WATER POLLUTION CONTROL RESEARCH SERIES
WP-20-16
Strainer/Filter Treatment
of
Combined Sewer Overflows
U.S. DEPARTMENT OP THE INTERIOR
FEDERAL WATER POLLUTION CONTROL ADMINISTRATION
-------
Strainer/Filter Treatment
of
Combined Sewer Overflows
Federal Water Pollution Control Administration
Storm and Combined Sewer Pollution Control Branch
Contract No. 14-12-17
Final Report
-by-
Stephen S. Blecharczyk
and
Edward L. Shunney
Research and Development
Fram Corporation
July 1969
-------
FWPCA Review Notice
This report has been reviewed by the Federal Water
Pollution Control Administration and approved for
publication. Approval does not signify tnat the
contents necessarily reflect the views and policies
of the Federal Water Pollution Control Administration,
-------
TABLE OF CONTENTS
Page
Index of Figures iv
Index of Tables v
Abstract vi i
Conclusions vii i
Kecommendations vi i i
I. Introduction
A. Nature of the Problem 1
B. Previous Investigations 3
C. Statement of the Problem 4
II. Experimental Program and Procedures
A. Outfall Analysis 5
B. Self-Cleaning Strainer 6
C. Self-Cleaning Filter 6
III. Experimental Kesults
A. Selection of Synthetic Contaminant 11
B. Outfall Analysis 13
C. Strainer Experiments 20
1. Flat Sheet Tests 20
2. Model Self-Cleaning Strainer 21
a. Synthetic Substrate 21
b. Sheridan Street Samples 23
c. Mechanical Reliability 24
d. Results with Fresh Solids 27
D. Self-Cleaning Filter 27
1. Johns-Manville Test System 27
a. Screening Tests 27
b. Filter Aid and Chemical Treatment 31
2. Self-Cleaning Strainer - Vacuum Modified 34
IV. Uiscussion of Results 36
A. Site Analysis 36
B. Self-Cleaning Strainer Effectiveness 37
C. Self-Cleaning Filter Effectiveness 40
V. References 41
VI. Appendix 43
-------
INDEX OF FIGURES
Figure Page
1 Cross-Section of a Circular Mold for Flat Sheet
Tester 7
2 Flat Sheet Test Fixture 8
3 System Schematic Flow Jiagram for Flat Sheet Testing. 9
4 15 GPM Model Self-Cleaning Strainer 10
5 Schematic Diagram of Model Self-Cleaning Strainer lOa
6 Johns-Manville Jiatomaceous Earth Filter Test System.12
7 System Schematic Flow Oiagram-Strainer-Filter 51
iv
-------
INDEX OF TABLES
Table Page
I Sheridan Street Overflow Samples - January 30, 1968 15
II Sheridan Street Overflow Samples - February 2, 1968 15
III Sheridan Street Overflow Samples - January 30, 1968 -
Effect of 12-Hour Settling 15
IV Sheridan Street Overflow Samples - April 15, 1968 16
V Sheridan Street Overflow Samples - April 24, 1968 17
VI Effect of Sample Size on Overflow Analysis 19
VII Flat Sheet Test Results 20
VIII Initial Model Strainer Results - 60 x 60 mesh, 230
micron Opening 21
IX Model Strainer Results - 60 x 60 mesh, Synthetic
Contami nant 22
X Model Strainer Results - Effect of Mesh Size on
Solids Removal 23
XI Model Strainer Results - Sheridan Street Overflow Samples 24
XII Model Strainer Results - Bucklin Point Sewage Treatment
Plant Influent, 60 x 60 mesh screen,230 Micron Opening... 25
XIII Model Strainer Results - Bucklin Point Sewage Treatment
Plant Influent, 100 x 100 Mesh Screen,150 Micron Opening. 26
XIV Dissolved Organic Concentration in Bucklin Point
Influent 26
XV Model Strainer Results - Fresh Sewage Solids, 60 x
60 Mesh Screen 27
XVI Effect of Submergence Time on Filter Performance 28
XVII Effect of Knife Advance on Filter Performance 28
XVIII Diatomlte Evaluation with Synthetic Substrate 29
XIX Effect of Diatomite Type on Filter Efficiency 30
-------
INDEX OF TABLES (Continued)
Table Page
XXI Sheridan Street Overflow - uiatomaceous Earth Filtration... 32
XXII Jiatomaceous Earth Filtration - Bucklin Point Primary
Effluent 33
XXIII Vacuum Modified Filter Runs with Model Strainer 34
XXIV Dissolved Organic Concentration in Filtered Effluent 35
XXV Modified Filter Runs - Hyflow Super Cel 35
XXVI Model Strainer Results - Bullock's Point Treatment Plant... 39
vi
-------
ABSTRACT
The primary objective of this feasibility study was to evaluate the prin-
ciple of a 'self-cleaning strainer, self-cleaning filter1 concept for the
treatment of combined sewer overflows. The anticipated goal was to de-
sign and construct a prototype system capable of handling up to 1000 gal-
lons per minute with a B.O.D. reduction near 60 percent, and with the
capability of automatic operation in remote locations.
A combined sewer overflow in Providence, Rhode Island, was sampled and
analyzed to determine the type and amount of contaminant discharged into
the receiving stream. The average concentration was determined to be
nearly equal to pure domestic sewage. It was also determined that the
analysis reported for overflows is very dependent on the exact sampling
method used. Automatic sampling devices utilizing small diameter tubing
do not take a representative sample since the suspended solids distribu-
tion is not uniform over the cross-sectional area of the discharging
stream. Based on overflow sample analysis data (samples taken manually),
a syntnetic substrate solution was prepared to evaluate a forced flow
self-cleaning strainer for significant operating variables.
The strainer and filter systems were evaluated using the syntnetic sub-
strate, primary influent to two separate municipal treatment plants, fresh
sewage solids and actual combined sewer flow. It was demonstrated that
the strainer model produced consistent suspended solids removal rates near
35 percent under highly varying load conditions, at a flux of 25 gallons
per minute per square foot.
The diatomite study showed operational success could be achieved at a
50 percent organic reduction rate at 4 gallons per minute per square foot
of area, but at a minimum estimated operating cost of $1.50 per 1000 gal-
lons.
This report was submitted in fulfillment of Contract 14-12-17 between the
Federal Water Pollution Control Administration and the Fram Corporation.
vii
-------
Conclusions
This feasibility study has shown that sampling methods
commonly used in evaluating the effect of combined sewer
overflows on receiving streams cannot be considered reliable.
Tne results indicate that most of the calculated loads that
are based on automatic sampling stations have most likely
understated the actual case. Visual observations have shown
that whole sections of toilet paper and most large suspended
solids are not sampled with small diameter, low velocity
sampling probes.
The applicability of a self-cleaning strainer in the
treatment of raw sewage has been demonstrated in terms of con-
sistent removal of suspended solids. The question of the total
level of organic removal that is possible has not been completely
answered since actual overflow samples were not available in
sufficient supply during the last seven months of the contract.
The results obtained with primary influent to a municipal treat-
ment plant and with fresh solids show that it should be possible
to remove at least 30 percent of the organic load in a combined
sewer overflow with a self-cleaning strainer.
The authors believe that a strainer-filter system of the type
originally envisioned in this project is not feasible from a cost
and operational point of view. It is their contention that if
additional treatment is necessary beyond that attainable with a
self-cleaning strainer, then a much simpler and less mechanically
complicated secondary system can be constructed.
Recommendations
It is recommended that a full-scale study be undertaken to
devise and establish a uniform approved method for the sampling of
combined sewer overflows. Since the design of proposed combined
sewer overflow treatment systems is based on data considered to be
questionable, the projects themselves must be considered questionable.
Primary evaluation was made with 60 x 60 (230 microns), 80 x 80
(190 microns) and 100 x 100 (150 microns) mesh screens. Best over-
all results were obtained with the 80 x 80 square weave screens.
These results combined with other studies indicate the need for a
more prolonged study of screen configuration and materials of con-
struction in full scale applications. Pore size to be studied
should be in the 170 - 200 micron range.
This project in conjunction with the Glenfield-Kennedy field
demonstration project should establish the efficiency and reliability
of self-cleaning strainers for combined sewer treatment.
vm
-------
I. Introduction
A. Nature of the Problem
It has been stated that during heavy rains, sewer systems that
carry combined storm-water and sewage can deliver up to 95 percent
of a community's raw sewage to a receiving stream without any form
of treatment. Storm-water runoff alone has been shown to contamin-
ate streams. On the other hand, it is clearly evident that not
enough is known about the highly variable nature of storm or com-
bined sewer runoff to permit clear-cut solutions to these water
pollution problems.
Combined sewer overflows are known to be an important source of
pollution (previously understated), but their intermittent nature
makes it difficult to obtain precise information about their total
effect and specific characteristics. Preliminary results of several
investigations suggest that stormwater overflows differ markedly in
character from what might and has been expected. Researchers have
expected a heavy runoff to dilute the sewage and cause light pollu-
tion. Instead, in some cases, the suspended solids concentration
increased as the intensity of runoff increased. With storm flow
three times that of dry-weather flow, samples taken during the first
five minutes at one station showed suspended solids 2.5 times of
average sewage. Samples taken more than thirty minutes later had a
solids concentration only 30 percent of the original value. In some
interceptors, therefore, it would appear that during dry-weather flow
solids settle out, which are then ultimately flushed during storms.
Older cities such as Providence, Rhode Island, have systems that
were installed in the late 1800's or 1900's. The population growth
has naturally produced increased sewage flow and spills can occur
during dry-weather conditions. Additionally, regulator malfunctions
can frequently cause unexpected discharge into a stream. The U. S.
Public Health Service has estimated that three to five percent of all
raw sewage is discharged to receiving streams by combined sewer over-
flows. This would mean at least 68 billion gallons of raw sewage enter
the nation's rivers and streams per year.
Complete separation of existing combined sewers is not considered
a practical solution since the cost and inconvenience is a burden the
taxpayer is not prepared to assume. It has already been pointed out
that stormwater alone is a source of pollution and thus this approach
is only a partial solution from the standpoint of pollution control.
A number of alternatives have been suggested: separation at the source;
separation in existing systems; express sewers; reduced stormwater input;
temporary storage; and point of discharge treatment. Analysis of re-
ports from various large cities in the United States indicate that al-
most all combined sewer systems will have to be engineered on a best-fit
basis and that more than one method will be used per system.
-------
This particular program is concerned with the point of
discharge treatment approach. Specifically the project con-
sidered solids separation without storage, and chlorination
of the effluent prior to discharge into a receiving stream.
The basic concept is one of using a self-cleaning strainer in
series with a self-cleaning diatomaceous earth filter capable
of operating in remote locations on demand without the need of
on-site operating personnel. The system would be designed to
process a variable flow up to 1000 6PM.
The basic function of the two components tested and evaluated
in this project are described below. The operation sequence of
the original total design concept is described and illustrated
on pages 47 - 51 in the Appendix.
In most cases, tne storm sewer overflow outlets would require
extensive modification to adapt for any treatment system. The
most practical approach would be the construction of a receiving
basin at the mouth of the discharge line. A pipe attached to a
sturdy cage resting on the bottom of tne basin would be the
source of contaminated water for the proposed strainer-filter.
Such a receiving basin, witn an appropriate cage guard on the pipe,
would prevent large objects such as animal corpes, tree branches,
construction timbers from blocking water flow to the strainer unit.
Periodic removal of such objects from the basin during times of no
discnarge would be the only maintenance required to keep the re-
ceiving basin operative. A level sensing device in the basin would
activate and deactivate the strainer-filter system. The system
would be provided with portable refuse bins to receive dewatered
solids discharged from the filter unit. At each storm sewer dis-
charge installation, access for periodic truck pick-up of the refuse
bins would be required.
Self-Cleaning Strainer
The strainer is a modified version of the current Fram self-
cleaning device. It was proposed that the unit would utilize
permanent screening on the strainer support basket which would be
continuously rotating with periodic blowdown cycles to be determined
by a pressure differential across the strainer screen. The back-
wash pump would operate continuously at sufficient pressure to back-
wash tne strainer screen. An internal baffling arrangement directs
the backwash contaminated liquid into the vessel sump from which it
would be discharged on the blowdown cycle.
It was proposed that the screen would be a permanent structure
in the self-cleaning strainer with a particle selectivity of approxi-
mately 50 microns and would be designed to relieve the self-cleaning
filter from all coarse particles larger than 50 microns, thereby
permitting the diatomite filtration unit to operate more efficiently
in the removal of fine suspended particles.
-------
Self-Cleaning Filter - Vacuum Process
This unit is a new concept involving a non-pressurized rectangular
cross-sectional vessel incorporating a rotary drum filter, utilizing a
filter cloth material capable of receiving a precoat of diatomaceous
earth and various powered or liquid chemicals which would be desirable
for water purification or for filtration efficiency. The filter drum
would be operated in cycles, based on pressure differential, by a lever
control switch on the suction side of the effluent pump. The lower
two-thirds of the filtration basket would be submerged in the liquid
to be filtered with the upper one-third being exposed to warm dry air
circulation to enhance the ability to discharge a relatively dry filter
cake.
The proposed method of backwashing and discharging of the filter
cake is considered unique. At the present time, one of the basic faults
of practically all diatomaceous earth equipment is that a completely
satisfactory dry cake discharge system has not been developed. The
proposed filter cake discharge method should permit a much more satis-
factory operation. The backwash would occur by a forced hot air stream
slightly ballooning the filter cloth against an adjustable rubber
scraper. This is, in turn, followed by a high pressure water discharge
spray to remove any remaining traces of diatomaceous earth from the
pores of the filter cloth.
B. Previous Investigations
Large concentrations of sediment, gravel or other coarse contaminants
have been filtered with self-cleaning strainers of the heavy duty bar
screen type and with strainers capable of removing contaminants down
to 50 micron size. Although screen systems can be chosen to reduce the
maximum particle down to less than 15 microns, these self-cleaning
strainers tend to clog or blind-off with hard filter cakes that are
not easily removed by conventional backwashing techniques. Self-cleaning
strainers have been used to remove organic suspended solids from waste
streams. Boucher and Evans (7) reported 50 - 90% removal efficiency was
found to be greater as the feed suspended sol Ids concentration increased.
Hudson (8) studied the removal of partially decomposed organic solids by
metal screen strainers and found them to be very effective, but subject
to clogging. It was indicated that backwashing efficiency could be
increased by the use of chlorine or ultra-violet radiation. Evans (9)
has also reported that 70% incidental removal of collform bacteria has
occurred during micro-straining operations, by surface adsorption on
particles removed by the strainer. Actual field tests (10, 11) on the
straining of river water by the Fram self-cleaning strainer has confirmed
that its present design is capable of handling massive contaminant slugs
during erratic river flow periods.
-------
Based on results obtained by the FWPCA at Pomona, California,
on secondary effluent (12) considerable concern has been expressed
that a diatomaceous earth filter for such an application has serious
functional drawbacks: premature clogging ofthe filter; insufficient
reduction in effluent turbidity and high raw material requirements.
Using both a vacuum filter and a pressure filter, a number of filter-
aids were evaluated at various flow rates and body-feed concentrations.
Eighty-five percent reduction in turbidity was accomplished at flow
rates of 0.53 to 1.0 gpm/ft2 using Celite 545, Celite 503, and Hyflo
Super-Cel grades of diatomaceous earth. The only information regard-
ing organic removal was the statement that a 21 percent chemical oxy-
gen demand decrease was found using Celite 545 at a flow rate of
0.52 gpm/ft2.
It is well known (1, 2, 3, 4) that the amount of body feed, as
well as type of contaminant, is very important in diatomaceous earth
filtration when applied to municipal water systems. The AWWA Task
Group report, "Oiatomite Filters for Municipal Use", February 1965,
pinpoints many of the problems associated with diatomaceous earth
filters. On the other hand, the problem of concern here is quite
different from that of potable water production or swimming pool
clarification. It is within the scope of this proposal to use dia-
tomaceous earth as both a mechanical strainer and chemical absorber
without regard to absolute turbidity reduction. Primary considera-
tion would be given to B.O.D. removal.
Effective use of diatomaceous earth has been made in the treatment
of laundry wastes (5, 6) utilizing automatic backwash and precoat
cycles. The experience developed in these instances is quite rele-
vant to the storm sewer situation. Spade (6) listed typical results
showing B.O.D. reductions near 90% and a reduction in suspended solids,
for example, from 220 to 12 mg/1. The characteristics of the laundry
wastes reported are similar in B.O.D. and suspended solid levels to
what might be found in a combined sewer outfall.
C. Statement of the Problem
The purpose of this study was to conduct a feasibility investigation
to determine the relative effectiveness of the self-cleaning strainer -
filter concept in treating combined sewer overflows. The variables in-
volved in this solids separation concept were investigated and the dif-
ficulties to be expected in a prototype design were considered.
Analysis of a typical combined sewer overflow in Providence, Rhode
Island, was carried out in conjunction with this project.
-------
II. Experimental Program and Procedures
The program was divided into three separate parts conducted con-
currently whenever feasible: (1) Analysis of sewer overflow: (2)
Self-cleaning strainer effectiveness; (3) Self-cleaning filter
effectiveness.
A. Outfall Analysis
The primary purpose for including the analytical study of a com-
bined sewer overflow was to determine the level of contamination that
could be expected in the Providence area. This data was used in the
preparation of samples for the laboratory evaluation of the proposed
strainer-filter concept. It was anticipated that undue delays in the
experimental program would occur if the study was restricted to a
study of actual overflow samples.
The correlation of rainfall with overflow was attempted using
data collected by the State of Rhode Island Water Pollution Branch
at a site one air mile from the drainage area contributing to the
selected site overflow.
In cooperation with the City of Providence, Rhode Island, an over-
flow site on the Woonasquatucket River was selected for study. A
54-inch sewer feeds into a 60-inch semi-circular open top channel prior
to discharge into the river.
Following visual observation of sewer overflow during two storm
events, the following sampling procedure was used throughout the pro-
gram:
1. Sampling was performed manually. Whenever rain fell in the
drainage area, a technician was dispatched to the overflow site.
2. Whenever flow was detected visually, sampling was started.
3. Samples were taken at fifteen minute intervals during the first
two hours of flow, thereafter at 30 minute intervals for two hours.
Additional samples were taken as dictated by the particular over-
flow event.
4. Samples taken for analysis were discrete in nature, not composites
over each time interval and were taken in two quantities.
a. Two-gallon sample taken with a one-gallon pail.
b. One-gallon sample taken with a one pint wide mouth cup.
5. Samples were brought to the analytical laboratory within six hours
of the initial sampling time. The following analyses were performed
immediately: B.O.D., C.O.D., settleable solids, suspended solids,
coliform count and dissolved oxygen level. Total and volatile solids
determinations were performed within 18 hours after sample collection.
-------
6. All analyses were performed in accordance with the Twelfth
Edition of "Standard Methods for the Examination of Water and
Wastewater".
B. Self-Cleaning Strainer
This program was divided into two sections: (1) experiments
with strainer screen in the configuration of a flat sheet; (2) ex-
periments with a working model of the Fram self-cleaning strainer.
1. Flat Sheet Testing
This test procedure is based on the premise that a relationship
exists between the rate of accumulation of a solid material on a
screen and its ability to remove the same material in a continuous
cleaning system without blinding off, or the development of an ex-
cessive pressure drop across the screen.
Flat sheet samples were prepared using the following procedure
with a mold described in Figure 1. Two gaskets 1/8" thick are molded
from plastisol. One side^of each is painted with plastisol. One is
put back in the mold painted side up; the screen sample is placed on
this and the other gasket is placed with the painted side making con-
tact with the screen. A piece of 1/8" aluminum 3-1/4" in diameter is
placed on this gasket with a 300 gram weight put on top for compression.
It is then put in the oven for cure. Cure time in all cases is 8 -
10 minutes at 300°F.
The samples were tested in a fixture as shown in Figure 2, in the
mode illustrated in Figure 3. Various screens were tested with the
same contaminant at identical flow rates and solids concentration.
Time-pressure readings were taken until the pressure drop across the
screen reached 19 psig.
2. Model Self-Cleaning Strainer
The strainer used in this study is shown in Figure 4. The sche-
matic drawing in Figure 5 illustrates how the unit functions. This
model has a screen area of 40 square inches available for flow and
filtration. The housing is constructed of plexiglass and is limited
to an internal working pressure of 15 psig. Maximum rated flow of
the unit is 15 gallons per minute.
The flat sheet testing procedure was used to screen those wire
screens considered suitable for use in the model unit.
C. Self-Cleaning Filter
This part of the study was divided into two sections: (1) a 0.1
square foot filter area test system designed and built by the Johns-
Manville Company; (2) adaptation of the model strainer into a vacuum
filtration mode.
-------
Cross-Section of a Circular Mold
For Flat Sheet Tester
X
N
5-3/4"
2-1/8""
XXX
3/4"
-$r 3/8"
rr
Figure 1
-------
FLAT SHEET TEST FIXTURE
.015
C. Nipple
Sample Area 3.5 irr^
.1/8 - 27 NPTF
1/8 - 27 NPTF
SCALE = FULL
Figure 2
8
-------
SYSTEM SCHEMATIC FLO VI DIAGRAM FOR FLAT SHEET TESTING
PUN/\P
-------
15 qom Model Self-Cleaning Strainer
-------
Schematic jiagram Of
Model Self-ileaninq Strainer
A Inlet-
B Flow Control Deflector Baffle,
C. Backwash Nozzle-
D. Cartridge Screen-
E. Bottom Sump Area Where Heavy Contaminants
Are Stored Between Blow-downs-
F. Bottom Blow-down Connection For Removal Of
Heavy Contaminants-
G. Upper Sump Area Where Lightweight Contaminants
Are Stored Between Blow-downs,
H. Upper Blow-down Connection For Removal Of
Lightweight Contaminants.
TJ
Figure 5
-------
1. Johns-Manville System (13, 14)
This unit simulates tne function of an equal area on the surface
of a full-sized rotary vacuum precoat filter drum. The system is
schematically outlined in Figure 6. during the course of each revolu-
tion, the filter drum passes through submergence, drying and residue
removal phases; and the small filter test leaf is capable of such se-
quential operation on a timed basis.
This part of the experimental program was outlined as follows:
a. Preliminary screening
(1) Selection of representative samples of a diatomaceous
earth.
(2) Selection of suitable septa.
(3) Selection of a suitable synthetic contaminant solution.
b. Filter performance studies using:
(1) Three diatomaceous earth sizes
(2) Three diatomaceous earth slurry feed rates
(3) Various contaminant concentrations
(4) Various flow rates
(5) Various septa
c. Filter performance with chemical treatment or additional
filter aids
(1) Activated carbon
(2) Ion exchange
(3) Flocculating agents
III. Experimental Results
a. Selection of Synthetic Contaminant
The data obtained during the first two storm events sampled
was used to establish the following minimum characteristics:
(1) B.O.U. - 125 mg/1; (2) C.0.0. - 400 mg/1; (3) Suspended
Solids - 250 mg/1; (4) Settleable Solids - 2
Based on prior experience, a biodigestable dog food (Burgerbits)
was selected as a suitable approximation to the chemical composition
of human solid waste products. Various concentrations of the dog food
were tested to determine how well the above noted values could be at-
tained without tne necessity of using some additional material. Twenty
liter samples were prepared by blending the proper amount of dog food
witn one liter of water for 15 minutes, followed by one hour of aeration
and dilution to 20 liters with tap water. At a concentration of 0.4 gm/1
tne following average values were obtained:
11
-------
JOHMS-MANV1LLE DIATOMATEOUS EARTH FILTER TEST SYSTEM
VACUUK
VACUUM
SUPPLY -*
ro
Figure 6
-------
(1) B.O.D. - 172 mg/1; (2) C.O.D. - 433 mg/1; (3) Suspended
Solids - 264 mg/1; (4) Settleable Solids - 2.1
It should be noted that since the Fram self-cleaning strainer
operates with forced flow, particle size reduction of human feces oc-
curs in the feed pump. The settling rate of the blended dog food and
mechanically ground fresh human feces were found to be essentially
the same. Since the density of the two materials are essentially the
same, then the average particle size and distribution were considered
to be essentially the same, based on Stokes Law.
Typical single values obtained when evaluating the dog food chara-
cteristics at various concentration levels are shown in Table (a) in
the Appendix. The values obtained clearly show that reproducible prop-
erties of the resulting solution were readily achieved well within the
normal variations of the analytical methods. The blending time was
found to be the most important variable.
B. Outfall Analysis
The first complete outfall analysis was performed on samples
taken January 30, 1968, and the results are shown in Table 1. A second
overflow was sampled and analyzed on February 2, 1968, and these results
are shown in Table II. There is a distinct difference between the two
sets of samples which is probably due to the historical events in the
sewer system. Prior to the January 30, event, there hadn't been any
significant rain or overflow for fifteen days. High density solids,
such as sand, coffee grounds, etc., could have accumulated along the
sewer lines between January 15 and January 30 and were flushed out with
the high flow rates on the 30th. The first sample taken on January 30
had a very large quantity of readily settleable coffee grounds, etc.
On the other hand, the February 2 samples showed little or no readily
settleable material such as coffee grounds, in addition to the fact
that the total amount of settleable solids was significantly less at
the first flushing. On January 14-15, 1968, the recorded rainfall be-
tween 10 PM ana 6 AM was 1.20 inches, which resulted in rapid and com-
plete flushing of the sewer system. The system was quiet for fifteen
days before the first sampling, versus only two days before the second
overflow sampling.
As a guide for the evaluation of straining or settling, the first
set of samples was analyzed twice to determine the effect of twelve
hour settling. The results are shown in Table III. The difference ap-
pears to be significant only with samples containing abnormally high
settleable solids.
On March 17, 1968, a record amount of rainfall caused considerable
flooding of the Woonasquatucket River and washed out a foot bridge at the
Sheridan Street combined overflow sewer, used for sampling in this
project. Due to the hazardous nature of the area during and after the
storm, no samples were taken for analysis.
13
-------
Time Interval
Previous
Sample
Sample No.
1
2
3
4
5
6
7
8
January 30, 1968 - Rainfall - 0.40 inches
TABLE I - SHERMAN STREET OVERFLOW SAMPLES - JANUARY 30, 1968
B.O.D.
mg/1
-
2 hrs.
0.5 hrs.
0.5 hrs.
0.5 hrs.
2.0 hrs.
12.0 hrs.
0.5 hrs.
440
150
90
76
80
22
400
300
C.O.D.
mg/1
1243
428
317
214
222
113
531
562
Settleable
Solids
ml/1
35.0
6.5
2.0
2.0
2.2
1.2
3.0
4.0
Suspended
Solids
mg/1
968
310
172
110
80
76
400
310
Volatile
Solids
mg/1
3200
400
125
400
200
200
800
1000
Total
Solids
mg/1
4000
1600
1000
800
400
200
2800
2000
Coli form
MPN
11,000,000
2,400,000
240,000
2,400,000
11,000,000
11,000,000
4,600,000
11,000,000
TABLE II
Sample No. Time Interval B.O.D.
1
2
3
4
5
0.5 hrs.
0.5 hrs.
0.5 hrs.
0.5 hrs.
65
60
38
78
60
SHERIDAN STREET OVERFLOW SAMPLES
C.O.J. Settleable Suspended
162
190
200
106
74
380
373
-
264
217
3.0
3.5
2.3
0.8
1.3
FEBRUARY 2, 1968
Volatile Total
Coliform
2800
800
400
1200
800
3600
2800
2000
2400
1600
750,000
1,500,000
2,400,000
430,000
240,000
D.O.
mg/1
8.8
9.9
10.
10.
12.0
7.3
7.5
.9
.7
February 2, 1968 - Rainfall - 0.28 inches
TABLE III - SHERIDAN STREET OVERFLOW SAMPLES - JANUARY 30, 1968 - EFFECT OF 12-HOUR SETTLING
Settleable Solids
Sample No.
1
2
3
4
5
6
7
8
B.O.D.
Supernatant
175
130
70
62
70
18
310
275
Blended
440
150
90
76
80
22
400
300
35.0
6.5
2.0
2.0
2.2
1.2
3.0
4.0
-------
During April, two significant overflows were sampled and analyzed.
Table IV lists the data obtained in samples taken April 15, 1968. It
is significant that the greatest organic load occurred after the first
flush (Sample 3). At 3:17 P.M., a heavy flow of human waste and toilet
tissue was observed in the first four inches of water, just after Sample
No. 5 was taken. This condition lasted for approximately 20 seconds and
could not be adequately sampled; therefore, it is not included in the
tabulated results shown in Table IV. Table V lists the data obtained in
samples taken April 24, 1968.
It should be noted that, in both instances, the flow rate was cyclic
in nature. Although the rainfall during the sampling period was almost
identical in both cases, it probably is a coincidence that a change in
the flow pattern and waste occurred between Samples 5 and 6 both times.
15
-------
TABLE IV - SHERIDAN STREET OVERFLOW SAMPLES - APRIL 15, 1968
Sample No.
1
2
3
4
5
6
7
8
Time Interval
From Previous
Sample
Visual Start of
Flow
15 minutes
15 minutes
15 minutes
30 minutes
30 minutes
30 minutes
30 minutes
B.O.D.
mg/1
126
75
90
110
46
120
65
18
Settleable
Solids
ml /I
2.5
0.9
7.5
3.0
0.5
0.25
negligible
negligible
Suspended
Solids
mq/1
30
5
220
90
30
80
40
10
Volatile
Solids
mq/1
65
55
80
40
35
150
45
20
Total
Solids
mq/1
295
255
240
230
140
300
170
no
Col i form
MPN/
100 ml
2,400,000
750,000
930,000
930,000
240,000
4,600,000
430,000
430,000
D.O.
mq/1
7.8
7.2
6.7
7.2
7.1
7.6
7.8
7.8
Note:
(1) Total Rainfall from
8:10 A.M. to 3:00 P.M.
8:10 A.M. to 10:30 A.M.
10:30 A.M. to 3:00 P.M.
(2) Flow commenced at 1:45 P.M.
Flow rate at 1:50 P.M.
2:45 P.M.
3:45 P.M.
0.42 inches
0.07 inches
0.35 inches
1150 gal/min.
1600 gal/min.
400 gal/min.
-------
TABLE V - SHERIDAN STREET OVERFLOW SAMPLES - APRIL 24, 1968
Time Interval
From Previous
Sample No. Sample
1 Visual Start of
Flow
2 15
3 15
4 15
5 30
6 30
7 30
8 30
minutes
minutes
minutes
mi nutes
minutes
minutes
minutes
B.O.D.
mq/1
183
144
126
130
99
115
85
70
Settleable
Solids
ml/1
7
3
2
1
1
1
1
0
.0
.5
.5
.2
.0
.5
.5
.75
Suspended
Solids
mq/1
165
130
30
20
22
25
17
10
Volatile
Solids
mq/1
280
200
185
165
75
115
100
90
Total
Solids
mq/1
530
400
320
300
250
270
265
275
Coli form
MPN/
100 ml
11,000
2,400
2,400
2,400
240
930
11,000
240
,000
,000
,000
,000
,000
,000
,000
,000
d.O.
mq/1
6.7
7
7
6
7
7
7
7
.2
.4
.9
.2
.6
.7
.6
Note:
(1) Total Rainfall from 7:30 P.M. to 9:10 P.M. = 0.30 inches
(2) Flow commenced at 7:50 P.M. — First Sample Taken 7:55 P.M.
Flow Rate at 8:00 P..M.=1900 gal/min.
8:45 P.M.= 900 gal/min.
9:30 P.M.=1440 gal/min.
10:15 P.M.=less than 50 gal/min.
-------
From April 24, 1968, to November 12, 1968, there were no overflows
observed at the test site. On a number of occasions during this time
period there was rainfall equal to or greater than that which had pre-
viously caused overflows. This situation probably resulted from the
flooding that occurred on March 17, 1968, and the construction work
carried out in the vicinity of the overflow.
Visual observation of several overflows conclusively showed the
presence of fresh human feces (larger than one-half inch) and whole
pieces of toilet paper. Samples were collected using a wire-mesh screen
with one quarter inch openings. Comparison of the suspended solids in
the usual pail samples with those collected with the wire mesh, con-
sistently showed a variation in particle size. Only when a sample was
taken at the surface of the flowing stream did the maximum particle size
obtained with the pail equal that found with the wire mesh strainer.
On April 1, 1968, a very brief overflow occurred at 8:15 A.M. Only
one set of samples was taken; one with a one pint scoop, the second with
a one gallon pail. The samples were simultaneously taken by two people
at the same surface depth. The pail sample was found to have higher values
for eacn variable tested.
Suspended Total Volatile Settleable
B.O.D. C.O.D. Solids Solids Solids Solids
Scoop 190 444 315 580 350 3.25
Pail 210 495 825 1140 784 4.0
The eight samples shown in Tables IV and V above actually represent
16 samples. At each time period both a scoop and pail sample was taken
for comparative analysis. The C.O.D. values are listed on page 19 in
Table VI.
Although whole sections of toilet paper were noted in the overflow,
the sampling technique used did not produce or yield any paper in the
samples. A double sheet of toilet tissue weighs approximately 0.37 grams
and would yield a C.O.D. value of approximately 19,400 mg/1.
18
-------
TABLE VI - EFFECT OF SAMPLE SIZE ON OVERFLOW ANALYSIS
C.O.D.
Sample No.
1
1
2
2
3
3
4
4
5
5
6
6
7
7
8
8
Sampler
scoop
pail
scoop
pall
scoop
pall
scoop
pail
scoop
pail
scoop
pail
scoop
pail
scoop
pail
April 15, 1968
244
264
160
180
192
212
184
284
120
100
248
212
136
152
36
45
April 24, 1968
372
452
248
300
256
248
224
270
212
188
204
232
172
208
172
180
19
-------
C. Strainer Experiments
1. Flat Sheet Tests
The following wire screens were initially evaluated using the
synthetic substrate at a concentration of 0.4 grams per liter and
a flow rate of one gallon per minute: 50 x 250 mesh plain dutch
weave; 25 x 25, 36 x 36, 60 x 60, 80 x 80, and 100 x 100 mesh square
weaves.
The 50 x 250 mesh screen blinded off too rapidly to permit ac-
curate visual measurement of the time-pressure relationship. Also,
the solids became very tightly bound in the interstices of the screen
and were not readily removed by backwashing. On the other hand, the
25 x 25 and 36 x 36 mesh square weave screens did not retain suffi-
cient solids to reach the end point pressure of 20 psig even after
20 minutes.
A series of runs were performed with 60 x 60, 80 x 80, and 100 x
100 mesh square weave screens. Table VII contains data showing the
relative time it took for each of the screens to reach a pressure of
19 psig at different blending times. The longer the blending time,
the smaller the average particle size. These results showed that
reproducible results could be obtained with the synthetic contaminant
in terms of particle size and particle size distribution. It was
quite obvious that the physical characteristics of the suspended solids
could be controlled and varied by manipulation of mixing time and
temperature.
TABLE VII - FLAT SHEET TEST RESULTS
Run No.
Hire Screen
Screen
Openings
Microns
1263-19
1263-16
1263-17
1263-20
1263-24
1263-25
1263-36
1263-42
100
100
100
100
60
60
80
80
x 100
x 100
x 100
x 100
x 60
x 60
x 80
x 80
150
150
150
150
230
230
190
190
5 min.
5 min.
5 min.
10 min.
5 min.
5 min.
10 min.
10 min.
Blend Time Average Run Time
62 sec.
83 sec.
72 sec.
47 sec.
147 sec.
129 sec.
94 sec.
85 sec.
Each value is the average of five runs.
20
-------
2. Model Self-Cleaning Strainer
a. Synthetic Strainer
The results obtained with the flat sheet samples were fairly
reproducible and significantly different under varying conditions,
but the analysis time was deemed too short for a suitable expanded
program. The remainder of the project was carried out using the
model strainer as described previously.
The initial short duration run used a 60 x 60 mesh square weave
wire screen at a synthetic feed rate of 5.9 gallons per minute (0.4
grams per liter) and a clean water backwash of 2.2 gpm, with no sump
discharge. The initial results are listed below in Table VIII. The
influent values shown have been corrected for dilution caused by the
clean water backwash.
TABLE VIII - INITIAL MODEL STRAINER RESULTS -
60 x 60 MESH, 230 MICRON OPENING
Sample B.Q.D.
Influent 125
Effluent
1 min.
Effluent
3 min.
Effluent
5 min.
Eff1uent
Compos i te
80
115
60
95
C.O.D.
324
259
290
243
251
Suspended
Solids
161
105
110
80
100
Total
Solids
598
460
515
445
465
Volatile Solids
226
95
45
75
85
An attempt was made to establish a statistical program to
evaluate the optimum combination of parameters (feed rate, backwash
rate, drum speed, mesh size, etc.) on B.O.D. and suspended solids
removal. The initial experimental results are shown in Table IX
obtained with a 60 x 60 mesh screen with the synthetic substrate
concentration at 0.4 grams per liter. Each value listed per run
number represents a separate 50-gallon batch of synthetic sub-
strate that was prepared.
21
-------
TABLE IX - MODEL STRAINER RESULTS - 230 Micron
60 x 60 MESH - SYNTHETIC CONTAMINANT
Run No. Percent Removal of Suspended Solids Average
1 63 19 12 30 43 30 33
2 58 57 63 54 69 72 62
3 23 31 72 40 19 13 33
4 14 18 19 17 10 24 17
5 37 51 51 57 57 44 50
6 20 14 47 45 34 10 28
7 52 13 53 36 32 33 37
8 29 13 29 55 23 60 37
The raw data for these calculations are listed in Table (b) in the
Appendix. The eight runs listed in Table IX were made at four different
ratios of solution feed rate to backwash rate: drum speed constant at
8 rpm.
Run No. Feed Rate Backwash Rate
1,2 6 gpm 2 gpm
3, 4 6 gpm 1 gpm
5, 6 5 gpm 2 gpm
7, 8 5 gpm 1 gpm
The primary objective of this experiment was to evaluate the effect
of velocity across the screen on solids separation efficiency. The wide
variation between runs made at identical conditions caused a re-examina-
tion of the experimental conditions. It was determined that improper
metering of the synthetic suspension into the strainer resulted in widely
fluctuating solids concentration in the influent material. The numbers
calculated as shown in Table IX were based on an average inlet concen-
tration and not, therefore, on the true value at any time. The effluent
samples values as shown in the Appendix represent true values at that
particular sampling time. Under more carefully controlled conditions a
series of runs were performed using the synthetic substrate at various
flow rates and backwash rates with three different size screens. The raw
data is listed in Table (c) in the Appendix. The influent to the strainer
was sampled at the same time as the effluent. During a 30-minute test
cycle the influent varied 10 to 30 percent from the average values found
with this particular synthetic substrate. Since actual field trials would
entail at least this amount of variation no further changes were made in
the test procedure. Table X on page 23 shows the calculated results for
percent removal of suspended solids, B.O.D. and C.O.D. These values were
calculated using the specific values obtained for each influent effluent
pair of samples. The numbers shown in brackets were not used in calculating
the averages. When the 100 x 100 mesh screen was used, the basket speed
of rotation was doubled to prevent screen plugging.
22
-------
TABLE X - MODEL STRAINER RESULTS - EFFECT OF MESH SIZE ON SOLIDS REMOVAL
Run No. Mesh Size
B.O.D. C.O.D.
1263-48
1263-49
1263-46
1263-47
1275-1
1275-2
1275-3
1275-4
1275-16
1275-15
1275-17
1275-22
1275-18
1275-24
1275-23
1263-44
1263-45
1275-6
1275-7
1275-8
1275-9
1275-10
60 x 60
60 x 60
60 x 60
x
x
X
60
60
60
60
60
60
20
10
13
14
11
60
60
80
80
80
80
80
80
80
x 60
x 60
80
80
80
80
80
80
80
100 x 100
100 x 100
100 x 100
100 x 100
100 x 100
100 x 100
100 x 100
25
12
30
17
25
19
22
19
9
(60, 80, 100 mesh = 230, 190, 150 microns respectively)
Notes:
12.2
7.1
15.6
12.8
15.4
17,
15,
Average Percent Removal Data
Inlet Backwash Suspended
Flow Rate Rate Solids
6 2 27
6 2 28
6 1 43
6 1 33
5 2 35
5 2 32
5 1 32
5 1 31
6 2 54
6 2 61
6 1 31
5 2 57
5 2 44
5 1 37
5 1 56
6 2 43
6 2 51
6 2 70
5 2 55
5 2 58
5 1- 65
5 1 52
18.9
16.5
21.5
20.6
17.8
17.2
16.0
13.9
16.9
18.6
5)
.3)
19.
21.8
17.0
1) Synthetic contaminant concentration - 0.4 gm/1
2) Normal drum speed - 8 rpm
Additionally, it should be noted that the 100 x 100 mesh screen could
not be satisfactorily operated at 6/1 feed to backwash ratio. Excessive
plugging of the screen occurred, causing a rapid increase in system pressure,
which required frequent blowdowns. At the same flow ratio of 6/1, the 80 x
80 mesh screen also did not perform very well, only one value is listed in
Table (X).
b. Sheridan Street Samples
Table XI lists the results obtained using four fifty-gallon com-
posite samples taken at Sheridan Street. The samples were taken from
the sanitary sewer line during a rain storm when overflow did not occur.
23
-------
This sampling was performed in June, 1968, two months after the dis-
astrous spring flood. Prior to this date, the rainfall which occurred
during this sampling has previously caused overflows. The samples
were taken with a gasoline engine powered centrifugal pump rated at
80 gallons per minute.
These results were obtained at a flow rate of 5 gallons per
minute, backwash rate of 1 gallon per minute, screen - 80 x 80 mesh(190M- )
and the basket revolving at 8 revolutions per minute.
TABLE XI - MOOEL STRAINER RESULTS - SHERIDAN STREET OVERFLOW SAMPLES
FLUX RATE - 18 GPM/FT.2
Suspended Total Volatile
B.O.J. C.O.D. Solids Solids Solids
Sample
1 Influent 65
1 Effluent 50
2
2
3
3
Influent
Effluent
Influent
Effluent
4 Influent
4 Effluent
5 Influent
5 Effluent
6 Influent
6 Effluent
67.5
52.5
62.5
60
70
62.5
65
60
80
60
188
168
196
172
216
180
208
184
192
180
200
168
1040
620
920
600
1000
640
1000
500
700
480
680
420
310
305
250
225
290
250
230
200
285
260
115
120
The coliform count on all samples was greater than 11,000,000. The
settleable solids test was not performed since the sample had been passed
through a pump three times prior to testing.
c. Mechanical Ability
One of the major concerns regarding the proposed system was the
mechanical reliability of the strainer when operated semi-continuously
on a stream containing a large amount of sewage solids. For this pur-
pose the strainer was moved to the Bucklin Point Sewage Treatment Plant,
East Providence, Rhode Island. This particular plant, operated by the
State of Rhode Island, has only primary treatment and the influent con-
tains a very high proportion of industrial waste, resulting in high
dissolved organic concentrations. The treatment plant chemist estimated
that at times seventy percent of the influent is industrial waste.
24
-------
The influent for the strainer was taken at a point between grit
removal and the sedimentation tanks. During the first six hours of
operation, the pump suction line was not protected to exclude large
objects which might slip by the bar screens. As a result, a number
of times large pieces of paper and rags were pulled into the strainer
housing and plugged the discharge dump valve. This resulted in a
rapid increase in system pressure requiring a shut down to free the
discharge part. This was not considered to be a real problem, since
a full-scale unit would have at least two inch discharge line versus
the three quarter of an inch dump valve on the model. This problem
was eliminated by the installation of a perforated basket with one
half inch holes around the intake.
The unit was operated for 83 hours on 15 separate days. Opera-
ting data was taken at 30 and/or 60 minute intervale during 77.5 of
the 83 hours on 13 days. The specific nature of the results ob-
tained in terms of percent removal of suspended solids, B.O.D., and
C.O.U. are partially listed in Tables XII on page 26. The system was
operated at an inlet flow rate of 7 gallons per minute with continuous
solids discharge at a rate of two gallons per minute. The backwash
nozzle was operated at two gallons per minute with strainer effluent
as the backwashing fluid. There was absolutely no difficulty in
operating the unit on the sewage solids. Plugging did not occur on
either the 60 x 60 or 100 x 100 mesh screens used in the experiments.
TABLE XII - MOJEL STRAINER RESULTS - BUCKLIN POINT
SEWAGE TREATMENT PLANT INFLUENT 60 x 60 MESH SCREEN
230 MICRON OPENING
Running Time
Total Hours
0.5
1.5
3.0
4.5
6.5
18.0
19.0
20.0
21.0
22.0
23.0
28.0
31.0
38.0
41.0
Suspended Solids
Influent % Removal
B. 0. D.
315
140
90
60
110
130
180
185
305
150
150
100
55
465
310
25
43
45
83
50
23
19
48
65
33
40
70
18
69
72
Inf.
245
205
200
183
160
210
165
213
243
210
260
160
168
310
330
% Removal
8
24
18
4
63
14
7
25
14
4
16
16
11
22
27
C. 0. D.
Inf. % Removal
812
444
456
467
689
468
652
576
644
580
508
657
664
796
816
19
21
10
14
7
16
3
13
6
15
15
1
13
37
40
FLUX RATE = 25 GPM/FT.'
25
-------
TABLE XIII - MODEL STRAINER RESULTS - BUCKLIN POINT
SEWAGE TREATMENT PLANT INFLUENT 100 x 100 MESH SCREEN,
150 MICRON OPENING
Running Time
Total Hours
1.5
2.5
3.5
5.5
9.0
12.0
15.0
16.5
22.0
26.0
30.5
Suspended Solids B.O.D. C.O.U.
Influent % Removal Inf. % Removal Inf. % Removal
175
175
205
160
255
175
225
45
90
40
80
60
49
34
40
80
31
93
66
61
50
31
130
150
175
380
160
290
208
270
185
210
11
13
14
5
12
24
7
4
16
14
465
515
595
764
524
628
568
452
412
452
648
25
16
16
17
12
8
12
7
20
12
31
FLUX RATE = 25 GPM/FT.2
Since it was known that industrial wastes are a major part of the
influent received at the Bucklin Point Treatment Plant, analysis was
carried out to determine the extent of dissolved organics present in
the waste treated. A number of samples were analyzed as received, and
also after filtration through a 0.45 micron membrane filter. Some
typical results are shown in Table XIV below.
TABLE XIV
Sample No.
86100
86103
86110
86113
87110
87103
87110
87113
DISSOLVED ORGANIC CONCENTRATION IN BUCKLIN POINT INFLUENT
C.O.D.
Suspended Solids
140
170
110
490
460
210
155
250
As Received
708
844
648
836
604
560
568
728
Filtered
188
196
260
304
232
196
240
200
The unit was moved to another municipal treatment plant in East
Providence, R.I., which receives less industrial waste. Typical re-
sults obtained with a 60 x 60 mesh screen are listed in Tables (d) and
(e) in the Appendix. The data in Table (d) were obtained when the in-
fluent to the strainer was taken upstream of the bar screens. Inlet
flow was seven gallons per minute, backwash four gallons per minute and
sump discharge at three gallons per minute. Again, it was not possible
to operate the sump discharge ata lower flow rate without plugging the
three quarter inch discharge line.
26
-------
Table (e) data were obtained when the strainer influent was taken
down stream of toe bar screens. The operating conditions were maintained
tne same as listed above.
No data were taken witn either an 80 x 80 or 100 x 100 mesh screen
at this location (190, 150 Microns respectively).
d. Results with Fresh Solids
For the purpose of conparing the results obtained at the munici-
pal treatment plants with those to be expected at an overflow site, raw
sewage was collected from a sewer line containing only sanitary wastes.
The sewage was collected in 55 gallon drums, using an eductor. Use of
the vacuum system permitted collection without mechanical action or ma-
ceration of the solids. The solids were, therefore, presented to the
system in the same physical estate found at the overflow site analyzed.
The data obtained are listed in Table XV below. With this
type of feed it was possible to operate the sump discharge rate at a
much lower value than previously possible at the treatment plants. The
backwash rate had to be maintained at a high value similar to the runs
made at Bullock's Point noted above.
TABLE XV - MODEL STRAINER RESULTS - FRESH SEWAGE SOLIDS, 60 x 60 MESH SCREEN
230 Microns
Inlet Flow -
Running Time
Total Hours
1
2
3
4
Inlet Flow - 7 gpm
7gpm Effluent Flow - 5
Suspended Solids B.O.D.
Influent
150
180
50
35
Ef f 1 uent
55
15
5
8
Influent
170
170
200
230
gpm Backwash - 4 gpm
C.O.D.
Effluent Influent Effluent
110
90
120
140
604
584
261
288
449
451
188
235
Effluent Flow - 6.5 gpm Backwash - 4 gpm
5
6
7
8
20
68
28
50
5
28
12
20
90
115
95
145
60
65
55
80
174
334
337
358
140
240
321
305
0. Self-Cleaning Filter
1. Johns-Manvilie Test System
a. Screening Tests
The initial program was conducted with the 0.1 square foot
filter area test system described above, that was designed and built
by the Johns-Manville Company. Studies were carried out to determine
which variables were important to the development of a standard test
procedure.
27
-------
Using the strainer effluent from the initial laboratory screening
experiments, where the influent was 0.4 grams per liter of synthetic
suostrate, studies were made witn the filter test leaf, Hyflo-Super
Cel grade diatomaceous earth and Grade 2006 Polypropylene monofilament
septa. These preliminary experiments were made to determine optimum
submergence time (simulation of drum rotation speed) and optimum knife
advance (simulation of residue removal phase). The analytical results
are listed in Tables XVII and XVIII.
TABLE XVII - EFFECT OF SUBMERGENCE TIME ON FILTER PERFORMANCE
Suomergence Time
15 seconds
30 seconds
45 seconds
60 seconds
C.O.D.
Influent
302
302
342
342
(mg/1)
Effluent
48
48
40
36
Suspended Solids (mg/1)
Influent Effluent
30
30
227
227
negligible
negligible
negligible
negligible
TABLE XVIII - EFFECT OF KNIFE ADVANCE ON FILTER PERFORMANCE
Knife Advance
10 mil
20 mil
30 mil
302
242
242
48
36
32
30
175
175
negligible
negligible
negligible
It can be observed that the submergence time and knife advance
tnickness do not nave to be critically controlled in order to obtain
comparable effluent C.O.D. and effluent suspended solids levels on this
synthetic suostrate.
Based on the results shown in Tables XVII
ing standard test procedure was adopted for the
tomaceous earth grades:
and XVIII, the follow-
screening of six dia-
2006 Polypropylene
Thickness - 1.50 in.
Precoat Slurry concentration - 6% Septum - Type
Volume of Slurry/addition - 300 mis, Final Cake
Vacuum Range - begin at 5, end at 20 in. Hg.
Operating Temp - 25 - 300C, Filtering Vacuum - 20 in. Hg.
Submergence Time - 22 sec., Advance and Cake Removal - 8 sec.
Knife Advance - 0.020 inches
28
-------
The six diatomaceous earth's are graded on a porosity scale of one
to ten, where ten is the most porous. The results are tabulated in Table XIX
below:
TABLE XIX - DIATOMITE EVALUATION WITH SYNTHETIC SUBSTRATE
J-M Diatomite
Grade
Porosity
560 10
545 9
Hyflo Super Cel 5
512 4
Standard Super Cel 3
Filter Cel 1
C.O.D.
Influent
326
326
326
216
216
216
(mg/1)
Effluent
63
63
40
44
28
36
Suspended Solids (mg/1)
Influent Effluent
195
195
195
175
175
175
negligible
negligible
negligible
negligible
negligible
negligible
It can be noted that the variations in porosity did not drastically
change the effluent C.O.D. and suspended solids levels.
On the basis of the results from the initial screening, three diatomite
grades were selected for further testing. Because of the small differences
in removal levels, one grade was selected to represent each porosity range.
They were Johns-Manville #545 (high porosity), Hyflo Super Cel (medium) and
Standard Super Cel (low).
Continuing the studies further, another evaluation was carried out using
the above three diatomite materials, the standard test procedure and two dif-
ferent influent substrates. Primary effluent was obtained from the Bucklin
Point, East Providence, Rhode Island, municipal treatment facility. This
effluent was used as is and also mixed 1:1 with 0.4 g/1 dog food strainer ef-
fluent. Further, both these substrates were/used in their unadulterated form
and also with a 0.5 g/1 Darco G-60 powered activated carbon treatment. The
results of the tests employing these substrates and treatments are tabulated
in Table XX.
29
-------
TABLE XX - EFFECT OF DIATOMITE TYPE ON FILTER EFFICIENCY
co
o
Volatile
Sample
description
Bucklin Pt. - Inf.
B.P. - Eff. 545
B.P. - Eff. HSC
B.P. -Eff. SSC
B.P. (AC) Eff. 545
B.P. (AC) Eff. HSC
B.P. (AC) Eff. SSC
B.P.:S.E. - Inf.
B.P.rS.E. - Eff. 545
B.P.:S.E.-Eff.HSC
B.P.:S.E.-Eff.SSC
B.P.rS.E. (AC) -
Eff. 545
B.P.:S.E. (AC) -
Eff. HSC
B.P.:S.E. (AC) -
Eff. SSC
C.O
mg/1
501
432
332
228
380
336
352
383
244
240
228
196
176
180
.L).
% Red
13.8
33.7
54.5
24.2
32.9
29.7
36.3
37.3
40.5
48.8
54.1
53.0
B.O.U.
mg/1
260
230
140
175
180
125
100
210
155
90
90
115
45
% Red
11.5
46.2
32.7
30.8
51.9
62.5
26.2
57.1
57.1
45.2
78.6
Coli form
MPNxlO*
1100
43
2.4
0.04
15
2.4
0.15
460
93
4.6
0.23
43
0.75
0.04
% Red
96.1
99.8
99.9
98.6
99.8
99.9
79.8
99.0
99.9
90.7
99.8
99.9
Total
mg/1
510
385
455
425
395
470
425
485
270
410
445
225
370
250
Solids
% Red
24.5
10.8
16.7
22.6
7.8
16.7
44.3
15.5
8.3
53.6
23.7
48.5
Solids
mg/1
260
190
260
275
235
270
170
260
130
235
280
165
200
100
% Red
26.9
0
0
9.6
0
34.6
50.0
9.6
0
36.5
23.1
61.5
CODE:
B.P. - Bucklin Point Primary Effluent Sample.
B.P. - Eff. 545 - Sample after filtration through grade 545.
B.P. (AC) Eff. 545 - Sample after treatment with activated carbon and filtration through grade 545.
B.P.:S.E. - Inf. - A fifty-fifty mixture of Bucklin Point primary effluent and strainer effluent
from synthetic substrate feed.
% Red - Percent Reduction
-------
b. Filter Aid and Chemical Treatment Evaluation
(1) Activated Carbon Treatment
Continuing this study, the effect of diatomaceous earth fil-
tration alone and aided by activated carbon was evaluated using three
different composite samples of Sheridan Street overflow. This evalua-
tion was performed according to the standard test procedure outlined
in this report. Also, the activated carbon treatment was the same as
that used previously, namely, 0.5 g/1 Darco G-60 powered activated
carbon.
The three influent samples were composites of (1) April 15,
1968, overflow Samples #1 through 8 as reported on PagelSof this re-
port, (2) April 24, 1968, overflow Samples #5 through 8 as reported
on Page 17of this report. The results of this evaluation are tabulated
in Table XXI.
(2) Polyelectrolyte Treatment
An initial screening of various polyelectrolyte flocculants
was carried out employing nine different coagulants (3 each of anionic,
cationic, and non ionic types) and 0.4 g/1 dog food as substrate. Three
dosage levels between 1.0 and 10.0 mg/1 were tested with no visible coagu-
lation noted. The nine possible cationic-anionic combinations were also
evaluated at various dosage levels and visible coagulation was noted only
with the following systems:
System Cationic Polyelectrolyte Anionic Polyelectrolyte
1 20 mg/1 Calgon Cat-Floe 10 mg/1 Dow Purifloc A-23
2 20 mg/1 Dow Purifloc C-31 10 mg/1 Dow Purifloc A-23
3 20 mg/1 Alum 10 mg/1 Dow Purifloc A-23
Primary effluent from the Bucklin Point, East Providence, Rbode
Island municipal treatment facility was treated with the above three poly-
electrolyte systems and then filtered through the three diatomaceous earth
candidates previously selected. The filtration was carried out according
to the standard test outlined above. The results of this "Uiatomaceous
Earth - Polyelectrolyte Study" are tabulated in Table XXII.
31
-------
TABLE XXI - SHERIDAN STREET OVERFLOW - DIATOMACEOUS EARTH FILTRATION
Sample
Description
4/15/68
4/15/68
4/15/68
4/15/68
4/15/68
4/15/68
4/15/68-Eff.(AC)SSC
Inf.
Eff.545
Eff.HSC
Eff.SSC
Eff.(AC)545
3-Eff.(AC)HSC
4/24/68(l-4)Inf.
4/24/68-Eff.545
4/24/68-Eff.HSC
4/24/68-Eff.SSC
4/24/68-Eff.(AC)545
4/24/68-Eff.(AC)HSC
4/24/68-Eff.(AC)SSC
4/24/68(5-8)-Inf.
4/24/68 Eff.545
4/24/68 Eff.HSC
4/24/68 Eff.SSC
4/24/68-Eff.(MC)545
4/24/68-Eff.(AC)HSC
4/24/68-Eff.(AC)SSC
C.O.D.
mg/1 % Red
127
53
36
16
52
40
51
158
92
64
40
68
48
60
109
64
44
48
56
60
56
58.3
71.6
87.5
59.0
68.5
59.0
41.8
59.5
74.8
57.0
69.6
62.0
41.3
59.6
56.0
48.6
45.0
48.6
B.O.D.
mg/1 % Red
62
28
12
20
16
75
50
30
16
26
18
55
20
18
20
25
22
55
80
68
74
33
60
79
65
76
45
67
64
55
60
Co li form
MPNxlO4 % Red
460
24
0.15
0.09
11
0.43
0.23
1100
24
0.43
0.04
15
2.4
0.09
43
15
2.4
0.23
11
4.6
0.11
94.8
99.9
99.9
97.6
99.9
99.9
97.8
99.9
99.8
98.6
99.7
99.9
65.1
94.4
99.5
74.4
89.3
99.7
Total
mg/1
250
150
135
145
155
110
no
280
165
100
60
215
155
120
210
170
80
60
80
90
120
Solids
% Red
40.0
50.0
42.0
38.0
56.0
56.0
41.0
64.4
78.5
23.2
44.7
57.2
19.0
62.0
71.5
62.0
57.2
42.9
Volatile
Solids
mg/1
145
70
80
90
75
50
75
175
85
50
35
135
80
80
125
105
30
40
30
70
70
% Red
51.7
44.9
37.9
48.3
65.5
48.3
51.5
71.5
80.0
22.8
54.3
54.3
16.0
76.0
68.0
76.0
44.0
44.0
ro
-------
TABLE XXII - DIATOMACEOUS EARTH FILTRATION - BUCKLIN POINT PRIMARY EFFLUENT
SamPle C.O.D. B.O.D. Coliform Total Solids Volatile Solids
Description mg/1 % Red mg/1 % Red mg/1 % Red mg/1 % Red mg/1 % Red
Inf. A 748 335 460 980 400
Eff. A 545 428 42.8 305 8.9 93 79.8 495 49.5 265 33.8
Eff. A HSC 328 56.2 4.6 99.0 415 57.7 210 47.5
Eff. A SSC 124 83.4 0.23 99.9 265 73.0 185 53.8
Eff. A 545 560 25.1 280 16.4 43 90.7 565 42.3 335 16.3
Eff. A"HSC 568 24.1 2.4 99.5 160 83.7 85 78.8
Eff. A!SSC 376 49.7 0.04 99.9 300 69.4 175 56.3
Eff. A2545 500 33.2 190 43.2 15 96.7 520 46.9 310 22.5
Eff. A2HSL 204 72.7 2.4 99.5 235 76.0 160 40.0
Eff. A2SSC 240 67.9 0.15 99.9 300 69.4 175 56.3
Eff. A3545 484 35.3 270 19.4 43 90.7 400 59.2 300 25.0
Eff. A3HSC 448 40.0 0.75 99.8 440 55.V 275 31.3
Eff. A3SSC 116 84.5 0.04 99.9 180 81.6 120 70.0
Code:
A], A2, A3 — signify polyelectrolyte systems 1, 2, or 3 were used as described on
Page 31
-------
2. Self-Cleaning Strainer - Vacuum Modified
Although the proposal and contract did not specify working model
evaluation of either a strainer or filter, changes were made in the
strainer model to permit additional evaluation of filter aid filtra-
tion under vacuum filtration.
Initially, three runs were made under continuous flow conditions
to evaluate three diatomite samples at the same load conditions. These
initial results are presented in Table XXIII below at the following flow
rates: (1) HSC at one liter per hour,(2) 545 at one gallon per hour, and
(3) 560 at one liter per hour. The values listed as filtered effluent
were obtained on the effluent samples after filtration through a 0.45
micron membrane filter.
TABLE XXIII - VACUUM MODIFIED FILTER RUNS WITH MODEL STRAINER
C. 0. D.
Type Running Time Influent Effluent Filtered Effluent
HSC 8 hrs. 396 170 125
16 hrs. 396 143 113
HSC 32 hrs. 554 131 97
545 4 hrs. 305 131 103
8 hrs. 580 165 111
16 hrs. 626 145 133
20 hrs. 626 143 117
545 24 hrs. 288 80 66
28 hrs. 288 92 74
560 8 hrs. 336 66 58
16 hrs, 304 65 47
24 hrs. 304 59 47
The results shown below were obtained using a constant body feed
of five percent with the cake thickness gradually increasing from 1/64"
to 3/32" during each run.
Flow Rate Run Length B.O.D. C.O.D. Coliform
gpm/ftz Filter Aid (Hrs) In Out Jn. Out In Out
3.70 Hyflo Super Cel 0.5 85 35 240 74 4,600,000 43,000
4.10 Cel He 545 0.5 68 32 133 70 2,400,000 430,000
1.85 Filter Cel 1.0 87 36 182 77 4,600,000 43,000
34
-------
A second series of runs were made at a constant-filter aid drum
thickness of 3/32 of an inch. In both instances the waste source was
obtained from the influent at the Bullocks Point Treatment Plant in
East Providence, Rhode Island, and was diluted with three parts of tap
water.
Flow Rate
gpm/ftz
3.70
3.70
0.92
Filter Aid
545
HSC
Filter Cel
Run Length
(Mrs
B.O.D.
In Gut
C.O.D.
In Out
Coli form
In
Out
110 42 282 98 4,600,000 240,000
90 38 253 86 11,000,000 750,000
124 49341 10211,000,000 93,000
As a measure of potential efficiency, the samples taken during the
runs shown above were filtered through 0.45 micron membrane filters. These
results are shown in Table XXIV.
TABLE XXIV - DISSOLVED ORGANIC CONCENTRATION IN FILTERED EFFLUENT
C.O.D.
Filter Aid
HSC-1
545-1
Filter Cel-1
HSU - 2
545 - 2
Filter Cel-2
Effluent
74
70
77
86
98
102
Filtered Effluent
70
61
63
78
75
98
Following the apparent success achieved in obtaining reasonable
flow rates under adverse conditions, a number of extended runs were
attempted at a fixed filter aid thickness of 3/32 of an inch using
Hyflow Super Uel.
TABLE XXV - MODIFIED FILTER RUNS - HYFLOW SUPER CEL
Flow Rate
gpm/ft*"
3.
3.
1.
1.
70
70
54
23
0.92
0.92
0.92
Run Length
(Hrs)
3
4
4
4
4
4
4
B.O.D.
In Out
In
51
51
51
51
93
93
93
27.0
18.0
17.5
16.0
20
28
14
144
144
144
144
380
380
380
c.
O.D.
Out
47
55
75
63
50
61
85
2
2
4
4
In
,400
,400
,600
,600
Coli
,000
,000
,000
,000
form
Out
240,
93,
430,
430,
000
000
000
000
35
-------
Additional runs were performed out a very rapid fall off in flow
rate was observed. In the space of 6 hours, in one instance, the flow
dropped from 3.70 gpm/ft2 to 0.123 gpm/ft2 when operating at a fixed
aid tnickness.
IV. Discussion of Results
A. Site analysis
Two significant factors were isolated concerning the characteristics
of tne overflows occurring at the particular site used for observation
and analysis. The first, that during the periods of significant organic
loadings that seventy to eighty percent of this load was represented by
suspended solids larger than one sixteenth of an inch. This was caused
by the presence of human feces (which had not been mechanically disin-
tegrated) individual pieces of toilet and facial tissue (not individual
fibers as found in the influent to most treatment plants), and kitchen
wastes.
This result snould not be surprising in view of what has been docu-
mented previously. As stated earlier, it had been found that the first
flush in a combined sewer overflow system could contain a very high
solids content as a result of settling during dry weather flow. This
type of result was indeed verified by the January 30, 1968 samples. In
contrast, the samples obtained on February 2, 1968, did not contain an
appreciably higher solids content in the first flush as was found with
the samples a few days earlier. The load contributed by material that
had oeen settled in the lines naturally will be almost entirely suspended
solids.
The physical characteristics of the solids obtained during tne first
flush should oe and were found to be quite different from those samples
later during an overflow. Bacterial action and particularly hydrolysis
reactions create solids which are readily disintegrated during and by the
turbulence created by the flow of water which sweeps them out of the sewer
system. Soluble organic compounds which are produced by bacterial action
in the settled sewage are continually removed by the water during dry-
weather flow conditions, therefore, only insoluble or suspended organic
solids are left behind waiting for the first rapid change in flow conditions.
After the first flush, the solids which reach the overflow are fresh
solids, such that little or no time has elapsed for hydrolysis reactions
to occur to any appreciable extent. This was verified by comparing the
physical state of toilet paper at the overflow site with that found at
the Bullock's Point Treatment Plant 1n East Providence, Rhode Island. One
of tne main influent sewer lines reaching this plant does not contain any
pumping installations, so that any mechanical action on solids is entirely
due to tne hydraulic situation. Very careful examination of this particular
stream showed practically no toilet tissue in discernable form. On the
other hand, most of the overflows contained a great deal of whole pieces
of toilet tissue.
36
-------
The second most important factor determined relates to sampling
methods for the collection of data on combined sewer overflows. The pre-
vious discussion points out that the characteristics of the suspended
solids present in an overflow can change markedly with time. The vertical
distribution of solids in the flowing stream changes with time for a
particular flow rate. During the first flush most of the solids are below
the surface, whereas most of the fresh solids are near the surface. The
nature of the solids and their distribution across a cross section of flow
would appear to preclude the usual type of automatic sampling device. Any
system which uses a sampling tube approximately one-half inch in diameter
cannot be expected to provide a suitable representative sample for analysis.
Fresh solids and toilet paper which represent a very high load per unit
volume are most certainly missed by most automatic sampling methods used
to date.
Additionally, there are two conflicting factors to consider when evalua-
ting the merits of a sampling system. First it is important to obtain the
sample without mechanical action. Second, because of high flow rates a
large sample should be taken in order to have any hope for a "representative"
sample - which almost certainly implies the use of a pump. This project has
only raised these two points - it has not solved them.
With regard to the exact load contributed to a receiving stream by an
overflow, this paper can only provide a guideline. For an overflow in an
area which is 80 percent (or greater) residential, the total load can be
approximated by multiplying the total overflow volume by an average B.O.D.
value of 120 mg/1.
B. Self-Cleaning Strainer Effectiveness
The flat sheet testing and analysis as described on Page 20,
Table VII, statistically showed that: (1) the synthetic substrate could be
reproducibly prepared; (2) the 80 x 80 and 100 x 100 mesh screens gave essen-
tially the same result; (3) the 60 x 60 mesh screen would be significantly
different at the 1 percent level from the 80 x 80 mesh screen in suspended
solids removal, (60, 80, 100 rnesh=230, 190, 150 microns, respectively).
The data in Table X, page 23, was statistically analyzed and it was
shown that the influent flow rate to backwash flow rate ratio was not signifi-
cant for those tested, when using the synthetic substrate. The result found
with the flat sheet tester was also true with model strainer. No significant
difference between the 80 x 80 and 100 x 100 mesh screens, but a definite
statistical difference between the 60 x 60 and 80 x 80 mesh screens. While
these particular results are specific for the synthetic substrate, they do
relate to the results found with sewage plant influent, fresh sewage and
actual stormwater overflow.
Of the four sources of sewage tested, the influent to the Bullock's
Point Treatment Plant was the most difficult to treat. The data in Table e
in the Appendix was calculated to show the percent suspended solids and C.O.D.
removed. Additionally, the suspended solids C.O.D. was determined in the in-
fluent and effluent samples. Since the strainer is designed to remove only
suspended matter, its efficiency was calculated on this basis in Table XXVI.
37
-------
Data obtained in this project indicated that 90 percent of the B.O.D.
found in the overflow discharges was exerted by suspended matter. On the
other hand, the primary influent to the two nearby treatment plants have
only 50-70 percent of the total B.O.D. present in suspended form. The last
column in Table XXVI was calculated, therefore, on the basis of the C.O.D.
exerted oy the suspended solids retained by 0.45 micron membrane filter as
follows for line 1 in the Table.
Removed = (437-196) - (390-231)
Kemovea (437-196)
x 100 = 33
These results are generally more in line
moval efficiency than the raw data indicated.
with the suspended solids re-
The most significant difference found between these results and those
obtained with fresh solids was the ratio of effluent flow to sump discharge
flow that was permissable. At the same inlet to backwash flow ratio, the
inlet to effluent flow was 7/4 at Bullock's Point versus 7/6.5 with fresh
solids.
Overall, the model strainer showed very consistent results with each
type of waste under widely fluctuating conditions. The Bucklin Point data
show that with the 60 x 60 mesh screen an average of *6 percent removal of
suspended solids was accomplished over a 43 hour period when the level varied
from 60 to 465 mg/1 of suspended solids. The 100 x 100 mesh screen gave an
average of 53 percent removal over a 34.5 hour operating period when the
level varied from 40 to 255 mg/1 of suspended solids.
38
-------
TABLE XXVI - MODEL STRAINER RESULTS - BULLOCK'S POINT TREATMENT PLANT
60 x 60 MESH
(230 Microns)
C. 0. D.
Kunning Time
Total Hours
0.5
1.5
3.0
4.5
6.0
11.0
15.0
19.0
23.0
27.0
31.0
35.0
39.0
43.0
47.0
Suspended
Influent %
85
510
260
90
80
440
1415
425
660
345
255
205
165
235
175
Solids
Removal
35
45
56
48
32
52
56
25
68
29
47
59
43
60
29
Influent
As Is
437
768
621
504
482
525
841
792
790
625
655
600
666
545
Filtered
196
208
216
225
235
124
276
192
267
204
186
290
225
263
167
Effl
As Is Fi
390
574
510
480
394
414
1430
692
719
545
525
498
514
490
467
uent
Itered
231
208
225
255
223
118
267
225
202
225
186
222
223
218
218
% Removed
Filterable
33
35
30
20
30
26
28
45
23
25
22
32
34
39
-------
C. Self-Cleaning Filter Effectiveness
The most obvious and straight-forward results are those that were
obtained with the polyelectrolyte-ion exchange systems. It is quite
clear that the fluctuating flows and concentrations make the use of
such chemical pretreatment systems impractical with a diatomite sys-
tem. Even with carefully controlled laboratory systems, the results
were not sufficiently positive to encourage further work in this di-
rection.
The trends visible in C.O.D. and B.O.D. reduction shown in Table XX
suggest the use of low porosity diatomaceous earth for this type of
application. Excellent reductions in coliform level were obtained,
however, with all grades of diatomite. The results obtained with pow-
dered activated carbon indicate its applicability only with the more
porous diatomite. While this appears to be an anomaly, it is un-
doubtedly due to the resulting change in porosity of the filter cake
due to the carbon. The standard Super Lei has, according to the manu-
facturer, 50 percent by weight of its particles seven microns or less.
The Darco 6-60 has 30 percent of its particles larger than 44 microns
with the distribution between 44 and 7 microns unknown. The activa-
ted carbon, therefore, produces a more porous cake when mixed with
SSC or HSC grades of diatomite.
The formula provided by the manufacturer suggests that at the
operating conditions used to obtain the data in Table XXI, the cost
of operating the system would be greater than $1.50 per 1000 gallons
of water treated. The data shown on Pages34 and 35 indicate that the
costs could be lowered if the diatomite could be reused and it would
not disintegrate with repeated usage.
40
-------
V. References
1. Baumann, E.R; Cleasby, J.L; & LaFrenz, R.L. - A Theory of
Uiatomite Filtration. Journal AWWA, 54:1109 (September 1962).
2. Baumann, E.R; Cleasby, J.L; & Morgan, P.E. - Theoretical
Aspects of Diatomite Filtration. Water and Sewage Works,
111:229, 290, 331 (1964).
3. Bell, G.R. Design Criteria for Jiatomite Filters, Journal
AWWA, 54:1241 (October 1962).
4. Baumann, E.R. & LaFrenz, R.L. Optimum Economical Jesign for
Municipal Diatomite Filter Plant. Journal AWWA 55:48 (January
1963).
5. Eckenfelder, W.E. - Proceedings 21st Purdue Industrial Waste
Conference, Lafayette, Indiana, 1964, p. 427.
6. Spade, J.F; Treatment Methods for Laundry Wastes, Water & Sewage
Works, 109. 110 (1962).
7. Boucher, P.L; Evans, G.R. Micro-Straining - description and
Application, Water and Sewage Works, 1963.
8. Hudson, W., Performance of Wire Filter Cloth in Self-Cleaning
Strainers - unpublished internal report - Fram Corporation,
June 1966.
9. Evans, G.R; Treatment of Water Supplies by Micro-Straining, J.
New Hampshire Water Works Association, December 1962.
10. Fram Self-Cleaning Strainer Field Test - Weldwood of Canada,
Quenelle, British Columbia.
Test Duration: Spring and Summer, 1966
Operation: Straining of make-up water for paper board plant.
Water Source: Raw river water.
Contaminant: Small fish, dirt, and sediment.
Screen Area: 350 in.2 50 x 250 plain Dutch Weave.
Test Flow: 200 GPM
Contaminant Removal Efficiency: 100% 40 microns and larger.
No clogging of screen experienced.
41
-------
References (Continued)
11. Fram Self-Cleaning Strainer Field Test - Suntide Refining Company.
Corpus Christi, Texas.
Test Duration: 1968 - 1969
Operation: Straining of cooling tower water.
Contaminant: Airborne dirt and algae.
Screen Area: 1,000 in.2 50 x 250 plain Dutch Weave
Test Flow: 750 GPM
Contaminant Removal Efficiency: 100% over 45 microns
No clogging of screen during test, to date.
12. Summary Report - Advanced Waste Treatment (WP-20-AWTR-19), 1968
13. Bell, G.R.; Hutto, F.B.; Analysis of Rotary Precoat Filter
Operations - New Concepts, Chemical Engineering Progress 54:69 (1958)
14. Description of Johns-Manville Rotary Precoat Filter Test Leaf.
Published by Johns-Manville Research Center, Manville, N.J.
42
-------
VI Appendix
43
-------
Table a
Concentration
gms/s
1.0
1.0
1.0
0.4
0.4
0=4
0.4
Synthetic
C.O.D.
mg/1
990
1085
1069
416
423
439
455
Substrate Characteristics
Suspended
Solids, mg/1
550
563
613
248
270
265
276
Settleable B.O.D.
Solids, ml /I mg/1
6
5.8
-
2.25
-
1,90 162
182
44
-------
Table b
Run No.
1 - In
Out
2 - In
Out
3 - In
Out
4 - In
Out
5 - In
Out
6 - In
Out
7 - In
Out
8 - In
Out
Data
lean
60 x
1
169
63
120
50
240
185
133
115
143
90
125
100
154
105
109
75
For Table IX Calculated Results
ing Strainer - Synthetic Substrate
60 Mesh Square Weave Screen
Pore Size - 230 Microns
2
184
150
173
75
120
25
159
130
154
75
104
90
179
155
92
80
Suspended
Batch
3
165
145
109
40
73
20
197
160
193
95
132
70
138
65
163
100
Solids,
No.
4
165
115
98
45
180
105
206
170
186
80
154
85
125
80
213
95
mg/1
5
256
125
113
35
86
70
172
155
186
80
143
95
154
105
117
90
6
158
no
109
30
105
90
197
150
143
80
89
80
196
130
267
105
45
-------
Table c
Raw Daba For Table X Calculated Results
Self-Cleaning Strainer - Synthetic Substrate
Suspended Solids
B.O.D.
C.O.D.
•£»
CD
Run No.
1263-48
1263-49
1263-46
1263-47
1275-1
1275-2
1275-3
1275-4
In
Out
In
Out
In
Out
In
Out
In
Out
In
Out
In
Out
In
Out
1
113
105
176
125
129
70
172
145
175
100
129
90
154
70
191
140
2
146
115
165
135
137
70
176
65
154
100
125
95
171
115
224
155
Batch
3
161
125
143
90
163
95
193
140
161
115
132
70
171
125
204
125
No.
4
135
90
139
115
129
95
193
130
196
90
143
90
150
115
196
160
5
191
130
169
no
189
85
228
170
154
100
172
120
188
145
175
180
6
225
220
176
110
155
95
193
170
161
130
132
95
140
134
280
170
1 2
105 101
75 75
94 86
65 70
163 180
190 210
189 189
180 180
185 165
165 147
170 175
168 164
180 160
158 158
-
Batch No.
3 4
105 95
100 79
79 101
50 100
189 214
190 180
190 240
189 190
165 180
143 164
221 150
145 135
175 160
158 150
-
5
185
94
117
110
206
170
206
180
147
145
132
130
170
150
-
6
105
85
113
110
189
180
189
150
165
164
129
125
170
158
-
Batch No.
123456
365 403 368 306 255 345
251 310 317 302 344 332
327 351 375 246 378 381
328 328 312 320 333 352
378 408 444 398 401 432
361 333 337 314 357 368
357 396 373 388 432 388
337 302 341 356 349 345
423 392 321 317 304 392
332 308 272 280 288 304
258 349 338 361 417 358
300 284 300 316 300 284
405 375 392 405 385 375
344 344 256 348 324 352
418 452 448 418 392 395
336 336 336 328 336 340
-------
Operating Sequence of Self Cleaning Strainer-Filter System
The proposed combined filtration/purification system contains
a number of rather sophisticated parts. The function of each is
outlined in the following operational sequence description. This
review should be made with reference to the schematic drawing
attached. As previously stated, the system is completely automatic.
All components making up the system are commercially available.
However, modifications may be necessary in some cases to adapt the
particular components to the specific problem.
Power Supply
To use the proposed system, an adequate source of electricity
is required. This will require the use of the public utility
system and when required the construction of extension lines and
transformers. For a test site demonstration, a portable engine-
powered electrical generator would be rented to provide electrical
power.
Influent Supply System
It is proposed that the overflow water be pumped from the
supply source; in this case, the previously mentioned receiving
basin. Tne suction hose, equipped with a large opening strainer
screen, would be placed in the receiving basin. To prevent clogging
by large debris such as tree limbs, timbers, rags, etc., the screen
strainer would be surrounded by a large mesh or bar screen cage.
Tne influent pump, of a centrifugal type, provides the supply
water to the self-cleaning strainer. Level controls placed in the
basin reservoir activate the influent pump motor, and in turn, the
remainder of the filtration equipment at a pre-selected level in
the basin. As the water level declines to normal, the influent
pump stops, thereby placing the remainder of the system on a standby
status.
Self-Cleaning Strainer
The water would be pumped into the self-cleaning strainer, in
the normal manner, in which the strainer screen support basket would
be continuously rotating and backwashing the deposited solids of 50
microns or greater. As the differential pressure builds up across
the strainer, the blowdown system would operate automatically and
discharge collected solids to a portable receiving bin which may be
removed from the test site and dumped at the municipal sanitary fill.
47
-------
Flow Control Mechanism (Valves VI. V2 and V5)
The flow control valves, as shown, would be throttled by
the pilot valve mechanism working off of the float level control
in the self-cleaning filter unit. In other words, if the liquid
level within the self-cleaning unit begins to rise above the de-
sired level in the filter case, the discharge from the influent
pump would oe throttled. At the same time the discharge of the
self-cleaning strainer would be throttled until such time as the
effluent pump could withdraw the liquid as fast as it is being
pumped into the unit. This balanced system would be established
to maintain a constant liquid level in the self-cleaning filter
unit downstream of the self-cleaning strainer.
Jiatomaceous Earth Injector System
A small portion of the flow stream from the effluent side
of the self-cleaning strainer would be continuously circulated
through an open funnel arrangement on the suction side of the
diatomaceous earth injection pump. The liquid level in the
funnel system would be automatically controlled by the float
mechanism operating Control Valve V-3. In this manner, the
injection system would be ready at all times to receive injected
portions of diatomaceous earth or activated carbon or any other
type of filter aid or powdered chemical treatment. If a liquid
chemical agent would be desirable, a Wallace-Tiernan type pump
would have to be added.
Self-Cleaning Filter
As shown in the schematic diagram, the self-cleaning filter
basket would be mounted on external bearings, which in this case,
are not required to seal against any high pressure and are not
required to maintain continuous rotation. The basket would be
covered with any cnangeable type of filter cloth such as Oacron,
Teflon, nylon or other conventional filter cloth materials which
can be readily sealed at the ends of the support basket. The
flow is directed into the filter body and the level controlled
as previously discussed. As the contaminated liquid enters the
filter housing, it will be drawn through the filter cloth when
tne liquid level reaches the float to open Valve 16. As there
would be no filter aid now in contact with the filter cloth, the
turbidity meter would sense a contaminated stream and the following
sequence would then take place:
a. The turbidity meter sensing the contaminated stream would
close Solenoid Valve V-7 and open Solenoid Valve V-8 to
direct the flow back to the inlet side of the filter case.
48
-------
b. Simultaneously with this operation, the signal from the
turbidity meter would also actuate the vibrating system for
the diatomaceous earth storage tank and open Valve V-4 to
inject the precoat material into the diatomaceous earth
injection system. This material would then be deposited
on the filter cloth. The filtration unit would continue
to bypass until such time as the filter precoat had been
established on the filter cloth sufficiently to permit a
clear effluent, at which time the Solenoid Valve V-8 would
close and V-7 would open discharging a clean effluent, through
the cnlorinator, to the water system. At the same time, the
signal from the turbidity meter would cut off the vibrating
hopper on the diatomaceous earth injection system and close
Valve V-4.
Self-Cleaning Filter Backwash Cycle
When the contamination level builds up across the filter cloth
in sufficient quantity, the suction pressure on the effluent pump
will decrease. The effluent control switch S-l, in the suction line
of this pump, will sense this condition and the following sequence
will take place simultaneously:
a. The rotary drum filter cloth drive system will be automatically
energized; rotating the basket through approximately 120° to
expose a clean section of the cloth to again permit full flow.
b. The high pressure nozzle will be energized, opening Vavle V-l,
"ballooning" the cloth outward against the adjustable rubber
scraper blade and directing the spent diatomaceous earth cake
to the discharge conveyor. The spent material is conveyed to
a portable receiving bin.
c. Following the scraper, Valve 9, actuated by the S-l switch, opens
to allow flow of high pressure water through the hydraulic nozzles
to remove any remaining traces of contaminant from the cloth. The
flow rate will immediately increase as the clean filter cloth is
exposed to the liquid. The effluent contamination will increase
causing tne turbidity meter to again energize the diatomaceous
earth feeder for additional precoat and bypass back to the un-
filtered side of the unit.
d. Tne backwash air stream will be directed through the nozzles (as
in Step 7b) by closing Valve 10 on the heater-blower unit and
opening Valve V-ll. In normal operation (not backwash) the
heater-blower unit will circulate a high-flow warm air stream through
the upper portion of the filter cloth which is above the liquid
level. The air flow will be in the outside-in direction to pre-
vent the premature rupture of the filter cake.
49
-------
Chlorinator
This would be a conventional device of the Wallace-Tiernan
type or equivalent and would be employed to feed sufficient
chlorine to maintain a desired residual chlorine content in the
water effluent.
50
-------
TREATMENT
OR
SETTLING TANK
IWLET FROM POLLUTED STREAM
OR SETTLING TANK (/F REQUIRED)
LI/VETO SLUSH PIT
CONTAMINANT ACCUMULATOR SUMP (BELOW)
DIATCMACEOUS
STORAUE
SELF CLEANING STRAINER
BACKWASH PUMP-MOTOR
STRAINER BASKET
DRIVE
DIATOMACEOUS EARTH
CHEMICAL IMJECTOR
FLOW CONTROL VALVE
/MANUAL DRAIN VALVE
TURBIDITY METER
(EFFLUENT PUR
EFFLUENT PUMP
CONTROL PANEL
ELECTPICALLY CONTROL!.
AUTOMATIC BY-PASS VALVE
BY-PASS RE-CIRCULATING LIME
BACK WASH RE-CIRCULATIMS LIME
ELECTRICALLY CONTROLLEO
EFFLUENT VALVE
—t^- CLEAN EFFLUENT LINE
SYSTEM SCHEMATIC FLOW DIAGRAM
-------
Table d
MODEL STRAINER RESULTS - BUI
TREATMENT PLANT - 60 x 60 MESH SI
RAW INFLUENT, FLUX R;
Kunning Time
Total Hours
0.5
1.5
3.0
4.5
6.0
9.0
11.0
15.0
19.0
23.0
27.0
31.0
35.0
39.0
43.0
47.0
Suspended
Influent %
85
510
260
90
80
180
440
1415
425
660
345
255
205
165
235
175
Solids
Removal
35
45
56
48
32
52
52
56
25
68
29
47
59
43
60
29
C. 0. D.
Influent
437
768
621
504
482
525
841
792
790
625
655
600
666
545
Effluent
390
574
510
480
394
414
692
719
545
525
498
514
490
467
52
-------
Table e
MOUEL STRAINER RESULTS - BULLOC
60 x 60 MESH SCREEN, 230 MICROI
BAR SCREENS, 2!
Running Time
Total Hours
48
49
50
51
52
53
54
55
56
Suspended
Solids
Influent Effluent
375
230
170
485
325
840
425
175
270
175
125
50
290
230
660
120
60
115
C. 0. D.
Influent
538
530
540
468
760
1504
1464
460
540
Effluent
436
474
450
358
704
852
850
400
456
53
------- |