XEA1
VAII
WATER POLLUTION CONTROL RESEARCH SERIES 17010EDO 06/70
II
PHOSPHORUS REMOVAL
USING CHEMICAL COAGULATION
AND A CONTINUOUS COUNTERCURRENT
FILTRATION PROCESS
ii
U.S. DEPARTMENT OF THE INTERIOR FEDERAL WATER QUALITY ADMINISTRATION
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Reports describe
the results and progress to the control and abatement
of pollution in our Nation's waters. They provide a
central source of information on the research, develop-
ment, and demonstration activities in the Federal Water
Quality Administration, in the U. S. Department of the
Interior, through inhouse research and grants and con-
tracts with Federal, State, and local agencies, research
institutions, and industrial organizations.
A triplicate abstract card sheet is included in the
report to facilitate information retrieval. Space is
provided on the card for the user's accession number and
for additional uniterms.
Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Head, Project Reports
System, Planning and Resources Office, Office of Research
and Development, Department of the Interior, Federal Water
Quality Administration, Room 1108, Washington, D. C. 20242.
-------
PHOSPHORUS REMOVAL USING CHEMICAL COAGULATION
AND A CONTINUOUS COUNTERCURRENT FILTRATION PROCESS
by
G. R. Bell
D. V. Llbby
D. T. Lordl
Johns-Manvilie Products Corporation
Research & Engineering Center
Manville, New Jersey 08835
for the
FEDERAL WATER QUALITY ADMINISTRATION
DEPARTMENT OF THE INTERIOR
Program #17010 EDO
Contract #14-12-154
FWQA Project Officer, J. J. Convery
Advanced Waste Treatment Research Laboratory
June, 1970
-------
FWQA Review Notice
This report has been reviewed by the Federal
Water Quality Administration and approved for
publication. Approval does not signify that
the contents necessarily reflect the views
and policies of the Federal Water Quality
Administration, nor does mention of trade
names or commercial products constitute
endorsement or recommendation for use.
-------
ABSTRACT
The Johns-Manville Moving Bed Filter, a continuous precipitation and
countercurrent filtration process, was evaluated for the removal of phos-
phorus from municipal wastewater.
Using alum and an anionic polyelectrolyte, the process was found to ef-
fectively reduce total phosphorus (TP), orthophosphate (OP) and condensed
phosphate (CP) over a wide range of influent phosphorus concentrations.
Preliminary work using jar tests established an alum dose of 200 mg/1
(17.4 mg/1 Al, molar ratio of Al/P is 27/31) as effective for removal of
90 per cent TP from the secondary clarifier effluent of a trickling filter
plant. This removal efficiency could not be sustained with an alum dose
of 200 mg/1 when higher TP levels were encountered. With total phosphorus
concentrations on the order of 25 to 28 mg/1 as P (Al/P - 0.6-0.7), the TP
removal efficiency averaged 90 per cent. With lower total phosphorus con-
centrations, removal efficiency averaged 95 per cent and ranged up to 99
per cent (Al/P - 1.2-2.6).
Substantial reductions in final effluent total suspended solids (TSS) and
5-day biochemical oxygen demand (6005) were also obtained. At an alum
dose of 200 mg/1, TSS reduction averaged 70 per cent and BOD5 reduction
80 per cent. If phosphorus removal were not a design consideration, the
reduction of TSS and 6005 could be achieved with lower alum doses.
The 200 mg/1 alum dose was also found to be equally effective for removal
of phosphorus from raw sewage and primary effluent with the added capa-
bility for removing substantial portions of the TSS and BODs- In short
studies on these streams, effluent as good or better than the final ef-
fluent from the trickling plant was obtained.
Costs for a 1.0 MGD plant are estimated to be $264,000 for capital and
12.0 cents per 1000 gal. for total operating cost. These costs would be
about the same for raw sewage, final effluent or the two intermediate
levels of prior treatment studied.
Ultimate disposal of the phosphorus-containing sludge could be achieved
by a dewatering and landfill operation. Dewatering by means of a rotary
vacuum precoat filter would require an estimated capital expenditure of
$30,000 and total cost would be 3 cents per 1000 gal. of original waste-
water treated.
This report was submitted in fulfillment of Contract No. 14-12-154 under
the sponsorship of the Federal Water Quality Administration (Program No.
17010 EDO).
-------
TABLE OF CONTENTS
ABSTRACT
TABLE OF CONTENTS ............................................. i*
List of Figures .......................................... iv
List of Tables ........................................... v
CONCLUSIONS [[[ 1
RECOMMENDATIONS ............................................... 3
INTRODUCTION .................................................. 5
Objectives ............................................... 5
Background ............................................... 3
MOVING BED FILTER PROCESS ..................................... 7
Description of Process Operation ......................... 7
MBF Pilot Plant .......................................... 9
MBF PILOT PLANT STUDIES ....................................... 13
Test Location ............................................ 13
Experimental Procedures , ................................. 13
Pilot Plant Operation ................................. 13
Chemicals Added ....................................... 15
Sampling .............................................. 13
Analytical Tests and Procedures .......................... 15
PHASE I - MBF TREATMENT OF FINAL EFFLUENT ..................... 19
Experimental Plan ........................................ 19
Experimental Results ..................................... 20
Jar Tests ............................................. 20
Selection of MBF Alum Feed Level ...................... 20
Jar Tests vs MBF Performance .......................... 20
Phosphorus Removal at 200 mg/1 Alum ................... 23
Other Parameters ...................................... 27
Extended Pilot Plant Runs ................................ 27
Objectives ............................................ 27
Limitations ........................................... 27
Experimental Conditions .............................. , 27
Experimental Results .................................. 29
-------
TABLE OF CONTENTS
PHASE II - MBF TREATMENT OF UNSETTLED TRICKLING FILTER EFFLUENT 37
Experimental Plan 37
Experimental Res.ults- ...-.....'. 37
Phosphorus Removal 37
Other Parameters 39
PHASE III - MBF TREATMENT OF PRIMARY EFFLUENT 41
Experimental Plan 41
Experimental Result 41
Phosphorus Removal 41
Other Parameters 45
PHASE IV - MBF TREATMENT OF RAW SEWAGE 47
Experimental Plan 47
Problems and Limitations 47
Experimental Results 47
Phosphorus Removal 47
Other Parameters 47
Physical Observations 49
Required Alum Dose for Phosphorus Reduction 49
COST ESTIMATES 53
General Considerations of MBF Capital Costs 53
Amortization 53
Operating Costs 53
Chemicals 53
Power 54
Operation and Maintenance 54
Cost Summary 54
Capital 54
Process Cost 54
Ultimate Disposal of Sludge 55
Basis 55
Cost Summary 56
Capital 56
Process Costs 56
BIBLIOGRAPHY 57
ill
-------
LIST OF FIGURES
Page
FIGURE 1: Schematic Flow Diagram of the MBF Process 8
FIGURE 2: MBF Pilot Plant Under Construction 11
FIGURE 3: Flow Diagram of Bernards Township Sewerage Author-
ity Plant, Liberty Corner, New Jersey 14
FIGURE 4: Results of Jar Tests Showing Phosphorus Removal by
Alum Coagulation 21
FIGURE 5: Phosphorus Removal During Run I 30
FIGURE 6: Phosphorus Removal During Run II 31
FIGURE 7: Phosphorus Removal During Run III 32
FIGURE 8: Comparison of Untreated and MBF Processed Primary
Effluent 44
FIGURE 9: Relationship of Alum Dosage to Removal Efficiency .. 50
lv
-------
LIST OF TABLES
Page
TABLE 1: Summary of Phosphorus Removal from Final Effluent .. 22
TABLE 2: Comparison of the MBF Process and Jar Tests for
Phosphorus Removal 24
TABLE 3: Phosphorus Removal from Final Effluent
Conditions: Alum Dose - 200 mg/1
Anionic Polyelectrolyte Dose - 0.5 to
0.75 mg/1 25
TABLE 4: Summary of Treatment Performance - Final Effluent .. 28
TABLE 5: Plant Flows During Extended Runs on Final Effluent . 33
TABLE 6: Summary of Phosphorus Removal for Extended Runs .... 35
TABLE 7: Summary of Treatment Performance from Extended Runs, 36
TABLE 8: Summary of Treatment Performance - Unsettled Trick-
ling Filter Effluent 38
TABLE 9: Comparison of Jar Tests and MBF Processing of Pri-
mary Effluent 42
TABLE 10: Summary of Treatment Performance - Primary Effluent. 43
TABLE 11: Summary of Treatment Performance - Raw Sewage 48
-------
CONCLUSIONS
From the foregoing reported data, the following are concluded:
1. The MBF in combination with alum precipitation and an anionic poly-
electrolyte effectively removed phosphorus from raw sewage, primary
effluent, unsettled trickling filter effluent and final clarifier
effluent.
2. The MBF under the same conditions effected substantial removals of
TSS and BODcj from each of the above feed streams.
3. Preliminary data for raw sewage and primary effluent treatment indi-
cate the capability for producing a filtered effluent as good or
better in quality than the existing trickling filter plant with the
added advantage of phosphorus removal.
4. The study program arrived at an alum dose of 200 mg/1 which resulted
in effective precipitation of phosphorus over a wide range of in-
fluent concentrations; however, even this dose was not sufficient to
insure complete precipitation of phosphorus at the highest concen-
trations encountered and this is reflected in the residual phosphorus
content of the MBF effluent.
a. Initial TF concentrations below about 10-12 mg/1 were reduced on
the average about 95 per cent.
b. Initial TP concentrations between about 12-15 and 20-25 mg/1 were
reduced at least 90 per cent with values to 95 per cent.
c. Initial TP concentrations above 20-25 mg/1 were reduced an aver-
age of 90 per cent with all values above 85 per cent for initial
concentrations ranging up to 28 mg/1.
5. Removal of CP even at fairly high concentrations appeared to be par-
ticularly effective.
6. While an alum dose of 200 mg/1 effectively reduced phosphorus over
a wide range of Al/P ratios, it was found quite consistently for the
limited number of data available that lower alum doses and the same
Al/P ratios were much less effective.
7. In the treatment of final effluent to remove phosphorus, approxi-
mately 70 per cent of the TSS and 80 per cent of the BODs were also
removed. Average values in the treated effluent were 15 and 6.4
mg/1, respectively.
-------
8. Alum doses as low as 100 tng/1 removed on the average 50 per cent of
the TSS and 70 per cent of the BOD5 from the settled trickling filter
effluent.
9. Capital costs of a 1.0 MGD installation in a mild climate location
are estimated to be $264,000. Capital costs per million gallon would
decrease with increasingly larger installations.
10. Total operating cost for a 1.0 MGD plant with a 200 mg/1 alum dosage
is estimated to be 12.0 cents per 1000 gal. treated. These costs would
apply for treatment of raw sewage and primary effluent as well as trick-
ling filter effluent.
11. Final concentrations of the phosphorus-containing sludge were not well
established, but using a defensible 1.5 per cent solids in 10,000 gal.
per day (from a 1.0 MGD plant), the phosphorus^containing solids can
be reduced to a damp solid suitable for landfill by rotary vacuum pre-
coat filtration for an estimated total cost of 3.0 cents per 1000 gal.
of original wastewater treated.
-------
RECOMMENDATIONS
The data from the processing of raw sewage and primary effluent appear to
represent a significant advance in treatment capability with substantial
removals of TSS and BOD5 obtained simultaneously with phosphorus removal
without prior biological treatment. However, the data base contained in
this report for each of these types of wastewater feed is very limited.
The following are therefore recommended:
1. That additional one-month periods of operation of the pilot plant be
carried out for primary effluent and for raw sewage to provide infor-
mation on an extended range of influent conditions.
2. Moving Bed Filtration followed by granular carbon adsorption is a
complete physical-chemical treatment system which could provide a
high quality effluent. An additional study should be made in which:
a. The effluent from a test series such as that proposed above be
passed through beds of granular activated carbon to determine
the extent of removal of the remaining organic matter,
b. The filtering medium in the MBF be changed to granular activated
carbon to determine its potential for a simultaneous continuous
filtration and sorption medium. The treatment system described
in Recommendation 2 represents a potentially important non-
biological treatment alternative for the treatment of municipal
wastewater.
-------
INTRODUCTION
Objectives
This study was undertaken to evaluate the Johns-Manville Moving Bed Filter*
process as a means for the continuous removal of phosphorus from typical
municipal wastewater. Previous pilot plant studies by the contractor indi-
cated the possibility for better-than-usual utilization of alum for phos-
phorus removal, possibly as the result of a depth-filtration mechanism, and
also indicated a capability for handling wide variations in influent quality
Accordingly, the objectives of this program were:
(1) to operate on a pilot plant scale for a sufficient time
to optimize precipitation, coagulation, flocculation and
filtration for removal of phosphorus,
(2) to determine at what levels of prior wastewater treatment
the process would be feasible, and
(3) to provide bases for estimates of capital and operating
costs.
Background
Nutrients discharged to waterways are coining to be recognized as a major
pollution problem. Traditionally "pollution" has been regarded as bio-
degradable materials which deplete oxygen in receiving waters and mate-
rials which have toxic or unsightly effects on these waters. More recently,
as the emphasis has increased on providing better treatment to overcome
this traditional pollution, there is a growing awareness on the role of
nutrients as a different form of pollution generally not removed by present
sewage treatment practices.(!)
"Nutrients" usually mean compounds of phosphorus and nitrogen which tend
to fertilize receiving waters and thereby cause wild and uncontrollable
growths of algae and other aquatic biota in a process called eutrophication.
This not only results in unsightly appearance but causes important changes
in the ecological balance in such waters and greatly decreases their util-
ity. There is a growing demand for removal or at least reduction of the
amount of these being discharged to waterways,(2» 3)
*Johns-Manville patent pending.
-------
Removal of phosphorus compounds has received initial emphasis for two reasons;
(1) phosphorus is easier to remove since it forms insoluble com-
pounds, and the chemistry seems straightforward with a rela-
tively limited number of options, and
(2) some blue-green algae have the capability of fixing nitrogen
from the atmosphere so that phosphorus becomes the limiting
nutrient in their life cycled3)
To date there appears to be no general agreement as to the level to which
phosphorus must be reduced before the effluent can be discharged. Dryden
originally proposed a limit of 2 mg/1 as PC>4 (0.65 mg/1 as P) for a recre-
ational lake supply but later revised this down to 0.5 mg/1 as PO^ (0.16
mg/1 as P) as necessary to prevent regrowth of algae. On the other hand,
an agreed upon clean-up program for Lake Michigan) recommended removal
of at least 80 per cent of all phosphorus from treated wastewater. This
permits substantially higher levels to be discharged than those cited above,
presumably by allowing for dilution by the lake water to reduce overall con-
centrations below nuisance levels.
A recent summary effectively covers the current state of the art with
respect to both biological and chemical methods of phosphorus reduction
placing the various options in perspective. The present studies were con-
cerned only with chemical precipitation and more particularly with alum,
i.e., aluminum sulfate, as the precipitant and the subsequent separation
of the precipitated material and other solids which may have phosphorus
associated with them from the substrate by filtration. A polyelectrolyte
was used to facilitate the filtration step of the process.
-------
MOVING BED FILTER PROCESS
Description of Process Operation
A treatment process employing the Moving Bed Filter, hereafter referred
to as MBF, includes chemical precipitation and coagulation directly fol-
lowed by filtration through a filter bed. The key component in the pro-
cess is a unique filter designed to provide continuous countercurrent fil-
tration even under high solids loading. Figure 1 is a schematic flow dia-
gram of the MBF process. The principle of operation is as follows:
The chemicals are added directly to the influent wastewater line. Pre-
cipitation, coagulation and flocculation of solids occur in the head tank.
The dosage of coagulant can be varied depending upon the nature of the
waste to be treated and the quality of effluent desired. The head tank
provides for four functions:
(1) it serves as a precipitation and flocculation chamber;
(2) it provides the necessary head for gravity feed through
the filter bed;
(3) it provides some surge capacity; and
(4) the conical bottom of the head tank provides for collec-
tion of the spent filter medium and sludge.
The filter medium, usually sand, is contained in a tubular shell and is
driven in one direction while the wastewater being treated passes through
the filter bed in the opposite direction. The filtering action occurs
throughout the depth of the bed as well as at its face. Filtered water
flows out of the bed through discharge screens located on the side of the
filter shell. Solids which have been removed are driven along with the
filter medium toward the head tank and subsequently are removed from the
filter face as rapidly as is required by their buildup. Movement of the
filter medium is accomplished by means of an hydraulically operated dia-
phragm. The diaphragm pushes the bed as a plug toward the inlet, or fil-
ter face, end. As the diaphragm relaxes, clean sand feeds by gravity into
the void left in the bed in front of the diaphragm. This cycle is then
repeated with the next pulse of the diaphragm. The frequency of pulsing,
and thus the rate of sand drive, can be varied depending upon optimizing
the various filtration factors and is controlled by the rate of change of
level in the head tank.
The sludge-sand mixture hydraulically or mechanically removed from the
filtering face falls down into the hopper bottom of the head tank. This
mixture is transferred to a washing column for cleaning. The clean sand
-------
INFLUENT |>
oo
SAND
R ECYCLE
T
CHEMICALS
CLEAN SAN
NOPFE
I LTEREO
WATER
SAND
WASHER
SAND RED DRIVE
SYSTEM
SAND-SLUDGE
WASH WATER
FIOURI 1.
ASIC CONCEPT Of MOVINO RED FILTER
-------
is then returned by gravity to the hopper of the MBF. The removal and
washing of the sand may be intermittent or continuous. Final washing of
the sand is accomplished by means of filtered effluent. Since the sand
is removed, cleaned and then returned to the system, the filtration pro-
cess is not interrupted for backwashing as it is in conventional practice.
The waste wash water flows into a settling tank where the sludge is con-
centrated prior to dewatering. The overflow, or supernatant, is recycled
to the influent line of the MBF.
MBF Pilot Plant
The pilot plant unit used in this study consists of a single tubular fil-
ter bed having the following dimensions: diameter at water inlet end -
30 inches; average diameter of tube - 27 inches; total length of bed - 68
inches; bed depth, face to screen center - 50 inches; maximum cross-
sectional area - 4.9 square feet; volume of sand contained within the
shell - 20 cubic feet.
The filter medium used in all these studies was a commercially available
filter grade sand. Grain size distribution of this sand was determined
by screening through U. S. Standard sieves, and the sand was found to have
a 10 per cent effective size of approximately 0.8 mm and a uniformity co-
efficient of 1.2. This is much coarser than the sand generally used in
fixed bed rapid sand filtration.
Wastewater was delivered directly to the pilot plant unit by means of a
self-priming centrifugal pump. The flow was determined by means of a
variable area flow meter and the rate of flow was controlled by means of
a valve on the discharge side of the pump. The waste was pumped to the
top of the head tank where alum was injected into the influent line. Mix-
ing and flocculation of the waste occurred within the head tank, prior to
filtration. Figure 2 shows the MBF pilot plant during installation and
prior to erection of the building around it.
A polyelectrolyte was used as a filter conditioner to control the depth
of floe penetration. The flocculated waste passed down through the fil-
ter bed and out through the water outlet screens. The water outlet con-
sisted of two screened slots, each 5-1/4 inches wide with a curved length
of 17 inches, located along the sides of the filter shell.
Various auxiliary equipment included in the pilot plant installation were
chemical solution tanks, mixers and feed pumps which allowed the appli-
cation of two separate chemicals simultaneously. Also included was a 200-
gallon storage tank for filtered water, part of which was reused for wash-
ing the spent sand and for the face cutting mechanism. The sand washing
system was initiated by the frequency at which the bed was pulsed. The
-------
various steps within the wash cycle, that is, withdrawal of the sand and
sludge, washing of the sand and sludge, return of the clean sand to the
filter feed hopper, were automatically controlled by means of a program
timer. The waste wash water was discharged to a 500-gallon steel tank for
settling.
10
-------
Figure 2. MBF PILOT PLANT UNDER CONSTRUCTION
11
-------
MBF PILOT PLANT STUDIES
Test Location
The MBF pilot plant was set up at the Township of Bernards Sewerage
Authority Treatment Plant at Liberty Corner, New Jersey. Placed in
service in 1964 with a design flow of 0.55 MGD, this plant consists
of coarse bar screens, comminutors, Clarigester*, high rate trick-
ling filter, final settling tank and chlorine contact tank. During
the study period (3/69 to 10/69), the flow in the plant ranged from
0.3 to 2.5 MGD. Heated sludge digestion is provided with disposal
of the digested sludge to sand drying beds and then as landfill.
This plant was selected from the several available in the area on the
basis of fairly typical trickling filter plant performance as well as
its convenience to necessary analytical facilities at the Johns-Man-
ville Research & Engineering Center.
A flow diagram of the Bernards Township Sewerage Authority Plant is
shown in Figure 3. Also included is the location of the pilot plant
and the alternative locations at which its feed streams were obtained
for the study program. Most of the work was done with effluent from
the final settling tank, which for convenience has been designated
"final effluent."
Experimental Procedures
Pilot Plant Operation. Operation of the pilot plant was substantially
as described in the preceding process description. After an initial
period for shakedown of the equipment and testing of the various ana-
lytical procedures, the pilot plant was put into routine operation
during April of 1969. Initially operated only on the day shift, it
was later placed on 24-hour, around-the-clock operation with personnel
present only during the day shift. The general operating procedure
was to maintain constant influent flow conditions and set the frequency
of push at a predetermined interval in order to maintain relatively con-
stant hydraulic head.
Data regarding operational parameters, such as flow, cycle time for for-
ward sand movement and diaphragm retraction, level fluctuations in the
head tank and within the filter bed, and turbidity and pH of feed and
filtered effluent, were recorded during each experimental run.
^Trademark of Dorr-Oliver, Inc.
13
-------
TO RIVER
SECONDARY SLUDGE RETURN LINE
BY PASS LINE
RECIRCULATION LINE
NORMAL FLOW LINE
MBF PILOT PLANT SUPPLY POINTS!
I - FINAL EFFLUENT
II - UNSETTLED TRICKLING FILTER EFFLUENT
III- PRIMARY EFFLUENT
IV ~ RAW SEWAGE
CM0R*4f
CONTACT
CHAMBER
FINAL
CLARIFIER
COMMINUTOR
PARSHALL
FLUME
itiiiRis TOIISNIP SEIIBE AUTHORITY PLIIT
-------
Chemicals Added. Alum was used as the primary coagulant throughout these
studies. This was commercial filter grade aluminum sulfate with an A1203
concentration of 16.2 per cent by analysis. This corresponds to the for-
mula Al2(804)3"16H20. To avoid possible problems due to evaporation or
aging, alum feed solution of 5 per cent (wt) concentration was prepared
approximately every other day. This solution was fed directly into the
influent flow just before it discharged into the head tank. No attempt
was made in this study to test other points of addition or means of mix-
ing of the alum solution with the feed stream.
As noted earlier, a polyelectrolyte was used as a filter conditioner to
improve floe retention within the filter. Several polyelectrolytes were
evaluated; however, most of the reported results were obtained with one
product.* Each polyelectrolyte was made up as an unheated stock solution
of 0.25-0.5 per cent by weight and refrigerated. The stock solution was
stored for no longer than one week. Each day a fresh 0.01 per cent feed
solution of the particular polyelectrolyte was prepared as feed solution.
Polyelectrolytes were added to the body of the liquid in the head tank
at a point just in front of the filter face.
Sampling. Initially, composite samples of the influent and effluent of
the MBF were collected daily by means of an automated sampling technique.
These samples were composited over a 2-hour period and were taken during
different times of the day in order to determine diurnal fluctuations in
performance. In addition, manually composited samples were taken from
the head tank and from the waste wash water settler overflow. Turbidity
measurements were routinely made on grab samples taken from within the
filter bed at 12-inch intervals and from the filter effluent. After the
basic operating characteristics were established for a given experiment,
the system performance was continuously monitored for periods of up to
three days. During these monitoring studies, analyses were made on 6-
hour composite samples. All samples were refrigerated until transpor-
tation to the laboratory for analysis.
Analytical Tests and Procedures
Summary of the analytical tests performed and the procedures used through*
out this investigation is as follows:
Biochemical Oxygen Demand - 5-day BOD was determined in accordance
with the recommended procedure in Standard Methods.&'
Alkalinity - Total alkalinity titrations were made using mixed
bromcresol green methyl red indicator. Results are expressed as
CaC03.
*Magnifloc 860A - Product of American Cyanamid Company.
15
-------
Filtration. - Laboratory filtrations were performed using 9-cro Reeve
Angel glass fiber filter pads.
j>H - pH measurements were made using a Leeds & Northrup Model 7664 pH
meter.
Suspended and Volatile Solids - All suspended solids and volatile solids
analyses were carried out according to Standard Methods with the fol-
lowing modification: the solids were dried at 104C for 14-16 hours.
Turbidity - Turbidity measurements were made at the pilot plant site on
all samples which were collected. Turbidity measurements were made with
a Hach Model 1860 laboratory turbidimeter and expressed as Jackson Tur-
bidity Units (JTU).
Phosphorus - The stannous chloride method for orthophosphate was used
for all phosphorus determinations. Ten minutes were allowed for color
development and readings were taken at a wavelength of 690 mu using a
Bausch and Lomb Spectronic 20 colorimeter.
Total phosphorus was determined by acidification and digestion of the
samples to convert the condensed phosphates and organic phosphorus to
orthophosphates. Fersulfate was used as a catalyst. The following
analytical scheme was used for the delineation of phosphorus forms of
the various effluents:
Complete Sample
A Filter sample through 9-cm Reeve Angel glass
Acid hydrolysis fiber filter pad
with persulfate
then Filtrate
orthophosphate B C
unfiltered Acid hydrolysis Orthophosphate
determination with persulfate filtered
then determination
orthophosphate
determination
The analytical results are designated throughout this report as:
TP Total Phosphorus Scheme A
TFP Total Filterable Phosphorus Scheme B
OP Orthophosphate, Soluble Scheme C
CP Condensed or Polyphosphates Schemes B-C
The stannous chloride method for determination of phosphate was used
because of its sensitivity at low levels (< 0.1 ppm) of orthophosphate.
This method, however, is also highly sensitive to temperature, humidity
16
-------
and pH of the sample. As a result, two separate standard curves had
to be generated; one for digested samples, one for undigested sam-
ples. Even with separate curves, there were in many instances dis-
crepancies between the ortho and total filterable phosphorus values
with OP greater than TFP (and in some cases greater than TF).
17
-------
PHASE I - MBF TREATMENT OF FINAL EFFLUENT
Experimental Plan
Each of the various unit operations which make up the MBF process was
given consideration in the initial experimental plan. Since this study
is primarily concerned with phosphorus removal, the plan was set up in
terms of this objective. However, other characteristics of the MBF feed
and product (filtered) water were also determined to provide a broader
evaluation of the process for treatment of trickling filter treated and
untreated wastewater.
For this study, phosphorus removal was thought of as a series of inter-
related steps:
(1) precipitation,
(2) coagulation-flocculation,
(3) filtration,
(4) removed solids-filter medium separation, and
(5) solids concentration for ultimate disposal.
A thorough analysis of phosphorus insolubilization and removal was pos-
sible with the pilot plant equipment. Factors of scale, however, pre-
cluded an intensive investigation of sludge characteristics and disposal.
The experimental plan for studying phosphorus removal from final effluent
included several operations, some of which proceeded simultaneously, and
which covered a period of four to five months:
1. Jar testing to:
a. Determine the probable levels of alum addition required for
various degrees of phosphorus insolubilization.
b. Compare the pilot plant precipitation with that expected
from laboratory procedures.
c. Provide baseline data for various levels of the phosphorus
forms designated in this report as TFP, OP, TP and CP.
d. Determine probable levels of anionic polyelectrolyte addi-
tion required to produce filterable "floe" or filter con-
ditioning needed for effective removal of the insoluble
phosphorus.
19
-------
2. Pilot plant operation to determine the level of phosphorus removal at
various levels of alum addition.
3. Pilot plant operation at an alum feed level selected from Step 2 for
a period of time to insure effective removal of phosphorus over a wide
range of influent conditions.
4. Continuous operation of the pilot plant for two to three days to:
a. Generate information on diurnal variations in the various plant
and pilot plant parameters.
b. Generate data for estimation of operating and capital costs.
Experimental Results
Jar Tests. Figure 4 illustrates the typical, and expected, relationship
between levels of alum addition and OP and TP precipitation from final ef-
fluent. On the basis of an influent IP level of 8.2 mg/1, an alum dosage
at least 150 mg/1 (13 mg/1 as Al) would be the expected requirement to
reduce P to less than 1 mg/1, even though the alum demand was found to be
somewhat lower to effect substantially complete precipitation of OP.
Also, as expected, other Jar test and MBF operating data show that the
alum requirement is influenced to some extent by the levels of phosphorus
in the final effluent used as the MBF process feed. These levels were found
to cover quite a wide range as an uncontrollable variable, as will be shown
in later discussion.
Selection of MBF Alum Feed Level. Based on the jar tests, a number of rep-
licate pilot plant runs were made covering alum feed levels from 100 to
250 mg/1. These runs, summarized in Table 1, also demonstrate the expected
relationship between alum dose and phosphorus reduction. From these data,
an alum dose of 200 mg/1 (17.4 mg/1 as A1+++) was selected for longer-term
study.
Jar Tests Versus MBF Performance. A number of simultaneous jar tests were
made during MBF pilot plant runs at various alum feed levels. Purposes of
these tests were:
(1) to compare MBF precipitation and removal with that obtained
by laboratory jar test precipitation, and
(2) to provide some basis for delineation of phosphorus in the
precipitation process as TP, TFP and OP.
20
-------
7/21/69
(SINGLE SIT)
HOURE 4. JAR TESTS ON SECONDARY CLARIHER
EFFLUENT
21
-------
ro
Table 1.
SUMMARY OF PHOSPHORUS REMOVAL FROM FINAL EFFLUENT
Phosphorus Concentration -
Secondary Effluent
Range
Alum Dosage
Total
Filterable
Ortho
Alum Dosage
Total
Filterable
Ortho
Alum Dosage
Total
Filterable
Ortho
Alum Dosage
Total
Filterable
Ortho
- 100 mg/1;
4.2 -
3.4 -
2.8 -
- 150 TO/I:
13.5 -
13.0 -
12.3 -
- 200 mg/l;
6.3 -
4.0 .-
4.1 -
- 250 og/1 ;
16.1 -
11.9 -
11.4 -
Anionic
19.4
17.8
14.6
Anionic
21.6
19.8
17.1
Anionic
15.0
13.1
11.6
Anionic
21.9
13.9
13.9
Avg.
rag/1 as
P
MBF Effluent
Range
Polyelectrolyte -
12.0
10.5
8.5
0.35
0.13
0.07
Polyelectrolyte -
16.5
15.5
13.6
2.2
1.7
1.4
Polyelectrolyte -
9.4
8.0
7.8
0.09
0.01
0.01
Polyelectrolyte -
17.8
13.2
12.9
0.41
0.11
0.08
0.5 mg/1
- 9.5
- 7.2
- 7.1
0.5 mg/1
- 10.0
- 7.1
- 6.9
0.5 mg/1
- 1.2
- 0.30
- 0.30
0.5 mg/1
- 3.4
- 0.31
- 0.26
Avg.
Reduction
Avg.
a)
to 0.75 mg/1
5
3
2
to
5
3
3
to
0
0
0
1
.40
.29
.88
0.75 mg/1
.35
.25
.07
0.75 mg/1
.51
.11
.10
.24
0.23
0
.20
59
63
72
67
79
75
94
98
98
93
98
98
.0
.0
.2
.2
.0
.7
.6
.6
.9
.4
.2
.5
Avg.
Ai/P
No. of Molar
Obs. Ratio
17
0
0
1
21
0
0
1
34
2
2
2
7
1
1
1
.83
.94
.16
.90
.96
.09
.10
.47
.54
.38
.87
.91
-------
The data summarized in Table 2, while showing the previously discussed
relationship between alum dosage and phosphorus removal, also show sig-
nificantly lower levels of TP in the MBF effluent than in the jar tests.
This is intriguing, but the range of influent phosphorus concentrations
where comparative data are available is too limited to permit any con-
clusions as to mechanism.
Phosphorus Removal at 200 tng/1 Alum. Table 3 presents data for TP, TFP
and OP reductions over a broad range of conditions encountered during a
period of almost four months. These data cover only the single alum dos-
age of 200 mg/1 (17.4 mg/1 as Al) and the same anionic polyelectrolyte*
throughout, dosed at 0.5 to 0.75 mg/1.
As shown in Table 3, the reduction in total filterable phosphorus (TFP)
might be considered as the degree to which the influent TFP was insolu-
bilized by the alum addition. For 34 composite samples, the average per
cent removal was 98.6, the maximum 99.8 and the minimum reduction 95.9,
respectively. Since alum feed was held constant, the range of 4.00 to
13.10 mg/1 of influent phosphorus reflected Al/P ratios between 1.33 and
4.35. These levels could be expected to cause high levels of precipita-
tion.
A more important measure of overall process performance, perhaps, is
total phosphorus (TP) removed. In addition to effective precipitation,
low values of TP require retention in the MBF of both precipitated phos-
phorus and that associated with solids in the final effluent. For the
34 composites tabulated, the average TP reduction was 94.6 per cent while
the maximum was 98.82 and the minimum 88.35. To give an idea of distri-
bution, only two of the 34 values were less than 90 per cent.
When Al/P ratios are considered in the light of TP content, rather than
TFP, the range of 1.16 to 2.76 to 1 represents efficient chemical utili-
zation.
The two samples which had TP reductions of less than 90 per cent illus-
trate that retention in the filter rather than precipitation was less than
optimum. Composite 3-6A had a TFP removal of 98.8 per cent but the corres-
ponding TP removal was only 88.35 per cent. Similarly, composite 3-7A had
99.0 per cent of the TFP removed and only 89.5 per cent of the TP. This
indicates that TFP was almost completely precipitated, but the resulting
precipitate was not as adequately retained in the filter. The reasons for
the decreased removal in these instances are not known, but they could be
associated with either a chemical parameter such as the polyelectrolyte
feed or with some mechanical problem. Despite these lapses, the overall
performance of the MBF process was very high.
*Magnifloc 860A - Product of American Cyanamid Company.
23
-------
Table 2.
COMPARISON OF THE MBF PROCESS AND JAR TESTS FOR PHOSPHORUS REMOVAL
TEST SERIES 1
Clarifier Effluent
MBF Effluent
Jar Test Effluent
Jar Test Effluent
TEST SERIES 2
Clarifier Effluent
MBF Effluent
Jar Test Effluent
TEST SERIES 3
Clarifier Effluent
MBF Effluent
Jar Test Effluent
Alum
PPBJ
0
100
100
150
0
200
200
0
200
200
860A
ppm
0
0.75
0.75
0.75
0
0.75
0.75
0
0.75
0.75
PH
7.0
6.8
6.8
6.7
7.2
6.3
6.2
7.1
6.5
6.4
OP
1.09
1.09
1.63
2.64
2.64
2.18
2.18
Al/P
TFP
0.91
0.91
1.35
2.64
2.64
2.05
2.05
TP
0.80
0.80
1.19
2.64
2.64
1.79
1.79
OP
ffiK/1 P
8.0
2.0
1.9
0.45
6.6
0.02
0.04
8.0
0.04
0.24
TFP
mjs/1 P
9.6
2.3
2.1
0.58
5.8(1)
0.04
0.08
8.5
0.09
0.30
TP
BBK/1 P
10.9
4.4
7.5
2.1
6.6
0.30
0.42
9.7
0.50
1.44
* See note on page 16
-------
10
in
Table 3.
PHOSPHORUS REMOVAL FROM FINAL EFFLUENT
Conditions: Alum Dosage - 200 mg/1 (17.4 mg/1 Al); Anionic Polyelectrolyte - 0.5-0.75 mg/1
Total Phosphorus (me/1)
Composite
3-3
3-4
3-5
3-6
3-7
3-13
3-18
3-19
3-20
4-8
AM
PM
PM
AM
PM
AM
PM
AM
PM
AM
AM
IM
AM
PM
AM
IM
AM
IM
Infl.
7.93
15.00
8.50
7.10
9.30
7.70
10.30
6.85
8.00
7.90
7.50
10.00
6.60
9.70
7.20
8.40
7.70
11.30
MBF
0.75
0.45
0.56
0.32
0.77
0.27
1.20
0.48
0.84
0.74
O./O
0.49
0.30
0.50
0.09
0.15
0.29
0.82
T. Red.
90.54
94.33
93.41
95.49
91.72
96.49
88.35
92.99
89.50
90.63
90.67
95.10
95.45
94.85
98.75
98.21
96.23
92.74
Al/P
2.19
1.16
2.05
2.45
1.87
2.26
1.69
2.54
2.17
2.20
2.32
1.74
2.64
1.79
2.42
2.07
2.26
1.54
Total Filterable Phosphorus (mg/1)
Influent MBF T, Red.
7.18
10.00
7.70
6.30
8.20
7.10
9.50
5.97
7.20
7.30
6.60
8.70
5.80
8.50
6.30
8.10
7.00
10.30
0.25
0.29
0.24
0.12
0.10
0.05
0.11
0.06
0.07
0.18
0.04
0.05
0.04
0.09
0.01
0.03
0.01
0.02
96.52
95.96
96.88
98.10
98.78
99.30
98.84
98.99
99.03
97.53
99.39
99.43
99.31
98.94
99.84
99.63
99.86
99.81
Orthophosphate
Influent MBF
7.93
8.90
8.80
6.70
7.20
8.10
9.20
4.50
4.80
8.00
7.80
8.80
6.60
8.00
6.70
7.60
7.30
9.00
0.25
0.22
0.22
0.10
0.09
0.05
0.09
0.06
0.04
0.16
0.03
0.05
0.02
0.04
0.01
0.04
0.01
0.01
(mg/1 P)
7. Red.
96.85
97.23
97.50
98.51
98.75
99.38
99.02
98.67
99.17
98.00
99.62
99.43
99.70
99.50
99.85
99.47
99.86
99.89
-------
O*
Table 3. (Cont'd)
PHOSPHORUS REMOVAL FROM FINAL EFFLUENT
Conditions: Alum Dosage - 200 mg/1 (17.4 mg/1 Al); Anionic Polyelectrolyte - 0.5-0.75 mg/1
Total Phosphorus (me/I)
Composite
4-9
4-10
4-15
4-16
4-17
6-17
6-18
6-19
Avg.
Min.
Max.
AM
PM
AM
PM
AM
PM
AM
PM
AM
PM
AM
PM
AM
PM
AM
PM
Infl.
6.80
9.30
7.00
8.00
11.50
12.30
9.20
11.20
6.30
9.50
7.80
13.20
8.30
12.20
14.00
14.90
9.37
6.30
15.00
MBF
0.09
0.11
0.10
0.10
0.46
0.88
0.50
0.63
0.31
0.31
0.30
0.90
0.40
0.90
0.50
1.00
0.51
0.09
1.20
7. Red.
98.68
98.82
98.57
98.75
96.00
92.85
94.57
94.37
95.08
96.74
96.15
93.18
95.18
92.62
96.43
93.29
94.60
88.35
98.82
Al/P
2.56
1.87
2.49
2.17
1.51
1.41
1.89
1.55
2.76
1.83
2.23
1.32
2.10
1.43
1.24
1.17
1.97
1.16
2.76
Total Filterable Phosphorus (mg/1) Orthophosphate (mg/1 P)
Influent MBF % Red. Influent MBF 7. Red.
5.60
8.80
4.00
5.00
9.20
10.90
8.60
10.40
5.60
7.30
6.60
12.80
7.30
11.00
8.90
13.10
8.03
4.00
13.10
0.02
0.02
0.01
0.15
0.28
0.22
0.15
0.06
0.06
0.10
0.10
0.10
0.10
0.30
0.30
0.11
0.01
0.30
99.64
99.50
99.80
98.37
97.43
97.44
98.56
98.93
99.18
98.48
99.22
98.63
99.09
96.63
97.71
98.60
95.96
99.86
6.60
9.80
4.10
5.00
9.20
10.00
8.60
8.50
4.90
7.20
7.50
8.70
7.60
9.20
10.80
11.60
7.80
4.10
11.60
0.02
0.02
0.02
0.01
0.09
0.23
0.20
0.10
0.02
0.02
0.10
0.10
0.20
0.10
0.20
0.30
0.10
0.01
0.30
99.70
99.80
99.51
99.80
99.02
97.70
97.67
98.82
99.59
99.72
98.67
98.85
97.37
98.91
98.15
97.41
98.86
96.85
99.89
-------
Other Parameters. In Table 4 data for parameters other than phosphorus
removal at various levels of alum addition are summarized.
These data show that on the average 100 mg/1 alum in final effluent re-
duced total suspended solids (TSS) about 47 per cent while doubling the
alum dose to 200 mg/1 increased average TSS removal to about 67 per cent.
Table 4 also indicates that the remaining TSS in the MBF effluent are sig-
nificantly different in character having a much higher level of fixed solids
at all levels of alum addition, indicating that there was a substantial re-
moval of volatile matter.
Additions of 100 mg/1 alum reduced 8005 on the average about 71 per cent,
and at 200 mg/1 the reduction was about 80 per cent.
This indicates that where phosphorus removal is not an objective substan-
tial improvement in the quality of the final effluent could be attained with
much lower lower chemical consumption.
Extended Pilot Plant Runs
Objectives. Extended runs were made on final effluent to demonstrate the
performance and reliability of the MBF system under variable load condi-
tions and to develop cost information.
Limitations. The cyclic nature of sewage flows and strength is well known
and does not require explanation here. However, any pilot plant program
which can also be substantially affected by uncontrollable factors, such
as the weather, may end up well outside of the planned experimental limits.
Such problems did arise. First, when after several months of "normal" rain-
fall the period preselected for an extended run turned out to be one of the
wettest periods in many years causing very high flows and diluted wastewater
through the plant; second, when a new extended run was attempted to have
substantially higher phosphorus values than had been encountered except for
an occasional sample in the previous several months of the study; and fi-
nally when a third attempt was made to find phosphorus values still substan-
tially higher than expected. The resulting data from the three extended
runs provide a broader diversity of results than had been planned. They
also raise several questions, noted below, for which answers are not ap-
parent.
Experimental Conditions. An alum feed level of 200 mg/1 was used during
all three extended runs. During the first extended run (Run I), 0.2 mg/1
of a special anionic polyelectrolyte* was used. The dose of Magnifloc
*Atlas Flocculant 110 - Product of Atlas Chemical Industries, Inc.
27
-------
Table 4.
SUMMARY OF TREATMENT PERFORMANCE - FINAL EFFLUENT
Alum
Dosage
100
150
200
250
100
150
200
250
100
150
200
250
Secondary Effluent
Ranee
6.8
7.0
6.9
7.3
105
189
134
174
41
33
22
31
- 7.2
- 7.3
- 7.4
- 7.4
Total
- 270
- 238
- 237
- 225
Total
- 135
- 67
- 118
- 55
Avg.
7.1
7.2
7.2
7.3
MBF Effluent
Ranee
PH
6.6
6.7
5.7
6.6
Alkalinity -
178
206
180
189
55
114
13
59
- 7.1
- 7.2
- 7.1
- 6.9
me/1 CaCOg
- 212
- 177
- 168
- 104
Ave.
6.9
7.0
6.7
6.8
134
141
95
80
Ayg.
Reduction
Suspended Solids - tng/1 ^
64
48
50
44
Fixed Suspended
100
150
200
250
1.0
0.1
0.5
11
- 38
- 26
- 46
- 23
19.9
13.5
15.6
18.1
8
9
1
5
Solids
10
0.1
0.0
33
Turbidity -
100
150
200
250
100
150
200
250
28
24
13
22
40
--
14
--
- 60
- 40
- 54
36
- 75
--
- 97
42
30
33
27
BOD,
58
--
65
--
5.0
5.0
1.4
2.8
- 64
- 46
- 57
- 12
34
20
15
8
46
58
66
73
.9
.4
.6
.0
- 7. of Total
- 83
- 67
- 85
- 67
JTU
- 36
- 18
- 27
- 13
40.4
36.6
42.8
48.3
19
8,7
6.4
4.5
54
71
78
83
.8
.0
.5
.2
= - mg/1
5
-
3
.
- 29
--
- 30
--
17
--
12
--
70
80
.8
.2
28
-------
86QA was 0.5 to 0.75 mg/1 during all of the other reported work on final
effluent.
The pilot plant was manned around the clock during these runs to insure that
adequate operating and analytical data would be obtained. Chemical feed
rates, head loss profiles along the filter bed and influent and effluent
turbidity values were monitored hourly. Six-hour composites of the final
effluent and MBF effluent were taken by automatic samplers.
These samples were analyzed for OP, TFP and TP. In addition, pH, total al-
kalinity, total suspended solids (TSS) and fixed suspended solids (FSS) also
were determined.
Attempts were made also to collect the settled sludge from the sand wash
water settling tank. Since the volume of sludge accumulated in any reason-
able sampling period turned out to be quite small (probably 10-15 gallons
or less in a 400-gallon tank) it proved impractical to withdraw sludge with-
out substantial volumes of supernatant causing dilution. Subsequent work
has established that solids concentrations of 1.5 per cent by weight are
readily attainable and that concentrations of 2.0 per cent or more are pos-
sible.
Experimental Results. Phosphate removal data for the three extended runs
on final effluent are presented in Figures 5, 6 and 7. Corresponding total
plant flows are given in Table 5.
Run I was initiated just after an extended period of excessive rainfall.
Flow through the plant had decreased from over six times the design flow
to about three times the design flow (1.65 MGD). Relatively low phosphorus
levels were observed during this period of high flow. There does not ap-
pear to be any explainable reason for the relatively high levels of TFP ob-
served and the correspondingly high levels of "condensed phosphates" (TFP-
OP). It would appear that almost no conversion of CP occurred at the high-
est level of TP encountered during this run, but in the absence of CP values
for the raw sewage this cannot be verified.
With 200 mg/1 alum feed, the MBF reduced TP from an average of 5.67 to 0.28
mg/1 or a reduction of 94.8 per cent despite the relatively high proportion
of CP. OP was reduced from an average of 2.97 to 0.005 mg/1 or 99.8 per
cent.
Phosphorus data for Run II (Figure 6) present quite a different picture.
The expected diurnal variations are evident and the peaks correspond to a
lag of 5 to 6 hours behind the lift pumps at the wet well. Unexpected,
however, were peak TP concentrations ranging up to 25.7 mg/1, and also the
relatively large amount of TFP, particularly toward the end of the run.
The high degree of removal (90 per cent) of TFP and TP at an Al/P ratio of
about 0.7 appears to be particularly worthy of note (1700 hr, 9/23, Fig. 6).
29
-------
ALUM 200 mg/l
gl ANIONIC POLYELECTROLYTE 0.2 mg/l
7/30/69 to
INFLUENT ($CE)
TP
TFP
PHOSPHORUS LEVELS RUN I
30
-------
22
20
It
16
14
O
k
-------
ALUM 2OOmf/l
ANIONIC POIYSLICTROLYTI O.5 ntf/l
» 10/10/69
O
MiF IPPIUINT
PHOSPHORUS LEVELS RUN III
TIMI
-------
Table 5.
PLANT FLOWS DURING EXTENDED RUNS ON FINAL EFFLUENT
Time
0200-0800
0800-1400
1400-2000
2000-0200
0200-0800
0800-1400
1400-2000
2000-0200
0200-0800
0800-1400
1400-2000
2000-0200
0200-0800
Run I
(7/30 - 8/1/69)
1.87
2.41
2.08
1.78
1.32
1.52
1.29
1.23
0.89
1.12
FLOWS (MGD)
Run II
(9/23 - 9/26/69)
0.28
0.66
0.54
0.58
0.27
0.61
0.50
0.57
0.26
0.59
0.50
0,56
0.23
Run III
(10/8 - 10/10/69)
0.30
0.67
0.53
0.62
0.31
0.70
0.55
0.62
0.30
33
-------
The constant alum feed of 200 mg/1 and high influent phosphorus concentra-
tions resulted in residual phosphorus concentrations which were considerably
higher than those obtained during the preliminary studies. This difference
can be attributed to the low Al/TP ratios which averaged 0,85 and ranged be-
tween 0.68 and 1.01. The average TP removal was 87 per cent. The removal
of TFP (89.9 per cent) and OP (88.2 per cent) was slightly better than the
removal of TP, as shown in Table 6.
In Run III, influent phosphorus levels were still quite high compared with
earlier baseline work but slightly lower than the levels of Run II. Higher
degrees of removal were obtained for TP (91 per cent), TFP (92 per cent)
and OP (92 per cent), as shown in Table 6. These removals were achieved
with average Al/P ratios of 1.08 for TP, 1.22 for TFP and 1.28 for OP, again
indicating quite efficient utilization of alum.
The results of these experiments demonstrate the performance reliability of
the MBF system under a wide range of influent conditions. During this run,
the TP removal efficiency was never less than 85 per cent and only two OP
and two TFP values were less than 90 per cent removed.
Other Parameters. Table 7 summarizes data on alkalinity, pH, TSS and FSS
for all three extended runs. 8005 data are included for two of these runs.
Total suspended solids removals were comparable to those previously reported
with average reductions ranging from 60 to 75 per cent and TSS values in the
processed final effluent averaging between 9 and 13 mg/1.
Average 6005 reductions were 88 and 72 per cent with average residual
values of 3 and 8 mg/1. It was observed that the highest percentages of
removal, as might be expected, were associated with the largest amounts of
TSS and BODs available in the final effluent.
34
-------
Table 6.
SUMMARY OF PHOSPHORUS REMOVAL FOR CONTINUOUS RUNS
Run No.
Run I
Run II
Run III
Avg.
Mln.
Max.
Avg.
Mln.
Max.
Avg.
Min.
Max.
In
5.67
2.92
8.91
20.9
17.2
25.7
16.75
11.59
21.57
Al/P
3.07
1.95
5.96
0.83
0.68
1.01
1.08
0.81
1.50
TP
Out
0.28
0.14
0.40
2.8
1.7
4.6
1.50
0.61
2.75
TL Red.
94.8
88.6
99.0
86.8
78.5
91.8
91.0
85.1
96.7
TFP
In
5.09
2.11
8.38
16.9
12.4
21.0
13.88
11.14
17.72
Al/P Out
3.42 < 0.005
2.08
8.25
1.03 1.7
0.83 0.7
1.40 2.8
1.22 1.22
0.98 0.40
1.53 2.31
TL Red.
99.9
89.9
84.3
94.5
92.0
87.0
96.4
OP
In
2.97
1.45
4.88
14.3
11.4
15.9
13.73
11.73
16.30
Al/P Out
5.86 <0.005
3.57
12.00
1.22 1.7
1.09 0.8
1.53 2.7
1.28 1.15
1.07 0.39
1.53 2.34
7. Red.
99.8
88.2
82.6
93.0
91.9
85.6
96.5
-------
Table 7.
SUMMARY OF TREATMENT PERFORMANCE FROM EXTENDED RUNS
PH
Run No.
Run I
Run II
Run III
Avg.
Min.
Max.
Avg.
Min.
Max.
Avg.
Min.
Max.
In
6.7
7.1
7.3
7.2
7.4
7.2
7.1
7.3
Out
5.9
6.3
6.8
6.7
7.0
6.8
6.7
6.8
Alkalinity
In Out
90
79
98
170
155
194
186
164
220
17
12
24
98
83
116
111
89
139
Total
Suspended Solids
In
36
11
77
32
18
40
37
19
65
Out
10
8
12
13
5
25
9
3
20
7. Red.
61
0
84
60
34
83
75
50
91
7. Fixed
Suspended Solids
In
32
20
45
11
3
30
11
0
22
Out
60
30
100
41
0
63
17
0
46
In
25
12
46
31
13
40
BOD
Out
3
2
4
8
4
14
7. Red.
88
80
96
72
46
84
-------
PHASE II - MBF TREATMENT OF UNSETTLED TRICKLING FILTER EFFLUENT
Experimental Plan
The satisfactory filtration of secondary effluent with 200 mg/1 of alum,
without presettling per se, demonstrated the capability of the MBF system
for handling high solids loadings. Filtration of unsettled trickling fil-
ter effluent was evaluated to achieve greater utilization of this removal
capability. If feasible, comparable removals of phosphorus, TSS and 8005
could be achieved with greater economy through elimination of the need for
final clarifiers in the design of future trickling filter plants.
The pilot plant was run for two 1-week periods in August and September.
Jar tests indicated that the alum dosage could be maintained at 200 mg/1.
This also permitted comparison of the performance and operation with the
results obtained on final effluent. Except for some minor difficulties
in maintaining constant flow for extended periods of time, no exceptional
operation problems were encountered.
Experimental Results
Phosphorus Removal. The phosphorus removal results and other performance
parameters evaluated during the two weeks of operation are summarized in
Table 8. As expected, the concentrations of wastewater constituents were
generally higher than had been observed in the final effluent. The total
influent phosphorus averaged 19.1 mg/1 P and had a maximum concentration
of 28.6 mg/1 P. These values are comparable to those observed during Runs
II and III. The concentration of particulate phosphorus in the unsettled
trickling filter effluent (4 mg/1 P) was significantly higher than it was
in the settled effluent (1.4 mg/1 P). The trickling filter effluent data
also appear to differ from the settled effluent data in that a higher de-
gree of hydrolysis of CP may have occurred, based on the average values
for OP and TFP. However, when TP and TFP concentrations were unusually
high, the degree of hydrolysis may have been less complete. Absence of
data on the corresponding levels of TFP, OP and CP in the raw sewage pre-
clude more definite conclusions on this. Regardless of the form of the
phosphorus, however, MBF removals were good.
During this limited period of study, the average phosphorus removal ef-
ficiencies were better than those obtained during treatment of the final
secondary effluent. As summarized in Table 8, phosphorus removals ranged
between 90 and 98 per cent with averages of about 95 per cent.
Comparison of the removal of the filterable phosphorus at approximately
equivalent influent levels to those in final effluent showed that treat-
37
-------
Table 8.
SUMMARY OF TREATMENT PERFORMANCE - UNSETTLED TRICKLING FILTER EFFLUENT
TF Effluent
Parameter
Phosphorus - mg/1 P
Total
Filterable
Ortho
pH
Total Alkalinity -
mg/1 as CaCO-j
Total Suspended Solids
Fixed Suspended Solids -
Z of Total
Turbidity - JTU
BODc
Ranee
10.1
8.2
7.2
7.1
151
63
3
22
30
- 28.6
- 21.3
- 16.7
- 7.3
- 208
- 212
- 22
- 83
- 223
AVK.
19.1
14.9
12.4
7.2
183
86
8.3
43
55
MBF Effluent
Ranee
0.25
0.01
0.01
6.3
29
0.3
9
1.0
1.0
- 2.70
- 2.60
- 2.00
- 7.0
- 121
- 17
- 50
- 8.8
- 9.0
AVR.
0.99
0.62
0.53
6.7
89
7.1
24.7
3.73
3.8
Avg.
Reduction
a)
95.1
96.2
96.1
91.4
91.3
91.9
-------
ment of the unsettled effluent was slightly more effective. The average
filterable phosphorus concentration was reduced from 14.9 mg/1 P to 0.62
mg/1 P in the treatment of the unsettled trickling filter effluent. The
same degree of removal was evident for the ortho fraction.
Thus, it is indicated that the higher solids loading did not interfere with
the chemical reaction of the aluminum and phosphorus. It is possible that
a slight benefit may have been derived from formation of a greater amount
of floe which may also have served to remove some of the phosphorus through
sorption.
Other Parameters. As shown in Table 8, the amount of alkalinity consumed
by the aluminum was slightly higher during the period of treatment on the
unsettled trickling filter effluent. On the average, total alkalinity was
reduced by 94 mg/1 as CaCOs as compared to 70 mg/1 as CaC03 from the final
effluent data. This is also shown in the slightly lower pH values of the
MBF effluent although the difference is only 0.1 unit.
As expected, higher suspended solids were encountered in the unsettled
trickling filter effluent, averaging 86 mg/1 whereas the final secondary
effluent values averaged about 50 mg/1. A greater portion of the unset-
tled effluent solids was organic in nature with the fixed suspended solids
being only 8,3 per cent of the total. The suspended solids were reduced
to an average of 7 mg/1 with the fixed suspended solids about 25 per cent
of the total.
The reduction of BOD5 in unsettled trickling filter effluent had been ex-
pected to be greater than for final effluent because of the larger amount
of suspended solids with which BOD could be associated. However, the de-
gree of removal, an average of 92 per cent versus an average of 80 per
cent for final effluent, was better than expected, and average 800$ values
in the MBF processed effluent were 3.8 mg/1 for unsettled and 12 mg/1 for
final effluent, respectively. It should be noted that the figure for final
effluent encompasses a substantially larger number of samples and wider
range of wastewater treatment plant flows and strengths.
During one 24-hour period of operation, a record was kept of the volumes
of sludge pumped and p.er cent solids of the resulting sludge. Based upon
these measurements, the sludge volume was 2.8 per cent of the treated flow.
The per cent solids of the sludge was 0.48 and the fixed solids content was
66.6 per cent of the dry solids. However, these values are subject to the
sane limitations discussed earlier in connection with sludge collection and
are not suitable for design purposes.
39
-------
PHASE III - MBF TREATMENT OF PRIMARY EFFLUENT
Experimental Plan
Success of MBF treatment of unsettled trickling filter effluent gave added
interest to the study in which primary settled wastewater was processed.
This was a two-week study, during which the influent to the MBF was taken
from the effluent channel of the primary settling tank. The objective of
this study was to determine the feasibility of applying the MBF ahead of
the biological secondary treatment. The plan included evaluation of the
other treatment parameters as well as the removal of phosphorus. Besides
operational feasibility, the difference in the relative character of the
various phosphorus forms would also be evaluated.
Again, the pilot plant was operated at a constant alum dosage of 200 mg/1
and an anionic polyelectrolyte dosage of 0.5 to 0.75 mg/1. These condi-
tions were determined by comparing MBF pilot plant operation with jar
testing shown in Table 9. They also permitted comparison with the earlier
work on unsettled trickling filter and final effluent.
Experimental Results
Phosphorus Removal. The data collected from this period of operation are
summarized in Table 10. The general levels of phosphorus, with the average
of 14.6 mg/1, were within the same range as those found for final effluent.
However, the relative forms were markedly different. Orthophosphate was
about 92 per cent of the TFP in the final effluent. In the primary ef-
fluent, OP comprised only 73 per cent of the TFP. This is to be expected
due to some subsequent hydrolysis of CP as it passes through the trickling
filter. The relative percentage of phosphorus associated with the solids
was similar to that found in the final effluent.
Total phosphorus removal averaged 93.1 per cent under a wide range of in-
fluent concentrations (7.0 to 20 mg/1 as P). At influent concentrations
of less than 10 mg/1, the TP removal averaged 96.6 per cent. The data
showing average removals of 95.8 per cent for OP and 96.1 per cent for
TFP are equally promising. The latter result indicates fairly complete
precipitation and removal of a relatively high proportion of CP. These
removals are better than those obtained in the treatment of the final ef-
fluent although there is no obvious reason that this should be so.
If appearance is any indication, the MBF was substantially effective in
the clarification of primary effluent as shown in Figure 8.
The character of the primary effluent varied widely during the test with
TSS ranging between 51 and 114 mg/1 and BOD5 between 50 and 103 ing/1.
41
-------
Table 9.
COMPARISON OF JAR TESTS AND
MBF PROCESSING OF PRIMARY EFFLUENT
Primary
Effluent
Date - 8/21/69
Alum Dosage - mg/1
Anionic Polyelectrolyte
Dosage - mg/1
pH 7.3
Total Alkalinity - 164
mg/1 CaCO^
Turbidity - JTU 36
Orthophosphate - 6.31
mg/1 P
Per Cent Removal
Total Phosphate - 9.7
MBF
Effluent
200
0.5
6.8
51
0.6
0.03
97.9
0.12
Jar Test Effluent
1
200
0.5
6.7
84
1.5
0.04
--
0.25
2
200
0.5
6.7
84
2.5
0.05
--
0.34
3
225
0.5
6.7
76
1.5
0.02
--
0.19
tng/1
Per Cent Removal -- 99.6
42
-------
Table 10.
SUMMARY OF TREATMENT PERFORMANCE - PRIMARY EFFLUENT
Primary Effluent
Parameter
Phosphorus - mg/1 P
Total
Filterable
Ortho
PH
Total Alkalinity -
mg/1 CaC03
Total Suspended Solids
Turbidity - JTU
BOD5 - ng/1
BOD5 (filterable)
Range
7.0
5.5
4.1
7.0
127
51
33
50
10
- 20.3
- 18.1
- 15.0
- 7.3
- 229
- 114
- 83
- 103
- 36
AYR.
14.6
13.2
9.76
7.2
179
77
53
67
23
MBF Effluent
Ranjje
0.08
0.02
0.02
6.5
51
1
0.65
3
- 4.00
- 2.80
- 2.80
- 6.8
- 145
- 27
- 8.5
- 25
--
AYR.
1.13
0.58
0.38
6.7
99
11
3.8
12
Avg.
Reduction
m
93.1
96.1
95.8
--
44.6
86.5
92.5
81.9
-------
Figure 8. COMPARISON OF UNTREATED AND MBF
PROCESSED PRIMARY EFFLUENT
44
-------
Other Parameters. As shown in Table 10, the average suspended solids and
BOD. of the primary effluent were 77 and 67 mg/1, respectively, which may
be considered to be lower than the levels generally found in primary set-
tled domestic sewage. The resulting MBF effluent had a suspended solids
concentration of 11 mg/1 and a 6005 of 12 mg/1. These removals are ex-
tremely encouraging for a chemical coagulation-filtration system.
In considering removals of suspended solids and 8005, filterable BODs was
run on the primary effluent and these data are also included in Table 10.
Since the MBF effluent BOD was lower than the filterable BOD of the in-
fluent to the MBF, it seems likely that some of the BOD associated with
colloidal materials was being removed by the chemical treatment.
Within the limits of measurement, these results were obtained with no ap-
parent decrease in flow rate through the MBF pilot plant system, which
indicates that the alum-reaction products (i.e., Al-colloid precipitates,
Al (011)3, Al PO^, etc.) are probably still the rate-governing factor.
Attempts to accumulate and measure the volume and concentration of sludge
were again unsuccessful so a new tack was taken. Estimations of the vol-
umes of the resultant alum sludge were made from settling tests using 2-
liter graduated cylinders carried out on the waste wash water. On the
basis of these tests, the volume of sludge was approximately 0.9 per cent
of the treated volume. After one-hour settling, the sludge density was
about 0.65 per cent with a 48.9 per cent of the dry solids being ash.
This concentration of solids is probably slightly lower due to wall ef-
fects and other inhibitions than that which would be obtainable from
steady-state operation, but it appears to have better validity than the
intermittent data from the pilot plant.
45
-------
PHASE IV - MBF TREATMENT OF RAW SEWAGE
Experimental Plan
The successful treatment of primary effluent created increased interest
in establishing the potential of the MBF to process raw sewage. While the
primary emphasis was still on phosphorus removal, evaluation of suspended
solids and BOD removal in the light of the experience with primary efflu-
ent assumed almost equal importance.
Since alum at the 200 mg/1 level had proven effective for primary efflu-
ent, this same level was used for the work with the raw wastewater. This
had the advantage of making data from all of the levels of wastewater treat-
ment studied comparable. The same anionic polyelectrolyte also was used at
the 0.5 to 0.75 mg/1 level.
Problems and Limitations. Because of limitations requested by the host
plant staff, MBF operation during this period was limited to 8 to 12 hours
when raw sewage was available. Filtration of raw sewage was attempted for
two weeks, although usable data were only obtained for the second week of
operation.
During the initial operation on raw sewage, difficulties in obtaining a
reliable flow to the pilot plant unit were encountered. The main problem
was the clogging of the suction lines and pump. After rearrangement of the
piping and shifting the point of withdrawal, operation of the unit was main-
tained over a period of one week. The raw sewage was taken from the influ-
ent channel after the comminutor and just ahead of entrance to the Parshall
flume. A coarse screen was placed across the suction inlet.
Experimental Results
Phosphorus Removal. The performance data collected over the week of oper-
ation on raw sewage are presented in Table 11. The total phosphorus level
of the raw sewage varied between 11.6 and 41.1 mg/1 P with an average of
21.5 mg/1 P. Only 57 per cent of the TP was in the OP form and about 25
per cent was associated with the suspended solids. Removal of TP averaged
91 per cent and for OP was 95.2 per cent. A single high effluent value of
5.40 mg/1 was observed. Of this, 4.0 mg/1 was associated with the effluent
suspended solids. For no apparent reason, very poor flocculation was noted
during this period. The average ratio of Al/TFP was 1.01. The Al/TP ratio,
which was even lower at 0.81, looks extremely promising.
Other Parameters. A very substantial reduction in BOD was concurrently ob-
tained as shown in Table 11. The influent level was reduced from an aver-
age of 115 mg/1 down to 19 mg/1. The maximum influent level of 153 mg/1
47
-------
p-
00
Table 11.
SUMMARY OF TREATMENT PERFORMANCE - RAW SEWAGE
Raw Sewage
Parameter
Phosphorus - mg/1 P
Total
Filterable
Ortho
PH
Total Alkalinity -
rag/1 CaC(>3
Total Suspended Solids
Fixed Suspended Solids -
% of Total
Turbidity - JTU
BOD5 - mg/1
BODc (filterable)
Ranee
11.6
9.1
8.8
7.1
170
110
6.9
63
84
19
- 41.1
- 39.7
- 20.4
- 7.2
- 217
- 214
- 25.0
- 200
- 153
- 41
Avg.
21.5
18.6
13.2
--
185
156
12.6
119
115
32
MBF Effluent
Range
0.49
0.07
0.04
6.8
82
7
14.3
4.5
7
--
- 5.40
- 1.50
- 1.30
- 7.1
- 131
- 69
- 50
- 48
- 46
Avg.
2.16
0.79
0.57
--
106
27
35.4
16
19
--
Avg.
Reduction
(7.)
91.1
95.8
95.2
--
42.7
82.6
--
86.5
83.5
--
-------
was reduced by the MBF treatment to 28 mg/1. In all samples, the chemical
filtration process removed more than just the BOD associated with the sus-
pended solids. The filterable BOD of the raw sewage, which averaged 32
mg/1, was about 60 per cent higher than the MBF effluent BOD. It is be-
lieved that most of the additional BOD removal was associated with col-
loidal materials which were coalesced by the chemical addition and subse-
quently removed by filtration.
Comparable reductions in TSS were obtained with the influent average of
156 being reduced to 27 mg/1. The character of the solids in the latter
was also different with 35 per cent fixed solids as opposed to 12.6 per
cent in the influent, again demonstrating a substantial removal of organic
matter.
These levels of removal were of considerable interest since they were at
least as good as those being effected by the present trickling filter plant
with the additional benefit of better than 90 per cent TP removal.
Physical Observations. A very significant difference in the operational
characteristics of the system was observed with the raw sewage as compared
to operation on either the secondary or primary effluents. The variations
in filter outflow and head in relation to the face cutting sequence were
very pronounced. Upon removal of the clogged filter interface, the outflow
rapidly increased and the head sharply decreased. An outflow of 3 gpm to
17 gpm was observed and resultant changes in head of as great as 12 inches
were common. As the frequency of cutting (as often as every two pushes of
the filter bed) was increased, the fluctuations appeared to level off. The
clogging of the filter bed under these conditions was probably a result of
the buildup on the interface of the macerated paper and other fibrous solids
contained in the raw influent.
During this short period of operation, there was no visible buildup of
grease within the system.
Required Alum Dose for Phosphorus Reduction. It has been reported(5) that
Al/P ratios of 1.25 to 1.75 mg/1 are needed to precipitate soluble phos-
phorus and that, therefore, the required dose of alum is reasonably pro-
portional to the soluble phosphorus concentration. Other literature ref-
erences support this theory.
The work at Bernards Township tends to show that this theory is somewhat
oversimplified. It was noted in Figure 4 that the orthophosphate concen-
tration was 4.4 mg/1 as P. Taking an Al/P ratio of 1.5 as a median value,
an alum dose of 75 mg/1 should have been sufficient to remove 90 per cent
or more of the soluble phosphorus. However, 90 per cent removal did not
occur until a dose of 125 mg/1 (Al/P - 2.5) alum was applied.
Figure 9 shows the relationship of Al/P ratio to orthophosphorus removal
at various alum doses. From this it can be seen that removals of 90 per
49
-------
I
100.
o
X
act.
o
70!
60|
Q 2OO mg/| ALUM MBF
ISO mg/| ALUM MBF
ISO ntf/| ALUM JAR TEST
100 mfl/l ALUM MBF
mf/| ALUM JAR TEST
.* IjO
FIGURE 9.
1.2
2.0
2.2
1.4 1.6 1.8
Al/P RATIO
RELATIONSHIP OF ALUM DOSAGE TO REMOVAL EFFICIENCY
2.4
2.6
-------
cent or better at an Al/P ratio of 1.25-1.5 cannot be achieved with a dose
of 100 mg/1 alum while 90 per cent removal is achieved at a ratio of 1.2
with 200 mg/1 alum.
This phenomenon was observed with all of the sources, i.e., final effluent,
primary effluent, etc., at the Bernards plant. Parenthetically, it can be
noted that similar effects have been observed at other treatment plants
having either trickling filters or the activated sludge process.
It would appear, then, that two conditions must be met to insure some spe-
cific level of soluble phosphorus precipitation with alum: (1) an Al/P
ratio of 1.2 or more for 90 per cent reduction, and (2) a minimum alum
dose of the order of 200 mg/1 (for Bernards). These data lead to a hypoth-
esis that sewage has an "alum demand" which must be satisfied before soluble
phosphorus can be effectively removed.
51
-------
COST ESTIMATES
General Considerations of MBF Capital Costs
All cost estimates in this report are based on a 1.0 MGD MBF system. Typ-
ically, the cost per million gallon of installed capacity will be less in
larger installations and more in smaller installations.
A "system" consists of all of the components, i.e., moving bed filters,
pumps, chemical feed systems, sand washers, wash water reclaim tank, etc.,
to make an operating system, F.O.B. shipping point. Sludge dewatering
equipment, when required, is supplied as a separate equipment package on
the same basis.
While the cost of system components can be established with precision as
of the date of this report, installed costs are of necessity less precise.
The equipment has been designed to minimize installation costs. Even so,
these have been found to vary widely based on local construction labor
costs; whether housing is needed or not and, if so, the type building spe-
cified by the owners; and the cost of connecting into the rest of the sys-
tem, i.e., whether wet wells and lift pumps are required, etc.
The cost of housing is probably the most difficult to estimate since in
some climates no housing except a shed covering of the main control panel
and chemical feed system would be required. In intermediate climates,
these components should be housed and some exposed piping and metal sur-
faces would require insulation or heat tracing. In climates having long
periods of extreme cold weather full housing would be required. For this
report, the first alternative has been used for estimating installed costs,
Amortization
Recent publications of the FHQ&(7) have used amortization of capital costs
at 4-1/2 per cent and a 25-year period. This same basis has been used so
as to permit direct comparison with other processes.
Operating Costs
Chemicals. Similar considerations arise with respect to operating costs
but these are more easily resolved. Where it is available commercial liquid
alum (48 per cent solution) may be as much as $10 to $12 less per ton of
alum content than dry alum. At the 200 rag/1 level, this represents savings
of 0.8 to 1.0 cents per 1000 gal. processed. For the purposes of this re-
port, the cost of alum has been assumed to be $0.025 per pound.
53
-------
Anionic polyelectrolytes also have a broad cost spread. Two of the several
polyelectrolytes used In this study were found to work well, but the one
which could be applied at 0.2 mg/1 and thereby halve the polyelectrolyte
costs is not yet commercially available. So, costs are based on use of 0.5
mg/1 of commercially available material at $1.50 per pound.
Power. The installation has approximately 65-hp connected load, excluding
two 10-hp and one 5-hp standby pumping units, and the equivalent of 42-hp
continuous load. Power cost has been assumed at $0.01 per kw-hr.
Operation and Maintenance. Since the equipment is automated on a fail-safe
basis, a 1.0 MGD installation requires only two man-hours per day. Chemical
tanks must be refilled, recorder charts changed and pumps and general oper-
ating conditions reviewed daily. All labor costs were estimated on a $3
per hour rate.
Non-routine maintenance requirements were estimated to be one man-hour per
day. Maintenance material is estimated at 30 per cent of operating and
maintenance labor.
Cost Summary
Capital, dollars
MBF Process Equipment $188,000
Installation (erection, piping, wiring on slab
with shed roof covering key components) 36.000
MBF Process Total $224,000
Outside Work (lift pumps, piping, etc.) 40.000
TOTAL ESTIMATED COST $264.000
Process Cost, cents/1000 gal.
Amortization - 4%£ and 25 yr 4.9c
Alum - 200 mg/1 @ 2.5c/lb 4.2
Polyelectrolyte - 0.5 mg/1 @ $1.50/lb 0.6
Power 0.8
(Subtotal) 10.5
54
-------
Process Cost, cents/1000 gal. (cont'd)
(Subtotal)
Operating and Maintenance Labor @ $3/hr
Supervision and Payroll Overhead
Maintenance Material
TOTAL ESTIMATED PROCESS COST
Ultimate Disposal of Sludge
With approximately 95 per cent of the phosphorus originally present in the
wastewater fixed and concentrated in the underflow from the wash water re-
claim tank, an unusual opportunity is provided for "ultimate" disposal of
phosphorus. One such method involves dewatering with a rotary precoat fil-
ter to a damp solid which can then be used as landfill, for example.
Data from the present study have been supplemented with data from other
larger-scale operations to enable preparation of preliminary estimates of
the cost of such a dewatering operation. All of the operating costs are
given in terms of the volume of the original wastewater. To these esti-
mates would have to be added the cost of trucking and landfill operations
which will be governed by local considerations.
Basis. This estimate assumes the dewatering of 10,000 gpd (1.0 per cent
of throughput) of sludge containing 1.5 per cent solids by weight.* The
dewatering requires a 75-sq ft rotary vacuum precoat filter on a 24-hr
per day basis with two .hours off stream for washup and precoating and
requires an operator only during that time. HYFLO SUPER-CEL** is the
filter medium and will be consumed at the rate of 0.156 lb/hr/ft2. The
damp cake, containing about 30 per cent solids, can be deposited directly
in a hopper body for trucking to a disposal site. The filtrate from the
dewatering operation is suitable for addition to the MBF processed efflu-
ent stream without recycling.
*These estimates will be favorably influenced by increases in feed solids
concentration and decreases in volume of sludge since rotary vacuum pre-
coat filtration rates are not linearly decreased with increased solids
concentration. The above estimates are based on the best currently avail-
able information.
**Product of Johns-Manville.
55
-------
Cost Summary
Capital, dollars
Equipment $22,000
Installation 8.000
TOTAL INSTALLED COST $30.000
Process Cost, cents/1000 gal.
Amortization - 4%7. and 25 yr 0.55C
Filter Aid @ 4.5c/lb ,. 1.3
Power 0.2
Operating and Maintenance Labor 0.6
Supervision and Payroll Overhead 0.2
Maintenance Material 0.1
TOTAL ESTIMATED PROCESS COST 2.95C
56
-------
BIBLIOGRAPHY
' 'cleaning Our Environment^ The Chemical Basis for Action. American
Chemical Society, pp 131-2 (1969).
' '"Clean-Up Program Approved for Lake Michigan," Environmental Science
& Technology. Vol. 2, p 397 (1968).
^Sawyer, C. N., "The Need for Nutrient Control," Journal Water Pol-
lution Control Federation. Vol. 40, pp 363-70 (1968).
' 'Dryden, F. D. and Stern, G., "Phosphate Reduction for Limiting Algae
in Lakes of Renovated Wastewater," Presented at the 152nd Na-
tional Meeting of the American Chemical Society, New York, N.Y.
(September 1966).
^ 'Convery, J. J., "The Use of Physical-Chemical Treatment Techniques
for the Removal of Phosphorus from Municipal Wastewaters,"
Presented before New York WPCA (January 29, 1970).
^ 'Standard Methods for the Examination of Water and Wastewater." 12th
Edition (1965).
^ 'Smith, Robert and McMichael, W. F., "Cost and Performance Estimates
for Tertiary Wastewater Treating Processes," FWQA Publication
TWRC-9 (1969).
57
-------
BIBLIOGRAPHIC:
ACCESSION NO.
Johns-Manville Products Corporation, Phosphorus Removal Using Chem-
ical Coagulation and a Continuous Countercurrent Filtration Process,
Final Report FWQA Contract No. 14-12-154, June 1970
ABSTRACT:
The Johns-Manville Moving Bed Filter, a continuous precipitation
and countercurrent filtration process, was evaluated for the removal
of phosphorus from municipal wastewater.
Using alum and an anionic polyelectrolyte, the process was found to
effectively reduce total phosphorus (TP), orthophosphate (OP) and
condensed phosphate (CP) over a wide range of influent phosphorus
concentrations. Preliminary work using jar tests established an alum
dose of 200 mg/1 (17.4 mg/1 Al, molar ratio of Al/P is 27/31) as ef-
fective for removal of 90 per cent TP from the secondary clarifier
effluent of a trickling filter plant. This removal efficiency could
not be sustained with an alum dose of 200 mg/1 when higher TP levels
were encountered. With total phosphorus concentrations on the order
of 25 to 28 mg/1 as P (Al/P * 0.6-0.7), the TP removal efficiency
averaged 90 per cent. With lower total phosphorus concentrations,
removal efficiency averaged 95 per cent and ranged up to 99 per cent
(Al/P - 1.2-2.6).
Substantial reductions in final effluent total suspended solids (TSS)
and 5-day biochemical oxygen demand (6005) were also obtained. At an
alum dose of 200 mg/1, TSS reduction averaged 70 per cent and BOD5
reduction 80 per cent. If phosphorus removal were not a design con-
sideration, the reduction of TSS and BOD5 could be achieved with
lower alum doses.
The 200 mg/1 alum dose was also found to be equally effective for
removal of phosphorus from raw sewage and primary effluent with the
added capability for removing substantial portions of the TSS and
BOD5. In short studies on these streams, effluent as good or better
than the final effluent from the trickling plant was obtained.
Costs for a 1.0 MGD plant are estimated to be $264,000 for capital
and 12.0 cents per 1000 gal for total operating cost. These costs
would be about the same for raw sewage, final effluent or the two
intermediate levels of prior treatment studied.
Ultimate disposal of the phosphorus-containing sludge could be
achieved by a dewatering and landfill operation. Dewatering by
means of a rotary vacuum precoat filter would require an estimated
capital expenditure of $30,000 and total cost would be 3 cents per
1000 gal of original wastewater treated.
This report was submitted in fulfillment of Contract No. 14-12-154
under the sponsorship of the Federal Water Quality Administration.
KEY WORDS:
Phosphorus
Removal
Moving Bed
Filter
Sand Filtration
Solids Removal
BOD Removal
Alum Coagulation
Chemical Treat-
ment
Treatment
Costs
* U. S. GOVERNMENT PRINTING OFFICE ; 1970 O - 405-lit
------- |