15 Proposed Screening Facility, Site Plan 39
16 Proposed Screening Facility, Floor Plan and Section 40
17 Wash Mode 1, Proposed Screening Facility 42
18 Summary of Control System for Proposed Screening Facility 43
-------
FIGURES (Continued)
NUMBER TITLE PAGE
Cl Experiment B, Original Form 56
C2 Experiment C, Modification 1 60
C3 Experiment D, Modification 2 63
C4 Experiment D2> Modification 2 65
C5 Hydraulic Efficiency Versus Time, Modification 3 69
C6 Hydraulic Efficiency Versus Time, Modification 4 69
vii
-------
SECTION 1
CONCLUSIONS
1. High-rate, fine-mesh screening is an economically feasible method of treating
combined sewage overflows. When compared to conventional primary
sedimentation, however, the selection of a screening facility as a treatment method
is dependent on the value and availability of land, the design capacity of the
treatment facility, the character of rainfall and runoff, and the available means of
disinfection.
2. The characterization of storm-caused combined sewage and dry-weather combined
sewage did not reveal any unusual constituents which could affect the long-term
effectiveness of the screening unit. These characterizations were compiled on the
basis of several composite samples.
3. The short-term effectiveness of the screening unit is significantly reduced by the
presence of oil and grease in the combined sewage. Oil slugs were observed at least
once a day for a duration of approximately 5 minutes, and were of a concentration
substantial enough to make the sewage appear black in color. The presence of an oil
slug reduces the hydraulic capacity of the screening unit by as much as 50 percent.
Frequent backwashing during the presence of an oil slug will minimize this problem.
4. The vibratory horizontal screen is not required in screening combined sewage
overflow. The presence of the vibratory horizontal screen reduces the hydraulic
capacity of the unit and, in some cases, results in lower removal efficiencies (see
Appendix C).
5. The overall performance of the screening unit is a function of the mesh size of the
collar screen, the rotational speed of the collar screen, the strength and durability of
the collar screen material, and the backwash operation.
6. The removal efficiencies of the screening unit increases as the mesh of the collar
screen becomes finer, and as the volume of the feed applied to the screen increases.
For example, 31 percent removal of total suspended solids was observed at an
influent flow rate of 1200 gpm (86 gal/min/ft^) with a 105 mesh screen (167
micron opening), while 35 percent removal was observed at a flow rate of 2000 gpm
(143 gal/min/ft ) with a 230 mesh screen (74 micron opening).
7. The removal efficiencies of the screening unit are independent of the rotational
speed of the collar screen.
-------
8. The hydraulic efficiency of the screening unit increases as the rotational speed of
the collar screen increases, as the mesh of the collar screen becomes coarser, and as
the velocity of the feed approaching the screen increases.
9. The life of the collar screen decreases as the velocity of the feed approaching the
screen increases and as the mesh of the screen becomes finer. For example, the
screen life observed at an influent flow rate of 1200 gpm (86 gal/min/ft2) with a
105 mesh screen (167 micron opening) was more than four hours, while the screen
life at a flow rate of 2000 gpm (143 gal/min/ft2) with a 230 mesh screen (74 micron
opening) was less than four hours.
10. Approximately 90 percent of the screen failures were mechanical failures caused by
hydraulic overloading of the screen. The remaining 10 percent of the failures were
caused by punctures from objects present in the feed.
11. It is possible to produce a 165 mesh screen (105 micron opening, 45 percent open
area) with a probable life of 500 hours while operating at a flow rate of 1750 gpm
(2.5 mgd or 128 gal/min/ft2).
12. The use of a solution of hot water and liquid solvent,in lieu of steam, was found
necessary to obtain effective cleaning of the screens.
13. Of the solvents tested, a caustic solution was the most efficient solvent for
backwashing the collar screen.
14. Screen blinding decreases as the velocity of the feed approaching the screen
increases, as the mesh of the screen becomes coarser, as the frequency of backwash
increases, and as the rotational speed of the collar screen increases.
15. A minimum of approximately 4.5 feet of fluid head above the downstream water
surface of the screening unit is required for gravity flow operation.
16. Based on the intensity and duration of rainfall in the Seattle area, a screening
facility in the Pacific Northwest can be expected to be in operation approximately
1000 hours a year.
17. The collar screen material is the limiting component of the screening unit. When a
stronger and more durable screen material is developed, it will be possible to
increase the hydraulic and removal efficiency of the screening unit.
-------
18. The estimated construction cost for a 25 mgd screening facility is $496,000. The
estimated annual cost of operation and maintenance is $18,500. Based on a
25-year bond issue, with an interest rate of 6-1/2 percent, the total annual cost is
estimated to be $59,500, which puts the cost of treatment at 22 cents/1000
gallons assuming 271 million gallons of overflow a year are treated. These cost
figures are based on a preliminary design of a screening unit for Seattle,
Washington, which is presented in this report.
19. Based on the scale-up design of the Seattle facility, a screening facility will
require 1/10 to 1/20 the land that a conventional primary sedimentation plant.
-------
SECTION 2
RECOMMENDATIONS
1. It is recommended that a full-scale screening facility be designed and constructed to
demonstrate the feasibility of utilizing high-rate, fine-mesh screens in the treatment
of combined sewer overflows.
The results of this study have established the feasibility of the high-rate, fine-mesh
screens. The performance of the screens should now be demonstrated through the
design and operation of a full-scale facility. Based, in part, on the results of this
study, the equipment supplier has developed and tested a second generation unit.
The new unit is operated at 3 mgd (2100 gpm, 150 gal/min/ft2) with a 165 mesh
(105 micron opening) stainless steel screen with little or no deterioration in the
performance observed at the 2.5 mgd level. The equipment supplier has also
developed a new screen that has a probable life of about 500 hours. This represents
a hundredfold increase in life over that observed in this study.
During a period of demonstration, these new units could be tested and further
optimized with regard to inlet conditions, hydraulic capacity, screen life,
back washing technique, and control systems. The period of demonstration would
also yield firm cost and operational data.
2. As part of a final design effort for a full-scale facility, it is recommended that a
systems analysis be performed to investigate the compatibility of the electrical and
hydraulic machinery.
In the preliminary design of the full-scale facility presented in this study, it was a
relatively simple matter to design a control system to operate the facility. Likewise,
it was also relatively simple to design the hydraulic machinery required of the
facility. The compatibility of the two systems, however, is very difficult to predict.
It is therefore recommended that an analog simulator be employed to simulate the
operation of a screening facility. The results of this study may reveal some basic
problems in control that can be resolved prior to the completion of a final design.
3. It is recommended that flow measurement and sampling facilities be installed at all
combined sewage outfalls where installation of treatment facilities is anticipated.
Based on the experience of this study, continuous flow recording at an overflow
point prior to the design of a treatment facility would be of significant value in
determining both the design capacity of the facility and the total use of the facility.
In addition, sampling facilities would be helpful ki determining the character of the
-------
waste to be treated. Composite samples would yield a general description of the
waste, and grab samples could be collected to determine the quality and frequency
of any unusual constituents that may be present in the waste. If the installation of a
screening facility was anticipated, this information would be required for sizing of
screen materials and estimating the frequency and quality of backwashing.
4. It is recommended that a comprehensive study be conducted to determine an
efficient method of contacting a disinfectant with a treated effluent.
A major advantage in developing high-rate treatment equipment, like the proposed
screening facility, is the ability of the equipment to treat large volumes of waste in a
small area. This advantage would be negated, however, if conventional chlorine
contact times are required to provide disinfection. Based on the findings of this
study, the land required to provide conventional chlorination is 3 to 4 times that
required of the screening facility. In some cases, this requirement can be reduced or
eliminated by utilizing an existing outfall downstream of the facility for the contact
channel; however, this is normally the exception rather than the rule. Therefore, in
order to maintain the space advantage of high-rate treatment equipment, a high-rate
method of disinfection must be developed.
Currently, there is considerable research available describing the bactericidal
mechanism of several different disinfectants. Several of these studies indicate that
acceptable bacterial kills can be obtained with conventional disinfectants at contact
times of 10 minutes or less. Based on these observations, it is recommended that
additional research be performed to develop a contact chamber that will reproduce
these laboratory results in the field. It is believed this research will lead to a contact
chamber with two compartments. The first compartment would be a
mechanically-mixed rapid-mix tank with a detention time of less than 3 minutes.
This complete-mix tank would provide rapid and intimate contact between
disinfectant and effluent. The rapid-mix tank effluent would then enter a period of
quiescent contact provided by a plug-flow type tank with a detention time of less
than 15 minutes. It is this combination of two distinct flow regimes that approaches
many of the laboratory procedures used in bactericidal studies, and it is a type of
flow regime that may provide a more efficient and economical method of
disinfection.
-------
SECTION 3
INTRODUCTION
NATIONAL IMPORTANCE OF STORM WATER OVERFLOWS
The majority of the existing combined sewers throughout the nation do not have
adequate capacity during heavy storm periods to transport all waste and storm-caused
combined flows to a treatment facility. The overflow is bypassed to a receiving stream,
thus causing pollution in the nation's watercourses.
In 1967, the FWPCA published a report prepared by the American Public Works
Association titled, "Problems of Combined Sewer Faculties and Overflows."1 This
report reviewed the effects and means of correcting combined sewer overflows on a
national basis. Of the 200 million people now residing in the U. S., approximately 125
million are served by combined or separate sewer systems. Of the 125 million,
approximately 29 percent are served by combined sewers.
Combined sewers are designed to receive all types of waste flows, including storm
water. In determining the size of the combined sewer, it has been common engineering
practice to provide capacity for 3 to 5 times the dry-weather flow. During intensive
storm periods, however, the storm-caused combined flow may be 2 to 100 times the
dry-weather flow, making overflow conditions unavoidable. To compound the problem,
most treatment facilities are not designed to handle the hydraulic load of the
combined sewer and, therefore, are required to bypass a portion of the storm-caused
combined flow to protect the treatment facility and treatment process from damage.
The nation's treatment facilities bypass flows an estimated 350 hours during the year,
or about 4 percent of the total operation time. The pollutional impact of the
storm-caused combined overflow on the waters of the nation has been estimated as
equivalent to as much as 160 percent the strength of domestic sewage biochemical
oxygen demand (BOD). This amount creates a major source of pollution for the
nation's watercourses.
The cost to physically separate the storm water from the sanitary wastes through the
use of separate conduits has been estimated to be S48 billion. The development of an
alternative means of treatment could conceivably reduce this cost to one-third.
"Problems of Combined Sewer Facilities and Overflows 1967," Water Pollution
Control Research Series WP-20-11, U. S. Dept. of Interior, FWPCA, December 1, 1967.
2 Ibid.
-------
SECTION 4
DEMONSTRATION PROCEDURE
SITE DESCRIPTION
The screening facility is located adjacent to the Sullivan Gulch pump station in Portland,
Oregon, as shown on Figure 1. The Sullivan station serves a drainage basin of about
25,000 acres of Portland's metropolitan area, from which it pumps up to 53 million
gallons a day (mgd). The drainage basin is a residential area, with about 30,OQQ
single-family residences within its boundaries. A broad spectrum of services are available
within the basin to support the population. However, the automobile related services are
the most heavily represented in the drainage basin. This became visually apparent when
periodic dumps of waste oil appeared at the screening facility.
The 72-inch trunk sewer that drains the basin has a capacity of 53 mgd. The capacity of
the Sullivan station is adequate to handle this volume; therefore, bypassing storm-caused
combined flows is forbidden.
The daily flow variations at the Sullivan station are illustrated on Figure 2. Since
overflows are not allowed at the Sullivan station, the presence of storm-caused combined
sewage had to be estimated by the height of water in the pump sump. The assumed
overflow condition of 27.5 mgd and greater, shown on Figure 2, is represented by a
known level of water in the pump sump. This level of water was attained only during a
storm event; therefore, when this level was reached or exceeded it was assumed that
storm-caused combined sewage was present in the sewer.
Monthly rainfall records for the period of operation are shown on Figure 3.
PILOT PLANT OPERATION
GENERAL LAYOUT-Figure 4 illustrates the general layout of the screening facility and
its relation to the Sullivan pump station. The combined sewage flow comes to the station
in the 72-inch horseshoe trunk sewer. Before reaching the pump station, a portion of the
flow is diverted to a bypass channel where it passes through a coarse bar screen prior to
reaching the screening facility's feed pump sump. This diverted flow, which is now
defined as combined sewage overflow, is lifted to the screening units by two 2,100 gallon
per minute (gpm) vertical turbine pumps. After passing through the screening units, the
treated effluent and solids concentrate, or untreated effluent, are both returned to the
trunk sewer. In an actual installation, the treated effluent will be bypassed to the
receiving stream, and only the solids concentrate will be returned to the interceptor.
8
-------
<
SULLIVAN GULCH
PUMP STATION
SCREENING
FACILITY
FIGURE 1
EXPERIMENTAL PILOT PLANT
SULLIVAN GULCH PUMP STATION
PORTLAND. OREGON
-------
•
"ASSUMED STORM-CAUSED
COMBINED SEWAGE CONDITIONS'
B 1O IB 2O 28
JANUARY
B IO IB 2O 29
FEBRUARY
B IO IB 2O 2B
MARCH
B 10 IB 20 23
APRIL
B IO IB 2O 2B
MAY
B 10 IB 20 28
JUNE
FIGURE 2
DAILY FLOWS AT SULLIVAN GULCH PUMP STATION
PORTLAND, OREGON - 1969
-------
11
1C
m
M
U
z
z
< 3
cc
2 E
1 *.i*_V*"*
•t*">" *" "
JAN. FEB. MAR. APRIL MAY JUNE JULY AUG. SEPT. OCT. NOV. DEC
MONTHLY AVERAGE
MONTHLY 1969 TOTALS
FIGURE 3
MONTHLY RAINFALL
PORTLAND, OREGON - 1969
-------
SCALE 1" - 20-
SULLIVAN
GULCH
PUMP
STATION
2,000 G.P.M. VERTICAL
TURBINE PUMPS
24 INCH DIAMETER
DISCHARGE
DISCHARGE
CHANNELS
SULLIVAN
GULCH
SCREEN
BUILDING
SCREENING FACILITY
- 8 INCH SCREENINGS DISCHARGE
12 INCH INFLUENT PIPES
SCREENS
72 INCH HORSESHOE TRUNK
FIGURE 4
GENERAL LAYOUT
SULLIVAN GULCH PUMP STATION
AND SCREENING FACILITY
PORTLAND, OREGON
12
-------
One of the specific goals of this study was to define the complete hydrograph of flow in
the combined trunk sewer during storm periods. During the initial stages of testing,
however, it became apparent that the flow in the trunk sewer was significantly influenced
by the on-off operation of the Sullivan Gulch pump station and the general layout of the
test site. All attempts to install a flow measuring device above this zone of influence
proved impractical. Because of these factors, the flow in the sewer could not be described
with any reasonable degree of accuracy; therefore, the definition of the storm hydrograph
was not attempted.
DESCRIPTION OF SCREENING EQUIPMENT-A perspective view of a single screening
unit, as it existed in its original form, is shown on Figure 5. The unit is fed through the
influent line with the feed changing direction from vertical to horizontal over the
stationary distribution dome. The flow over the dome is ideally laminar. Upon leaving the
dome, the flow strikes the rotating collar screen at a velocity of 5 to 15 feet per second,
depending on the diameter of the influent line and the flow. The speed of the collar
screen can be varied between 30 and 60 rpm by adjusting a variable drive unit at the 1/2
horsepower drive motor. Depending on the velocity of the feed, and the fineness,
condition, and speed of the collar screen, approximately 70 to 90 percent of the feed will
penetrate the screen. The remaining 10 to 30 percent, with the retained solids, drops onto
the vibrating horizontal screen for further dewatering. The dewatered solids, through the
vibrating action of the horizontal screen, migrate toward the center of the screen where
they drop through an opening in the screen to a solids discharge pipe. This solids flow is
returned to the interceptor sewer and subsequently to a sewage treatment plant. The
screened flow is discharged to a receiving water body as treated effluent.
During the course of the investigation, the unit has evolved into a more elementary piece
of hardware than that described above. The changes that occurred will be described in
Section 5.
ASSOCIATED EQUIPMENT-The screens are continuously cleaned with a solution of
hot water and concentrated household detergent. The wash water is heated to
approximately 170 degrees F. with a gas-fired, commercial water heater. The detergent is
injected into the hot water piping by a 10 gpm positive displacement pump. The
detergent is diluted about 800:1 at the spray nozzles, and is discharged at a rate of 1.8
gpm per nozzle at a pressure of SO pounds per square inch (psi). The collar screen has two
stationary nozzles directed at the outside of the screen, and the horizontal screen has four
nozzles mounted on a rotating bar directed at the underside of the screen.
During the course of the investigation, the backwash operation was changed from a
continuous operation to an intermittent backwash operation. The development of this
procedure, and a description of the changes made to accomplish it, are discussed in
Appendix C. The range of backwash water ratios investigated is indicated in Table 2 of
the text.
13
-------
CONTINUOUS
DETERBENT BACKWASH
WATER
VARIABLE DRIVE
COLLAR ICRECN MOTOR
STATIONARY
DISTRIBUTION DOME
S FOOT DIAMETER
ROTATING COLLAR
SCREEN [DACRON CLOTH)
VIBRATING HORIZONTAL
SCREEN
ROTATINS MORIZONTAt
SCREEN BACKWASH SPRAT
COLLAR SCREEN OETER8ENT
BACKWASH SPRAY
SCREENED EFFLUENT
BYPASSED TO RECE1VINB
WATER AFTER DISINFECTION
FIGURE 5
ORIGINAL SCREENING UNIT
14
-------
OPERATION OF SCREENING FACILITY-A specific goal of this study was to perform
all test runs during storm-caused combined sewage conditions. However, after
approximately one-third of the testing was accomplished, the rainy season came to an end
and the project was faced with a possible delay. To avoid this possible one-year delay, it
was decided to complete the study using dry-weather flow. In making this decision, it was
assumed that the differences between dry-weather flow and storm-caused flow were not
great enough to affect the objective of this study.
SAMPLING PROGRAM
SAMPLING TECHNIQUE AND FREQUENCY-When the screening operation began, it
was observed that the character of the waste frequently changed in concentration and
color over very short periods of time. This was expected, and it was a specific goal to
detect and characterize these changes with a grab sampling technique. During the course
of the investigation, however, it became desirable to minimize the very short-term
interferences associated with the variability of the sewage so that the long-term
performance of the unit could be evaluated. To do this required composite sampling.
During the testing program, the duration of any one test ranged from a minimum of one
hour to a maximum of twelve hours. In most tests, composite samples were collected
every hour, with each composite consisting of three grab samples of equal volume
collected in the middle of each one-third of that hour. The flow rate to the unit during
any one test was constant. It was this type of composite sampling that was used to
evaluate the long-term effectiveness of the screening unit, and also to obtain a general and
representative description of the sewage being applied to the unit.
Grab sampling was used to describe the more unusual constituents of the sewage that
affected the short-term performance of the screening unit. These unusual constituents,
and their affect on performance, were noted and are discussed in the text.
It was a specific goal of this study to take discreet samples at specific points of the runoff
hydrograph to determine both the character of sewage, and the treatment effectiveness of
the unit as the storm passed through the sewer. Since the hydrograph could not be
accurately described, there was no purpose in this mode of sampling.
OBSERVATIONS—A schematic diagram of the screening facility, the process streams
sampled, and the observations made on each stream are shown on Figure 6.
15
-------
SULLIVAN GULCH
PUMP STATION
COMBINED SEWAGE
COMBINED SEIA6E OVEIFLOI INFLUENT
SOLIDS (TOTAL, VOLATILE,
SETTLEABLE)
B.O.D., D.O., C.O.D.
pH, TEMPERATURE
CI2 DEMAND
MPN
NH4
FLOW RATE
SAMPLE POINT
INTERCEPTOR SEWER TO
SEWAGE TREATMENT PLANT
SCREENING UNIT
SCREENED SOLIDS CONCENTRATE
SOLIDS (TOTAL & VOLATILE)
B.O.D.
C.O.D.
FLOW RATE
WASH WATER
TEMPERATURE
PH
FLOW RATE
CONCENTRATION
OF DETERGENT
BYPASS TO RIVER
V '
TW•"
SCREEN EFFLUENT
SOLIDS (TOTAL, VOLATILE, SETTLEABLE)
B.O.D., D.O., C.O.D.
CI2 DEMAND
MPN
FLOW RATE
WASH WATER EQUIPMENT
FIGURE 6
COMBINED SEWAGE OVERFLOW SCREENING
SAMPLING PROGRAM
-------
All laboratory tests were performed according to Standard Methods * with the exception
of COD. All samples, except settleable solids, were blended in a Waring blender prior to
analysis to improve the precision of the results. Settleable solids determinations were
made by the Imhoff cone procedure.
The COD test was performed according to the "rapid method" as described by Dr. John
S. Jeris in the May 1967 issue of "Water and Wastes Engineering." The rapid method
COD test made routine collection of organic strength data very reliable because it
minimized the possibility of loss of data, which may have been experienced if only the
5-day BOD test was performed.
During the initial stages of the testing program, parallel tests of BOD and COD were
performed on all process streams to establish a BOD/COD ratio for each stream. During
subsequent tests, only the rapid COD test was run and the BOD/COD ratio was used to
provide a BOD value when this appeared desirable.
EXPERIMENTAL DESIGN AND DATA REDUCTION
EXPERIMENTAL DESIGN-During startup of the screening unit, several variables were
noted in its construction and operation that would affect its performance. These included
influent flow rate, the velocity at which the feed strikes the collar screen; rotational speed
of the collar screen; mesh size and material of the collar and horizontal screen; duration
and frequency of the backwash; and type of detergent used in the backwash. With this
many variables, a means of experimentation was required that would efficiently evaluate
the relative influence each variable had on the overall performance of the unit. This
required an experimental procedure which could investigate several variables
simultaneously, and reveal what the exact effect of each variable was on the performance
of the unit.
To accomplish this, a form of factorial experimental design was used for each
investigation of the testing program. Figure 7 illustrates the initial experiment, which was
designed to investigate the three variables that, at the time, were believed to have the
most effect on performance. This experiment design is statistically termed a 2^ Factorial
Design, Multiple Response Experiment, which means that two levels of three variables are
simultaneously investigated. If all combinations are tested, the experiment requires eight
test runs. Under these particular set of conditions the experiment can be visualized as a
cube in which each corner of the cube represents a unique combination of the variables to
be tested.
*• "Standard Methods for the Examination of Water and Wastewater." 12th Ed., Amer.
Pub. Health Assn. New York (1965).
17
-------
SETTLE ABLE SOLIDS REDUCTION ~\
EFFICIENCIES I
86%
EXPERIMENT NO.
62%
93%
RESPONSES
1. SETTLEABLE SOLIDS
REMOVAL
2. TSS REMOVAL
3. VSS REMOVAL
4. B.O.D. REMOVAL
5. DURATION OF TEST RUN
6. CONDITION OF SCREEN
7. SOLIDS CONTENT OF
SCREENINGS.
B. HYDRAULIC CAPACITY
EXP.
NO.
1
2
3
4
5
6
7
8
RUN*
NO.
5
3
7
8
2
4
1
6
COLLAR
SCREEN
175
110
175
110
175
110
175
110
HORIZONTAL
SCREEN
175
175
110
110
175
175
110
110
FLOW
(GPM)
700
700
700
700
1200
1200
1200
1200
'TEST RUNS ARE RANDOMIZED TO MINIMIZE
EFFECT OF A TIME TREND WHICH MAY EXIST
DURING TESTING PERIOD.
FIGURE 7
EXPERIMENTAL DESIGN AND DATA REDUCTION
18
-------
At the completion of the experiment, a cursory evaluation can be made by plotting any
one, or all, of the responses observed at their respective positions on the cube. In most
cases, the observer can immediately determine, by visual inspection, which of the three
variables is contributing the most and/or least to the particular response observed.
DATA REDUCTION—While in most cases a visual interpretation of the data is sufficient
during the early stages of an investigation, the limitations of the eye are soon realized. A
mathematical method is used to further inspect the data.
In reference to Figure 7, the effect that any one variable has on a particular response is
calculated by subtracting the average of the four observations at the lower level of the
variable from the average of the four observations at the higher level of the variable. For
example, the observed reductions in settleable solids of the first experiment are plotted at
their respective positions on the experimental diagram of Figure 7. The following
calculation was made to determine the effect that changing the horizontal screen from
175 (105 micron opening, 52 percent open area) to 110 mesh (150 micron opening, 42
percent open area) had on the efficiency of settleable solids reduction.
Average of higher level (110 mesh) = 79 + 75 + 86 + 62 = 76
4
- Average of lower level (175 mesh) = 81 +92 + 85 + 93 = 8g
4
Effect =-12 percent
From this calculation, one can conclude that: "When the horizontal screen was changed
from 175 mesh (105 microns) to the coarser 110 mesh (150 microns), the settleable
solids reduction efficiency was decreased by 12 percent, from 88 percent to 76 percent."
Using the same calculation for the collar screen variable and influent flow rate variable,
the results of the first experiment for settleable solids reduction efficiencies can be
summarized as follows:
Effect On
Variable Settleable Solids Reduction
Changing horizontal screen from 175 to 110 Decreased 12 percent
Changing collar screen from 175 to 110 Decreased 2 percent
Changing flow rate from 700 gpm (50 gal/min/ft2)
to 1200 gpm (86 gal/min/ft2) None
19
-------
From this summary, one can conclude that the size of the horizontal screen most affects
settleable solids removal, and the flow rate applied to the unit least affects settleable
solids removal. If the next experimental design was based on only these results, a finer
horizontal screen would be selected to obtain better results. Likewise, since increasing the
flow rate to 1200 gpm (86 gal/min/ft^) had little effect on the performance, it would also
be natural to try a higher flow rate, since this would increase the hydraulic capacity of
the unit. This type of analysis and reasoning was applied throughout the testing program;
however, for any one experiment, several responses were evaluated before a change was
made in the variables. A review of all the evaluations, collectively, provided most of the
information necessary to evaluate the overall performance of the unit and to modify the
unit to improve its performance.
20
-------
SECTION 5
INVESTIGATIONS
The chronology of the investigations, and the clarifying data, will be discussed in this
section. Information of a more analytical nature will be found in Appendix C.
CHARACTERIZATION OF COMBINED SEWAGE OVERFLOW
Several composite samples were taken from the trunk sewer during storm periods for the
purpose of characterizing storm-caused combined sewage. The sampling technique,
frequency, and analysis is described in Section 4. A summary of results is presented in
Table 1 and a complete discussion is included in Section 6.
TREATMENT CAPABILITIES OF SCREENING UNIT
Several levels of the known variables were tested. The results of these tests led to several
equipment modifications in the course of developing the screening unit as it now exists. A
list of the known variables, the range at which each was tested, and the level at which the
best results occurred are presented in Table 2. The evolution of the screening unit from
its original form to its present form is illustrated in Figure 8.
The major modifications included removing the vibrating horizontal screen, improving the
backwash procedures, selecting an effective detergent, changing the screen materials, and
reducing the size of the influent pipe to increase the velocity of the feed striking the
screen. A discussion of these modifications and the accompanying data are presented in
Appendix C.
21
-------
TABLE 1
SUMMARY OF CHARACTERIZATION OF COMBINED SEWAGE
SULLIVAN GULCH PUMP STATION
PORTLAND, OREGON
FEBRUARY - APRIL. 1969
CHARACTERISTIC
PH
TEMPERATURE. °F
DISSOLVED OXYGEN. MG/L
SETTLEABLE SOLIDS, ML/L
TOTAL SUSPENDED SOLIDSrMG/L
VOLATILE SUSPENDED SOLIDS. MG/L
% VOLATILE SUSPENDED SOLIDS
B.O.D.. MG/L
C.O.D.. MG/L
B.O.D./C.O.D.
AMMONIA NITROGEN. MG/L
ORGANIC NITROGEN. MG/L
TOTAL NITROGEN. MG/L
NUMBER OF
OBSERVATIONS
26
25
16
25
28
28
28
14
24
14
7
7
7
MEAN
5.0
48.7
8.0
3.1
146
90
67
105
199
.51
5.1
8.2
13.3
STANDARD
DEVIATION
+ .4
+ 6.5
+ 2.2
+ 1.0
+ 59
+ 25
+ 17
+ 36
+ 50
+ .08
+ 1.4
+ 3.1
+ 4.3
MINIMUM
4.5
34.0
3.7
1.5
70
57
36
57
138
.35
3.7
5.10
9.5
MAXIMUM
6.0
56.0
10.4
5.0
326
166
93
156
324
.64
7.0
14.0
21.0
22
-------
TABLE 2
RANGE AND LEVEL OF VARIABLES TESTED
VARIABLE
HORIZONTAL SCREEN MESH SIZE
COLLAR SCREEN MESH SIZE
COLLAR SCREEN MATERIAL
COLLAR SCREEN ROTATIONAL SPEED
INFLUENT FLOW RATE
COLLAR SCREEN HYDRAULIC LOADING
VELOCITY OF FEED WATER STRIKING
COLLAR SCREEN
TYPE OF OPERATION
BACKWASH RATIO {GAL. BACKWASH
WATER/1000 GAL. APPLIED WASTE)
RANGE INVESTIGATED
110 (150 MICRON OPENING)
TO 175 (105 MICRON OPENING)
105 (167 MICRON OPENING)
TO 230 ( 74 MICRON OPENING)
DACRON CLOTH. MARKET GRADE
STAINLESS STEEL FABRIC, TENSII
BOLTING CLOTH.11'
30 RPM TO 60 RPM
700 TO 2000 GPM
50 GAL/FT2/MIN. TO
143 GAL/FT2/MIN.
3 TO 12 FT/SEC.
INTERMITTENT TO CONTINUOUS
.200 GAL/1000 GAL
TO 25.6 GAL/1000 GAL.
LEVEL OF BEST
PERFORMANCE
REMOVAL OF HORIZONTAL SCREEN
165 (105 MICRON OPENING,
47.1% OPEN AREA)
TENSILE BOLTING CLOTH
60 RPM
1700 GPM
122 GAL/FT2/MIN.
11 FT/SEC.
4% MIN. ON, V, MIN.
OFF FOR BACKWASH
.235 GAL/1000 GAL.
(1) SEE APPENDIX A FOR SCREEN SPECIFICATIONS.
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
SECTION 6
DISCUSSION OF RESULTS
CHARACTERIZATION OF COMBINED SEWAGE
A summary of the characterization of storm-caused combined sewage was presented in
Table 1 of Section 5. This characterization was based on the average of several composite
samples collected during the early stages of the test program. The composite samples
consisted of three grab samples collected over a one-hour period during a test run.
Composite sampling was used in lieu of discreet sampling to obtain a more representative
description of the sewage being applied to the screens over an extended period of
operation. A review of the characterization did not reveal any unusual constituents in the
sewage that could affect the long-term operation of the screening unit.
During this period of characterization, however, it was observed that there were several
unusual constituents in the sewage which markedly reduced the short-term effectiveness
of the screening unit. These include waste oil dumps, waste paint dumps, and the cleanup
wastes associated with a fish packing plant. All of these waste dumps were of high
concentration, low frequency and .short duration, and significantly reduced the hydraulic
capacity of the screening unit by their presence. When these constituents were
encountered, grab samples were collected and analyzed.
The waste oil dump appeared about 3:00 p.m. every day and lasted for a period of
approximately five to ten minutes. The oil was present in sufficient concentration to turn
the sewage to a black color. The waste paint dumps were less frequent occurring only
once or twice a week about the same time of day. The duration of the paint's presence
was about the same as the oil and was also of sufficient concentration to change the color
of the sewage. In the case of the paint, it was either a brilliant red or green. Both of these
waste dumps also had a strong volatile odor associated with them.
The dump from the fish packing plant was observed a total of five times and each time
for a period of approximately 15 minutes. No color change was noticeable by its
presence. However, a strong odor of decayed fish made its presence known. The pH of
the sewage during this period was 8.5, considerably above the normal of 5.0.
In each of these waste dumps, the hydraulic capacity of the screening unit was
significantly reduced through grease-blinding of the collar screen. If the screens were not •
backwashed during this period, the hydraulic capacity was reduced to a point where only
40 percent of the feed would pass through the screen, down from the normal 80 to 90
percent passing the screen. After the waste dump would pass, the screens would not
recover until they were backwashed. When the screens were backwashed during the waste
dump flows, the reduction in hydraulic capacity was minor.
25
-------
As previously discussed, it became necessary to complete a major portion of the testing
with dry-weather sewage for the lack of storm-caused combined sewage. The dry-weather
sewage was characterized in the same manner as the storm-caused combined sewage. A
comparison of the two sets of data are included in Table 3. For all practical purposes, the
two wastes are similar in character with regard to the affect they have on the long-term
performance of the screening unit. The short-term reductions in hydraulic capacity,
however, were more severe under dry-weather sewage conditions than under wet-weather
sewage conditions.
TREATMENT CAPABILITIES OF SCREENING UNIT
The performance of the screening unit is ultimately evaluated by its ability to remove
organic material from a wastewater stream, and by the volume of wastewater that it can
process. These performance parameters are directly dependent on variables within the
screening unit. The mesh size of the screen, the strength of the screen, the velocity at
which the feed strikes the screen, and the backwash operation are among the most
important variables. The final experiment which was designed with these variables in
mind, clearly defined the capabilities and limitations of the screening unit.
The final experiment consisted of six 3-hour tests. Each was performed on a different
day. Four of the six tests investigated two levels of influent flow rate and screen-mesh
size. The remaining two tests were duplicated at the intermediate levels to obtain an
estimate of the day-to-day variances in operating the unit and in the character of the feed
water. The tests at the intermediate levels also helped to interpret the final results. The
design of the final experiment and the observations during the experiment are presented
on Figure 9.
An examination of all the observations reveals that each response is dependent on both
the flow rate and the mesh size of the screen: No response is completely independent of
either flow rate or mesh size; however, the unit's efficiency in removing organic material
is more dependent on the screen-mesh size than on the flow rate. The dependency of
removal efficiency on screen-mesh size was expected. If a finer screen is installed on the
unit, one could expect higher removal efficiencies. Other variables, however, tend to bias
this dependency. In most instances, as the flow rate was increased, slightly poorer
removal efficiencies were observed. It is believed the higher flow rates are fracturing the
more friable solids at the surface of the screen and forcing them through the screen. The
slight reduction in removal efficiency observed at the higher flow rate, however, is more
than offset by the increase in hydraulic efficiency.
The hydraulic efficiency, as measured by the percentage of screened effluent and the
condition of the screen, also shows a very strong interdependence on flow rate and
screen-mesh size. As seen on Figure 9, the best hydraulic efficiency and most stable
26
-------
TABLE 3
COMPARISON OF STORM - CAUSED COMBINED FLOW
AND
DRY-WEATHER FLOW
CHARACTERISTIC
SETTLEABLE SOLIDS.
ML/L
TOTAL SUSPENDED
SOLIDS, MG/L
C.O.D., MG/L
STORM-CAUSED COMBINED FLOW
NUMBER
OF
OBSERVATIONS
25
28
24
MEAN
3.1
146
199
STANDARD
DEVIATION
± 1.0
± 59
± 50
MIN.
1.5
70
138
MAX.
5.0
325
324
DRY-WEATHER FLOW
NUMBER
OF
OBSERVATIONS
35
35
25
MEAN
4.8
129
345
STANDARD
DEVIATION
i 1.1
± 44
± 138
MIN.
2.5
50
144
MAX.
7.0
244
696
to
-------
EXPERIMENTAL DESIGN
230 (74 MICRONS)
z
ui
£"
E IM
CJ —
M CO
o
u
165 (105 MICRONS)
105 (167 MICRONS)
©
93%
86 114 143
INFLUENT FLOW RATE
(GPM/FT2)
OBSERVATIONS
100% 41% 35% 31.3%
27.5%
99%; 98%
28%; 24.5%
21%; 23.4%
92% 89%
SETTLEABLE SOLIDS REMOVAL
31%
32%
26.5%
19.6%
TOTAL SUSPENDED
SOLIDS REMOVAL
79.1%
89.8% GOOD
C.O.D. REMOVAL
FAILED. REPLACED
AND FAILED AGAIN
87.1*
91.1; 92.8
32.5% GOOD
GOOD; FAILED. REPLACED
AND SURVIVED
(GOOD
SCREENED EFFLUENT
AS % OF INFLUENT
CONDITION OF SCREEN AT
END OF FOUR-HOUR TEST RUN
FIGURE 9
EXPERIMENTAL DESIGN AND OBSERVATIONS
OF FINAL EXPERIMENT
28
-------
performance occurs at the higher flow and coarser screen condition. The hydraulic
efficiency declines as both the flow rate decreases and the screen becomes finer. This is
illustrated more vividly on Figure 10, where the actual flow recorder charts are displayed
at their respective positions on the experimental design. The graphs were generated
continuously by a four-hour flow recorder that pneumatically sensed the head over a
90-degree V-notch weir. The screened effluent flow and the unscreened flow were
recorded simultaneously. The total influent flow was found by summation. The graphs
are discontinuous because the screening unit was shut off for the backwash cycle.
For this final series of tests, the screening unit was operating 4-1/2 minutes on and 1/2
minute off. During the 1/2 minute, the flow was shut off and the screens were
backwashed with an 800:1 dilution of hot water and liquid detergent. At the end of a
20-minute cycle, the flow was shut off, and the screens were backwashed with a 10:1
dilution of water and liquid detergent. The distinction between the two backwash cycles
is easily seen on the flow charts. Frequent backwashing is necessary, as seen on the flow
charts, at the 1200 gpm (86 gal/min/ft2) flow level by the rapidly rising level of the
unscreened flow graph. This need for backwashing diminishes at the higher flow level, and
therefore the frequency of backwashing could have been reduced. Further examination of
the flow charts shows that the flow rate, or velocity of flow, to the various screen-mesh
sizes has a significant effect on hydraulic efficiency and performance stability.
High velocities and flow rates are limited, however, by the strength of the screen. Figure 9
shows that the 165 mesh screens (105 microns, 47.1 percent open area) started failing at
1600 gpm (114 gal/min/ft2). Failure of the 230 mesh screen (74 microns, 46.0 open area)
was persistent at 2000 gpm (143 gal/min/ft2). Screen life is also approximated on Figure
10 by the relative length of chart run. The photographs on Figure 11 illustrate typical
screen failures.
The failure of the steel screens was attributed to the tremendous live load applied to the
screens during high-flow conditions. The forces contributing to the failure include the
velocity head of the flow striking the screen, the centrifugal forces associated with the
rotation of the screen, and the mass of water carried along on the inside of the screen. By
calculating the velocity head and G-force at 2000 gpm and 60 rpm, and assuming a
thickness of water on the inside of the collar screen, the equipment supplier found that
the steel wires of the screens were stressed beyond their yield point soon after the 2000
gpm (143 gal/min/ft2) was applied. Since the calculation of the estimated fiber stresses
involved some very general assumptions, a tabulation of these results is not presented in
this report.
A failure of this kind was termed a mechanical failure, and the situation was corrected to a
degree in reducing the effective live load on the screen by reducing the unsupported span
29
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
AT LEFT AND BELOW:
165 MESH TBC AT 114 GPM/FT2
FAILURE AFTER 6 HOURS
, ' - 3"*
1
165 MESH TBC AT 122 GPM/FT2,
FAILURE AFTER 6 HOURS
SHOWN AT RIGHT.
165 MESH TBC AT 122 GPM/FT2,
FAILURE AFTER 12 HOURS
SHOWN AT LEFT
-
FIGURE 11
TYPICAL SCREEN FAILURES
-------
of the screen. Recent developments in extending screen life by the equipment supplier
have produced a 165 mesh screen (105 micron opening) that now has a probable life of
500 hours when operated at 1750 gpm (128 gal/min/ft2). If operated at 2500 gpm (178
gal/min/ft2), the probable life will drop to 300 hours.
Based on the results of the last experiment, a final test was performed to gather data on
extended operation of the unit. The previous tests indicated that the unit operated best at
1700 gpm (122 gal/min/ft2) on a 165 mesh screen (105 microns, 47.1 percent open area).
To further stabilize the performance, backwash operation was changed to a 30-second
wash to 40:1 solution of water and liquid detergent at the end of 4-1/2 minutes of
operation. A discussion of the various solvents tested is included in Appendix C. The test
lasted for six hours and ended with the failure of three screens. A summary of the
operating conditions, performance data, and character of flow streams are presented in
Table 4. An Imhoff cone comparison of the flow streams is presented on Figure 12.
The results of the final test show that the unit's ability to remove organic material from
the wastewater stream is good, and is comparable to the efficiency of a primary clarifier.
The hydraulic efficiency of the unit is excellent; however, failure of the three screens
shows that the unit is operating beyond its capacity. The screen is the limiting component
of the entire unit.
32
-------
TABLE 4
SUMMARY OF EXTENDED TEST
OPERATING CONDITIONS PERFORMANCE DATA
INFLUENT FLOW RATE - 1700 GPM 100% REMOVAL FLOATABLE SOLIDS
(122 GAL./MIN./FT.2) 98% REMOVAL SETTLEABLE SOLIDS
COLLAR SCREEN SPEED - 60 RPM 34% REMOVAL TOTAL SUSPENDED SOLIDS
COLLAR SCREEN - 165 MESH TBC 27% REMOVAL C.O.D.
'VM*1 OPEN1 AREA)NG' ^ °F INFLUENT AS A SOLIDS CONCENTRATE
BACKWASH RATIO - .235 GAL/1000 GAL. ^*™ ^ * "^ ^ ™
AVERAGE CHARACTER OF FLOW STREAMS
SCREENED UNSCREENED
INFLUENT - ^ EFFLUENT + EFFLUENT
____ _ ______
FLOW (GAL/MIN.) 1700 1570 130
(% OF INFLUENT) 100% 92% 8%
SETTLEABLE SOLIDS
(ML/L) 5.7 <0.1 73
TOTAL SUSPENDED SOLIDS
(MG/L)
(POUNDS/MIN.)
C.O.D.*
(MG/L)
(POUNDS/MIN.)
122
1.73
302
4.30
87
1.14
240
3.15
542
0.59
1060
1.15
* B.O.D./C.O.D. = 0.5
-------
100% REMOVAL
OF FLOATABLE SOLIDS
SHOWN AT LEFT
. ..•. .
SCREENED EFFLUENT AT LEFT;
AT CENTER SOLIDS CONCENTRATE
AND INFLUENT COMBINED SEWAGE
SHOWN AT RIGHT.
FIGURE 12
IMHOFF CONE COMPARISON
OF FLOW STREAMS
34
-------
SECTION 7
PRELIMINARY DESIGN OF A SCREENING FACILITY
The final performance data of the screening unit shows that fine-mesh screening could be
used for treating combined sewage overflow; therefore, a preliminary design of a full-scale
facility was warranted.
DESIGN CRITERIA
The proposed project site is located in Seattle, Washington. The site is in the heart of the
business district of Seattle and is located within a valuable parking lot at the intersection
of King Street and the Alaskan Way viaduct. The entire area surrounding the site is
constructed on fill material, and almost every structure is supported on piling. The site is
also close to the waterfront of Elliott Bay; therefore, high tide comes to within a few feet
of the ground surface. Construction in this area is difficult and expensive.
The drainage basin above the site includes about 190 acres and is served by a 48-inch,
pile-supported, concrete sewer. Since the drainage basin is almost entirely pavement or
building roofs, a runoff coefficient of .95 was assumed to determine storm water
volumes.
The rainfall pattern in the City of Seattle was studied to determine the design capacity of
the screening facility. The intensity and duration of rainfall in the area received particular
attention so that it could be estimated how long the facility would operate at a certain
capacity. The study revealed that measurable precipitation occurred approximately 1,000
hours each year. While the rainfall occurrences were relatively frequent, they were of low
intensities. Rainfall intensities up to .055 inches/hour produce a runoff of 10 mgd, and
account for 75 percent of the rainfall occurrences. A summary of this study is shown on
Figure 13.
Runoff in excess of 2.75 mgd will produce combined sewage overflow. This flow is based
on the capacity that the dry-weather flow of the drainage basin requires in the interceptor
sewer which carries this flow to the Seattle treatment plant. With the base flow of 2.75
mgd and the runoff pattern shown on Figure 13, the total volume of combined sewage
overflow would be 282 million gallons a year. Based on the runoff pattern of this
particular drainage basin, a design capacity of 25 mgd was chosen for the screening
facility. With this capability, 96 percent of the total volume of overflow would receive
treatment before being discharged to Elliott Bay. The added cost to achieve 100 percent
treatment capabilities cannot be justified, as this would require a 40-mgd facility.
Approximately 40 percent of the 40-mgd facility's capacity would be idle 95 percent of
the time rainfall occurred.
35
-------
60
v
AMOUNT AND RATE OF STORM-CAUSED COMBINED OVERFLOW
AT
KING STREET REGULATOR DRAINAGE BASIN
SEATTLE, WASHINGTON
BYPASS 11 MILLION GALLONS
(4% OF TOTAL OVERFLOW VOLUME)
DESIGN CAPACITY OF SCREENING
FACILITY (96% OF TOTAL OVERFLOW VOLUME)
TOTAL VOLUME OF OVERFLOW
282 MILLION GALLONS
TOTAL HOURS OF OVERFLOW
AT 2.75 MGD OR GREATER >
100
200
300 400 500 600 700
TOTAL DURATION OF OVERFLOW AT R OR GREATER (HOURS/YEAR)
800
900
FIGURE 13
-------
PRESENTATION OF PROPOSED SCREENING FACILITY
A 25-mgd screening facility requires a structure approximately 30 feet wide and 75 feet
long. A perspective of the proposed facility is shown on Figure 14. The elevated facility is
an attempt to illustrate what may be done to conserve the valuable parking area and still
provide an attractive and functional treatment facility. The configuration of the elevated
facility also offers the possibility of its becoming an integral part of an elevated parking
facility. This would provide more parking than is now available, which is an asset to be
considered. The facility does not provide disinfection.
Underground construction of the facility was investigated; however, this presented several
hydraulic problems, and would be more costly than the elevated structure. A ground level
structure for the Seattle facility was not investigated because conservation of the parking
was a major design consideration.
A site plan of the proposed facility is shown on Figure 15. The combined overflow comes
to the facility in the 48-inch sewer and would pass through a Parshall flume prior to
reaching the facility. The flume would provide the primary control for the operation of
the facility. After passing the Parshall flume, the flow would drop into a pump sump
where it would be lifted to the screening units by a single 250 hp, vertical turbine,
mixed-flow pump. The pump speed would be automatically controlled so that the pump
discharge matches the flow in the incoming sewer. After the flow has passed the screening
units, the screened effluent would be returned to the 48-inch interceptor downstream of
the pump sump, and would be discharged to Elliott Bay. The unscreened flow would be
returned to the trunk sewer where it would continue on to the treatment plant. It is
assumed that the influent flow will be adequately disinfected upstream of the screening
facility.
A design capacity of 25 mgd requires the use of ten 2.5 mgd screening units. The floor
plan and sections on Figure 16 illustrate a proposed layout of such a facility. The
arrangement of the units, with the provisions of the center aisle, lends itself to easy
operation and maintenance of the screening facility. Also shown on the floor plan is
office space, a restroom, storage for spare parts and screens, a boiler room, and control
panels. The foundation piers would be used for entrance to the facility at one end, and
for housing of the pump suction at the other end. Storage of the backwash solvent is not
shown on the plan as it would be provided in an underground tank.
The screening facility will be designed to operate automatically. The primary control for
the facility would be located in the interceptor at the Parshall flume. The flume would
monitor the depth of flow in the sewer, and screening units would be turned on and off
in increments of 2.5 mgd as the depth of flow in the sewer rises and falls. Because
37
-------
LU
H
FIGURE 14
PROPOSED SCREENING FACILITY
PERSPECTIVE
-------
PAGE NOT
AVAILABLE
DIGITALLY
-------
F IQOR PLAN
SECTION B
FIGURE 16
PROPOSED SCREENING FACILITY
FLOOR PLAN AND SECTION
-------
occasional backwashing is necessary, a secondary control system is required to sense this
need and to initiate the process. This would be accomplished by installing a flow meter
on the screened effluent line. The flow signal from the effluent meter would then be
compared to the flow signal from the flume to detect a decrease in hydraulic efficiency
and, therefore, a need to backwash. It is anticipated that when the ratio of screened
effluent flow/influent flow falls below .80, the backwash cycle will be initiated.
Because of the complexity of these two control systems working together over the entire
flow range, two modes of operations were developed. Wash Mode 1, illustrated on Figure
17, would be in control when the influent flow is 15 mgd or less. This requires the use of
five screening units and accounts for 85 percent of the total operating time. In Wash
Mode 1, the ten screening units are divided into two blocks of five screening units each.
Each unit in Block 1 has a partner in Block 2 to which the flow can be switched when it
becomes necessary to backwash. Therefore, if the flow is 2.5 mgd and backwash is
necessary, the flow to Unit 1 is transferred to Unit 1' and Unit 1 is backwashed. Unit 1'
remains in operation until it becomes necessary to backwash. The primary control of
adding or deleting units in operation would occur in the block of units that is in
operation at the time addition or deletion is required.
As shown on Figure 17, the units would be washed two at a time. The period of
backwash would be approximately 20 seconds. The duration of the backwash cycle
would vary from 20 seconds to one minute, depending on the number of units in
operation.
When the flow increases beyond 15 mgd, control is switched to Wash Mode 2 and seven
units automatically go into operation. In the event of a backwash in Wash Mode 2, each
individual unit is shut off and backwashed while the remaining six units absorb the excess
flow. This procedure is repeated until all seven units have been backwashed. The period
of each backwash remains the same, but the backwash cycle will now vary from 140
seconds to 200 seconds, depending on the number of unit's in operation. The maximum
capacity of the facility is 10 units, or 25 mgd, and all flows above 25 mgd will be
bypassed, untreated, to Elliott Bay. A summary of the control system is shown on Figure
18.
The control system for the screening facility, while complex, is relatively easy to build,
and the components necessary to build it have a history of high reliability. It is believed
the control system can be built reliably and that the screening unit will perform
satisfactorily; however, there are several unknowns associated with the two systems
working together. Because of the unknowns, it is recommended that a systems analysis be
performed prior to final design. This can be done in the form of an analog computer
simulation which would be used to investigate the compatibility of the hydraulic and
electrical machinery under various flow situations.
41
-------
/BLOCK NO.
(4) X
C5) /' (£* C5
"> XBLOCK NO. 2
S^
6 MGDi Q-=9 MGD
BEGIN BACKWASH
CYCLE
to
UNITS 1 AND 2
ARE REPLACED
BY UNITS 1' AND
2' - UNITS 1 AND
2 ARE WASHED
FOR A PERIOD OF
20 SECONDS.
UNIT 3 IS REPLACED
BY UNIT 3'. UNIT 3
IS BACKWASHED FOR
A PERIOD OF 20
SECONDS.
END OF BACKWASH
CYCLE.
FIGURE 17
WASH MODE NO. 1
PROPOSED SCREENING FACILITY
-------
INFLUENT
FLOW, Q
(MGD)
1.5 3 Q - 3
3 = Q - 6
6 S Q - 9
9 S Q - 12
12 S Q * 15
15 S Q - 17.5
17.5 S Q - 20
20 = Q - 22.5
22.5 = Q - 25
25 = Q - P
SCREENING
UNITS IN
OPERATION
«
©
©
@®0
©
P-25 BY-
PASSED
FLOW
PER
SCREEN
(MGD)
1.5-3
1.5-3
2-3
2.3-3
2.4-3
2.15-3
2.2-2.5
2.2-2.5
2.25-2.5
i
UJ ^
— °C°
UJ P <
2 "Jill
° S •£
li<
^25
CM Z
• IU
fill
^0*^0
$ OT < 00
1 w
o o
m O
»
Jl
o
BACKWASH ,
1' = SCREENED EFFLUENT FLOW
Q - INFLUENT FLOW
FIGURE 18
SUMMARY OF CONTROL SYSTEM
FOR
PROPOSED SCREENING FACILITY
43
-------
ESTIMATED CONSTRUCTION COST OF SEATTLE FACILITY
The cost of constructing the Seattle screening facility is estimated to be $538,000. The
construction cost estimate is based on estimated 1970 prices and assumes that all work
will be performed by private contracting firms. The estimate also includes an allowance
for design engineering, field surveying, soil exploration, construction supervision and
inspection, legal fees and contingencies. The estimate does not include the cost of land or
the cost of disinfection.
ESTIMATED ANNUAL OPERATION AND MAINTENANCE COSTS
A summary of the annual operation and maintenance costs is shown in Table 5. Annual
labor costs are based on one man-hour for each hour of operation. This is based on the
experience with the pilot unit, and is only an estimate of what may be experienced in a
full-scale facility. Annual maintenance costs are based on 3 percent of the major
equipment costs. Power and utility costs are based on present rates. Screen replacement
costs are based on a predicted life of 500 hours. Costs for cleaning agent are based on the
use of concentrated sodium hydroxide, purchased in bulk lots. The total annual operation
and maintenance cost is estimated to be $18,500.
Table 5
Estimated Annual
Operation and Maintenance Costs
Item Cost
Labor $ 5,600
Equipment Maintenance 3,000
Screen Replacement 3,500
Power 3,000
Gas 1,200
Cleaning Agent 700
Vehicle Operation and Maintenance 1,500
Total Annual Operation and Maintenance $ 18,500
ESTIMATED TOTAL ANNUAL COST
A total annual cost figure provides the best basis on which an economic comparison can
be made, provided the items to be compared are relatively equal in basic design
considerations. The construction cost estimate for the Seattle facility violates this premise
in that the total cost includes provision for special foundation consideration and special
architectural treatment.
44
-------
In order to compensate for this in the total annual cost figure, another cost estimate was
prepared for a more conventional type screening facility. In effect, the Seattle facility was
moved to an assumed site that did not have any special foundation problems or did not
require any special aesthetic considerations. It was assumed that this new structure would
be of concrete block walls supported by a concrete wall footing. The floor would be a
concrete slab on grade, and the roof would be of a timber joist system. All other
mechanical and electrical items would be the same as the Seattle facility. These changes
reduced the estimated total construction cost to $496,000 and it is this figure which is
used in the total annual cost figure to represent a more typical screening facility.
The total annual cost summary is presented in Table 6. All costs shown in Table 6 have
been adjusted to assumed 1970 prices and include an allowance for design engineering,
legal fees, administrative costs, and contingencies. The cost of land and disinfection is not
included. The construction costs are amortized over a period of 25 years assuming an
interest rate of 6-1/2 percent. The cost per 1000 gallons is based on treating a total of
271 million gallons per year.
Table 6
Estimated Total Annual Cost
Estimated Total Construction Cost $496,000
Annual Debt Service 41,000
Annual Operation and Maintenance 18,500
Estimated Total Annual Cost $ 59,500
Estimated Cost Per 1000 Gallons = 22 Cents
DISCUSSION OF FEASIBILITY
In order to get a feel for the economic position of this type screening facility relative to
other possible methods of treatment, a brief economic comparison was made. Particular
attention was paid to conventional primary sedimentation; however, since a detailed cost
comparison was beyond the scope of this study, no cost figures will be presented. The
brief comparison did reveal that screening can be a feasible treatment method depending
on particular conditions present at the site.
A major advantage of conventional primary treatment is that disinfection, by means of
conventional chlorination, can be accomplished within the primary clarifier. This
eliminates the need for a separate chlorine contact chamber which, at the present state of
the art of chlorination, would be required at a screening facility. This, of course,
represents a considerable added cost when disinfection is found to be either desirable or
mandatory.
45
-------
This advantage, however, could be offset with a new method of disinfection that could be
as efficient as chlorination and at the same time eliminate the long contact time that is
presently required.
Another advantage of conventional primary clarification is that the volume of the
primary clarifier would be large enough to completely hold the storm-caused combined
sewage of the short-duration, low-intensity storm events. After the storm has passed and
the peak flow in the sewer has subsided, the impounded sewage could be returned to the
sewer at a reduced flow rate. This advantage is enhanced when there is a high percentage
of short-duration, low-intensity storms such as in the Seattle area.
The most important disadvantage of conventional primary clarification is the large
amount of land required. It has been estimated, by preliminary layouts, that conventional
primary clarification requires 10 to 20 times more land area than a screening facility. The
actual difference is dependent on the design capacity chosen for a primary treatment
plant, and how much reserve capacity of a primary clarifier is actually used to meet the
flow requirements of a particular drainage basin. This disadvantage becomes more severe
as the size of the drainage basin increases, and as the value of the land increases. The
Seattle site is an example of a site where conventional primary clarification would most
likely not be feasible.
In summary, the screening unit can be an economically feasible method of treating
combined sewage overflows when compared to conventional primary clarification. The
selection of the screening unit as a method of treatment at a particular site, however, will
require the review of at least four factors. These are:
1. The value and availability of land.
2. The size of the drainage basin, and therefore, the design capacity of the treatment
facility.
3. The character of rainfall and the pattern of runoff.
4. Available means of disinfection.
Other factors that would require review also would include other methods of treatment,
aesthetic considerations, and ancillary use of the facility, such as surrounding the Seattle
facility with a parking structure. In all, it must be emphasized that each point of overflow
is unique, and all these factors must be reviewed before the most economical and efficient
method of treating combined sewage overflow is selected.
46
-------
SECTION 8
ACKNOWLEDGMENTS
The firm of Cornell, Rowland, Hayes & Merryfield acknowledge the City of Portland,
Oregon, and SWECO, Inc., of Los Angeles, California, for their cooperation and assistance
in conducting this study for the Federal Water Pollution Control Administration.
47
-------
SECTION 9
GLOSSARY OF TERMS
AVERAGE DAILY FLOW—The flow from a complete sewer system, or a defined portion
thereof, measured in total gallons throughout a 24-hour period (expressed in millions of
gallons per day, mgd).
BACTERIA—Primitive plants, generally free of pigment, which reproduce by dividing in
one, two, or three planes. They occur as single cells, groups, chains, or filaments, and do
not require light for their life processes. They may be grown by special culturing out of
their natural habitat.
BAR SCREEN—A screen composed of parallel bars, either vertical or inclined, placed in a
waterway to catch floating debris, and from which the screenings may be raked. Also
called a rack.
BOD—Biochemical oxygen demand is a measure of the oxygen necessary to satisfy the
requirements for the aerobic decomposition of the waste. This provides an indication of
the organic content and pollutional strength of the waste (expressed in parts per million,
ppm).
CLARIFIER—A tank or basin in which wastewater is retained for a sufficient time, and in
which the velocity of flow is sufficiently low to remove by gravity a part of the
suspended matter.
COD-Chemical oxygen demand is a measure of the oxygen necessary to stabilize most of
the oxidizable compounds in a waste.
COMPOSITE SAMPLE-Integrated sample collected by taking a portion at regular time
intervals, with sample size varying with flow; or taking uniform portions on a time
schedule varying with the total flow.
DESIGN FLOW—Sewage flow for which facility is designed.
DIGESTION-The anaerobic or aerobic decomposition of organic matter resulting in
partial gasification, liquefaction, and mineralization.
DISSOLVED OXYGEN-Usually designated as D.O. The oxygen dissolved in sewage or
other liquid usually expressed in parts per million or percent of saturation.
DISSOLVED SOLIDS-Solids which are present in solution.
48
-------
EFFICIENCY-The ratio of the actual performance of a device to the theoretically
perfect performance usually expressed as a percentage.
EFFLUENT-Liquid flowing out of a basin or treatment plant.
EFFLUENT WEIR-A weir at the outflow end of a sedimentation basin or other
hydraulic structure.
GREASE-In sewage, grease includes fats, waxes, free fatty acids, calcium and magnesium
soaps, mineral oils, and other nonfatty materials.
GRIT-The heavy mineral matter in water or sewage, such as sand, gravel, cinders, etc.
IMHOFF CONE-A conically shaped graduated glass vessel used to measure
approximately the volume of settleable solids in sewage.
INFLUENT-Liquid flowing into a basin or treatment plant.
MILLIGRAMS PER LITER(mg/l)-The weight in milligrams of material in one liter of
'liquid. ^ ^
MGD—Million gallons per day.
OUTFALL SEWER-The outlet or structure through which sewage is finally discharged.
OVERFLOW RATE—One of the criteria for the design of settling tanks in treatment
plants; expressed in gallons per day per square foot of surface area in the settling tanks.
PRIMARY TREATMENT-The removal of settleable organic and inorganic solids by the
process of sedimentation.
RAW SEWAGE SLUDGE-The accumulated suspended and settleable solids of sewage
deposited in tanks or basin mixed with water to form a semi-liquid mass
SECONDARY TREATMENT-Treatment of sewage by biological methods following
primary treatment.
SEDIMENTATION-The process of subsidence and deposition of suspended matter
carried by water, sewage, or other liquids, by gravity. It is usually accomplished by
reducing the velocity of the liquid below the point where it can transport the suspended
material.
-49-
-------
SETTLEABLE SOLIDS—Suspended solids which will settle in sedimentation tanks in
normal detention periods.
SEWAGE TREATMENT PLANT—Man-made structures which subject sewage to
treatment by physical, chemical, or biological processes for the purpose of removing or
altering its objectionable constituents, and rendering it less offensive or dangerous.
SLUDGE DIGESTION—A process by which organic or volatile matter in sludge is
gasified, liquefied, mineralized, or converted into more stable organic matter through the
activities of living organisms.
STANDARD METHODS-Methods of analysis of water sewage and sludge approved by a
Joint Committee by the American Public Health Association, American Water Works
Association, and Federation of Sewage Works Association.
STORM SEWER—A sewer which carries storm water and surface water, street wash and
other wash waters or drainage, but excludes sewage and industrial wastes. Also called a
Storm Drain.
SUSPENDED SOLIDS-Solids which can be mechanically filtered from the sewage
(expressed in parts per million, ppm or milligrams per liter, mg/1).
TOTAL SOLIDS -The solids in water, sewage, or other liquids. It includes the suspended
solids (largely removal by filter paper) and the nonfilterable solids (those which pass
through filter paper).
VOLATILE SOLIDS -The quantity of solids in water, sewage or other liquid lost on
ignition of the total solids.
VOLATILE SUSPENDED SOLIDS (VSS)-The quantity of solids in wastewater that are
lost on ignition of the total suspended solids.
-50-
-------
SECTION 10
APPENDIXES
A. Screen Specifications
B. Detergent Specifications
C. Experimental Data Presentation and Interpretation
51
-------
APPENDIX A
SCREEN SPECIFICATIONS
SCREEN MATERIAL
AND MESH NUMBER
DACRON CLOTH
110
175
MARKET GRADE
STAINLESS
STEEL FABRIC
100
120
150
200
TENSILE BOLTING
CLOTH
(STAINLESS STEEL)
105
165
200
230
OPEN
(MICRONS)
150
105
149
125
105
74
167
105
88
74
ING
(INCHES)
.0059
.0041
.0055
.0046
.0041
.0029
.0065
.0042
.0034
.0029
WIRE
DIAMETER
(INCHES)
.0031
.0016
.0045
.0037
.0026
.0021
.0030
.0019
.0016
.0014
%
OPEN
AREA
42
52
30.3
30.5
37.9
33.6
46.9
47.1
46.2
46.0
CLOSEST
STANDARD
SIEVE
100
140
100
120
140
200
80
140
170
200
52
-------
APPENDIX B
DETERGENT SPECIFICATIONS
The primary detergent used throughout the testing program was a product called "Zif."^
The product, manufactured by Bestline Products, Inc., of San Jose, California, is a
biodegradable detergent containing solvents and coupling agents, chelating agents, and
corrosion inhibitors. The product is completely soluble in water, contains 13 percent
solids, has a pH of 10.2 and a specific gravity of 1.035. The cost of the product is $5.00 a
gallon, in 55 gallon lots.
1 Mention of commercial products does not imply endorsement by the Federal Water
Pollution Control Administration.
53
-------
APPENDIX C
EXPERIMENTAL DATA PRESENTATION AND INTERPRETATION
The primary goal of an experimental program is to define the variables that affect the
performance of a system and the level at which each variable produces the best
performance. The design of the first experiment is, at best, an educated guess based on
preliminary information of the system. The variables investigated, and the levels at which
they are investigated, are those which are believed to most affect performance of the
system. The results of the first experiment, however, yields information on the direction
to be taken to improve the performance of the system. This information may direct the
investigator to test the same variables at different levels, eliminate a particular variable
and concentrate on the remaining variables, or test a new variable. With each succeeding
experiment, the information becomes more specific until all known variables have been
accurately defined at their best level of performance.
This basic philosophy was followed during the course of the screening unit development.
The experimental data compiled during the investigation is presented in the following
Appendix in a format similar to Figure 8 in the text.
54
-------
ORIGINAL FORM
OPERATING CONDITIONS
Influent Flow Rate 700 to 1200 gpm (combined sewage)
(50 to 86 gpm/ft2)
Collar Screen Speed 30 rpm
Collar Screen 110 and 175 mesh dacron cloth
(150 and 105 micron opening)
Horizontal Screen 110 and 175 mesh dacron cloth
(150 and 105 micron opening)
Operation Continuous flow and continuous backwash
Backwash Ratio 12.0 to 20.6 gal/1000 gal.
EXPERIMENTAL DESIGN
See bigure Cl
PRESENTATION OF DATA
See Tables Cl and C2
INTERPRETATION OF DATA
A survey of the magnitude of the removal efficiencies of Table Cl indicated the screening
unit was capable of removing large quantities of solids and organic material from
combined sewage. The data of Table Cl also revealed several weaknesses, particularly in
the value of the vibratory horizontal screen.
We observed that the horizontal screen produced a solids concentrate, or volume of
unscreened effluent, that was dry enough to shovel. This made it necessary to dilute the
concentrate so it could be carried out of the system. This represented a waste of
concentrating effort because the concentrated solids were intended to flow from the
units.
Unstable performance of the screening unit, attributed to the horizontal screen, was also
indicated in the data of Table C2. This is particularly evident by the relatively large
magnitude of the two-factor interaction between the collar screen and horizontal screen.
This strong interaction indicates that the performance of one is strongly influenced by
the other. It also indicates that the main effects, which were observed individually, were
no longer valid and that their effects must be investigated separately.
55
-------
EXPERIMENT NO.>
EXP.
NO.
1
2
3
4
5
6
7
8
RUN*
NO.
5
3
7
8
2
4
1
6
COLLAR
SCREEN
175
110
175
110
175
110
175
110
HORIZONTAL
SCREEN
175
175
110
110
175
175
110
110
FLOW
(GPM)
700
700
700
700
1200
1200
1200
1200
•TEST RUNS ARE RANDOMIZED TO MINIMIZE EFFECT OF A
TIME TREND WHICH MAY EXIST DURING TESTING PERIOD.
FIGURE C1
EXPERIMENT B. - ORIGINAL FORM
56
-------
TABLE C1
SUMMARY OF RESPONSES
EXPERIMENT B
EXP.
NO.
1
2
3
4
5
6
7
8
COLLAR
SCREEN
175
110
175
110
175
110
175
110
VARIABLES
HORIZONTAL
SCREEN
176
175
110
110
176
175
110
110
ORIGI
FLOW
GPM
700
700
700
700
1200
1200
1200
1200
MAL FORM
(1)
SETTLEABLE
SOLIDS
REMOVAL
81%
92%
79%
75%
85%
93%
88%
62%
RESPONSE
(2) T5.S.
REMOVAL
28%
14%
23%
33%
10%
10%
§5%
24%
C.O.D. (2)
REMOVAL
12%
12%
13%
13%
5%
8%
5%
10%
(1) (ML/L INFLUENT - ML/L EFFLUENTJ/ML/L INFLUENT X 100%
(2) (MG/L INFLUENT - MG/L EFFLUENT)/MG/L INFLUENT X 100%
TABLE C2
SUMMARY OF MAIN EFFECTS AND INTERACTIONS
EXPERIMENT B - OWGINAL FOK*
VARIABLES
MAIN EFFECTS
COLLAR SCREEN FROM 175 MESH TO 110 MESH
HORIZONTAL SCREEN FROM 175 MESH TO 110 MESH
FLOW RATE FROM 700 GPM TO 1200 GPM
TWO FACTOR INTERACTIONS
COLLAR SCREEN X HORIZONTAL SCREEN
COLLAR SCREEN X FLOW RATE
HORIZONTAL SCREEN X FLOW RATE
THREE-FACTOR INTERACTION
COLLAR SCREEN X HORIZONTAL SCREEN X FLOW RATE
KTTLEA8LE
tpUDS REMOVAL
•-2.2% * 6.1%
•12.3% * 6.1%
•0.2% i 6.1%
•11.8% * 6.1%
-6.7% * 6.1%
-2.7% * 6.1%
•4.2% * 6.1%
TSS.
REMOVAL
+3.4% * 3.8%
47.1% * 3.8%
-10.9% * 3.8%
49.4% * 3.8%
+4.4% * 3.8%
-0.9% ± 3.8%
•1.6% * 3.8%
C.O.O.
REMOVAL
+2.0% * 3.0%
+1.0% ± 3.0%
•5.5% * 3.0%
+.5% i 3.0%
+2.% * 3.0%
0% * 3.0%
.5% ± 3.0%
EFFECT * STANDARD DEVIATION
•A minus sign W indicates a decree in efficiency caused by changing the level of the variable.
A plus sign (+) indicates an increase in efficiency caused by changing the level of the variable.
-------
Based on these two observations, it was decided to remove the vibrating horizontal screen
from the unit and rerun the experiment. This meant that one variable was eliminated, but
in doing so it added another response. The new response was that of maximizing that
portion of the flow passing through the collar screen. The removal of the horizontal
screen enabled the equipment supplier to reduce the overall height of future units by
approximately 21 inches to a new height of 63 inches.
The new response was defined as hydraulic efficiency or flow split. If 80 percent of the
influent flow passed through the collar screen, the collar screen had a hydraulic efficiency
of 80 percent, or a flow split of 80/20. The remaining 20 percent of the influent flow was
retained by the collar screen and became part of the solids concentrate flow.
58
-------
MODIFICATION 1
REMOVE VIBRATING HORIZONTAL SCREEN
OPERATING CONDITIONS
Influent Flow Rate 700 to 1200 gpm (combined sewage)
(50 to 86 gpm/ft2)
Collar Screen Speed 30 to 45 rpm
Collar Screen 110 and 175 mesh dacron cloth
(150 and 105 micron opening)
Operation Continuous flow and continuous backwash
Backwash Ratio 3.0 to 5.1 gal/1000 gal.
EXPERIMENTAL DESIGN
See Figure C2
PRESENTATION OF DATA
See Table C3
INTERPRETATION OF DATA
After two tests, it was apparent that the hydraulic efficiency of the unit could not be
maintained at the desired 80 percent level under the present operating conditions. It was
also observed that the higher collar screen speeds produced a better hydraulic efficiency,
but poorer solids removal efficiency. The improved hydraulic efficiency was attributed to
the increased screen area exposed to the flow at the higher rotational speed of the screen.
The poorer solids removal efficiencies were probably caused by more frequent screen
failures observed at the higher collar screen speed. Failure of the screen ranged from small
tears in the dacron material, which could be repaired, to large rips which necessitated the
replacement of the entire screen.
Based on these observations, the dacron screen was replaced with a stainless steel screen.
It was expected that this new material would improve the durability of the collar screen.
It was also decided to investigate the use of a finer screen and a higher collar-screen speed
with the objective of improving removal efficiencies and hydraulic efficiency.
59
-------
EXP.
NO.
1
2
3
4
5
6
7
8
RUN
NO.
2
8
4
7
1
3
6
5
COLLAR
SCREEN
175
110
175
110
175
110
175
110
FLOW
(GPM)
700
700
1200
1200
700
700
1200
1200
SPEED
(RPM)
30
30
30
30
45
45
45
45
FIGURE C2
EXPERIMENT C. - MODIFICATION 1
60
-------
TABLE C3
SUMMARY OF DATA
EXPERIMENT C - MODIFICATION 1
RESPONSE
RUN NO. 2
30RPM
RUN NO. 6
46RPM
HYDRAULIC EFFICIENCY
OF COLLAR SCREEN
SETTLEABLE SOLIDS
REMOVAL
TOTAL SUSPENDED
SOLIDS REMOVAL
C.O.D. REMOVAL
67/33
86%
23%
14%
73/27
59%
19%
14%
NOTE: EXPERIMENT TERMINATED DUE TO EXCESSIVE SCREEN
FAILURES.
61
-------
MODIFICATION 2
STAINLESS STEEL COLLAR SCREEN
OPERATING CONDITIONS
Influent Flow Rate 700 to 1200 gpm (raw sewage)
(50 to 86 gpm/ft2)
Collar Screen Speed 30 to 60 rpm
Collar Screen 150 and 200 mesh, market-grade, stainless-steel fabric
(105 and 74 micron opening)
Operation Continuous flow and continuous backwash
Backwash Ratio 3.0 to 5.1 gal/1000 gal.
EXPERIMENTAL DESIGN
See Figure C3
PRESENTATION OF DATA
See Tables C4 and C5
INTERPRETATION OF RESULTS
The data of Tables C4 and C5 revealed two important developments. First, the variables
affecting removal efficiencies became well defined, and the performance of the -unit
stabilized, as related to removal efficiencies. Second, the hydraulic performance of the
screening unit became more unstable, and the variables affecting this performance became
more difficult to isolate. This was again evident in the examination of the two-factor
interactions of Table C5. Most of the two-factor interactions associated with removal
efficiencies are small compared to their respective main effects, but the two-factor
interactions associated with hydraulic efficiency are large compared to the main effects.
The latter suggests that the main effects are meaningless in interpreting the effects that
the collar screen mesh size, speed, and flow rate had on hydraulic efficiency. It also
suggests that other variables may be influencing the performance.
Based on these observations, Experiment D2 was performed to investigate the effect that
flow rate and collar screen speed had on hydraulic efficiency, without regard to collar
screen mesh size. The design of the experiment and a summary of the results are
presented in Figure C4. While the hydraulic efficiencies were poor, the data revealed an
overall improvement in performance when the speed of the collar screen was increased to
62
-------
(9
EXP.
NO.
1
2
3
4
5
6
7
8
RUN
NO.
1
3
2
4
1A
3A
2A
4A
COLLAR
SCREEN
MESH
200
150
200
ISO
200
150
200
150
FLOW
(GPM)
700
700
1200
1200
700
700
1200
1200
RPM
30
30
30
30
45
45
45
45
FIGURE C3
EXPERIMENT D. - MODIFICATION 2
63
-------
TABLE C4
SUMMARY OF RESPONSES
EXPERIMENT D - MODIFICATION 2
EXP.
NO.
1
2
3
4
5
6
7
8
VARIABLES
COLLAR
SCREEN
MESH
200
150
200
150
200
150
200
150
H.OW
RATE
(GPM)
700
700
1200
1200
700
700
1200
1200
COLLAR
SCREEN
SPEED
(RPM)
30
30
30
30
45
45
45
45
RESPONSES
SETTLEABLE
SOLIDS
REMOVAL
98%
92%
99%
99.5%
98%
98%
99%
99%
TOTAL
SUSPENDED
SOLIDS
REMOVAL
33%
11.3
29
21
34
22
33
19
C.O.D.
REMOVAL
12.5
6
13
10
11
10
11
9
HYDRAULIC
EFFICIENCY
74/26
61/39
64/36
63/37
67/33
70/30
46/54
60/40
TABLE C5
SUMMARY OF MAIN EFFECT AND INTERACTIONS
EXPERIMENT D - MODIFICATION 2
MAIN EFFECTS
COLLAR SCREEN FROM 200 TO 150 MESH
FLOW FROM 700 TO 1200 FPM
RPM FROM 30 TO 45 RPM
TWO FACTOR INTERACTIONS
COLLAR SCREEN X FLOW
COLLAR SCREEN X RPM
FLOW X RPM
THREE-FACTOR INTERACTION
COLLAR SCREEN X FLOW X RPM
EFFECT ± STANDARD DEVIATION
SETTLEABLE
SOLIDS
REMOVAL
-1.4% ± .55%
+2.6% ± .55%
+1.4% ± .55%
+1,6% ± .55%
+1.4% ± .55%
-1.6% ^ .55%
•1.6% i .55%
TOTAL
SUSPENDED
SOLIDS REMOVAL
-13.9% ± 3.1%
+0.4% ± 3.1%
+3.4% ± 3.1%
+ 2.9% ± 3.1%
+ 0.9% ± 3.1%
-2.4% + 3.1%
-3,9% ±3.1%
C.O.D
REMOVAL
-3.0% ± 2.7%
+1.0% ± 2.7%
0.% ± 2.7%
+.8% ± 2.7%
+1.8% ± 2.7%
-1.3% ± 2.7%
-1.0% ± 2.7%
HYDRAULIC
EFFICIENCY
+ .8% ± 2.2%
-9.7% ± 2.2%
•4.7% ± 2.2%
+5.7% ±2.2%
+7.7% ± 2.2%
-5.7% ± 2.2%
-.2% ± 2.2%
-------
OPERATING CONDITIONS
COLLAR SCREEN - 200
MESH MARKET GRADE
STAINLESS STEEL FABRIC
OPERATION - CONTINUOUS
FLOW AND CONTINUOUS
BACKWASH
EXPERIMENTAL DESIGN
60
o
ui
5
8
-RUN NO.
660
FLOW RATE (6PM)
900
SUMMARY OF OBSERVATIONS
96%
98%
INCREASED
EFFICIENCY
93%
99%
SETTLEABLE SOLIDS
REMOVAL
23%
x
16%
INCREASED
EFFICIENCY
TOTAL SUSPENDED
SOLIDS REMOVAL
16%
19%
48/62
39/61
INCREASED
EFFICIENCY
\
41/59
36/64
HYDRAULIC EFFICIENCY
FIGURE C4
EXPERIMENT D2 - MODIFICATION 2
-------
60 rpm. The maximum speed that could be tested was 60 rpm; however, tests run by the
equipment supplier up to speeds of 100 rpm also indicated a peak efficiency at speeds
between 60 and 65 rpm. It was decided at this point to eliminate this variable and run all
future tests at 60 rpm.
Due to the poor hydraulic efficiencies of Experiment D2, several changes in the basic
screening unit were made to directly improve the hydraulic efficiency. A new distribution
dome was installed. The design of the new dome was to provide more uniform
distribution of the flow to the screen, and also to direct the flow to the screen at a more
perpendicular angle. A set of backwash sprays were also added to the unit. The new set of
sprays, installed on the inside of the collar screen, made it possible to wash both sides of
the screen simultaneously or alternately.
The backwash procedure was also changed from continuous backwashing to
intermittently backwashing the screens, whereby, at the end of every 4-1/2 minutes of
operation, the flow would be shut off and the screens would be backwashed with a hot
soap solution for 30 seconds. Finally, a new screen material, stainless steel tensile bolting
cloth, was installed. The new material has 50 percent more open area than the market
grade material first used.
After all the changes were installed, testing resumed as shown in Modification 3.
66
-------
MODIFICATION 3
MODIFIED DISTRIBUTION DOME
ADDITION OF BACK SPRAY
MODIFIED BACKWASH PROCEDURE
IMPROVED COLLAR SCREEN MATERIAL
OPERATING CONDITIONS
Influent Flow Rate 100 to 114 gpm/ft2
Collar Screen Speed 60 rpm
Collar Screen 105 mesh, tensile bolting cloth
(167 micron opening)
Operation 4-1/2 minutes on, 1/2 minute off for backwash
Backwash Ratio .50 to .57 gal/1000 gal.
PRESENTATION AND INTERPRETATION OF EXPERIMENTAL RESULTS
Prior to a formal experiment, a 45-minute test was performed to shake down the
modified unit. The hydraulic efficiency results of this experiment are shown on Figure C5
as a function of time. At first, the flow split was relatively good at 80/20; however, since
a coarser 105 mesh screen was used, a flow split of 90/10 was expected. As the test
progressed, it became apparent that the new washing procedure, while a big improvement,
was still unsatisfactory.
After the test, the screens were inspected and found to be severely blinded with what
appeared to be waste oil products. This explained the rapidly decreasing hydraulic
efficiencies. To remedy the blinding, a stronger solvent was periodically used to cut the
grease buildup on the screens and to renew their hydraulic capacity. A bench test was
performed in which several solvents were used to wash portions of one of the blinded
screens.
The solvents tested included gasoline, acetone, a liquid household detergent, a
commercial liquid cleaner (concentrated KOH), and "Mr. Clean." Each solvent was
applied to a screen panel blinded by the waste oil and allowed to "soak" for 30 seconds.
The screen was then sprayed with hot water and with the aid of a microscope, the
cleansing ability of the solvents was evaluated. The alkali-based detergents, such as "Mr.
Clean" and the concentrated KOH, were found to be the most effective cleaning agents.
Based on this information, the unit was again modified to include a new solvent pump
capable of injecting a concentrated detergent into the backwash piping for 30 seconds of
every 20 minutes of operation.
67
-------
MODIFICATION 4
ADDITION OF CONCENTRATED DETERGENT BACKWASH WATER
OPERATING CONDITIONS
Influent Flow Rate 1400 gpm (raw sewage)
(100gpm/ft2)
Collar Screen Speed 60 rpm
Collar Screen 105 mesh TBC
(167 micron opening)
Operation 4-1/2 minutes on, 1/2 minute off for normal backwash,
with a 1 /2-minute concentrate backwash every 20 minutes
Backwash Ratio .25 gal/1000 gal.
PRESENTATION AND INTERPRETATION OF EXPERIMENTAL RESULTS
Again, prior to a formal experiment, a short test was performed to monitor hydraulic
efficiency. The results of that test are presented on Figure C6. Grease blinding remained a
problem, but the concentrated detergent backwash was effective in renewing the collar
screen's hydraulic capacity and, therefore, improving the unit's average hydraulic
efficiency.
At this point, the original equipment supplier was consulted. Independent and concurrent
tests by the supplier showed that the hydraulic efficiency of the Portland unit should be
better than that which was observed at Sullivan Gulch. A comparison of the Portland unit
to the supplier's units revealed that the Portland unit had a 10-inch influent pipe while
the supplier's unit had an 8-inch influent pipe. Under a flow condition of 1500 gpm, this
meant that the velocity in the Portland influent pipe was 6.1 fps compared to 9.6 fps in
the supplier's influent pipe. This represented a significant difference. To test this
difference, an 8-inch orifice plate was installed in the Portland screening unit and another
short test was performed.
The addition of the orifice plate proved to be the turning point in the investigation. All
variables were now defined and the level of each variable was well established. An
experiment was performed on the screening unit in its final form. The results of that
experiment are presented and discussed in the text.
68
-------
15
30 45
DURATION OF TEST (MINUTES)
FIGURE C5
HYDRAULIC EFFICIENCY VS TIME
MODIFICATION 3
12 16
DURATION OF TEST (MINUTES)
D AFTER CONCENTRATED DETERGENT BACKWASH
A AFTER NORMAL BACKWASH
O BEFORE BACKWASH
FIGURE C6
HYDRAULIC EFFICIENCY VS TIME
MODIFICATION 4
69
20
24
28
-------
r-
BIBLIOGRAPHIC: Cornell, Howland, Hayes & Merry field. Rotary
Vibratory Fine Screening of Combined Sewer Overflows FWPCA
Publication No. DAST 5, 1970.
ABSTRACT: The objective of this study was to determine the
feasibility, effectiveness, and economics of employing high-rate,
fine-mesh screening for primary treatment of storm water overflow
from combined sewer systems. The final form of the screening unit
stands 63 inches high and has an outside diameter of 80 inches. The
unit is fed by an 8-inch pipe carrying 1700 gpm (122 gal/min/ft2)
which is distributed to a 60-inch diameter rotating (60 rpm)
stainless steel collar screen having 14 square feet of available screen
area and a 165 mesh (10S micron opening, 47.1 percent open area).
The screen is backwashed at the rate of 0.2 35 gallons of backwash
water per 1000 gallons of applied sewage. Based on final
performance tests run on dry-weather sewage, the unit is capable of
99 percent removal of floatable and settleable solids, 34 percent
removal of total suspended solids and 27 percent removal of COD.
The screened effluent is typically 92 percent of the influent flow.
On the basis of a scale-up design of a 2$ mgd screening facility, the
estimated cost of treatment is 22 cents/1000 gallons.
BIBLIOGRAPHIC: CorneU, Howland. Hayes & Merryfield. Rotary
Vibratory Fine Screening of Combined Sewer Overflows FWPCA
Publication No. DAST 5, 1970.
ABSTRACT: The objective of this study was to determine the
feasibility, effectiveness, and economics of employing high-rate,
fine-mesh screening for primary treatment of storm water overflow
from combined sewer systems. The final form of the screening unit
stands 63 inches high and has an outside diameter of 80 inches. The
unit is fed by an 8-inch pipe carrying 1700 gpm (122 gal/min/ft2)
which is distributed to a 60-inch diameter rotating (60 rpm)
stainless steel collar screen having 14 square feet of available screen
area and a 165 mesh (105 micron opening, 47.1 percent open area).
The screen is backwashed at the rate of 0.235 gallons of backwash
water per 1000 gallons of applied sewage. Based on final
performance tests run on dry-weather sewage, the unit is capable of
99 percent removal of floatable and settleable solids, 34 percent
removal of total suspended solids and 27 percent removal of COD.
The screened effluent is typically 92 percent of the influent flow.
On the basis of a scale-up design of a 25 mgd screening facility, the
estimated cost of treatment is 22 cents/1000 gallons.
BIBLIOGRAPHIC: CorneU, Howland, Hayes & Merryfield. Rotary
Vibratory Fine Screening of Combined Sewer Overflows FWPCA
Publication No. DAST S, 1970.
ABSTRACT: The objective of this study was to determine the
feasibility, effectiveness, and economics of employing high-rate,
fine-mesh screening for primary treatment of storm water overflow
from combined sewer systems. The final form of the screening unit
stands 63 inches high and has an outside diameter of 80 inches. The
unit is fed by an 8-inch pipe carrying 1700 gpm (122 gal/min/ft2)
which is distributed to a 60-inch diameter rotating (60 rpm)
stainless steel collar screen having 14 square feet of available screen
area and a 165 mesh (105 micron opening, 47.1 percent open area).
The screen is backwashed at the rate of 0.235 gallons of backwash
water per 1000 gallons of applied sewage. Based on final
performance tests run on dry-weather sewage, the unit is capable of
99 percent removal of floatable and settleable solids, 34 percent
removal of total suspended solids and 27 percent removal of COD.
The screened effluent is typically 92 percent of the influent flow.
On the basis of a scale-up design of a 25 mgd screening facility, the
estimated cost of treatment is 22 cents/1000 gallons.
ACCESSION NO:
KEY WORDS:
Screens
Overflow Treatment
Economic Analysis
Storm Water
Separation
Rotary Screens
Waste Water
Treatment
ACCESSION NO:
KEY WORDS:
Screens
Overflow Treatment
Economic Analysis
Storm Water
Separation
Rotary Screens
Waste Water
Treatment
ACCESSION NO:
KEY WORDS:
Screens
Overflow Treatment
Economic Analysis
Storm Water
Separation
Rotary Screens
Waste Water
Treatment
:—I
i
U. S. 1-.OVKRNMENT PRINTING OFFICE t 1"0 O - 308-937
------- |