EPA-R2-73-124
JANUARY 1973
Environmental Protection Technology Series
Microstraining and Disinfection
of Combined Sewer Overflows
- Phase II
Office of Research and Monitoring
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, Environmental Protection Agency, have
been grouped into five series. These five broad
categories were established to facilitate further
development and application of environmental
technology. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL
PROTECTION TECHNOLOGY series. This series
describes research performed to develop and
demonstrate instrumentation, equipment and
methodology to repair or prevent environmental
degradation from point and non-point sources of
pollution. This work provides the new or improved
technology required for the control and treatment
of pollution sources to meet environmental quality
standards.
-------
EPA-R2-73-124
January 1973
MICROSTRAINING AND DISINFECTION OF
COMBINED SEWER OVERFLOWS - PHASE II
by
George E. Glover
George R. Herbert
Project 11023 FWT
Project Officer:
Richard Field, Chief, Storm and Combined Sewer Technology Branch
Edison Water Quality Research Laboratory NERC - Cincinnati
Edison, New Jersey 08817
prepared for:
OFFICE OF RESEARCH AND MONITORING
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
and the
PHILADELPHIA WATER DEPARTMENT
PHILADELPHIA, PA.
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402
Price $2.10 domestic postpaid or $1.75 GPO Bookstore
-------
EPA Review Notice
This report has been reviewed by the Environmental Protection Agency
and approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the Environmental Protection
Agency, nor does mention of trade names or commercial products constitute
endoresment or recommendation for use.
11
-------
ABSTRACT
A microstrainer using a screen with 23 micron apertures reduces the sus-
pended solids of the combined sewer overflow from 50 to 700 mg/1 down
to 40 to 50 mg/1 levels operating at flow rates of 35 to 45 gpm/ft^ of \
submerged screen. The organic matter as measured by COD and TOC was
reduced 25 to 40%. Coliform concentrations were 0.1 to 9 million cells
per 100 ml and no reduction was brought about by Microstraining (C).
The coliform concentrations of both overflow and microstrained overflow
were reduced by four or more orders of magnitude by disinfection with
5 mg/1 chlorine in specially built, high rate contact chambers of only
2 minutes contact time.
The drainage area served by the combined sewer comprises 11.2 acres
of a residential area in the City of Philadelphia having an average dry
weather sanitary flow of 1,000 gph. The overflow rates recorded were
generally 100 times, with a maximum of 400 times, the average dry
weather flow.
The limitations of conventional analytical methods for organic matter in
the highly variable combined sewer overflow are discussed. Modifications
of the analytical methods are described.
The extreme importance of very low - 2 minute - residence volume equip-
ment for suspended solids removal and for disinfection in the very high
instantaneous rates encountered with stormwater is shown.
The cost of a microstrainer - special chlorine contact chamber installation
is cited as $6,750/cfs of peak flow rate capacity less land and engineer-
ing. On the basis of 2 cfs instantaneous design overflow per acre, this is
$!3,lOO/acre.
The observed performance of the microstrainer was reduced to a
permeability parameter. The influence of turbulence on chlorine dis-
infection rate is shown.
This report, submitted in fullfillment of Project #11023 FWT, is a
continuation of the work previously reported in #11023 EVO 06/70.
(C) Copyrighted Trade Name - Crane Co.
iii
-------
CONTENTS
Section Page
I CONCLUSIONS 1
II RECOMMENDATIONS 5
III INTRODUCTION 9
IV EXPERIMENTAL EQUIPMENT 13
V SAMPLING 19
VI ANALYTICAL PROCEDURE 27
VII DISCUSSION OF RESULTS 29
VIII GLOSSARY 89
IX ACKNOWLEDGMENTS 91
X REFERENCES 93
XI PUBLICATIONS 95
XII APPENDICES 97
v
-------
FIGURES
Page,
1 Isometric Drawing of a Microstrainer 14
2 Intensely Mixed 4 Min. Chlorine Contact Chamber 16
3 Discrete Sampler With Flow-through Chamber 20
4 Outlet Chamber of Microstrainer Showing
Three Effluent Sample Lines 21
5 Salt Tracer Test - Microstrainer Effluent 23
6 Salt Tracer Test - Microstrainer Influent 24
7 Outfall 67th & Callowhill 30
8 Drainage Area 31
9 Rainfall Intensity - Philadelphia, Pennsylvania 1903-1951 32
10 Hourly Variations , Sanitary Flow, 67th &
Callowhill Sts. Sewer (City of Philadelphia Data) 33
11 Observed Runoff Coefficient Based Upon
10 Minute Intensity 37
12 Suspended Solids in Overflow vs Overflow Rate 38
13 Differential vs Applied Solids
(Permeability Parameter & Boucher Index) 45
14 Permeability Parameter vs Radial Flow Rate 49
15 Permeability Parameter vs Influent Suspended Solids 50
16 Effluent Suspended Solids vs Influent Suspended Solids 51
VI
-------
Page
17 Fraction Solids Retained vs Influent Suspended Solids 53
18 Typical Plot of Mass Retained vs Mass Applied 54
19 Combined Sewage Solids Retained vs
Combined Sewage Solids Applied 55
20 Fraction Solids Retained vs Permeability Parameter 58
21 Fraction Volatile Suspendeds Solids Retained vs
Fraction of Suspended Solids Retained 64
22 BODs vs BOD5 on Macerated Sample 67
23 COD vs COD on Macerated Sample 68
24 Fecal Coliform Count on Macerated Sample vs
Fecal Coliform Count on Untreated Sample 73
25 Total Coliform Count on Macerated Sample vs
Total Coliform Count on Untreated Sample 74
26 Fecal Coliform Survival vs Time
15 mg/1 Chlorine 79
27 Total Coliform Survival vs Time
15 mg/1 Chlorine 80
28 Fecal Coliform Survival vs Time
10 mg/1 Chlorine 81
29 Total Coliform Survival vs Time
10 mg/1 Chlorine 82
30 Fecal Coliform Survival vs Time
5 mg/1 Chlorine 83
31 Total Coliform Survival vs Time
5 mg/1 Chlorine 84
32 Comparison of Effluent Disinfection Curves
5, 10, 15 mg/1 Chlorine 86
VII
-------
TABLES
Page
1 Hydrographic Data 34
2 Microstrainer Dynamic Performance Data 40
3 Organic Removal Performance 60
4 Effect of Microstraining on Organic Matter
as Indicated by Several Test Methods 62
5 Ultimate BOD and Oxidation Rate Studies 65
6 Total Coliform Removal by Microstraining 70
7 Fecal Coliform Removal by Microstraining 71
8 Comparison of Disinfection Devices 77
Vlll
-------
SECTION I
CONCLUSIONS
1. Suspended solids removal performance tends to confirm previous
work. The microstrainer operated at flow rates of 35-45 gpm/ft^
with differentials of 24 inches of water. At these rates, the
suspended solids in combined sewer overflow were reduced from
50-700 mg/1 to 40-50 mg/1 and below. At the higher influent levels
of suspended solids, removal performance was enhanced, yielding
effluent concentrations of approximately 10 mg/1. The conventionally
used, percentage removal performance criteria are not valid for
Microstraining (C) of combined sewer overflow.
Volatile suspended solids reduction paralleled the reduction of
total suspended solids.
2. The highest concentration of suspended solids frequently occurs
when the overflow rate is highest. The concurrence of high suspended
solids concentration and high overflow rate results in a very high
potential contaminant loading per unit time to the receiving stream.
The microstrainer was unusual in that it removed a much greater
percentage of the suspended solids when the concentration of these
was higher.
A reasonably sized microstrainer can then limit the pounds of
suspended solids per unit time entering the stream from a combined
sewer outfall. Availability of this equipment will permit the
regulation of this important aspect of pollution from combined sewer
overflows.
3. The organic matter in combined sewer overflow was highly variable -
ranging from 10 to 2,000 mg/1 as BOD5 and 20 to 4,000 mg/1 as COD
and TOG. The microstrainer reduced organic matter by some
25 to 40%.
4. The total coliform content of combined sewer overflow at our site was
variable from 0.1 to 3 million cells/100 ml. The fecal coliform
content of the combined sewer overflow was also variable from 1,000
to 100,000 cells per 100 ml. The microstrainer demonstrated little
or no ability to remove total or fecal coliform from combined sewer
overflow.
5. Chlorine contact time of only 2 minutes under relatively high turbulence
-------
conditions with chlorine dosages as low as 5 mg/1 reduced total
coliform from 1 million levels in combined sewer overflow down to
5 to 10 cells per 100 ml. Fecal coliform was similarly reduced under
the same conditions from 100,000 levels in the combined sewer over-
flow down to 5 to 10 cells per 100 ml.
6. Bacteria counts made on macerated samples were generally over two
times higher than on the unmacerated samples.
7. The capital cost of a microstrainer-high rate chlorination preceded
by bar screening, previously reported (1), would be $13,100 (in 1972
dollars) per acre not including engineering or the 1/5,000 acre land
required per acre drainage area. This capital cost is based upon
1.94 cfs design runoff per acre and upon an intensely mixed chlorine
contact chamber such as used herein. The present work confirms the
high rate performance of the microstrainer and shows 99.99% bacterial
kills in 20 gpm pilot size, high rate chlorine contact chambers. The
conclusion is repeated here that a high rate Microstraining unit and
high rate chlorination are practical and the least expensive of the
alternate treatment methods considered (20). The single exception
is surface impounding where very low cost land is available.
8. The design criteria used for the construction of the special, high
mixing intensity, pilot size contact chambers, appear to be appropriate
and practical for full size units.
9. The special character of solids in combined sewer overflow (and
stormwater) is vastly different from other municipal and industrial
wastewaters and from the potable water sources to which a micro-
strainer is normally applied. The high level (liquid) differentials
across the screen required to achieve the high flow rates necessary
to be practical in stormwater service are far beyond those conventionally
used. Thus, empirical relations and experience useful in conventional
applications are misleading in the Microstraining of combined sewer
overflows.
A performance parameter was developed which is closely akin to the
permeability of the deposited solids on the screen. This permeability
parameter was calculated from the operating conditions of each storm.
This parameter is believed to characterize most of those properties
of the deposited particles which determine resistance to flow. There-
fore, it can be used, for similar particles, to predict head loss at
other flow rates, drum speeds and temperatures.
10. The permeability parameter was also found to increase with higher
-------
suspended solids concentrations indicating that the additional solids
in the runoff changed the properties of the average particle. Thus,
with the conditions at our site, the permeability parameter can be
used to predict head loss at other runoff suspended solids concen-
trations as well.
11. It was further found empirically that the suspended solids removal
also increased as the permeability parameter increased. Conversely,
then, permeability can be used to qualitatively predict solids
removal.
12. The allocation of capital costs of facilities for the treatment of
combined sewer overflow is not uniform from study to study. The
difference in method is particularly evident when capital costs; are
reported as dollars-per-acre of drainage basin area. There is a
broad range for the cost of the various proposed treatment methods
on dollar per cu ft/sec of design through-put basis. There is also
a very broad range of design runoff or sewer overflow rates specified
per acre of drainage area served. The overflow rates vary from
2 cu ft/sec-acre as used in item 7 above and in our previous report (1)
down to 0. 2 cu ft/sec-acre for very large drainage areas where
considerable flow equalization within the trunk sewers and absence
of simultaneous rain fall over the entire area are considered (2).
Less understandable is the use of low overflow rates such as 0.25
cu ft/sec-acre for the smaller drainage areas that might be served by
satellite overflow plants (3) .
-------
SECTION II
RECOMMENDATIONS
1. Microstrainers employed in stormwater service should be operated
at high differentials of approximately 24 inches water. At this
differential and with the deposited (suspended) solids permeabilities
we encountered, flow rates of 35-45 gpm/ft^ can be obtained with a
deep enough mat of solids to produce an effluent quality of 40 mg/1
suspended solids. To achieve lower effluent levels with the
suspended solids characteristic of this test period, slower drum
speeds, possibly lower flow rates and/or the use of polymers will
be required. Additional storms should be monitored at a broader
range of conditions to confirm the possibility of cleaner effluents
without polymer addition, as was illustrated in a few instances in our
earlier work (1).
2. A full-scale, in-line microstrainer and chlorination facility, without
an impounding basin, should be built as a demonstration plant at a
site where the benefit of stormwater treatment can be measured by
comparison with a previous stream quality study.
3. Additional work is required to confirm the disinfection rate on both
untreated and microstrained combined sewer overflow with lower
chlorine dosages, at low ozone dosages, and with low two-stage
ozone-chlorine dosages.
4. In order to bring more realism to the treatment of combined sewer
overflows, regulatory agencies should establish guidelines for the
performance of such treatment in the many different overflow
situations. In view of the range of almost two orders of magnitude
in both combined sewer overflow rate per acre and in suspended
solids concentrations, a generally applied percentage removal guide-
line would be totally ineffective in one case and excessively punitive
in others. Further, in view of the rapid variations in flow rate and
pollutant concentration within a given storm, as compared to the
relatively steady state flow of domestic sewage, the measurement
of percentage removed would be a difficult and expensive enforcement
procedure. Performance guidelines should be based upon average
effluent quality or stream quality or pounds pollutant added per unit
time and should include residual chlorine levels.
5. In the interest of both economy and the public health, the influence
of mixing intensity in the disinfection of municipal and other waste-
waters should be recognized. This recognition might take the form
-------
of specifying a minimum mixing intensity and actual contact times
in conventional chambers. Ultra high rate chambers of suitable
design can be especially applicable in such cases as combined
sewer overflow service.
6. With the additional potential utility of the permeability parameter to
qualitatively predict suspended solids removal, it is recommended
that additional storms be monitored without the use of polymers to
further establish this base, and that this parameter be included-as
a means to aid in evaluation of polymer addition.
7. The treatment of combined sewer overflow has only recently received
serious attention. Rational design of facilities to treat this very
high, variable flow rate, variable composition wastewater is seriously
handicapped by the use of inappropriate performance criteria and
analytical methods borrowed from the relatively steady-state domestic
sewage treatment art. Performance criteria should be adopted which
recognize both the unusual nature of this waste and the equipment
employed to treat it. For example, the mgd dimension of flow rate
is frequently used because of its familiarity to sanitary engineers.
It is strongly recommended, however, that they be aware that it is the
instantaneous capacity of the equipment that is the governing parameter.
8. In view of the peak instantaneous capacity limitation of the equipment,
the benefit of flow equalization within the sewers themselves should
be fully evaluated. Also, the benefit of aboveground flow equalization
basins should be evaluated to determine their capability to reduce the
peak flow requirement of the equipment.
9. Similarly analytical methods should be developed and adopted which
recognize the widely variable amount of and the unusual nature of
the solid and dissolved matter in the combined sewer overflow. The
unusual nature of the organics in combined sewer overflow is exemplified
by the refractory nature of the solids and the short time interval
between wetting and discharge via overflow. The BOD test is basically
inaccurate, unrepeatable, and unreliable .
The problems with the BOD test are magnified in combined sewer
overflow by the probable presence of toxic metals and the difficulty
of maintaining a culture of acclimatized micro-organisms for the
decomposition. The use of non-biological analytical methods such as
COD and TOO for combined and separate sewer overflows is recommended
The maceration pretreatment improves the COD as well as the
methods and should be adopted.
-------
Also, the effect of cell clumping, occlusion of cells in solids
particles, and delayed growth character of recently wetted spores
strongly indicates the use of the maceration pretreatment of sample
prior to cell counting methods for truer bacteria counts. The
maceration technique is of recent origin and additional work is
required to define and optimize the several parameters - particularly
duration of maceration and mixing intensity.
10. Additional storms should be monitored to determine the advantage
of the use of polymers in suspended solids and organic removal from
combined sewer overflow. Particle size data would be helpful in
such an effort.
-------
SECTION III
INTRODUCTION
This work is a continuation of the study of a commercial size microstrainer,
with chlorination and/or ozonation, treating combined sewer overflows.
The original work was reported as "Water Pollution Control Research Series ",
Report #11023 EVO (1) and in abbreviated form elsewhere (4).
Previous Work
Briefly, the original work covered 26 storms during the period 1/69 to 9/69
from an 11.2 acre residential drainage area in Philadelphia, Pa. A Crane-
Glenfield microstrainer fitted with 47 sq ft of first Mark I (35 micron
openings) screen and later with Mark 0 (23 micron) screen was used. The
suspended solids content of the storm overflow ranged from 20 to 500 mg/1,
more or less proportional to the storm overflow rate. The flow rate through
the microstrainer was controlled at a variable rate proportional to the storm
overflow rate and average flow rates ranged from 1 to 45 gpm/ft2 of sub-
merged screen area. In order to achieve the higher flow rates through the
screen with existing pumping equipment, 4/5 of the screen area was
masked during the later tests with Mark 0 screen.
About 3/4 of the storms were unattended. The Microstraining equipment
idled continuously and the drum speed was adjusted at the onset of storm-
water flow by automatic controls. The performance was monitored by
recording instruments and by flow insensitive composite samplers . The
outlet weir level was set initially for a 6 inch of water maximum differential
and for the later storms the weir was set for 24 inches of water.
The suspended solids removal ranged from 20% to 98% with the higher
removals generally occurring at the higher flow rates and always at the
higher inlet suspended solids concentrations. Removal of suspended
volatile solids (SVS) paralleled the removal of suspended solids (SS).
The microstrainer performance, as opposed to other treatment techniques,
was insensitive to wide and rapid changes in the composition of the
combined sewer overflow (e.g. , the content of suspended solids or the
temperature).
The BOD5 of the sewer overflow ranged from 5 to 500 mg/1 averaging
50 mg/1. The sanitary contribution was less than 2,000 gph from this
area while measured storm overflows were frequently over 100,000 gph.
BOD5 removals by the microstrainer ranged from 80% down to 10% on 16
of the storms . On the other 10 storms, increases in BOD were reported
and many were of significant magnitude.
-------
Similarly, the total and fecal coliform counts on the sewer overflow
ranged from 0.1 to 3 million cells per 100 ml. Coliform removals by the
Microstraining unit ranged from 90% down to 20% on 16 storms while on
the other 10 storms there were increases in coliform count. Many of the
increases were significant and, incidently, did not occur altogether on
the same storms as the BOD increases.
Bench scale chlorination of the microstrained combined sewer overflow
showed some 99% bacterial kill with 5 mg/1 dosage and 10 minute contact
time. Ninty-nine percent kill was also obtained in one test in 0.5 minute
contact time using 15 mg/1 chlorine. Ozonation in 20 gpm pilot scale
equipment using 4 mg/1 ozone dosage and 12 minute contact time showed
some 99% bacterial kill.
An economic study was made of the cost of combined sewer overflow
treatment by a Microstraining filter and disinfection as well as treatment
by other proposed methods. This study (1) indicated that all methods,
except for surface impounding which on a design basis of 2 cfs runoff per
acre surface impounding requires 1 acre basin area of 10 ft depth for each
30-50 acres drainage area, would require a capital cost of at least
$lO,000/acre excluding land and engineering (1969 dollars). The other
methods included some proposed (14) treatment trains comprised of
conventional and advanced sewage treatment processes such as screening
flotation, bio disc contact, activated carbon, primary clarification, etc.
This study indicated that the Microstraining filter and disinfection at high
rate was the least expensive and most compact of the other methods then
proposed and, moreover, required less than 1 acre of treatment area per
5,000 acres of drainage area.
Present Work
The second phase of the work was conducted to confirm the performance
of the Microstraining unit and disinfection at high rate under more closely
controlled conditions. The same facility was used. It was adapted by
the addition of a third microstrainer feed pump, and by the substitution
of automatic vacuum type discrete samplers (Serco Model SG-15 manufactured
by Sanford Products) and the necessary flow-through sample chambers for
the composite sample pumps, which withdrew directly from the microstrainer
inlet and outlet chambers. Also, two specially-designed chlorine contact
chambers of the flow-through type were substituted for the batch type
glassware previously used. The 5 ft diameter by 3 ft microstrainer with
9.4 ft2 of open Mark 0 (23 micron) screen was used as before. The inlet
and outlet chambers were reduced in size to reduce residence time within
these chambers.
10
-------
The controls were changed so that only storms producing more than 0.66
mgd (30,000 gph) of overflow were treated (30 times mean dry weather
flow). The flow rate through the microstrainer was preset and maintained
at that constant rate during a storm, independent of variations in sewer
overflow rate.
The disinfection procedure used in the second phase was developed to
confirm the economically attractive high rate chlorination work of the
first phase. Also, the procedure was designed to confirm or disprove
the postulated improved ease of disinfection of the microstrained water.
To this end the raw overflow as well as the microstrainer effluent were
disinfected simultaneously in separate contact chambers during an
attended storm. The chlorine dosage of the raw water was set higher to
compensate for the greater chlorine demand so that both raw and micro-
strained water flows had the same residual after 195 seconds.
Project time and occurrence of suitable rainfall have not permitted
expansion of the original data for disinfection via ozonation or with both
ozone and chlorine.
11
-------
SECTION IV
EXPERIMENTAL EQUIPMENT
Microstrainer
The test system uses a Microstraining filter for the removal of suspended
solids followed by sodium hypochlorite disinfection. The microstrainer is
a rotating drum filter, 5 ft in diameter by 3 ft long, covered with specially
woven Mark 0 stainless steel wire fabric having nominal openings of
23 microns.
Eighty percent of the screen area is blanked, with three layers of poly-
ethylene sheet so that 9.4 sq ft is open and effective in the straining
process.
The water enters the drum through the open end, flows radially through
the drum into the outlet chamber and deposits suspended solids on the
inside of the drum screen (see Figure 1). As the drum rotates, the dirty
screen passes up under backwash jets which wash the deposited solids
off the inside of the screen into the wash water hopper and then to waste.
The mat, of deposited solids on the screen causes the resistance to flow to
increase; i.e. , causes the differential to increase for any flow rate
through the unit. The mat of deposited solids are a much finer filter than
the screen itself and the thicker deposits improve the removal of suspended
solids.
For the purpose of monitoring the unit's performance on short duration
storms, the inlet and outlet chambers were reduced to minimum volume to
reduce response time. The sample streams were withdrawn from the inlet
chamber at the entrance to the drum and from the outlet chamber at the
center of the open screen panel about 2 inches from the bottom of the drum;
i.e., at mid-point of screen travel during submergence.
The microstrainer ran at its idling speed continuously, under UV light,
with backwash jets flowing city water during dry weather. During storms,
the jets could have been supplied with microstrained water by means of a
pump, but city water was used here too for convenience. The backwash
flow was less than 2 gpm (less than 1/2% of combined sewer flow through
the microstrainer). Normally, the small (approximately 1-1/2 times the
mean dry weather) flow would be directed to the interceptor sewer down-
stream of the overflow regulator but for our convenience, it was directed
to the creek. At the onset of a storm, automatic controls started the
feed pump(s) when the overflow rate reached 0.66 mgd (30,000 gph) . The
selection of 0. 66 mgd (30,000 gph) as the trigger flow rate was to avoid
13
-------
Figure 1
Isometric Drawing of a Microstrainer
-------
storms of insufficient intensity or duration to yield meaningful results.
The flow rate of water to the microstrainer was controlled at a constant
rate, predetermined by the selection of the number of pumps to be
activated by the storm overflow rate control. The flow rate through the
microstrainer was then constant and independent of the combined sewer
overflow rate. The flow of water through the microstrainer caused a
difference in level between inlet and outlet chambers. The automatic
drum speed control sensed these levels and increased the drum speed
proportional to the differential. The proportional control band maintained
the drum speed at a maximum (6 rpm) speed with a maximum differential
of 24 inches of water and proportionately slower speeds at smaller
differentials. It might be noted here that the preferred mode of operation
for continuous operation is to control drum speed to maintain the maximum
differential of 24 inches by use of a reset-proportional band. This mode
of control was abandoned during the previous phase of this work on
intermittent flows of stormwater because of the effect of so-called "re-set
windup" during long idle periods between storms.
The automatic sampler controls were also triggered by the overflow rate
control. The inlet sampler was delayed 2 minutes and the outlet sampler
was delayed 5 minutes after the start of the feed pumps. The residence
time between inlet and outlet sample points was 3 minutes, and plug flow
was obtained to a remarkable extent as indicated by salt tracer studies to
be described in Section V, SAMPLING.
Chlorine Disinfection Equipment
The chlorine disinfection equipment consisted of two specially-designed
chlorine contact chambers each equipped with a Schutte and Koerting,
variable area, rate-of-flow meter for the inlet water, a Lapp, variable
capacity, hypochlorite, metering pump, a container of hypochlorite ;
solution and a homemade inorganic cation exchanger.
The equipment was manually operated and one of the chambers - the
one treating the microstrainer effluent - was equipped with a feed pump.
The microstrainer inlet pressure was sufficient to supply the other chamber.
The field test equipment used to monitor and control the disinfection
conditions consisted of an orthotolidine comparator set for free and
combined available chlorine, a millivolt meter with oxidation-reduction
potential (ORP) cells, a pH meter, and a liquid level gauge.
The contact chamber shown in Figure 2 provided 4 minutes residence
(contact) time at a flow rate of 20 gpm. The large number of baffles
insured true plug flow so that full residence time was utilized for chlorine
contact.
15
-------
Figure 2
Intensely Mixed 4 Min.
Chlorine Contact Chamber
16
-------
The large number of baffles resulted in narrow passages and high liquid
velocity. The high liquid velocity (.27 ft/sec) together with the corrugated
configuration of the baffle panels resulted in a head loss of 1-1/2 inches
over the 64 ft long travel of liquid. This energy dissipation created a
relatively high velocity gradient throughout the liquid.
Sample taps at 1/4, 1/2 and 3/4 the length of the liquid path provided
samples at 75, 135, and 195 sec contact time, respectively at 20 gpm.
The top of the chamber was open so that samples could be collected at
the surface after the inlet chamber (15 sec) and at the end of the path
(255 sec) .
All disinfections were performed at a constant flow rate of 20 gpm through
the chambers. Except where it was not possible, both the raw combined
sewer overflow (microstrainer influent) and the microstrained effluent were
disinfected simultaneously during the storm. A pre-selected initial
concentration (dosage) of chlorine was controlled by the setting of a
positive displacement Lapp pump and by the concentration of the hypo-
chlorite solution. No attempt was made to control the residual concen-
tration of the microstrainer effluent stream. However, it was measured
after 195 seconds contact and the dosage to the raw combined sewer over-
flow with its higher chlorine demand was adjusted to yield the same
195 second residual. No attempt was made to control pH; however, since
the combined sewer overflow was already in pH 7-8 range and the hypo-
chlorite was alkaline, the hypochlorite pump discharge was passed on
later storms through an inorganic cation exchanger in hydrogen form to
avoid raising the stream pH to a higher range favoring less effective forms
of chlorine. The cation and alkalinity content of the hypochlorite solutions
and the buffering action of the combined sewer overflow were such that
instead of raising the pH about .3 - .4 without the cation exchanger - the
pH was lowered - .4 - .5 units.
After equilibrium was established as indicated by constancy of pH and
ORP at all points, and equal chlorine residuals at the 195 second sample
points, samples were taken from the flowing sample cocks and from the
surface of the flowing liquid at locations representing 15, 75, 135, 195,
and 255 seconds contact time. Samples for bacteria analysis were
collected in sterile bottles containing several ml of sodium thiosulfate
solution to quench the chlorine and were shaken vigorously and covered.
Chlorine residuals were measured with the orthotolidine comparator set.
The comparator equipment was arranged to permit the analysis to be
performed in about 15 seconds. With the necessarily short reaction time
between sample and the orthotolidine reagent, the Standard Methods
Procedure indicates that for practical purposes only free chlorine was
17
-------
measured, although at the recommended 5 minute reaction time the
orthotolidine will reveal both free and combined chlorine.
The hypochlorite stock solutions were made up based upon the label
concentration of household bleach solution and stored until needed.
In one case, the actual concentration of the hypochlorite stock solution
was measured at the time of use. The calculated chlorine dosage based
upon this measurement was within a few percent of the preselected dosage,
18
-------
SECTION V.
SAMPLING
The inconsistencies in BOD and bacterial results noted in the first phase
of this work (1) prompted a greatly expanded analytical program and
alterations in the sampling procedures for this the second phase of work.
Unfortunately, many storms occurred and were sampled before any
analytical results were reported by August 24, 1971, at which time it was
evident that there were faults in the sampling and/or analytical procedure.
It was then found that the automatic sampler bottles were being incorrectly
sterilized in the field. Accordingly, neither the organic content nor
coliform results prior to October 10 when the automatic sampler was
involved, are considered in any of the discussions.
Prior to October 10, the sample bottles placed in the automatic vacuum
sampler (Figure 3) were sterilized with methanol and were not completely
dried. At the same time, the procedure for sterilizing and cleaning the
sample tubes was revised to include flushing the individual sampler tubes
with hypochlorite followed by a high velocity rinse with city water.
Many of the storms were of short duration and all were of varying, some-
times rapidly varying, composition. The response time of the automatic
sampler and its auxiliary flow-through chamber to changing composition
of the flow through the microstrainer is important. As previously mentioned,
the volume of the microstrainer inlet and outlet chambers had been reduced
to improve response time. The microstrainer inlet sample was withdrawn
from a point in the inlet chamber 2 inches axially out from the inlet end of
the drum and about 18 inches radially to the emerging side of the drum
center.
The inlet to the effluent flow-through sample chamber was located at the
center of the open screen panel about 2 inches from the bottom of the drum.
Figure 4 shows the microstrainer effluent chamber with the plywood wall
reducing the chamber volume and with three sample points. The visible
opening collects from a point near where the screen enters the water.
The sample chamber inlets are then located less than 2 seconds from the
inlet and outlet of the drum. The flow rate through the sample chambers
was approximately 10 gpm, at which rate the 6.5 and 13 gallon volumes
of the influent and effluent flow-through chambers could cause a response
lag of .65 and 1.3 minutes respectively.
Figure 3 shows the effluent sampler with 24 evacuated bottles; the 13
gallon flow-through sample chamber and the sample pipes from the
microstrainer outlet chamber.
19
-------
ro
Figure 3
Discrete Sampler With Flow-through Chamber
-------
Figure 4
Outlet Chamber of Microstrainer Showing
Three Effluent Sample lines
-------
The cumulative effect of all these contributions to the time lag in the
response of sample composition to changing overflow water concentration
is shown by a typical salt tracer test (Figures 5 and 6).
Note that Figure 5 shows a lag of 5 minutes - the theoretical residence
time of the volume between the pump suction, inlet line, inlet chamber,
drum volume and the outlet chamber. Similarly, Figure 6 shows a lag of
2 minutes - the theoretical residence time of the volume between pump
suction, inlet line, and inlet chamber.
Figure 5 and Figure 6 also show that there was surprisingly little back
mixing (departure from plug flow) through the microstrainer. Also, these
curves show that about 1 to 1-1/2 minutes additional lag was introduced
by the flow-through the sample chambers.
The inlet sampler was started automatically 2 minutes after the feed pumps
(flow of water) were started at the beginning of a storm. Similarly the
effluent sampler was started 5 minutes after the pumps started, so that
the first sample taken of inlet and outlet represented more or less the same
increment of water. The samplers then collected instantaneous, discrete
300 ml samples of the two flows every 2 minutes and stored them in a
refrigerated chamber.
The full analysis program for composite samples required approximately
2,700 ml sample volume which is the contents of at least nine automatic
sampler bottles. Where sufficient number of samples were collected
both a composite sample covering a period of the storm and discrete
samples were analyzed.
In general, composite samples were made up by plotting the collection
time of the instantaneously collected sample increments on the storm
overflow rate meter chart and selecting those increments representative
of a period of more or less uniform storm overflow rate.
The sampling procedure used in each case is noted in an abbreviated
manner in Table 2. For example, the "8 minute automatic composite"
notation for the storm of 7/1/71 indicates that the sample was composited
from the contents of 4 or 5 instantaneous samples collected over an
8 minute period, beginning 5 minutes after the onset of the Strom at
7:55 p.m. The "hand discrete" notation for the storm of 7/19/71 at
11:11 p.m. indicates that a large volume sample was collected by hand
at time 11:11 p.m. and was analyzed (in addition to the sample composited
from the automatically collected instantaneous samples over an 18 minute
period during the 10:33 p.m. to 11:11 p.m. storm) .
22
-------
CO
CO
700 --
600 --
a 500
S-i
O
O 400
o
>300
+->
o
CD
O
O
200.:
5 min 1
residence]
time
10
^Microstrainer Effluent
A Effluent Sample Chamber
Time - Minutes
Figure 5
Salt Tracer Test - Microstrainer Effluent
-------
700.. 1
Q)
to
j2 min
residence time
O Micro strainer Influent
Influent Sample Chamber
30
Time - Minutes
Figure 6
Salt Tracer Test - Microstrainer Influent
50
-------
The "auto discrete" notation for the storm of 7/30/71 indicates that the
automatic samplers collected an insufficient volume during this short
duration storm and the collected volume was submitted for analysis under
the abbreviated program for discrete samples.
In some, cases, due to a failure of one of the samplers, a composite of
the bottles of the working sampler was submitted for analysis along with a
large sample of the contents of the flow through sample chamber of the
faulty sampler collected long after the storm. These samples almost always
yielded anomalous results and were discarded in the Discussion of Results
Section.
25
-------
SECTION VI
ANALYTICAL PROCEDURE
The anomalous increases in BOD and coliform counts on passage through
the microstrainer noted in the first phase of this work prompted a greatly
expanded analytical program in this second phase.
The additional analytical treatment had two general objectives. The first
was to get more reliable data on microstrainer performance. The second
was to determine if the Microstraining operation had an effect on the
combined sewer overflow that might influence the results of conventional
analytical methods.
It was postulated (1) that the anomalous increases in BODs and coliform
count observed on passage through the microstrainer reflected a significant
change in the character of the particulate matter by the Microstraining
process .
Specifically, it was postulated (1) that the organic matter in particulate
form was reduced in size so that it became more available to biological
oxidation during the period of the 5 day BOD test. To investigate this
possibility in depth a manifold approach was adopted. Long term
oxidation rate analyses were made so that the ultimate BOD could be
predicted. Further, the suspected effect of particle size reduction by
Microstraining was simulated by maceration of both influent and effluent
samples in a Waring blender prior to the BOD test. Finally, the BODs
test was supplemented by the more accurate COD and TOO tests for
organic matter.
Similarly, it was postulated that many of the coliform in combined sewer
overflow existed in clumps which were broken up by passage through the
microstrainer and were counted as individual cells in the effluent sample.
To explore this possibility, the particle size reduction effect of the
microstrainer was simulated by maceration of an aliquot of both influent
and effluent samples in a Waring blender.
Also, it was postulated that passage through the microstrainer would so
reduce particle size that many of those cells shielded by inclusion in a
particle of suspended solids would be exposed to disinfection. To explore
the possibility that the Microstraining resulted in greater ease of dis-
infection a two-fold approach was adopted. The combined sewer overflow
both before and after Microstraining were subjected to chlorination
simultaneously in identical chambers. The coliform counts were made on the
chlorination study samples after maceration in a Waring blender to normalize
27
-------
any difference in particle size of samples before and after Micro straining,
The maceration technique recommended by Geldrich (5) and the other
analytical procedures are described in Appendix A. The apparent effects
of the maceration operation on the BOD, COD and coliform analyses are
described in the Discussion of Results, Section VII, Organic Removal by
Microstraining.
28
-------
SECTION VII
DISCUSSION OF RESULTS
Drainage Area - Rainfall_and Runoff
The 11.2 drainage area shown in Figures 7 and 8 is served by a combined
sanitary and storm sewer. Rainfall Intensity-Return Frequency curves
for the drainage area, as furnished by the Philadelphia Water Department,
are shown in Figure 9. The drainage area can be characterized by a
maximum calculated runoff coefficient of 61% using the rational method.
The dry weather sanitary flow is shown in Figure 10.
Table 1 shows the hydrographic characteristics measured for 19 storms
in 1971 and 1972. Rainfall intensities were measured at a rain gauge
located in our test drainage area, about 100 yards from the test site.
The rain gauge was a 4.8 inch dual traverse, 6 hour Universal Rain Gage,
manufactured by Belfort Instrument Co.
The flow through the slot-like regulator to the interceptor sewer was
approximately 0.35 mgd for all storms based upon the observed opening
and a calibration curve for this particular regulator furnished by the
Philadelphia Water Department. (Note that in contrast to normal practice
the interceptor flow was unintentionally 15 times the average dry weather
flow.) The excess combined sewage overflows into a concrete trough
in which are located the feed pumps of the experimental combined sewer
overflow treatment facility. During a storm these pumps feed a part of
the overflow through the treatment facility and the remainder of the
overflow bypasses the facility and falls over a combination trapezoidal-
rectangular weir directly to a creek. The trough and weir impounds
approximately 4,000 gallons.
The flow rates into the treatment facility and the flow rate over the weir
are continuously measured and recorded. The storm flow noted in Table 1
is the sum of the bypass flow plus the flow through the treatment facility
plus the constant 0.35 mgd flow through the regulator minus the dry
weather sanitary flow. The storm flow listed in Table 1 was the peak
5 to 10 minute flow observed during the indicated duration. The average
storm flow, at least during the shorter storms,during microstrainer
operation was within 0-20% of the peak flow listed. As previously
mentioned the microstrainer operated at a constant flow rate during any
storm although the overflow rate varied greatly.
29
-------
+144.5
Paved
Channel
+150.7 Intercepting
Figure 7
Outfall
67th & Callowhill
30
-------
Figure 8
Drainage Area
-------
10.0
8.0
6.0
a 4.0
o
ffi
l-l
CD
a
w
I 2.0
CO
CD
+->
C
II
i-H
r-H
iS
C
1.0
.8
.6
.4
.2
0.1
Note
Frequency Analysis by Method of
Extreme Values, After Gumbel
Return Period (Years)
[III
5 10 15 20 30 40 50 60 2 3 <
Minutes Hours
Figure 9
Rainfall Intensity - Philadelphia, Pennsylvania
1903-1951
32
-------
18001
CO
CO
Average Flow - 1000 gph
10 noon 2 4
Hours of the Day
Figure 10
Hourly Variations, Sanitary Flow, 67th & Callowhill Sts. Sewer
(City of Philadelphia Data)
10 midnight
-------
Table 1
Hydrographic Data
Date
7-29-71
7-30-71
7-30-71
7-31-71
8-27-71
9-10-71
9-11-72
9-11-71
9-11-71
9-11-71
9-13-71
9-13-71
9-21-71
9-21-71
10-10-71
10-10-71
11- 2-71
11-29-71
2-13-72
Time
Start
7:10p
Il:l5a
I2:44p
I2:l8a
ll:35a
5:34p
7:43a
9:00a
9:50a
6:l4p
I2:22p
l:41p
2:56a
3:34a
6:38a
I0:08a
7:23p
5:25p
I:l8p
Time
Stop
7:24p
I2:17p
I2:57p
I2:3la
12:36p
5:52p
8:05a
9:36a
I0:56a
6:37p
I2:30p
2:35p
3:08a
3:53a
7:37a
I0:40a
7:41p
5:47p
l:44p
5 Min
Intensity
in/hr
1.80
1.33
0.96
1.20
1.80
1.80
2.32
2.39
4.67
2.45
0.84
1.80
2.04
2.82
0.75
0.96
1.80
0.84
0.60
10 Min
Intensity
in/hr
1.68
1.14
0.72
0.65
1.50
1.32
2.10
1.65
4.37
2.04
0.84
1.20
1.20
1.95
1.03
0.66
1.44
0.78
0.48
Storm
Flow
MGD
2.3
4.4
2.8
2.8
3.2
4.8
2.9
2.8
10.7
2.8
1.5
5.0
2.5
3.1
1.9
1.5
2.8
2.4
1.2
34
-------
Table 1
Hydrographic Data
Actual
Runoff
Coefficient
.185
.530
.534
.588
.294
.500
.188
.229
.334
.185
.248
.576
.292
.220
.258
.314
.265
.422
.349
S.S.
Influent
mg/1
27
115
39
52
29
190
90
130
80
96
231
72
33
64.
751
14
357
115
208
Permeability
Parameter
gpm/ft2 x cp
-
.00164
.00127
.00051
.00116
.00086
.00158
.00258
.00066
.00133
.00335
.00091
.00042
.00124
.01252
.00101
.00465
.00192
_
BOD5
Influent
mg/1
30
260 +
N.D.
N.D.
880
820
2240
2280
1100
3300
2150
1730
1620 +
885
1910
8
55
13
_
Total Inf Inf Fecal
Coliform Coliform
million cells/100 ml
2.80
N.D.
N.D.
N.D.
0.85
0.28
8.40
8.70
2.30
0.88
N.D.
0.53
N.D.
0.46
0.20
0.90
0.32
0.57
0.07
.0740
.0010
N.D.
N.D.
.0023
.0840
.0500
.0060
.0270
.0120
N.D.
.1330
N.D.
.0030
.0068
.0170
.0040
.0470
.0640
ND - Not Determined
35
-------
The actual runoff coefficients listed in Table 1 are based upon the 5 to 10
minute peak overflow rate and the maximum 10 minute rain intensity
occurring during a period 30 minutes before microstrainer started and 15
minutes before it stopped.
Measurements of the suspended solids, EOD^, total coliform and fecal
coliform characterized the quality of the combined sewer overflow.
Calculation of the permeability of the solids after deposition on the screen
yielded an overall characterization of the type of solids, including such
particle properties as average particle size, breadth of the range of
particle sizes, the shape of particles and their rigidity.
Figure 11 shows combined sewer overflow versus the 10 minute storm
intensity. Guidelines showing runoff coefficients of 0.1 and 0.6 are
shown for convenience in interpreting the data. The line shown on
Figure 11 represents the best straight line fit of the storm flow and rain
intensity data as obtained by regression analysis .However, any factor
which influences the ability of a surface to accept rainfall (such as
rainfall pattern, rainfall intensity, geography, antecedent dry period,
etc.) must affect the runoff coefficient.
The relationship between suspended solids content and overflow rate is
shown in Figure 12. The data here is much too diffuse to draw any
conclusion. However, the possibility of high suspended solids content
occurring at high overflow is strongly indicated. No explanation is
available for the scatter of the data during these late summer and fall
storms. In the previous phase of our work during winter and early summer
of 1969, the reported (1) relationship was definitely that of increasing
suspended solids content with increasing overflow rate. In any event,
it is probable that the higher suspended solids will sometimes occur at
the higher overflow rates and the resulting high solids rate must be
considered in equipment selection.
The practice of taking discrete samples, when possible, during a storm
was adopted even though the major source of information regarding the
character of the untreated and treated combined sewer overflow was to be
a composite sample. The purpose of the discrete samples was to obtain
additional information on the change in composition of combined sewer
overflow within a storm. The storms were of such short duration that it
was generally impossible to obtain more than one discrete sample because
of the press of other duties. A single discrete sample, by itself, or in
conjunction with a composite sample could not and did not accomplish the
secondary objective of characterizing the variation in composition of the
overflow during a storm.
36
-------
7.2 --
00
w
CD
S-,
<4.8
Csl
a
o
Q
O
2.4 __
03
6
Runoff
Coefficient =
.Storm Flow =0.25 + 1.95
R = .796
Std Error of Estimate = 1.3
Runoff
Coefficient
O
= . 1
234
- Ten Minute Rainfall Intensity - Inch per Hour
Figure 11
Observed Runoff Coefficient Based Upon 10 Minute Intensity
-------
CO
CO
tn
E
I
1 i
4-1
S-i
0)
6
t~,
-------
It appears that the characterizing of the variation in composition of the
overflow during a storm cannot be achieved by a catch as catch can
schedule imposed upon an already busy program. To accomplish this end
only the discrete sampling program must be carried out on a number of
storms. It is suggested that this work be undertaken as a separate study.
The Importance of Differential
The screen area available for straining is determined by the water level
inside the drum. This level is kept at least an inch below the lip of the
backwash water hopper but otherwise as high as possible. A maximum
differential is predicted or selected and the water in the outlet chamber
is maintained, by the position of the outlet weir, at such a level that the
water inside the drum will not overflow the washwater hopper at the
differential anticipated at the design flow rate.
In conventional applications such as the straining of algae from lake water,
the flow rates are kept low to avoid shearing fragile particles through the
screen, and design differentials are about 6 inches of water (6).
'The physical conditions of our tests are shown on Table 2.
For combined sewer overflow service, the design differential used is
24 inches of water. The differential varies directly as the radial flow rate
(gpm/sq ft of submerged screen) , the viscosity of the water, the weight
of solids retained (i.e. , the thickness of the mat) and with the characteristics
of the solids. It might be noted here that the head loss of clean water at
58ฐ F through a clean Mark 0 screen at 45 gpm/sq ft is less than 5 inches
of water. The additional differential is due to the resistance to flow of
(a) the interaction of screen and solids, and (b) the mat of deposited
solids. Furthermore, the differential increased by not only an increased
mat thickness but by any decrease in the permeability of the deposit.
Such decreases in permeability (i.e. , compression of the once-formed
deposit) are the general rule in filtration of algae, etc. , as will be seen
later by the exponential form of the widely used Boucher's equation in this
practice (6) .
Differential
The prediction of differential has been achieved in the past by the use of
Boucher's equation. The wide acceptance of this relationship suggests
that it is applicable for the low concentrations of the algae-like solids at
the low flow rates where it was used. Limited tests performed in the
previous work (1) , however, indicated that the Boucher Index did not pre-
dict differential for higher concentrations of combined sewer overflow solids
39
-------
Table 2
Microstrainer Dynamic Performance Data
Date
of
Storm
7-1-71
7-19-71
7-29-71
7-30-71
7-30-71
7-31-71
8-27-71
8-27-71
8-28-71
9-10-71
9-11-71
Time
Start
7:55p
I0:33p
ll:llp
7:lOp
Il:l5a
I2:05p
I2:44a
I2:50a
I2:l8a
3:42a
5:50a
ll:35a
2:35a
3:lOa
5:34p
7:43a
9:00a
9:50a
9:50a
Time
Sto_p
8:l3p
ll:llp
-
7:24p
I2:l7p
-
I2:50a
I2:57a
I2:3la
4:32a
6:22a
I2:36p
2:48a
-
5:52p
8:05a
9:36a
I0:56a
I0:00a
Trough
Over-
flow
MGD
.94
none
none
1.4
3.6
3.6
1.9
1.9
1.9
1.4
2.3
2.3
1.4+
-
4.3
2.4
2.3
10.1
10.1
Flow-
thru
MS
MGD
.52
.52
.52
.49
.44
.44
.52
.12
.52
.47
.52
.48
.43
.41
-
.12
.12
.12
.12
.12
Vis-
cosity
c.p.
-
1.150
1.150
1.000
1.005
1.005
1.005
1.005
0.920
1.035
1.035
1.035
1.020
-
0.955
0.992
0.992
0.992
0.992
Drum sub-
mergence
% ft 2 (a)
83
82
82
81
85
85
77
77
83
82
82
78
82
-
66
59
59
62
62
7.8
7.7
7.7
7.6
8.2
8.2
7.3
7.3
7.8
7.7
7.7
7.3
7.7
-
6.2 '
5.5
5.5''
5.8'
5.8
(a) Total unmasked drum area is 9.4 sq ft of 23 micron
aperture Mark 0 screen.
40
-------
Table 2
Microstrainer Dynamic Performance Data
Differ-
ential
inch
24.0
23.5
23.5
23.0
24.0
24.0
21.0
21.0
24.0
23.5
23.5
21.0
23.5
13.5
8.0
8.0
10.0
10.0
Drum
Speed
rpm
1.75
6-7
6-7
stopped
7.00
7.00
6.00
6.00
7.00
7.00
7.00
7.00
6.70
3.25
1.70
1.50
2.70
2.70
Radial
Flow
Rate
gjjm/'ft2
46.0
46.5
46.5
45.0
37.4
37.4
50.0
50.0
42.0
47.0
43.0
40.0
37.0
13.5
15.2
15.2
14.5
14.5
Suspended Solids
Influent Effluent Removal Sample
mg/1 mg/1 % Procedure (b)
165
47
317
27
115
72
39
11
52
66
29
effluent
68
190
65
31
33
15
26
26
23
28
16
1
7
sampler
1
37
60
34
90
45
78
64
41
Incr .
69
98
76
failure
98
80
8 min auto.comp.
18 min auto.comp.
hand discrete
16 min auto.comp.
hand discrete
30 min hand comp.
hand discrete
auto, discrete
auto, discrete
10 min auto.comp.
50 min auto.comp.
20 min auto. eff.
& hand infl. comp,
@ 8:00a
64 min hand comp.
hand comp.
15 min auto.comp.
90 45 50 15 min auto.comp.
130 15 88 33 min auto.comp.
insufficient sample volume for a composite
80 36 55 10 min auto.comp.
(b) Refer to Section V for explanation of sample procedure
41
-------
Table 2
Microstrainer Dynamic Performance Data
Date
of
Storm ...
9-11-71
9-13-71
9-21-71
10-10-71
11- 2-71
11-29-71
2-13-72
Time.
Start .
6il4p
I2:22p
1:4 lp
2:56a
3:34a
6:38a
7:06a
. lG-:08a
7:23p
5:25p
I:l8p"
time
Stop
6:37p
I2:30p
2.35p
3:08a
3:53a
7:06a
7:3la
I0:40a
7:41p
5:47p
l:44p
Trough
Over-
flow
MGD .
2.3
.72
4-3
1.8
2.3
1.2
1.2
.72
1.4
10.7
.72
Flow-
thru
MS
MGD
.12
.42
.36
.35
.44
.36
.36
.41
.41
.41
.12
Vis-
cosity
c.p.
1.005
1.005
1.005
1.005
1.035
1.065
1.065
1.050
1.020
1.308
1.440
Drum sub-
mergence
% ft2(,
60
82
81
78
81
78
78
65
83
83
79
5.6
7.7
7.6
7.4
7.6
7.4
7.4
6.1
7.8
7.8
7.4
(a) Total unmasked drum area is 9.4 sq ft of 23 micron
aperture Mark 0 screen.
42
-------
Table 2
Micro strainer Dynamic Performance Data
Differ-
ential
9.0
23.0
22.5
21.0
22.5
21.0
21.0
13.0
24.0
24.0
22.0
Drum
Speed
1.90
7.00
6.80
6.10
6.50
5.50
5.50
4.00
7.00
7.00
2.0
Radial
Flow
Rate
15.0
38.0
36.5
32.5
41.0
34.0
34.0
46.5
36.4
36.4
16.8
Suspended Solids
Influent Effluent Removal
ma A ma/1 %
96
231
72
33
64
751
330
14
357
115
208
27
6
46
17
32
1
4
9
113
77
106
72
97
36
48
50
99 +
99
36
69
33
49
Sarnple^
Procedure (b)
20 min auto.comp.
8 min auto.comp.
50 min auto.comp.
8 min auto. discrete
16 min comp.
28 min auto.comp.
25 min auto.comp.
32 min hand comp.
18 min auto.comp.
22 min auto.comp.
24 min hand comp.
(b) Refer to Section V for explanation of sample procedure,
43
-------
and another index is now under study. This index shows some promise of
predicting performance from known conditions of combined sewer overflow
characteristics and environmental factors. Additional work is planned on
this new performance which we have termed "the permeability parameter".
The Boucher equation is generally stated as ln(H/Ho) - IV where V is the
volume in gallons passed per sq ft of screen. H is the differential in
inches and Ho is the head obtained by extrapolating a plot of In H vs V
back to zero V. I is a Filtrability Index which includes the effect of flow
rate used, suspended solids concentration, screen size, particle size and
other characteristics of the solids such as bulkiness, etc. In order to
illustrate the Boucher equation on Figure 13, the Volume term V is shown
as the ordinate (Ibs solids applied/sq ft of screen) divided by the
suspended solids content of the feed in pounds/gallon (mg/1 x 1/120,000).
As shown on Figure 13, the determination of Boucher's Filtrability Index
requires a laboratory measurement of the differential after passage of at
least two increments of water volume using the one water of interest at
the one flow rate of interest through a stationary section of the one screen
of interest (6). Boucher's Filtrability Index will then permit prediction
of differential at other volumes of the same water and the same flow rate.
The laboratory investigation is often extended enough to include the
"break point" where differential starts to fall. At this break point, the
shear forces due to differential reach such a level that much of the pre-
viously retained solids are forced through the screen with a sharp
reduction of removal efficiency.
The Boucher Filtrability Index, however, cannot be determined directly
from data obtained from a rotating drum. Estimates of Boucher's Index
from rotating drum data requires the use of a transcendental equation
based upon implicit assumptions which do not seem to hold for the
suspended solids in combined sewer overflow. The Boucher relation is,
therefore, unsuitable for combined sewer overflow and was not used in
this phase of the study.
We calculated the permeability parameter of the combined sewer overflow
solids in the manner shown on Figure 13. It might be noted here that the
true permeability would be a function of the head loss due to passage of
a constant flow through a uniform thickness mat of the deposited solids
only, corrected for viscosity, and would not include the solids-screen
interaction resistance. The elimination of the effect of solids-screen
interaction was not possible with our shallow deposits (approximately
0.003 inch). Nor could the effect of changing (increasing) thickness
of deposited solids during a rotation cycle be excluded. Also the effect
of varying flow rate through an increment of screen area during a
44
-------
flD
i-t
ฃ
0)
S HoJ,
Hcs-
Boucher Equation
In H/Ho = IV where V = f solids/ft2
SS in feed
Break point
"H-Hcs = #solids/ft2
centipoise density of solids
gpm/ft2
permeability parameter
x
ID Resistance to flow of test
water at flow rate extrapolated to
zero volume on log scale
Resistance of clean screen to flow
of filtered water at flow rate
Screen
interaction
resistance
Clean
screen
resistance
Weight Solids (Dry Basis) Applied per sq ft Screen
Figure 13
Differential vs Applied. Solids
(Permeability Parameter & Boucher Index)
45
-------
submergence cycle could not be excluded. The permeability parameter,
then, is the average flow rate in gpm/ft^ of submerged screen at which
the total head loss less the clean screen loss (corrected for viscosity)
is one inch of water per inch of maximum or final deposit thickness .
The thickness of the deposit was calculated as follows:
Thickness = (flow-qpm) (influent SS-mg/1 or lb/120,0.00 gal.) =
(drum area - ft2/rev.) (drum speed - rpm) (bulk density)
Ib solids applied/ft2 _
bulk density lb/ft 3
ft
The bulk density of the heavier combined sewer overflow solids is expected
to range from 60 for clay to 110 Ib/ft3 for sand; other components as leaves,
etc., will be lighter. A single measurement of our combined sewer overflow
solids showed 85 Ib/ft3 and a bulk density of 60 lb/ft3 was used in
calculating the permeabilities in Table 1. It might be noted that other sorts
of wastewater solids are much bulkier. For example: activated sludge
solids range from 1 to 0. 25 lb/ft3 settled solids as determined by the
Imhoff Cone Analysis. Hydrated alum floe has been reported (7) to be
less than 0.2 lb/ft 3 settled solids and algae is probably as light. The
bulk density of the retained solids on the screen can be approximated by
a measurement of settled solids density similar to Sludge Volume Index
(SVI) and is equal to 62.4/SVI. Sludge Volume Index (SVI) is the ratio of
the particulate solids content as measured by settling in an Imhoff Cone
and reported as ml/1 to the weight of suspended solids reported as grams/1.
A survey was started late in this phase of the study and will be continued
in the next phase to collect information on the bulk density of various
solids found in combined sewer overflow. The calculated permeability
parameters for each storm are shown on Table 1, Hydrographic Data.
The permeability of the deposit, of course, determines the permissible
flow rate through the microstrainer for any water with its suspended solids
content within the established maximum differential and drum speed
limitations. Higher permeability permits higher flow rates and/or operation
at lower drum speeds which should promote better solids removal.
The range of permeability calculated was 0.005 to 0.0005, a 10 to 1 range.
The permeability parameter in our units of (average gpm/ft2) x (centipoise)
is numerically equivalent to 70 Darcy permeability units as defined by
the American Petroleum Institute.
46
-------
Permeability as an Index of the
Character of the Particulate Solids
Permeability by definition combines as a single value the effect of all
the properties of the individual particles and of the group of these particles-
on their resistance to flow. Also, we found empirically that particle
retention by the screen increased at higher permeabilities. It can be
inferred that many of the particle properties which determine permeability'
also influence particle retention as well.
The relation of permeability to those particle properties of interest in the
liquid-solid separation techniques used in water supply and wastewater
treatment forms the basis of the Membrane Refiltration Method of Coagulant
Testing (8).
In the next phase of this study, the Membrane Refiltration Method has
been selected for evaluating the effect of polymer addition and other pre-
treatments of combined sewer overflow prior to a microstrainer.
The characteristics of solids which govern permeability of deposited solids
are expected to be important in further analysis of past and present work
and in evaluating the effect of polymer addition in future work. It should
be informative, then, to consider the art relating to flow through porous
and granular beds.
The reported (9) equations for pressure loss through granular media in our
range of interest (Reynolds Number 1) indicate that pressure loss is a
function of the average particle size, the particle shape, the depth of the
media, and the void volume of the solids which in turn is governed
primarily by the breadth of the range of particle sizes.
Specifically, then, permeability would be expected to vary directly as i
(average particle diameter)2, (I/a shape factor)2, and a complex function
of void volume. The permeability parameter can be expected to be
influenced by other factors as well; for example, oil in liquid, etc.
The 10 to 1 range of permeabilities observed in Table 1 could be explained
by a 3 to 1 range in average particle size.
Shape factors range from 1.0 for spheres'to 1.8 for very angular particles,
and thus alone could account for a 3 to 1 range in the permeabilities
observed.
47
-------
The void volume of beds of a narrow size range of any average size
particles, if generally spherical shape, is about 40-50%. The broader
size range reported in another combined sewer overflow study (10) could
reduce the static void volume to approximately 25%. Under flow conditions,
the void volume could be further reduced by distortion of individual
particles due to compressive forces. Figure 14, however, indicates that
our combined sewer overflow solids, during this period, were rigid, and
permeability did not decrease with increased compressive force due to
increased flow rate, even at the high (24 inches of water) differentials
used.
Shown on Figure 14 for reference are typical Permeability-Differential (i.e. ,
Flow Rate) relationships for compressible and non-compressible filter
cakes as obtained from Perry's Handbook of Chemical Engineering. A
regression analysis showed that the relationship of permeability to flow
rate was truly random; that is, the permeability was independent of the
radial flow rate as in the case of incompressible solids.
The exponential rise in differential shown on the left-hand curve in
Figure 13 with increased deposit thickness of deformable, algae-
like solids extensively reported by Boucher (6) indicates decreasing
premeability.
Permeability of the solids on the mat increased with increasing concen-
tration of suspended solids in the combined sewer overflow as shown in
Figure 15. This relationship, if confirmed by further work, will help
define the limiting condition when a microstrainer is applied to high
solids waters at high flow rate.
Suspended Solids Removal by a Microstraininq Unit
The suspended solids removal performance of the microstrainer is shown
in Table 2. The analysis of the first storm of July 1 was made in the field.
One or more of the samples during the storm of August 27 at 5:50 a.m. and
the storm of August 28 were collected from stagnant contents of the flow-
through sample chamber some 5 hours after the storm and, therefore, these
three results are not considered in this discussion.
Figure 16 shows that there is no uniform percentage removal nor is there
a simple relation between removal efficiency and influent solids
concentration.
Figure 16 illustrates the capability of a microstrainer to treat combined
sewer overflow; i.e. , it is able to reduce combined sewer overflow
suspended solids from any level down to generally 40-50 mg/1 and lower.
48
-------
CD
-ooaj-
CO
c!
a
O
X .006--
a
tn
I
u
0)
4-J
CD
s
(0
13
a,
.004-
a -002f
td
CD
Incompressible Solids
^Compressible
.Solids
O
O
O
10 20 30
Radial Flow Rate - gpm/ft^ Submerged
Figure 14
Permeability Parameter vs Radial Flow Rate
H-
40
-------
en
O
=0.00003 + .000013 (SS)
R=.967
Std Error of Estimate = .00075
100 200 300
Suspended Solids in Combined Sewer Overflow - mg/1
Figure 15
Permeability Parameter vs Influent Suspended Solids
-------
Cn
s 80
X
CD
<+-<
CD
60 --
en
o
!-i
O
40 --
d
-^
i
o
to
ra
0
X
i I I
CD
CD
0,
X
X
x x
X
x x
x x
X
100 200 300
Suspended Solids in Combined Sewer Overflow (Microstrainer Influent) - mg/1
Figure 16
Effluent Suspended Solids vs Influent Suspended Solids
-------
This is shown in another way by Figure 17 . This figure shows that the
percentage removal as well as the absolute removal increases with high
levels of influent suspended solids. Due to the rapidly changing flow
rate and character of combined sewer overflows (as compared to domestic
wastewaters), the conventionally used percentage removal is an unrealistic
parameter for considering both the benefit of any treatment and the
Microstraining technique. Fortunately, it is during the intense storms when
the potential pollution at the outfall is the greatest that the percentage
removal is also the greatest. It seems that a properly sized microstrainer
will then limit the pounds of suspended solids per unit time entering the
stream from a combined sewer outfall, and this is the ultimate criterion.
Mechanism of_Suspended Solid_s Removal
The actual mechanism of suspended solids retention by a fine screen has
been studied by papermakers (11) (12). They find that most of the first
increments of solids applied to the screen will leak through, but those
solids which are retained form a mat which in turn will reduce the leakage
of solids from subsequent increments of applied liquid with its suspended
solids.
Figure 18 from Abrams (11) shows the results of a test where successive
increments of a suspension of fiber-type solids were applied normal to a
stationary, fine, rectangular mesh screen. The initial, or bare screen,
retention (shown as tan 0) is the dependent variable which determines
overall retention for any loading. The initial retention is calculated
graphically by an incremental loading plot such as in Figure 18. Also,
the initial retention is the dependent variable that Abrams (11) found could
be best correlated by independent variables as flow rate, fiber length and
screen opening size, etc. As the cumulative mass of fibers reaching the
screen is increased, the cumulative percentage retained increases; i.e.,
the slope of the curve approaches the indicated 45ฐ line asymptotically.
In our work, we collected a composite of the entire volume of water
approaching and leaving an increment of screen area during a submergence
cycle. That is, we determined only a single point (shown as an open
circle on Figure 18 on the curve during any one storm. And since the
character of our suspended solids apparently varied substantially from
storm to storm, as well as during a storm as indicated by the differing
permeabilities in Table I, our points do not fall upon a single curve but
rather each point identifies one of a family of curves.
Figure 19 indicates that our combined sewer overflow suspended solids
on a moving microscreen behaved similar to paper fibers on a paper forming
screen. Thus with combined sewer overflow solids as well, the average
removal during a drum rotation cycle is determined generally by the initial
52
-------
en
GO
CD
C
2 i.o J.
8
o
CD
C
-i-l
(0
(->
CD
o
to
CD
T3
fi
CD
Q,
to
m
O
O
-in
-M
O
2
.6 ..
.4 ..
.2
X
X
X
X
X
X
X X
X
100 200 300
Suspended Solids in Microstrainer Influent - mg/1
Figure 17
400
Fraction Solids Retained vs Influent Suspended Solids
-------
Cn
I
T3
CD
-rH
(0
CD
.i
O
O
0
o
CD
H1
0
& I
3 ซ
3 2>
o ฃ
co "d
U_l ^
o Q
W T3
S 1
o
0
0
2
CD
w
O
CO
M-H
o
o
a
0
0
ra
0
S
0
o
G
Incremental Application of Solids (after Abrams)
Application of Solids During a Submergence Cycle
Mass Solids Applied ^
Figure 18
Typical Plot of Mass Retained vs Mass Applied
-------
tNI
o
o
o
w
X!
30 _
20--
O
CQ
T5
0)
fi
-rH
fO
10..
O
10
Applied Solids lbs/1000 ft2
Figure 19
Combined Sewage Solids Retained vs
Combined Sewage Solids Applied
20
55
-------
retention. However, from a single point on a removal curve it is possible
to only estimate the initial retention. The initial retention parameter is no
doubt a function of (a) particle size relative to screen opening size, (b)
in the case of rod shaped particles, of the orientation of the approaching
particle relative to the direction of flow of the water, (c) the radial velocity
of the water relative to the tangential velocity of the screen; i.e., the
(tangent of) the angle of approach of the water relative to the plane of the
screen, and (d) in the case of the twill mesh microscreen, to the direction
of the openings relative to the plane of the screen. Other factors such as
surface characteristics of particles may also influence retention.
In any event, the greater the weight of solids applied per sq ft of exposed
screen the greater the cumulative percentage removal as shown by the
stylized dashed line in Figure 19.
The weight of solids applied per sq ft of exposed screen is:
(flow-through microstrainer-qpm) x
(drum area - ft2/rev)
(suspended solids in feed-mg/1 or lb/120,000 gal.)
(drum speed - rev/min)
This explains the relationship wherein the percentage removal increases
with flow rate and as the concentration of suspended solids increase.
It has been previously reported (13) that the lowest drum speed obtainable
within the head loss (differential) limitations of the machine yielded the
best removal. It might be noted that the slower the drum speed, with all
other conditions being equal, the greater the weight of solids applied per
sq ft of screen. Thus, we too found, as shown in Figure 19, higher
removals at slower drum speeds.
The increase in percentage solids removed caused by higher solids con-
centration observed at our site is shown by Figure 17. The effect of higher
solids concentration is due to the contribution of more solids of any sort
being applied to the screen area. Also, it is believed that the additional
solids in the higher concentrations at our site are of a more strainable
nature; e.g., larger size.
There are indications (6), however, that flow rate per se, in contrast to
its desirable contribution to greater weight of solids applied per unit screen
area causes per cent removal to decrease. Flow rate causes shear forces
which tend to rupture the mat of deposited solids and force it through the
screen. This undesirable effect of flow rate is not nearly so pronounced
56
-------
on the removal of combined sewer overflow solids as with fragile solids
like algae or alum floe.
The factors which influence the retention of suspended solids in combined
sewer overflows (and other waters) need to be more closely identified.
A better understanding of these factors might permit a microstrainer to be
applied to a broader range of wastewaters.
In the previous section the flow rate, head loss, and other measurements
were reduced to an overall permeability parameter which characterized
the solids in relation to their resistance to flow. This flow resistance
parameter was empirically found to be loosely related to solids retention
performance as well (see Figure 20). Regression analysis showed that the
relation was not linear.
It has been shown (9) that the resistance to flow (i.e., I/permeability)
through beds of sand and similar particles is predictable from known
particle properties. As will be shown later, these properties can be
related to the smallest dimension of the individual particles. Since initial
solids retention by a screen is determined by the smallest dimension of a
particle, it is not surprising that solids retention is related to permeability.
Filter Head Loss Equation Analogy (9)
Solids
Retention
Permeability
Parameter
I ^_ (avq particle dia)^ (void vol) j
J Ov~" (shape factor) 2 (viscosity) J
Obviously- the smaller the average diameter of a group of particles, the
smaller the smallest particle will likely be, and the less likely the small
particles will be retained initially by a screen.
The shape factor is a property of an individual particle and is numerically
1.0 for spheres and 1.8 for very angular or rod-like particles. An
individual particle with a high shape factor has a smaller dimension
relative to its nominal diameter than a spherical particle and would be
less likely to be retained by a clean screen.
Void volume is a property of a group of particles . It is determined chiefly
by the breadth of the range of nominal particle diameters with the narrower
size range making for larger void volumes. The shape of the particle has
relatively little influence. A low void volume condition would indicate
a broad range of individual particle sizes. A broad range of sizes would
require that the smallest particle be smaller than the smallest particle in
a group of particles of the same average size but of a narrow size range.
57
-------
ง !-ฐ -
-i-i
G
2
o
-iH
^ .8 .
ฃ
0
c
rH
(0
-M
0:1 . 6 -
3
"o
CO
0
0 .4 -
Q.
w
CO
0
rH
1 1
D,
51 .2 -
m
O
c
0
-H
2
5-1 n
PL, 0
0 ฐ
0
o
0 ฐ
o
o
o
8 ฐฐ
o
o
., 1 I ........ i. , ... i L
.002 .004 .006 .008 .01
Permeability Parameter - gpm/ft^ x Centipoise
Figure 20
Fraction Solids Retained vs Permeability Parameter
58
-------
The smaller particles would be less likely to be retained by a clean screen.
In the next phase of this study, an extensive laboratory study will survey
the effect of polymer addition and other pretreatments on the removal of
solids by amicrostrainer. These pretreatments will be evaluated in laboratory
glassware by means of a Membrane Refiltration Test (8) which is essentially
a permeability measurement. Also, additional storms will be monitored in
full scale equipment, as before, without any pretreatment and Membrane
Refiltration measurements will be made on the actual stormwater. It is
expected that these measurements will reflect full scale performance so that
this test can be used to predict microstrainer performance.
Organic Removal by a Microstraining Unit
The organic matter in combined sewer overflow, as in other natural waste
streams, is made up of many compounds at varying concentrations. At the
arbitrarily selected conditions of the test method (e.g., 5 day incubation
period at 20ฐ C in the case of the BOD5 method), certain fractions of many
of these organic compounds will be sensed. In addition to organic matter,
other reducing (oxygen demanding) compounds are sensed to some extent
by the BOD5 and COD tests. In particular, ammonia, or organic nitrogen
forms which are frequently present in substantial concentrations, exhibit
an oxygen demand. The duration and other conditions of the 5 day BOD
tests are designed to exclude these nitrogen forms as much as possible.
Another measurement of organic matter, TOG (total organic carbon) senses
only true organic (carbon containing) compounds and thus furnishes an
indication of the total organic material present.
The possibility of residual methanol in the sample bottles rendered many
of the earlier samples suspect from the standpoint of interpreting organic
removal performance. These storms, and also the storm of August 28 for
the reason noted in the Suspended Solids Section, are not considered.
Table 3 shows the non-suspect data on organic removal performance of
the microstrainer. All of the data, including the suspect data, is shown
in Appendix B.
Table 4 summarizes the change in organic content on passage through
the microstrainer. As noted in Appendix A, measurements of the standard
deviation of the precision of the analytical results at Northeast Laboratory
of the City of Philadelphia showed ฑ 9% for BOD measurements, + 2% for
TOC and +5.7% for Suspended Solids measurements. The 13th Edition of
Standard Methods indicate deviations of + 6.5% for COD, ฑ6.5% for
Suspended Volatile Solids and + 5% for Dissolved Volatile Solids. In
Table 4 all changes of less than the deviation noted in the table are listed
as instances of no change.
59
-------
Table 3
Organic Removal Performance
Date
of
Storm
7-29-71
7-30-71
8-27-71
9-13-71
10-10-71
11-2-71
11-29-71
2-13-72
Radial
Flow
Rated
45.
45.
45.
37.
37.
37.
40.
40.
40.
36.
36.
34.
34.
34-
46.
46.
46.
36.
36.
36.
36.
36.
16.
16.
16.
0
0
0
4
4
4
0
0
0
5
5
0
0
0
5
5
5
4
4
4
4
4
8
8
8
s.s.
Infl
mg/1
27
27
27
115
115
72
29
29
29
72
72
330
330
330
14
14
14
357
357
115
115
115
208
208
208
SVS
In Out
mg/1
22
-
-
29
-
18
16
-
-
24
-
81
-
-
7
-
-
76
-
31
-
-
49
-
-
12
-
-
4
-
7
6
-
-
17
-
3
-
-
4
-
-
48
-
23
-
-
30
-
-
DVS
Removal In Out
% mg/1
45
-
-
87
-
69
63
-
-
29
-
96
-
-
43
-
-
36
-
26
-
-
39
-
-
34
-
-
18
-
-
10
-
-
ND
-
ND
-
28
ND
-
11
30
ND
ND
-
17
52
-
-
67
-
-
16
-
-
17
-
-
ND
-
ND
-
39
ND
-
16
32
ND
ND
-
18
19
-
-
Removal
%
Incr
-
-
15
-
-
Incr
-
-
-
-
-
-
Incr
-
-
Incr
Incr
-
-
-
Incr
63
-
-
ND - Not Determined
gpm/ft2
60
-------
5 Day BOD
Table 3
Organic Removal Performance
COD
TOG
In Out
mg/1
30
28
20
260+
348
ND
880
940
1180
1730
2150
129
112
106
8
7
7
55
69
12.
13.
-
6
5
2
30
28
ND
252+
256+
ND
550
440
540
1740
1250
112
98
124
7
7
7
26
34
5 26
5 27.5
-
16
15
8
Removal
%
no change
no change
-
-
_
-
37.5
53.0
54.0
Incr
42.0
13.2
12.5
Incr
12.5
no change
no change
52.5
51.0
Incr
Incr
-
Incr
Incr
Incr
In Out
mg/1
79
89
-
2350
2160
-
3760
1410
-
3930
4460
256
280
-
32
384
-
230
250
117
110
-
81
68
68
108
-
6270
10200
-
1830
990
-
3390
3570
216
248
-
200
416
-
170
180
117
61
-
76
41
Removal
%
14-0
Incr
-
Incr
Incr
-
51.5
30.0
-
13.5
20.0
15.6
11.4
-
Incr
Incr
-
26.0
28.0
no change
45.0
-
6.0
40.0
In Out
mg/1
ND
39
-
ND
864
-
ND
409
-
ND
1065
ND
78
-
ND
10
-
77
ND
ND
23
-
ND
24
-
ND
29
-
ND
3154
-
ND
224
-
ND
810
ND
54
-
ND
10
-
38
ND
ND
22
-
ND
21
-
Removal
%
25. Oa
c
_
Incr a
_ b
_
45- Oa
c
-
24. Oa
-
30. 8a
c
-
no change3
c
51.0
a
-
4.0a
c
_
12.53
c
Sample Procedure
amacerated composite
'-'discrete
cfiltered
ND - Not Determined
61
-------
Change in Content
of Organic Matter
by Microstraining
No Changes
Number of Instances
Table 4
Effect of Microstraining on Organic Matter
as Indicated by Several Test Methods
Total Organic
BOD5
COD
TOG
unmacer- macer- unmacer- macer- macer-
ated ated ated ated ated
Dissolved
Organic
DVS
Solid
Organic
SVS
BOD5
f i Itered
macer- macer- unmacer-
ated ated ated
Increase
Number of Instances
Average of Increases mg/1
Average of Increases %
Decrease
Number of Instances
Average of Decreases mg/1
Average of Decreases %
Deviation of
Precision of Test
2
12
138
2
180
45
2
12
126
4
380
46
2
2050
345
5
516
24
2
4010
145
5
290
33
1
2300
165
6
86
32
4
14
63
1
33
63
1
6
300
1
640
54
0
0
0
10
20
53
+ 9%*
+ 6.5%**
+ 5%** + 9%* + 6.5%**
All changes in concentration upon passage
through microstrainer of less than the
deviation are considered no change.
*Deviation of precision determined this work.
**Deviation of precision predicted by Standard Methods.
-------
With this assumption, the data in Table 4 indicates rather clearly that any
changes in total organic matter were in the direction of a significant
reduction of some 25-45%. The reductions in total organic matter are some
200 mg/1 BOD5, 400 mg/1 COD and 100 mg/1 TOG in magnitude. The
extraordinary high average increases shown are heavily influenced by the
single storm of 7/30/71.
The change in suspended (solid) organic matter on passage through the
microstrainer was in the direction of a decrease in every case. As shown
in Table 4, the decreases in Suspended Volatile Solids (SVS) averaged
21 mg/1 or 51%. Also, as shown in Figure 21, the reduction in SVS
paralleled the reduction in Total Suspended Solids (SS) . The basis for the
SVS test is so dissimilar from the bases of the BOD, COD and TOG tests
that no absolute comparison can be made. However, the magnitude of the
total organic reduction, as indicated by all tests except the Ultimate BOD,
is 5 to 20 times the suspended organic reduction as measured by the SVS.
The magnitude of this difference suggests, we believe erroneously, that
some forms of organic matter other than suspended organic are removed
by the microstrainer.
The change in dissolved organic matter on passage through the microstrainer
is shown in Table 4 by the Dissolved Volatile Solids (DVS) and Filtered
BOD5 tests.
No good explanation is available for the four instances of increase in
soluble organic matter indicated by the DVS test. The increases were,
however, small in magnitude, averaging 14 mg/1, and are perhaps due to
poorer than anticipated precision of the DVS test.
The filtered BOD5 data shows two instances of no change, one very small
increase, and one instance of a large decrease.
It seems from the above that there is little if any change in the soluble
organic matter on passage through the microstrainer.
The ultimate BOD and rate constant data from non-suspect, unmacerated
and macerated samples are shown on Table 5. These data follow the same
pattern as that previously discussed for BODs, COD, and TOG.
One storm showed practically no change (0 to 2 mg/1) on both unmacerated
and macerated samples. Two storms showed small increases (9 to 16 mg/1)
on both unmacerated and macerated samples. The other two storms showed
decreases (32 to 80 mg/1).
63
-------
i.u -
_, Q
TJ . o .
CD
C
-i-t
(0
CD
CO
.^j
3 -6 -
CD
fi
CD
a
W
co .4 -
CD
1 1
(->
"o
>
1 -2 -
t)
(0
1-.
PH
0
w
-------
Table 5
Ultimate BOD and Oxidation Rate Studies
CD
Cn
UNMACERATED SAMPLE
Ultimate BOD Rate Change
Out Change In Out Change
mq/l-day
MACERATED SAMPLE
Ultimate BOD Rate Change
In Out Change In Out Change
mq/1 % mq/l-dav %_
7/29
10/10
11/2
11/29
2/13
220
12
65
14
14
140
10
33
30
27
- 36%
- 15%
- 49%
+ 100%
+ 93%
.290
.327
.433
.090
.260
.234
.390
.195
- 11%
- 28%
- 10%
+116%
220
17
86
18
16
150
17
51
30
25
-31%
0%
-43%
+75%
+56%
.800
.237
.322
.092
.
.730
.176
.447
.145
- 9%
-26%
+14%
+58%
-------
The sample maceration technique is strongly recommended for any further
work on combined sewer overflow. On the storms showing a decrease in
organic matter across the microstrainer as measured by BODs and COD
the decrease shown was always higher with macerated samples. This is
attributed to the character of the suspended and colloidal organic matter
in combined sewer overflow and the apparent effect of maceration of the
influent sample.
It is believed that maceration renders the organic matter more available
to the analytical procedure.
It has been theorized that the same affect can be produced across a
microstrainer and is due to the very intense shearing (estimated G =
100,000+ sec"1) for a fraction of a second (.001 sec) caused by passage
of water through the mat of deposited solids on the microscreen. For
comparison, in the laboratory maceration process, the sample is subjected
to intense agitation (approximately 60,000 sec"1) for 60 seconds.
The possible magnitude of this increase in availability of suspended
organic material to oxidation due to shearing is indicated by the difference
in measurements between the laboratory macerated sample and the un-
treated sample. The laboratory macerated influent samples averaged 17%
more BODs than the untreated influent samples and generally the same or
perhaps 5% more COD. See Figures 22 and 23. The actual extent of the
effect of the Microstraining unit's operation on availability is indicated
by the fact that macerated effluent samples averaged only 14% more BOD5
than the unmacerated effluent sample. That is, the microstrainer operation
appeared to make available (to the BOD5 test procedure) some 18% of the
unavailable suspended solids organic matter in the combined sewer over-
flow. The effect of maceration of microstrainer effluent samples prior to
COD analysis is uncertain.
The rate constant is also directly related to the availability of organic
matter to an oxygen supply. This availability is increased by higher
particle surface to volume ratio. A tendency was noted for increased rate
constant with more severe straining conditions (i.e. , higher differential,
higher radial flows, etc.). Similarly, the ultimate demand exhibited a
pattern of decrease with more efficient removal of SVS.
Again, these observations did not hold in all cases. However, on the
average, the effect of a Microstraining filter under high differential, high
radial flow conditions was towards increased rate and decreased ultimate
BOD values .
Laboratory maceration has the same effect on BOD rate constant and
66
-------
1000 L_
en
e
i 500
CD
i
a
S
co
CO
T3
0)
)->
s
CD
O
(0
100
o LVJU
LO
Q
O
CQ
50
10
ฎ Influent
Microstrained
50 100 50
BOD5 on Untreated Sample - mg/1
Figure 22
vs BOD5 on Macerated Sample
1000
67
-------
5000
tn
S 1000
I
CD
1
a
S
(0
-O
0>
+j
2
CD
O
fO
Q
O
O
A Influent
A Microstrained
500
100
100 500 1000
COD of Untreated Sample- mg/1
Figure 23
COD vs COD on Macerated Sample
68
-------
ultimate value. The effect is more pronounced on influent samples than
on the effluent due to the maceration a sample is given by the micro-
strainer process. The fact that there were differences in all cases supports
the recommendation that samples be macerated prior to analysis.
In general it seems that the sample maceration pretreatment and the more
reliable, chemically determined, COD and TOG analytical methods
yielded the most consistent results. The rapidly changing and variable
average level of contaminants, together with the inherent time lag in the
method, make the BOD and related biological tests inappropriate for
combined sewer overflow service.
The maceration technique, as used here, has been useful and its use is
strongly recommended for all combined sewer overflow organic analyses
as well as those on other similar wastewaters .
The need for refinement of this technique, particularly when used prior to
bacteria counts, will be discussed later.
Coliform Removal by a Micros training Unit
The coliform data which are not suspect because of possible methanol
contamination are shown in Tables 6 and 7. The complete data including
suspect samples are shown in Appendix C and Appendix D.
Coliform counts were performed by the 12th Edition of Standard Methods.
In this method there are several difficulties. First, the coliform count
of combined sewer overflows varies unpredictably, as shown on Tables
6 and 7 over several orders of magnitude. It is difficult, then, to
preselect the proper sample dilution ratio in order to arrive at a countable
concentration of cells, which is a narrow range. Thus, as shown by
footnotes on Tables 6 and 7, the majority of the counts were made on
dilutions well outside the ideal counting concentration range. Second,
the diluted sample is passed through a very fine filter which entraps all
sorts of bacteria in addition to the type of interest; i.e. , fecal coliform.
The membrane filter with entrapped bacteria is then placed in an
incubating broth whose composition as well as the incubation temperature
and duration are selected to foster the growth of only the single type of
bacteria of interest. The actual count made is of the colonies produced
by each surviving cell. Thus, a clump of many cells in the original
sample would produce a single colony and would be counted as a single
cell. To reduce the possibility of clumping and of cells isolated from
nutrient solution by a solids barrier, an aliquot of the sample was sub-
jected to intense maceration in a Waring blender. Counts were made
upon both the macerated and the untreated samples.
69
-------
Table 6
Total Coliform Removal by Microstraining
Date of
Storm
7-29-71
7-30-71
8-27-71
9-13-71
10-10-71
11-2-71
11-29-71
2-13-72
Radial Flow
Rate gpm/ft^
45.0
37.4
40.0
36.5
34.0
46.5
36.4
36.4
16.8
S.S.
Influent
mg/1
27
115
29
72
330
14
357
115
208
Temp
OF
69
68
68
68
64
65
67
50
45
PH
(1)
7.1
6.9
7.1
7.3
7.2
6.7
7.7
7.5
7.5
7.3
7.4
7.3
7.0
7.5
Across Microstrainer
Influent
Cells/100
2.800
3.300
0.900
0.870
0.850
0.730
0.530
0.780
1.400(2)
2.100
0.900
1.100(2)
0.320
0.760
0.570
0.340
0.005
0.004
Effluent
ml x 1Q6
0.001(3)
0.180
0.130
0.880
0.092
0.550
0.400
4.200
0.004(2)
0.008(2)
0.050
1.360(2)
0.800
0.680
0.780
0.850
0.070
0.036(2)
Removal
%
99 +
94.5(4)
85.6
no change (4)
89.2
24.7(4)
24.5
Incr(4) (540%)
99 +
99+
94.4
Incr(4) (123%)
Incr (250%)
10 . 0 (4)
Incr (134%)
Incr (4) (250%)
Incr (1400%)
Incr (900%) (4)
(1) pH of filter influent
(2) Estimated count based on non-ideal colony count
(3) Zero plate count based on 0.1 ml sample
(4) Values are results on analyses of another aliquot
of same composite after intense maceration'
-------
Table 7
Fecal Coliform Removal by Microstraining
Date of Radial Flow
Storm Rate gpm/ft2
7-29-71
7-30-71
8-27-71
9-13-71
10-10-71
11-2-71
11-29-71
2-13-72
45.0
37.4
40.0
36.5
34.0
46.5
36.4
36.4
16.8
s.s.
Influent
mq/1
27
115
29
72
330
14
357
115
208
Temp
OF
69
68
68
68
64
65
67
50
45
PH
(1)
7.1
6.9
7.1
7.3
7.2
6.7
7.5
7.5
6.9
7.3
7.4
7.3
7.0
Across Micros trainer
Influent Effluent
Cells/100 ml x 106
0.0740
0.0610
0.0010
0.0010
0.0023
0.0220
0.1330(2)
0.0018(2)
0.0270
9.2000
0.0170
0.0530
0.0044
0.0088
0.0470
0.0480
0.0600
0.0140
.0010(4)
.0020(2)
.0001(3)
.0008
.0030(2)
.0060(2)
.0190
.0470
.0001(3)
.0001(3)
.0200
.0450
.0270
.0150
.0520
.0780
.0640
.0150
Removal
%
98.7
97.0(5)
90.0
20.0(5)
Incr (130%)
73.0(5)
85.7
Incr (5) (2 600%) (2)
99.6
99+(5)
Incr (118%)
15.0(5)
Incr (670%)
Incr (170%) (5)
Incr (110%)
Incr (164%) (5)
Incr (107%)
Incr (107%) (5)
(1) pH of filtered influent
(2) Estimated count based on non-ideal colony count
(3) Zero plate count based on 1 ml sample
(4) Zero plate count based on 0.1 ml sample
(5) Values are results on analyses on another aliquot
of same composite after intense maceration
-------
As can be seen, either clumping and/or occlusion of cells in solids could
have occurred (Figures 24 and 25). The macerated sample compared to the
untreated sample showed higher counts in 25 cases, a lower count in
10 cases, and the same count in one case. The increase ranged from 10
to 100%.
The effect of a Micro straining unit is shown by the fecal coliform data
(neglecting changes of 0.001 million cells/100 ml and changes of less
than 10%) by the standard count: three instances of no change, three
increases, and three decreases. Similarly, comparison of the micro-
strainer influent and effluent samples after maceration showed: two
instances of no change, three increases, and four decreases.
On the same basis, the total coliform data showed by the standard count;
no instances of no change, three increases, and six decreases.
Similarly, the macerated samples showed: two instances of no change,
four increases, and three decreases.
The conclusion regarding coliform removal drawn from the data is that
there is no significant reduction across the microstrainer.
Disinfection of Combined Sewer
Overflows by Chlorination
The extent of disinfection by chlorination has been shown to be a function
of chlorine residual, pH, temperature and contact time (14). The first
three of these parameters pertain to the chemical form and activity at
which the disinfecting agent exists. Contact time determines the extent
of contact made possible by the equipment in use.
The specific effect of some of the individual parameters, especially
contact time and residual chlorine, has been shown in many studies.
Until recently, however, there has been little information of the effect
of mixing intensity. Collins, et al (15) and Kruse et al (16) have clearly
shown the magnitude of the influence of mixing intensity by the comparison
of bacterial kills with two mixing intensities. In Collins1 work using
identical conditions - 5 mg/1 chlorine residual, 1 minute contact time,
pH 7, and 15ฐ C temperature - a tubular reactor yielded a survival ratio
of 6 x 10~4 while a mechanically stirred reactor yielded a ratio of
1 x 10~ *. In Kruse's work, the effect of a flash mixing compared with
gentle stirring at 5 mg/1 chlorine dosages and at 1 minute total contact
at pH 7 yielded survival ratios of 1 x 10~2 and 5 x 10"* respectively.
Also, it might be noted here that a conventional chlorine contact unit
under similar conditions might yield a ratio of 9 x 10"1 if, indeed, the
chlorine containing chemicals were dispersed at all in 1 minute.
72
-------
o
o
i-H
w
tI
II
CD
O
c
o
I
0)
I I
a.
S
(0
CO
s
CD
o
(0
fi
o
I
(0
o
0)
,05
01
1.005
,001
o Influent
O Microstrained
.001 .005 .01 .05 .1
Fecal Coliform Count on Untreated Sample - Million Cells/100 ml
Figure 24
Fecal Coliform Count on Macerated Sample vs
Fecal Coliform Count on Untreated Sample
73
-------
o
o
0>
O
a
S
to
CO
T3
CD
+J
S
CD
O
ro
3
O
O
o
O
.1
,05
,01
o
ป Influent
O Microstrained
.01 .05 .1 .5 1
Total Coliform Count on Untreated Sample - Million Cells/100 ml
Figure 25
Total Coliform Count on Macerated Sample vs
Total Coliform Count on Untreated Sample
74
-------
Conventional chlorination is typically carried out in units with residence
times of 15 to 30 minutes with a residual of 1/2 to 2 mg/1 available
chlorine. Such equipment requires high capital costs in land and concrete
especially when provided for the very high flow rates of combined sewer
overflow service. They are often inefficient in providing contact due to
poor mixing and the likelihood of short circuiting. Operating costs are
often magnified because higher chemical dosages are required to bring
'bacterial counts within acceptable standards. For the purpose of dis-
infecting combined sewer overflow, the provision of the customary 30
minutes contact time is especially inconsistent in combination with a
micros'trainer having 2 minute residence time.
Also as pointed out by Field (21), the temperature of combined sewer
overflow will likely be considerably hotter or colder than the influent to
a sewage plant. The lower temperature will reduce the rate of reaction
and require much higher chlorine residuals, and/or much longer equivalent
contact time. Substitution of more mixing intensity or more dosage will
be especially important in low temperature conditions.
The development of design criteria for short duration chlorine contact
processes is without precedent. However, in the related field of water
flocculation, where a high degree of contact must be obtained in short
mixing times, a design procedure has been developed which is gaining
acceptance in the industry (16) (17). This relationship was used to
design the 4 minute contact chambers.
An equation for the velocity gradient, "G", for mixing intensities in a
flowing open channel was developed (18).
Where /* = viscosity, centipoise
V = velocity, ft/sec
S = slope, ft head loss
ft travel
G = velocity gradient, sec"-*-
Camp and Stein (19) showed that velocity gradient is a measure of mixing
intensity and that it is in fact a measure of the number of opportunities
for particle collisions per unit time per unit volume.
Velocity gradient has units of sec"1. Then G x Volume x Time yields
a design number representing the total number of expected particle
collisions. For equal collisions in a batch process where volume is
held constant,
vl = V2 and' therefore,
GJ.TI =
75
-------
It is seen here that contact (batch) time may be reduced to the extent to
which mixing intensity is increased, and the same number of particle
collisions is achieved.
In the continuous (flow-through) process where residence time is generally
the controlled factor, equal collisions occur when:
= G2V2I2< TI = T2 and, thus,
= G2V2
Of course, contact volume is directly related to residence time for any
flow rate and, for convenience, the duration to which a flow is subjected
to disinfection is reported as a residence time. Again, residence time
(and thus residence volume) may be reduced as long as the velocity
gradient is proportionally increased.
Note in the defining equation for the velocity gradient that it is related
directly to the square roots of both velocity and head loss per unit
length of travel. Head loss is, in turn, a function of velocity. Therefore,
by increasing velocity and inducing a greater head loss per unit length
through the use of corrugated baffles and other techniques, the magnitude
of the velocity gradient may be increased to offset a reduced contact
chamber volume.
It should be noted here that all size chambers, not only high velocity
gradient chambers, should be designed to avoid short circuiting to the
maximum possible extent. That is "backmixing " must be avoided and
"plug flow" attained. The baffle arrangements which will promote plug
flow are ineffective for gross mixing. Thus an effective pre-mix with
disinfection chemicals is required before the water enters the actual
contact chamber.
Table 8 illustrates the relative efficiencies of the four disinfection
instruments discussed thus far. The velocity gradient in the tube reactor
used by Collins et al (15) has been calculated rather exactly from their
reported velocity, tube diameter and temperature (20). G for their batch
reactor has been estimated from scanty information (20). Actual velocity
and slope measurements have been taken to calculate the gradient for
both a local conventional basin and the unit designed by the authors for
disinfection of combined sewer overflows. Insufficient information is
given regarding the equipment to estimate the velocity gradient in the
work of Kruse et al (16). The results of a limited number of studies with
our corrugated baffle chambers are shown for comparison in Table 8.
76
-------
Table 8
Comparison of Disinfection Devices
Fraction of
Organisms
Surviving
Dimensionless
6 x 1CT4
1 x ID'1
3 x ID'4*
lx ID"3*
1 x 10~2 est.
Velocity
Gradient
Sec"1
6800
400
40
40
6
Contact
Time
Sec
60
60
240
120
1800
GT
Dimensionless
4 x 105
2.4 x 104
1 x 104
4.8 x 103
1.1 x 104
Chlorine
Residual
mq/1
5
5
3
3
_
Type Chamber
tube (15)
stirred batch (15)
corrugated baffles
corrugated baffles
conventional
*Based on limited tests.
-------
The high intensity mixing, high rate contact chamber used is shown in
Figure 2. Chlorine in the form of sodium hypochlorite solutions of pre-
determined strength was initially fed directly into a 15 second residence
mixing chamber by a variable capacity Lapp pump. Midway in the
disinfection study program a homemade column filled with an inorga-nic
cationic ion exchanger in the hydrogen form was added between the pump
discharge and introduction to the chamber inlet stream. Sodium
hypochlorite solutions are naturally alkaline (pH 9-10) in which state the
less active hypochlorite ion, OC1~, exists. Passage through the ion
exchangers lowers pH to the 6-7 range, coverting hypochlorite ion to
hypochlorous acid (HOC1), the more powerful oxidizing agent.
By varying both sodium hypochlorite feed strength and pump delivery rate
into a constant 20 gpm water feed, dosages between 0 and 15 mg/1
available chlorine may be added. Not shown in the figure are product ,
drawoff lines representing 1,2, and 3 minutes of contact time (or contact
volume). A dosage-time study was then simply a matter of setting for a
single dosage and sampling from the influent mixing chamber (at 15 seconds
residence), product drawoff lines, and effluent.
Two units are presently installed at the project site. Microstrainer
influent and effluent waters were treated simultaneously in order to
determine any differences in ease of disinfection brought about by the
removal of solids and other effects of the Microstraining unit. For this
study, the chlorine residual was matched at the 3 minute product drawoff
points for influent and effluent, compensating for the greater chlorine
demand exhibited by the influent combined sewer overflow.
Three studies have been made during three separate storms in the test
period, each at a velocity gradient of G = 40 sec"1.
Figures 26 through 31 show that the Microstraining pretreatment did not
increase the disinfection rate in the high rate chambers used in this study
as we had expected it would (1). The effect of Microstraining on "ease
of disinfection" in conventional rate equipment has not been determined.
The first of these was made on the storm of 11:30 a.m. , August 27, 1971,
using 15 mg/1 chlorine dosage, and results are depicted graphically in
Figure 26 and Figure 27. The 3 minute residuals for this test were each
12 mg/1.
A second study was made during the storm of 1:40 p.m. , September 13,
1971, using 10 mg/1 chlorine dosages. The results are illustrated in
Figure 28 and Figure 29. Three minute residuals on this study matched
at 7 mg/1 each.
78
-------
Date of Storm:
Nominal Chlorine Dosage:
Free Residual Chlorine
at 195 Seconds:
pH:
Temperature (est):
Sample Pretreatment
8-27-71 11:40 a.m
15 mg/1 as C12
12 mg/1 as C12
7.1
650 F
Not Macerated
10'
X Chlorinated Influent
Chlorinated Effluent
100
Contact Time - Seconds
200
Figure 26
Fecal Coliform Survival vs Time
15 mg/1 Chlorine
79
-------
o
o
1I
t-l
0>
a
0
U
03
105
Date of Storm:
Nominal Chlorine Dosage:
Free Residual Chlorine
at 195 Seconds:
pH:
Temperature (est):
Sample Pretreatment
8-27-71 11:40 a.m.
15 mg/1 as
12 mg/1 as C12
7.1
650 p
Not Macerated
10* I-
Chlorinated Influent
Chlorinated Effluent
io
c
(0
E1
O
Cn
102
10
100
Contact Time - Seconds
200
Figure 27
Total Coliform Survival vs Time
15 mg/1 Chlorine
80
-------
Date of Storm:
Nominal Chlorine Dosage:
Free Residual Chlorine
at 195 Seconds:
pH:
Temperature (est):
Sample Pretreatment
9-13-71 1:41 p.m.
10 mg/1 as C12
7 mg/1 as Cl2
9.5
65ฐ F
Not Macerated
O
Chlorinated Influent
Chlorinated Effluent
Surviving Organisms
i i
0
CO
\ \
10
100
Contact Time - Seconds
200
Figure 28
Fecal Coliform Survival vs Time
10 mg/1 Chlorine
81
-------
10
Date of Storm:
Nominal Chlorine Dosage:
Free Residual Chlorine
at 195 Seconds:
pH:
Temperature (est):
Sample Pretreatment
9-13-71 1:14 p.m.
10 mg/1 as Cl2
7 mg/1 as Cl2
9.5
65ฐ F
Not Macerated
X
ฉ
X
X Chlorinated Influent
ฎ Chlorinated Effluent
I
100
Contact Time - Seconds
Figure 29
Total Coliform Survival vs Time
10 mg/1 Chlorine
200
80
z
-------
Date of Storm:
Nominal Chlorine Dosage:
Total Residual Chlorine
at 195 Seconds:
pH:
Temperature (est):
Sample Pretreatment
2-13-72 1:18 p.m.
5 mg/1 as
3 mg/1 as
6.6
470 p
Macerated
X
Chlorinated Influent
Chlorinated Effluent
*
100 150 200
Contact Time- Seconds
Figure 30
Fecal Coliform Survival vs Time
5 mg/1 Chlorine
250
83
-------
Date of Storm:
Nominal Chlorine Dosage:
Free Residual Chlorine
at 195 Seconds:
pH:
Temperature (est):
Sample Pretreatment
2-13-72 1:18 p.m.
5 mg/1 as
3 mg/1 as
6.6
47ฐ F
Macerated
o
o
1I
S-i
Q)
a
tn
~~i
ii
0>
w
6
w
-rH
c
(0
B1
O
CO
X Chlorinated Influent
Chlorinated Effluent
10
100
Contact Time - Seconds
Figure 31
Total Coliform Survival vs Time
5 mg/1 Chlorine
200
84
-------
A third study made on collected effluent water only using 5 mg/1 chlorine
dosage is shown on Figure 30 and Figure 31. The 3 minute residual in
this case was 3 mg/1.
In this study, an aliquot of each of the five varying residence samples
was intensely blended before analysis in order to determine whether some
organisms may be occluded by remaining solids. This effect might be
more apparent on chlorinated influent samples. However, influent water
could be disinfected only when an operator was present during the actual
storm and this was not always possible in the case of shorter duration
storm events.
The coliform counts for the other disinfection studies at 15 and 10 mg/1
chlorine dosage were made on samples without prior maceration. It is
unfortunate that the coliform counts for all studies were not made on both
macerated and untreated samples.
Note as illustrated in Figure 32, that in each run the coliform levels drop
rapidly to their minimum within the first 2 minutes of contact. Interestingly,
the absolute number of surviving organisms is the same in each case,
though the fractional survival may vary with influent quality.
The disinfection phase of this project as outlined in the grant Statement
of Work was to apply sodium hypochlorite and ozone singly and in
combination. Also, the objective disinfection level as stated above was
to achieve 99.9% kill with influent total coliform levels of 100,000 cell/
100 ml or greater and 99.5% kill with influent levels of 100,000 cell/100 ml
or less; i.e. , to achieve effluent levels of 100 to 500 cell/mg total
coliform or less. As previously stated rain fall events during the contact
period allowed only three disinfection studies and these three studies
were performed with sodium hypochlorite dosages of 15, 10, and 5 mg/1
available chlorine. As can be seen by Figures 27, 29, and 31, the
disinfection performance in the high gradient chambers was much better
than anticipated and total coliform levels of 10 cells/100 ml and less
were achieved in every case. The minimum dosage to achieve the target
disinfection level of 100 cell/100 ml total coliform was not established
beyond the finding that it is less than 5 mg/1 at the conditions used; i.e. ,
50 sec at G = 40 sec"1, pH 6.6 and temperature of 47ฐ F.
Further studies should be planned at higher velocity gradients and at lower
chlorine dosages. The studies at lower residual chlorine values may
indicate differences in disinfection rates between microstrainer influents
and effluents. Maceration of test samples prior to assay is likely to aid
in revealing these differences.
85
-------
10'
o
o
(-1
a.
CO
ii
iH
Q)
u
CO
s
10
O)
c
CO
103
Storm Identification:
Nominal Chlorine Dosage-
mg/1 as Cl2*
Free Chlorine Residual at
195 Seconds - mg/1 as C^
pH:
Temperature (est) ฐF:
Velocity Gradient - sec"1:
10 =
O
15
12
7.7
65
40
A
10
7
9.5
65
40
D
3
6.6
47
40
8
100 200
Contact Time - Seconds
Figure 32
Comparison of Effluent Disinfection Curves
5, 10, 15 mg/I Chlorine
86
-------
The disinfection studies were all conducted at a constant design flow rate.
Our chambers are equipped, as are the conventional contact chambers,
with a horizontal outlet weir located at an elevation of approximately 10%
below the liquid level at design flow. Thus, at flows less than design
rate the residence time would be increased and the velocity would be
decreased almost proportionately to the flow rate reduction.
Studies (22) have shown that the velocity gradient varies as the (velocity) ' 5
when the effect of velocity on the slope term is considered. In chambers
with a horizontal outlet weir located near the working liquid level at design
rate, the constant time will vary inversely with the velocity. The net
effect is that the GT product will be reduced at less than peak design flow
and chamber disinfection performance can be expected to be reduced. The
use of a Sutro weir, if head is available - to maintain a level proportional
to flow; i.e. , constant velocity and constant residence time, and thus a
constant GT product are proposed (22).
87
-------
SECTION VIII
GLOSSARY
Boucher's Filtrability Index
The strainability of a specific water with suspended solids, through a
specific screen as defined by the equation
In (H/Ho) ~ IV (See Figure 13)
Break Point
The differential pressure across a specific screen with deposited solids
which is sufficient to rupture the bridge of solids across the screen
opening. It indicates the limiting thickness of solids for a given flow
rate and screen size. (See Figure 13)
Combined Sewer Overflow
The purposeful discharge of the contents of a combined sewer to a
receiving stream. The contents of a combined sewer at these times is
a mixture of stormwater and sanitary wastes. During dry weather the
contents of a combined sewer are exclusively sanitary flow. This flow
is normally directed to a sewage plant by a regulator adjusted to pass
1.5 - 3.0 times the mean dry weather flow, or on more sophisticated
regulators, the residual capacity of the interceptor sewer or sewage plant.
Combined sewer flows, which because of rain fall are in excess of the
above, are discharged to the stream.
Maceration
A pretreatment for samples undergoing analysis. It consists of intense
mixing-shearing at low temperature and is generally performed in a
Waring Blender. Detail of maceration technique appears in Appendix A.
Membrane Refiltration Test
A measure of the filtrability of suspended solids . It is normally used as
an analytical test to evaluate the effect of suspended solids particle
characteristics.
Mixing Intensity
The degree of turbulence created by the expenditure of mechanical and
89
-------
hydraulic energy. It is proportional to the rate of fresh liquid to liquid
interface created. The unit of mixing intensity is the velocity gradient.
Permeability
The flow rate in gpm-cp/ft2 promoted through a granular bed by a
differential pressure equal to one foot of liquid head per foot of bed
thickness . (cp - viscosity in centipoise)
Permeability Parameter
As applied to Microstraining, where flow rates and mat thicknesses are
variable within the filtering cycle, the permeability parameter is the
average flow rate in (gpm/ft2) (centipoise) through the final mat thickness
deposited on the screen. The thickness of the final mat is calculated from
the weight of suspended solids deposited and a bulk density which is
estimated or measured in an Imhoff cone.
Plug Flow
The passage of liquid through a chamber such that all increments of liquid
move only in the direction of flow and at equal velocity.
Runoff Coefficient
The fraction of the flow calculated to have reached the ground from rain
gauge data which reaches the sewer. The coefficient may be measured
from actual data or estimated from the topography of the drainage area.
Theoretical Residence Time
In a dynamic system, Theoretical Residence Time is the duration any
increment of liquid is within the boundaries of a system. It is calculated
from the liquid volume of the system divided by the volumetric flow rate
of the liquid.
When batch systems are used to simulate flow systems as in a beaker-type
disinfection study, the stirring time is sometimes referred to as Theoretical
Residence Time.
Velocity Gradient
The unit measure of mixing intensity. It is defined as the difference in
velocity of two parallel planes of fluid divided by the distance between
the two planes, and has the dimensions of ft/sec .
ft
90
-------
SECTION DC
ACKNOWLEDGMENTS
This work was supported by the Environmental Protection Agency and the
City of Philadelphia.
The overall planning and execution was under the general direction of
Richard Field, Chief, Storm and Combined Sewer Technology Branch, EPA,
Carmen F. Guarino and Joseph Radzuil, Co-Project Leaders for the City
of Philadelphia, William Wankoff, Co-Project Engineer for the City, and
William Keilbaugh, Manager of Crane-Environmental Systems Division's
R&D Department.
Thanks are extended to Martin Lozanoff, of the City of Philadelphia's
Northeast Water Pollution Control Laboratory and Paul Baker, Chemist,
who performed all of the analyses without which this work would have
not been possible. Thanks are also extended to the personnel of the
City of Philadelphia Water Department's R&D Group, who transported the
samples and tended the rain gauge.
Valuable contributions to the initial planning of the disinfection work
were received from C. W. Chambers, Robert Smith, and E. E. Geldrich,
of the EPA Advanced Waste Treatment Research Laboratory in Cincinnati,
Ohio.
91
-------
SECTION X
REFERENCES
1. Cochrane Division-Crane Co. (Glover, G.E. and Yatsuk, P.)
"Micros-training and Disinfection of Combined Sewer Overflows ",
Water Pollution Control Research Series, Report #11023 EVO 06/70.
2. Parthum, C.A. , "Building for the Future - The Boston Deep Tunnel
Plan", 42nd Annual Conference of the Water Pollution Control
Federation, Dallas (1969)7
3. Anonymous (from information furnished by Rex-Chain Belt),
"Three Solutions to the Problem of Combined Sewer Overflows ",
Water and Sewage Works (1972) .
4. Diaper, E.W.J. and Glover, G.E., "Microstraining of Combined
Sewer Overflows " , Journal of the Water Pollution Control Federation ,
v43 nlO (1971).
5. Geldrich, E.E., Water Hygiene Division, U.S. Environmental Agency.
private communication.
6. Boucher, P.L. , "Microstraining and its Application - Twelve Years
Experience", Transactions of the South African Institute of Civil
Engineers, v8, n4 (1958).
7. Camp, T.R. , "Floe Volume Concentration", Journal of the American
Water Works Association , 60:565 (1968).
8. Te Kippe, R. J. and Ham, R.K. , "Coagulant Testing: A Comparison
of Techniques, Part II", Journal of American Water Works Association,
(1970).
9. Leva, M. , et al, "Fluid Flow Through Packed Beds", U.S. Bureau
of Mines Bulletin #504.
10. Environgenics Division, Aerojet General (Feverstein, D.L.), "In-
Sewer Fixed Screening of Combined Sewer Overflows", Water Pollution
Control Research Series Report #H024FKJ 10/70.
11. Abrams, R.W. , "Retention of Fibers in Filtration of Fine Fibers",
Tournal of Technical Association of Pulp and Paper Industry - TAPPI,
(1964).
93
-------
12. Estridge, R. , "Initial Retention of Fibers by Wire Grids", Journal
of the Technical Association of the Pulp.and Paper Industry, (1962).
13. Lynam, B. , Ehelt, G. , and McAloon, T. , "Tertiary Treatment at
Metro Chicago by Means of Rapid Sand Filtration and Microstraining-",
Journal of the Water Pollution Control Federation, (1969).
14. Chambers, C.W. , "Chlorinations for Control of Bacteria and Viruses
in Treatment Plant Effluents", Journal of the Water Pollution Control
Federation, (1971),
15. Collins, H.F., Selleck, R.E., and White G-C ., "Problems in Obtain-
ing Adequate Sewage Disinfection", Proc. Sanitary Engrg. Division,
Am. Soc.. of Civil Engineers, (October 1971).
16. Kruse, C.W. , Olivieri, V. , and Kawata, K. , "The Enhancement
of Viral Inactivation by Halogens ", Water and Sewage Works ,
(1971).
17. Hudson, H.E., "Physical Aspects of Flocculation ", Journal of
the American Water Works Association, n57 p885 (1965).
18. Glover, G.E. , "Calculation of Average Velocity Gradient in Open
Channel Flow", Internal Report, Cochrane Division-Crane Co. (1967).
19. Camp, T.R. , and Stein, P.C. , 'Velocity Gradients and Internal Work
in Fluid Motion", Journal of Boston Society of Civil Engineers, 30 219
(1943).
20. Various Authors "Combined Sewer Overflow Seminar Papers", Water
Pollution Control Research Series, Report 11020 06/70.
21. Field, R. and Struzeski, E.J. , "Management and Control of Combined
Sewer Overflows ", Journal of the Water Pollution Control Federation.
(July 1972)
22. Glover, G.E. , "Influence of Turbulence on Disinfection Rate",
Unpublished paper.
94
-------
SECTION XI
PUBLICATIONS
The following publications and presentations have been based upon the
work reported herein and in the previous Report EVO 11023 06/70.
1. Keilbaugh, W.A., Glover, G.E. , andYatsuk, P., "Microstraining
and Disinfection of Combined Sewer Overflows", presented at the
FWQA Combined Sewer Overflow Symposium. Edison, N.J. (1969).
2. Cochrane Division-Crane Co. (Glover, G.E. andYatsuk, P.),
"Microstraining and Disinfection of Combined Sewer Overflow",
Water Pollution Control Research Series #EVO 11023 06/70.
published by E.P.A.
3. Diaper, E.W.J. and Glover, G.E., "Microstraining of Combined
Sewer Overflows", presented at 44th Annual Meeting of Ohio
Water Pollution Control Conference, Cincinnati, Ohio (1970).
4. Guarino, C., "Satallite Treatment Plants for Stormwater Overflow",
presented at the 8th Annual Conference of the Delaware River Basin
Water Resources Board, Buck Hill Falls, Pa. (1970).
5. Diaper, E.W.J. And Glover, G.E., "Microstraining of Combined
Sewer Overflows", presented at 43rd Annual National Conference
of the Water Pollution Control Federation, Boston, Ma. (1970).
6. Glover, G.E., "Discussion of Problems of Obtaining Adequate
Sewage Disinfection", Proc Paper 8430, Journal of San. Enqr.
Division of Am. Soc. of Civil jlnqineers, (1972).
7. Glover, G.E. , "Application of Microstraining of Combined Sewer
Overflows", presented at N.Y. State Dept. of Environmental
Consulting - E.P.A. Combined "Wastewater Training Course,
Rochester, N.Y., (November 1972).
8. Glover, G.E., "High Rate Disinfection of Combined Sewer Overflows",
Ibid.
No patents have been produced or applied for under this project.
95
-------
SECTION XII
APPENDICES
Page
A. Microstrainer Project 98
B. Organic Removal Performance 104
C. Total Coliform Removal by Micros training 110
D. Fecal Coliform Removal by Micros training 113
97
-------
Appendix A
Microstrainer Project: Analytical Data
This Appendix was prepared February 2, 1972, by Martin Lozanoff, Head
of the Central Water Pollution Control Laboratories of the City of
Philadelphia, as a memorandum to William Wank off, Co-Project Engineer
for the City of Philadelphia.
To assist in interpreting and evaluating the analytical data generated
during the Microstraining project, it is appropriate to review some of
the procedures practiced by the laboratory to insure valid results. These
procedures have two main objectives: first, to control the quality of the
results and second, to measure the quality of the results.
Procedures to control quality are applied to every phase of the analytical
work load including sampling techniques, sample preservation, aliquotting
procedures, dilution procedures, chemical or physical separations and
purifications, instrumental procedures, calculation and reporting of results,
etc. They take such forms as the use of standard analytical methods;
the production of high purity water; the periodic servicing and adjusting
of analytical balances; the use of filters on compressed air lines to trap
oil, moisture, and other contaminants; the use of constant voltage trans-
formers to minimize line voltage variations; the choice of appropriate as
well as properly cleaned glassware; the use of high purity reagents, solvents,
and gases, their proper storage and correct selection for particular
applications; the calibration and standardization of analytical instruments;
the proper use of these instruments and care of the many accessories; the
thorough and continual training of personnel; and countless other measures.
Procedures to measure analytical quality consist of daily checks on pre-
cision and accuracy and comparing them to predetermined limits.
Precision is determined by duplicate analyses and accuracy, where
applicable, by spiked or standard samples. Fully 20%, and frequently
more, of an analyst's time is spent checking precision and accuracy.
Following is a brief discussion of a few of the analytical considerations
involved in each set of samples from the microstrainer project.
Analytical Methods
Unless otherwise specified, all analyses are run according to the pro-
cedures set forth in Standard Methods, 12th edition. If modifications
are made, such changes are clearly stated. When procedures are used
that are not contained in Standard Methods, the details and source of
such procedures are given. The emphasis is on standardization of methods
-------
throughout the laboratory using a reference that is widely accepted and,
in fact, specified by Pennsylvania State Law. (The EPA Methods Manual
is used as a comparative reference.)
Pure Water
The laboratory utilizes the Millipore Super - Q (Super Quality) system to
produce high purity water. This system consists essentially of disposable
cartridges for prefiltration, organic adsorption, deionization, and 0.45
micron filtration. Water up to 18 megohms resistance is produced. The
distillate from an ordinary still is only 0.5 megohms.
The output from the Super - Q system is continuously monitored for purity
by a built-in resistivity meter. In addition, periodic checks are made on
a separate conductivity meter to verify the Super - Q meter reading.
Moreover, frequent checks on the organic quality of the effluent are made
by total organic carbon analyses.
A Super - Q system is installed on each floor with an outlet located in
each lab. Use of this system eliminates storage and transportation of
distilled water and the consequent likelihood of contamination.
ฃH_
The electrometric method is used. Prior to use the pH meter is standardized
with three buffer solutions and after use with one buffer solution in the
range of analyses.
Specific Conductance ,
1. For each set of samples the cell constant is determined with a
standard solution.
2. Every fifth sample is a duplicate; at least one duplicate is run with
each set of samples.
3. Periodically, duplicate readings are taken with two different cells.
4. All readings are reported at 25ฐ C.
Total Suspended Solids
1. The method used is filtration with Gooch crucibles using glass
fiber filter discs .
99
-------
2. Every fifth sample is a duplicate with at least one duplicate for
each set of samples.
3. There is no satisfactory procedure for determining the accuracy of
this method since the true concentration of suspended matter is
unknown.
Total Dissolved Solids
1. Total dissolved solids are obtained by difference between the
residue on evaporation (total solids) and total suspended solids.
This method is preferred to direct evaporation of the filtrate from
the total suspended solids determination in order to minimize the
effect of experimental error.
2. Every fifth sample is a duplicate with at least one duplicate for
each set of samples.
3. Accuracy is not measurable because there is no standard of
comparison.
Biochemical_Oxyqen Demand
1. Three different dilutions with duplicates of each dilution are set up
for each sample.
2. The BOD of each sample is reported as the average of the set of
duplicate dilutions that meets the criteria set forth in Standard
Methods.
3. Every tenth sample is a standard glucose - glutamic acid solution.
At least one standard solution is analyzed with each set of samples
Chemical Oxygen Demand
1. Each sample is done in duplicate and reported as the average.
2. Every tenth sample is a standard potassium acid phthalate solution.
At least one standard solution is run with each set of samples.
Total Organic Carbon
1. Carbon dioxide free water for standard solutions, blanks, and
dilutions is produced by the Millipore Super - Q system. Analysis
of this water on the TOG analyzer at the highest sensitivity in our
100
-------
working range indicates no discernible carbon.
2. Samples containing particulate matter are macerated at 23,000 rpm
for a minimum of 5 minutes in a Waring blender.
3. Bore size of the syringe used for sample injection is 154 microns
which is large enough to accommodate all the particles of a
thoroughly macerated sample.
4. Three consecutive peaks reproducible within 1% of full scale are
required for each sample.
5. Every fifth sample is either a duplicate to measure precision or a
spiked sample to measure accuracy. If less than five samples
are analyzed, then at least one duplicate and one spiked sample
are run.
Total and Fecal Coliform
1. Three different sample volumes are filtered for each sample.
2. At least two colony counts are made of each filter.
Rate Coefficient and Ultimate BOD
1. The procedure is outlined below.
2. The rate constant K and the ultimate BOD L are calculated by three
different methods:
a. Reed-Thereault
b. Log - Difference
c. Thomas Slope
3. Reported results are calculated by the Reed-Thereault method because
it has the lowest standard error. K is reported to the base e on a
daily basis.
Procedure for Rate Constant Analysis
1. Set up sample dilution according to the standard method.
2. Select dilutions that will result in a significant dissolved oxygen
depletion in 24 hours.
101
-------
3. Determine dissolved oxygen concentration daily-
4.* If more than half of the available dissolved oxygen has been depleted
reaerate the sample using a reaeration adaptor and an empty BOD
bottle.
5. Record the dissolved oxygen concentration before and after
reaeration.
6. Multiply the dissolved oxygen consumed during each period by the
dilution factor.
*Reaeration Procedure
1. The reaeration adaptor is a short piece of PVC pipe which has a taper
on each end which fits a BOD bottle.
2. Remove the stopper from the bottle to be reaerated and insert one end,
of the reaeration adaptor.
3. Invert a clean BOD bottle and place it on the other end of the adaptor.
4. Tilt the assembly so that 1/4 to 1/2 of the dilution flows into the
clean BOD bottle.
5. Shake the assembly vigorously so that the dilution is reaerated.
6. Set the assembly down in the same position it had been before it
was picked up.
7. When the BOD dilution has drained into the BOD bottle remove the
aeration adaptor and the second bottle.
8. Replace the stopper in the BOD bottle.
Procedure to Macerate a Sample
The maceration procedure for preparation of samples prior to determining
BOD, COD, TOO and coliform counts was as follows:
1. Cool sample to 40ฐ F.
2. Cool container of maceration device (Waring Blender)
by means of ice cubes wrapped in a towel for 15 minutes.
102
-------
3. Transfer cool sample to cooled blender container.
4. Blend at high speed (frappe setting) for 60 seconds.
5. Proceed at once with measurement.
The standard deviation of the precision (cv) of the tests based upon multiple
analyses of stormwater samples described above are listed in the follow-
ing table under "Northeast". Also, shown for comparison are the deviation
values obtained by surveys of many laboratories conducted by the
Standard Methods Committee and by the EPA.
TEST
NORTHEAST
Precision & Accuracy
STD. METHODS
EPA
BOD5 Cv = 9.0% Cv = 17%
Cone = 5-129 mg/1 Cone = 184 mg/1
COD
Cv = 6.5%
Cone = 200 mg/1
Recovery = 95-100%
TOG Cv = 1.9%
Cv = 5-10%
Cone = 16-182 mg/1 Samples with solids
Recovery = 97% Cv = 1-2%
Cone = 61-193 mg/1 No Solids
Cv = 20 . 6%
Cone = 194 mg/1
Cv = 10.2%
Cone = 270 mg/1
Recovery = 96%
Recovery = 85%
Cone = 4.9 mg/1
Recovery = 99%
Cone = 107 mg/1
SS
5. 7
3 3
Cone = 3-52 mg/1 Cone = 15 mg/1
Cv = 10%
Cone = 242 mg/1
Coliform: Cv = 25%
Ref: "MPN & MF Comparison", Sewage & Industrial Wastes.
31, 1, 78 (Jan. 1959).
103
-------
Appendix B
Organic Removal Performance
Date
of
Storm
7- 1-71
7- 9-71
7-19-7lB
7-29-71&
7-30-71A
7-31-71
8-27-71
8-28-71
9-10-71
Radial
Flow
Ratec
46.0
35.0
46.5
46.5
46.5
45.0
45.0
45.0
18.0
37.4
37.4
37.4
50.0
50.0
42.0
47.0
47.0
47.0
40.0
40.0
40.0
37.0
37.0
37.0
13.5
13.5
13.5
S.S.
Infl
mq/1
165
ND
47
47
317
27
27
27
ND
115
115
72
39
11
52
66
66
66
29
29
29
68
68
68
190
-
-
svs
In Out
mq/1
ND
ND
6
-
249
22
-
ND
29
-
18
21
6
24
17
-
-
16
-
-
36
-
-
49
-
_
ND
ND
16
-
27
12
-
_
ND
4
-
7
14
16
15
1
-
-
6
-
-
1
-
-
17
-
_
Removal
o/
/o
-
-
Incr
-
90
45
-
_
-
87
-
69
33
Incr
38
94
-
-
63
-
-
97
-
65
_
_
DVS
In Out
mg/1
ND
ND
41
-
_
34
-
_
ND
18
-
-
-
-
-
100
-
-
10
-
19
-
_
48
_
_
ND
ND
108
-
67
-
_
ND
16
-
-
-
-
57
-
-
17
-
_
25
_
_
98
_
_
Removal
o/
/to
Incr
-
_
Incr
- :'..
_ ; '
15
-
-
-
_
_
43
_
_
Incr
_
_
Incr
_
_
Incr
__
Sample Procedure:
macerated composite
"'discrete
^filtered
ND - Not Determined
Remarks:
A - Sample dumped
B - Effluent conductivity 350 close
to city water; influent cond. 100
C - gpm/ft2
104
-------
Appendix B
Organic Removal Performance
5 Day BOD
In Out
mg/1
ND
ND
ND
ND
ND
30
20
28
ND
260 +
348
ND
ND
ND
ND
ND
15
ND
880
940
1180
2650
2700
2000
820
920
880
ND
ND
ND
ND
ND
30
ND
28
ND
252+
256+
ND
ND
ND
ND
165
175
170
550
440
540
940
1150
ND
ND
3300
2550
Removal
Q/
/o
-
-
-
-
no change
-
no change
-
-
-
-
-
-
-
ND
Incr
ND
38
53
54
65
58
-
-
Incr
Incr
COD
In Out
mg/1
ND
ND
720
ND
-
79
-
89
ND
2350
2160
-
-
-
-
150
300
-
3760
1410
-
4680
5230
-
1820
1820
-
ND
ND
2880
ND
-
68
-
108
ND
6270
10200
-
-
-
-
310
320
-
1830
990
-
1740
1920
-
8350
7260
-
Removal
%
-
Incr
-
-
14
-
Incr
-
Incr
Incr
-
-
-
-
Incr
Incr
-
52
30
-
63
63
-
Incr
Incr
-
TOO
In Out Removal
mg/1 %
ND
ND
ND
212
-
ND
-
39
ND
ND
864
-
-
-
-
ND
34
-
ND
409
-
ND
1278
-
ND
470
-
ND
ND
ND
1048
-
ND
-
29
ND
ND
3154
-
-
-
-
ND
80
-
ND
224
-
ND
504
-
ND
1950
-
-
-
Incr
-
-
-
25
-
-
Incr
-
-
-
-
-
Incr
-
-
45
-
-
60
-
-
Incr
-
a
b
c
a
a
b
b
b
b
a
c
a
c
a
c
a
c
Sample Procedure:
amacerated composite
^discrete
cfiltered
ND - Not Determined
105
-------
Appendix B
Organic Removal Performance
Date Radial
of Flow
Storm Ratec
9-11-71 15.2
15.2
15.2
15.2
14.5
14.5
15.0
15.0
-
-
9-12-71
9-13-71 38.0
36.5
36.5
9-21-71 32.5
41.0
41.0
41.0
10-10-71 34.0
34.0
34.0
34.0
34.0
34.0
46.5
46.5
46.5
S.S.
Infl
mg/1
90
90
130
130
80
80
96
96
347
223
93
231
72
72
33
64
64
64
751
751
751
330
330
330
14
14
14
svs
In Out
mg/1
29
-
44
-
32
-
32
-
no
42
20
78
24
-
18
26
-
204
-
-
81
-
-
7
-
-
16
-
8
-
20
-
24
-
14
ND
14
6
17
-
10
18
-
_
0
-
-
3
-
-
4
_
-
DVS
Removal In Out Removal
% mg/1 %
45
-
66
-
38
-
25
-
87
_
87
93
29
-
45
31
-
100
-
-
96
_
-
43
_
-
24
-
19
-
19
-
16
-
-
ND
-
ND
_
_
23
-
_
ND
_
17
ND
_
28
ND
ป
11
51
-
61
-
79
-
89
-
-
_
ND
-
ND
_
_
26
_
_
ND
_
68
ND
_
39
ND
_
16
Incr
-
Incr
_
Incr '
-
Incr
\T
_ ;
_
-
_
Incr
_
_
_
_
Incr
_
_
Incr
_
_
Incr
Sample Procedure:
amacerated composite
^discrete
cfiltered
ND - Not Determined
Remarks:
A - Sample dumped
B - Effluent conductivity 350 close
to city water; influent cond. 100
C - gpm/ft2
106
-------
Appendix B
Organic Removal Performance
5 Dav BOD
In
mg/1
2240
ND -
2280,
ND -
11501
ND -
3350T
ND -
1580+
1580 +
1280
2150
1730
2150
1620+
885
1050
1640 +
1910
1940
1950
129
112
106
8
7
7
Out
11750
ND
23.50
ND
1350
ND
4100
ND
1580+
1580 +
1380
1750
1740
1250
800
1140
930
1640 +
86
82
70
112
98
124
.- 7
7
7
Removal
. %
Incr
-
Incr
-
Incr
--
Incr
_..
In ,
mg/1
3450
ND
3270
ND
3070
ND
7260
ND
COD
Out
19770
ND
3180
ND
3690
ND
. 4090
ND
Removal
Incr
-
03
-
Incr
-
44
-
In
mg/1
ND
1200
ND
1270
ND
730
ND
1900
TOG
Out
ND
7700
ND
1080
ND
840
ND
:2550
Removal
-
Incr
-
15
-
Incr
-
Incr
no change - - : -
no change
Incr
19
Incr
42
51
Incr
11
ND
3930
4460
-
2820
2650
ND
-
3390
3570
-
3360
3270
-
-
14
20
-
Incr
Incr
630
-
ND
1065
810
ND
850
640
-
ND
810
495
. ND
775
Incr
-
-
24
38
-
9
no change - - - -
95
96
97
13
13
Incr
13
no change
3840
3440
-
256
280
-
32
384
392
440
-
216
248
-
200
416
90
87
-
16
11
-
Incr
Incr
ND
980
-
ND
78
-
ND
10
ND
47
-
ND
54
-
ND
10
-
95
-
-
31
-
-
a
a
a
a
b
b
b
a
b
a
c
a
c
a
c
no change3
no change- - - -
c
Sample Procedure:
amacerated composite
^discrete
cfiltered
ND - Not Determined
107
-------
Appendix B
Organic Removal Performance
Date
of
Storm
11- 2-71
11-29-71
2-13-72
Radial
Flow
Ratec
36.4
36.4
36.4
36.4
36.4
16.8
16.8
16.8
S.S.
Infl
mg/1
357
357
115
115
115
208
208
208
svs
DVS
In Out Removal In Out Removal
mg/1 % mg/1 %
76
-
31
-
-
49
-
-
48
-
23
-
-
30
-
-
37
-
26
-
-
39
-
-
30
ND
ND
-
17
52
-
-
32
ND
ND
-
18
19
-
-
Incr
-
-
-
Incr
63
-
-
Sample Procedure:
amacerated composite
^discrete
cfiltered
Remarks:
A - Sample dumped
B - Effluent conductivity 350 close
to city water; influent cond . 100
C - gpm/ft2
ND- Not Determined
108
-------
Appendix B
Organic Removal Performance
5 Day BOD
In
ma/1
55
69
13
.14
. -
'.? 6
5
2
Out
26
34
26
28
-
16
15
8
Removal
%
53
51
Incr
Incr
-
Incr
Incr
Incr
COD
In Out
mq/1
230
250
117
110
-
81
68
-
170
180
117
61
-
76
41
-
Removal
%
26
28
no change
45
-
6
40
-
In
mg/1
77
ND
ND
23
-
ND
24
-
TOO
Out
38
ND
ND
22
-
ND
21
-
Removal
%
51
-
-
4
-
-
13
-
a
a
c
a
Sample Procedure:
amacerated composite
bdiscrete
cfiltered
ND - Not Determined
109
-------
Appendix C
Total Coliform Removal by Microstraining
o
Date of
Storm
7- 1-71
7- 9-71
7-19-71
7-29-71
7-30-71
7-31-71
8-27-71
8-28-71
Radial Flow
Rate gpm/ft2
46.0
35.0
46.5
45.0
18.0
37.4
50.0
42.0
47.0(5)
43.0
40.0
37.0
S.S.
Influent
mq/1
165
ND
47
27
ND
115
39
52
66
29
68
Temp
OF
NR
NR
68
69
NR
68
68
75
66
68
67
pH
(1)
ND
ND
7.6
7.1
ND
6.9
6.8
NR
7.2
7.4
7.1
7.3
6.8
6.8
Across Microstrainer
Influent Effluent
Cells/100 ml x 106
ND
ND
0
2.80
3.30
ND
0.90
0.87
ND
ND
0.39
2.00
0.85
0.73
2.40
4.80
ND
ND
0
0.001(4)
0.180
ND
0.130
0.880
ND
ND
0.100(3)
0.001(4)
0.092
0.550
0.290
1.500(3)
Removal
%
-
-
97.0
94.5(2)
-
85.6
no change (2)
-
-
74.5
99.0+(2)
89.2
24.7(2)
88.0
69.0(2)
(1) pH of filtered influent
(2) Values are results on analyses on
another aliquot of same composite
after intense maceration
(3) Estimated count based on non-
ideal colony count
(4)Zero plate count based on 0.1 ml
sample
(5) Analytical results are from a composite of
two storms, one at 3:42a and one at 5:50a
ND-Not Determined
NR-Not Reported
-------
Appendix C
Total Coliform Removal by Microstraining
Date of
Storm
9-10-71
9-11-71
9-12-71
9-13-71
9-21-71
Radial Flow
Rate gpm/ft2
13.5
15.2
15.2
14.5
15.0
probably none
36.5
41.0
S.S.
Influent
ma/1
190
90
130
80
96
93
72
64
Temp
op
72
69
69
69
68
67
68
66
PH
(1)
7.5
7.5
6.9
6.6
6.7
6.9
6.7
7.3
6.7
7.8
7.2
7.2
6.7
7.4
ND
Across Microstrainer
Influent
Cells/100
0.28
0.62
8 . 40 (3)
6.10
8.70(3)
21.00
2.30
6.60
0.88
0.89
0.06(3)
0.53
0.78
0.46
0.24
Effluent
ml x 106
1.300(3)
2.700
0.750
0.990(3)
0.079
0.072
0.081
0.090(3)
0.001(4)
0.001(4)
0.001(4)
0.400
4.200
0.013(3)
0.039
Removal
o/
/o
Incr
Incr (2)
91.0
84.0(2)
99.0+
99.0+(2)
96.5
98.6(2)
99.0 +
99.0 + (2)
98.3
24.5
Incr (2) (540%)
97.2
83.8(2)
(1) pH of filtered influent
(2) Values are results on analyses on
another aliquot of same composite
after intense maceration
ND-Not Determined
(3) Estimated count based on non-
ideal colony count
(4) Zero plate count based on 0 .1 ml sample
-------
Appendix C
Total Coliform Removal by Microstraining
Date of
Storm
10-10-71
11- 2-71
11-29-71
2-13-72
Radial Flow
Rate gpm/ft2
34.0
34.0
46.5
36.4
36.4
16.8
S.S.
Influent
mq/1
751
330
14
357
115
208
Temp
OF
64
64
65
67
50
45
PH
(1)
6.9
7.2
7.2
7.5
6.9
7-3
7.4
7.3
7.0
7.5
Across Microstrainer
Influent
Cells/100
0.20
0.10(6)
1.40(3)
2.10
0.90(3)
1.10(3)
0.320
0.760
0.570
0.340
0.005
0.004
Effluent
ml x 106
0.001(4)
0.001
0.004(3)
0.008(3)
0.050
1.360(3)
0.800
0.680
0.780
0.850
0.070
0.036(2)
Removal
99.0 +
99.0(2)
99.0 +
99.0 +
94.4
Incr (2) (123%)
Incr (250%)
10.0(2)
Incr (134%)
Incr (250%)
Incr (1400%)
Incr (900%) (2)
(1) pH of filtered influent
(2) Values are results on analyses on
another aliquot of same composite
after intense maceration
(3) Estimated count based on non-
ideal colony count
(4) Zero plate count based on 0.1 ml sample
(6) Zero plate count based on 0.001 ml sample
-------
Appendix ID-
Fecal Coliform Removal by Micros-training
CO
Date of
Storm
7- 1-71
7- 9-71
7-19-71
7-29-71
7-30-71
7-31-71
8-27-71
Radial Flow
Rate gpm/ft2
46.0
35.0
46.5
45.0
18.0
37.4
50.0
42.0
47.0(9)
43.0
s.s.
Influent
rng/1
165
NR
47
27
NR
115
39
52
66
Temp
Op
NR
NR
66
69
NR
68
68
75
66
PH
(1)
NR
NR
7.6
7.1
NR
6.8
6.8
NR
7.2
7.4
Across Microstrainer
Influent
Cells/100
NR
NR
0
0
0.0740
0.0610
NR
0.0010(8)
0.0010
0.0010
NR
NR
0.0130(2)
0.0600(3)
Effluent
ml x 106
NR
-
0
0
.00100(5)
.00200(2)
NR
.00010(6)
.00080
.00080
NR
NR
.00003(2)
.00001(2)
Removal
%
-
-
no change
no change3
98.7
97. Oa
-
90.0
20. Oa
20 . Oa
-
-
99.0 +
99.0+a
(1) pH of filtered influent
(2) Estimated count based on non-ideal colony
count
(3) At least 60 colonies in 0. 1 ml sample
(4) At least 60 colonies in 1.0 ml sample
(5) At least 60 colonies in 10 ml sample
(6) Zero plate count based on 1 ml sample
(7) Zero plate count based on 10 ml sample
(8) Zero plate count based on 0. 1 ml sample
(9) Analytical results are from a composite
of two storms, one at 3:42a and the other
at 5:50a
Values are results of analyses on another
aliquot of same composite after intense
maceration
NR-Not Reported
-------
Appendix D
Fecal Coliform Removal by Microstraining
Date of
Storm
8-27-71
8-28-71
9-10-71
9-11-71
Radial Flow
Rate gpm/ft2
40.
37.
13.
15.
15.
14.
15.
0
0
5
2
2
5
0
S.S.
Influent
mg/1
29
68
190
90
130
80
96
Temp
Op
68
67
72
69
69
69
68
pH
(1)
7.
7.
6.
6.
7.
7.
6.
6.
6.
6.
6.
7.
6.
7.
1
3
8
8
4
C
O
9
6
7
9
7
3
7
0
Across Microstrainer
Influent
Cells/100
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.0023
.0220
.0031
.0060(4)
.0840(2)
.0260
.0500
.0260
.0060(4)
.0270
.0270
.0140(2)
.0120(2)
.0060(4)
Effluent
ml x 106
.00300
.00600
.00600
.00060
.00010
.00011
.00400
.01200
.00200
.00020
.00230
.00060
.00001
.00001
(2)
(2)
(4)
(5)
(6)
(2)
(2)
(2)
(2)
(5)
(7)
Removal
%
Incr (130%)
73. Oa
Incr (192%)
90. Oa
99.0 +
99.0+a
92.0
54. Oa
67.0
99.0+a
91.5
95. 7a
99.0 +
99.0+a
(1) pH of filtered influent
(2) Estimated count based on non-ideal colony
count
(3) At least 60 colonies in 0. 1 ml sample
(4) At least 60 colonies in 1.0 ml sample
(5) At least 60 colonies in 10 ml sample
(6) Zero plate count based on 1 ml sample
(7) Zero plate count based on 10 ml sample
(8) Zero plate count based on 0 .1 ml sample
(9) Analytical results are from a composite
of two storms, one at 3:42a and the other
at 5:50a
aValues are results of analyses on another
aliquot of same composite after intense
maceration
NR-Not Reported
-------
Appendix D
Fecal Coliform Removal by Microstraining
Date of
Storm
9-12-71
9-13-71
9-21-71
10-10-71
Radial Flow
Rate gpm/ft2
probably none
36.5
41.0
34.0
34.0
46.6
S.S.
Influent
ma/1
93
72
64
751
330
14
Temp
oF
67
68
66
64
64
65
PH
(1)
7.2
7.2
6.7
7.4
NR
6.9
7.2
7.2
7.5
6.9
7.3
Across Microstrainer
Influent
Cells/100
0.0040(2)
0.1330(2)
0.0018(2)
0.0030
0.0033
0.0068
0.0170
0.0270
9.2000
0.0170
0.0530
Effluent
ml x 106
.00001(7)
.01900
.04700
.00001(7)
.00001(7)
.00010(6)
.00010(6)
.00010(6)
.00010(6)
.0200
.04500
Removal
%
99.0 +
85.7
Incra(2600%)(2)
99.7
99. 7a
98.5
99. 4a
99.6
99.0+a
Incr (118%)
15. Oa
(1) pH of filtered influent
(2) Estimated count based on non-ideal colony
count
(3) At least 60 colonies in 0.1 ml sample
(4) At least 60 colonies in 1.0 ml sample
(5) At least 60 colonies in 10 ml sample
(6) Zero plate count based on 1 ml sample
(7) Zero plate count based on 10 ml sample
(8) Zero plate count based on 0.1 ml sample
(9) Analytical results are from a composite
of two storms, one at 3:42a and the other
at 5:50a
aValues are results of analyses on another
aliquot of same composite after intense
maceration
NR-Not Reported
-------
Appendix D
Fecal Coliform Removal by Microstraining
Date of
Storm
11- 2-71
11-29-71
2-13-72
Radial Flow
Rate gpm/ft2
36.4
36.4
16.8
S.S.
Influent
mg/1
357
115
208
Temp
oF
67
50
45
PH
(1)
7.2
7.3
7.0
Across Micros trainer
Influent Effluent
Cells/100 ml x 106
0.0040 .02700
0.0088 .01500
0.0470 .05200
0.0480 .07800
0.0600 .06400
0.0140 .01500
Removal
%
Incr (670%)
Incr (170%P
Incr (110%)
Incr (164%)*
Incr (107%)
Incr (I07%)a
en
(1) pH of filtered influent
aValues are results of analyses on another
aliquot of same composite after intense
maceration
-------
, 1 Accession Number
2 Subject
Field Si Group
05D
SELECTED WATER RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
r 1 Organization
Environmental Systems Division-Crane Co., King of Prussia, Pa.
\TiUe
MICROSTRAINING AND DISINFECTION OF COMBINED SEWER OVERFLOWS
10-
Authors)
Glover, George, E.
Herbert, George, R.
, , Date
12
Pages
Jฃ I Project Number
Program
11023 FWT
21
i r Contract Number
Note
22
Citation
Environmental Protection Agency report
number. EPA-R2-73-12A, January 1973.
23 Descriptors (Starred First)
*Sewers, *Storm Runoff, *Filtration, *Water Pollution Control, *Cost Comparisons,
Water Quality, Ozone, Chlorine, Biochemical Oxygen Demand
25 I Identifiers (Starred First)
*Microstraining,
*Combined Sewer Overflow, *Suspended Solids Removal
Abstract ^ microstrainer using a screen with 23 micron apertures reduces the suspended
solids of the combined sewer overflow from 50 to 700 mg/1 down to 40 to 50 mg/1 levels
operating at flow rates of 35 to 45 gpm/ft2 of submerged screen. The organic matter as
measured by COD and TOC was reduced 25 to 40%. Coliform concentrations were 0.1 to 9
million cells per 100 ml and no reduction was brought about by Microstraining
(C)
Storm-
water service requires special analytical techniques which are described in detail. The
coliform concentrations of both overflow and microstrained overflow were reduced by four or
more orders of magnitude by disinfection with 5 mg/1 chlorine in specially built, high rate,
contact chambers of only 2 minutes contact time. The drainage area served by the combined
sewer comprises 11.2 acres of a residential area in the City of Philadelphia having an
average dry weather sanitary flow of 1000 gph. The overflow rates recorded were generally
100 times, with a maximum 400 times, the average dry weather flow. The extreme importance
of very low - 2 minute - residence volume equipment for suspended solids removal and for
disinfection in the very high instantaneous rates encountered with stormwater is shown. The
cost of a microstrainer - special chlorine contact chamber installation is cited as $6,750/cfs
of peak flow rate capacity less land and engineering. On the basis of 2 cfs instantaneous
design overflow per acre this is $13,100/acre. This report, submitted in fulfillment of
Project 11023 FWT, is a continuation of the work previously reported in 11023 EVP 06/70 =
(C)
Copyrighted Trade Name - Crane Coป
Abstractor
Authors
Institution
Environmental Systems Div. , Crane Co.
"R-102 (REV. OCT. 1968)
"RSIC
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U. S. DEPARTMENT OF THE INTERIOR
WASHINGTON, D. C. 20240
------- |