EPA-600/2-77-012
January 1977
Environmental Protection Technology Series
METHODS FOR IMPROVEMENT OF
TRICKLING FILTER PLANT PERFORMANCE
Part II • Chemical Addition
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
EPA-600/2-77-012
January 1977
METHODS FOR IMPROVEMENT OF
TRICKLING FILTER PLANT PERFORMANCE
PART I I
CHEMICAL ADDITION
by
James C. Brown
Linda W. Little
University of North Carolina
Chapel Hill, North Carolina 27514
Contract No. 14-12-505
Project No. I 1010 DGA
Project Officer
Richard Brenner
Wastewater Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
ICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Munic.ipal Environmental Research
Laboratory, U. S. Environmental Protection Agency, and approved for publication,
Approval does not sign!fy that the contents necessarily reflect the views and
policies of the U, S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recommendation
for use.
I I
-------
FOREWORD
The Environmental Protection Agency was created because of increasing public
and government concern about the dangers of pollution to the health and
welfare of the American people. Noxious air, foul water, and spoiled land
are tragic testimony to the deterioration of our natural environment. The
complexity of that environment and the interplay between its components re-
quire a concentrated and integrated attack on the problem.
Research and development is that necessary first step in problem solution and
it involves defining the problem, measuring its impact, and searching for
solutions. The Municipal Environmental Research Laboratory develops new and
improved technology and systems for the prevention, treatment, and management
of wastewater and solid and hazardous waste pollutant discharges from munici-
pal and community sources, for the preservation and treatment of public drink-
ing water supplies, and to minimize the adverse economic, social, health, and
aesthetic effects of pollution. This publication is one of the products of
that research; a most vital communications link between the researcher and
the user community.
The studies described herein were undertaken to demonstrate the feasibility
of upgrading the overall performance of a typical high-rate trickling filter
plant through the addition of liquid aluminum suIfate (alum) to the secondary
clarifier. A concomitant benefit of a I urn addition is phosphorus precipitation
The technology emanating from this project should be thoroughly considered by
those charged with the responsibility of upgrading existing or designing new
trickling filter facilities.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
i i i
-------
ABSTRACT
An experimental program to explore potential methods for removing phosphorus
and generally enhancing trickling filter plant performance was conducted at
the Mason Farm Wastewater Treatment Plant, Chapel Hill, North Carolina. Pre-
liminary investigations included characterization of quality and quantity of
plant flows, jar testing with several coagulants (lime, alum, and iron salts)
and coagulant aids, and pilot studies to determine the effect of the point of
alum addition on phosphorus removal in a high-rate trickling filter system.
Follow-up full-scale studies utilized the Chapel Hill high-rate trickling fil-
ter plant which consists of two parallel identical trains of main-stream treat-
ment units. From January 25 through October 6, 1972, alum was added to the
influent of one final settling tank. During the 18 experimental periods, alum
dosage and influent flow rates to the dosed train were varied and phosphorus
removal, general plant performance, sludge production, and sludge digestion
performance were monitored.
Alum addition effectively removed phosphorus and enhanced overall plant per-
formance. Optimization of alum precipitation will require a flow-paced a I urn
feed system, restriction of average dry weather final settling tank surface
loadings to 20.4 m^/day/m2 (500 gpd/ft2), and inclusion of tertiary fine
solids removal facilities.
Alum sludge decreased the alkalinity and pH in the primary anaerobic digester
and led to liquid/solids separation problems in the secondary digester.
Separate facilities may be necessary for handling alum-humus sludge from the
fina I sett I ing tank.
This report was submitted in partial fulfillment of Project No. I 1010 DGA,
Contract No. 14-12-505, by the University of North Carolina under the sponsor-
ship of the U. S. Environmental Protection Agency. Research work conducted
during Part II of this project and reported herein covers the general time
span of mid-1971 to October 1972. Studies undertaken prior to the Part II
chemical addition experiments of this project were previously reported in
EPA-670/2-73-047a entitled "Methods for Improvement of Trickling Filter
Plant Performance - Parti - Mechanical and Biological Optima," August 1973.
-------
CONTENTS
FOREWORD i i i
ABSTRACT i v
FIGURES vi
TABLES !x
ACKNOWLEDGEMENTS xi
SECT I ON
I INTRODUCTI ON I
II CONCLUSIONS 3
III RECOMMENDATION 7
IV PRELIMINARY PHOSPHORUS REMOVAL INVESTIGATIONS 9
Characterization of Plant Influent 9
Jar Tests I 3
Pilot Plant Studies of Phosphorus Removal with Alum 44
V ALUM ADDITION TO CHAPEL HILL MAIN PLANT 60
The Chapel Hill Treatment Plant 61
Preparatory Work for Alum Treatment 64
Faci I ities for Alum Treatment at Chapel Hill 66
Sampling and Analysis 69
Description of Experimental Program and Performance
Results 71
Discussion of Results 81
VI REFERENCES 102
APPENDIX A. Abstract of Publication Resulting from Project 104
APPENDIX B. Abstracts of Theses Resulting from Project 105
-------
FIGURES
Number f.?£i
I Relationship Between pH and Calcium Hydroxide Dose 18
2 Relationship Between pH and Residual Total Inorganic
Phosphate in Various Plant Samples 19
3 Effect of Laboratory Filtration on Residual Total
Inorganic Phosphorus as a Function of Calcium
Hydroxide Dose and pH Level (0.45y Filter) 21
4 Effect of Laboratory Filtration on Residual Total
Inorganic Phosphorus as a Function of Calcium
Hydroxide Dose and pH Level (0.22y Filter) 22
5 Effect of Cationic Polyelectrolyte (Cat-Floe) Addi-
tion on Residual Total Inorganic Phosphorus and
Turbidity at a Calcium Hydroxide Dose of 80 mg/l
and a pH of 9.3 24
6 Effect of Sodium Fluoride Addition on the Relation-
ship Between pH and Residual Total Inorganic
Phosphorus During Lime Precipitation 25
7 Polynomial Regression Models for Phosphorus Removal
from Influent as a Function of Alum Dosage 28
8 Polynomial Regression Models for Phosphorus Removal
from Primary Effluent with Alum Addition 29
9 Polynomial Regression Models for Phosphorus Removal
from Trickling Filter Effluent with Alum Addition 30
10 Polynomial Regression Models for Phosphorus Removal
from Secondary Effluent with Alum Addition 31
II Phosphorus Removal from Plant Influent with Ferric
Chloride and Ferric SuIfate Addition 37
12 Phosphorus Removal from Trickling Fl Iter Effluent
with Ferric Chloride and Ferric Sulfate Addition 38
13 Removal of Total Suspended Solids and Turbidity
from Plant Influent with Ferric Chloride and Ferric
Sulfate Addition 39
vi
-------
FIGURES (Continued)
Number
14 Removal of Total Suspended Solids and Turbidity from
Trickling Filter Effluent with Ferric Chloride and
Ferric Sulfate Addition 40
15 Effect of pH on Fe Capture During Iron Addition for
Phosphorus Removal 46
16 Flow Diagram Trickling Filter Pi lot Plant for Single-
stage FiItration 50
17 Partial Flow Sheet for Chapel Hill Wastewater Treat-
ment Plant 62
18 Typical Diurnal Variation in Total Phosphorus Loading
to Final Clarifier and Corresponding AI Dosage
Pattern 65
19 Elevation View of Alum Feeding System 67
20 Total Phosphorus Removal as a Function of Flow and
Final Clarifier Overflow Rate for Experimental
Periods I Through 4, and 6 Through 12 82
21 BOD Removal as a Function of Flow and Final Clarifier
Overflow Rate for Experimental Periods I Through 4,
and 6 Through 12 83
22 Total Suspended Solids Removal as a Function of Flow
and Final Clarifier Overflow Rate for Experimental
Periods I Through 4, and 6 Through 12 84
23 Effect of Flow on Total Phosphorus Removal from Sorting
Analysis 89
24 Scattergram for Alum Concentration Dosage versus Percent
Total Phosphorus Removal 90
25 Scattergram for Ah Influent TP (Mole) versus Percent
Total Phosphorus Removal 91
26 Scattergram for Flow versus Percent Total Phosphorus
RemovaI 92
27 Effect of Alum Concentration Dosage on Total Phosphorus
Removal from Sorting Analysis 93
VI
-------
TABLES
Number
I Characteristics of Plant Influent - September 1969-
February 1972 (Monthly Averages) 1°
2 Ratio of Total Inorganic to Total Phosphorus Concen-
tration in Plant Influent ''
3 Phosphorus Concentration at Various Points in Plant
Flow Scheme, Mason Farm Treatment Plant, April 1970-
March 1971 l2
4 Diurnal and Weekday Variations in Influent Orthophos-
phate Concentration and Loading '4
5 Analytical Procedures 15,16
6 Jar Test Results for Lime Addition to Trickling Filter
Effluent 20
7 Effect of Cationic Polyelectrolyte on Phosphate Removal
from Plant Effluent at Constant Lime Dose, pH 9.3 23
8 Polynomial Regression Models of Choice for Phosphorus
Removal with Alum Addition 27
9 Alum Dosages Required for 97-98 Percent Removal of Total
Phosphorus 27
10 Ratio of Aluminum to Total Inorganic Phosphorus Required
for 97-98 Percent Removal of Total Phosphorus 32
II Ratio of Aluminum to Total Phosphorus Required for 97-98
Percent Removal of Total Phosphorus 32
12 Effect of Cat-Floe and Magnifloc on Alum Precipitation
of Phosphorus from Trickling Filter Effluent 34
13 Effect of Calgon WT-300 on Alum Precipitation of Phos-
phorus from Trickling Filter Effluent and Secondary
Effluent 34
14 Effect of Natron Floe Aid on Alum Precipitation of Phos-
phorus from Secondary Effluent 35
vi i i
-------
TABLES (Continued)
Number
15 Phosphorus Removal with Ferric Sulfate: Comparison of
NaOH and Ca(OH)2 for pH Control at pH 6 and pH 9 41
16 Phosphorus Removal with Ferric Sulfate: Comparison of
NaOH and Ca(_OHl2 for pH Control at pH 7 42
17 Phosphorus Removal with Ferric Sulfate; Comparison of
NaOH and Ca(OH)2 for pH Control at pH 8 43
18 Effect of pH on Phosphorus Removal and Colloidal Iron
Capture with Ferric Sulfate and Lime 45
19 Effect of Order of Addition of Iron and Lime on Phos-
phorus Removal and Colloidal Iron Capture 47
20 Comparison of Ferric Sulfate and Alum for Phosphorus
RemovaI 48
21 Design Conditions for Trickling Filter Pilot Units 49
22 Comparison of Alum Addition (100 mg/l) to Primary Clari-
fier Influent, Trickling Filter Influent, and Secondary
Clarifier Influent Ahead of Recirculation Takeoff Point 52,53
23 Comparison of Alum Addition (150 mg/l) to Primary Clari-
fier Influent, Trickling Filter Influent, and Secondary
Clarifier Influent Ahead of Recirculation Takeoff Point 54,55
24 Comparison of Alum Addition (200 mg/l) to Primary Clari-
fier Influent, Trickling Filter Influent, and Secondary
Clarifier Influent Ahead of Recirculation Takeoff Point 56,57
25 Comparison of Alum Addition (200 mg/l) to Primary Clari-
fier Influent and Secondary Clarifier Influent After
Recirculation Takeoff Point 58
26 Characteristics of and Design Parameters for Units in
Chapel Hill Wastewater Treatment Plant 63
27 Points of Sampling and Analyses Conducted for Main-Plant
Alum Addition Studies 70
28 Experimental Periods for Full-Scale Phosphorus Removal
Studies 72
29 Alum Dosage Programs 73
30 Main Plant Phosphorus Removal 74
ix
-------
TABLES (Continued)
Number Page
31 Main Plant BOD5 Removal 75
32 Main Plant Total Suspended Solids Removal 76
33 Main Plant Total Organic Carbon Removal 77
34 Quality of Primary Effluents from 1/27/72 to 8/30/72 85
35 Volumes and Characteristics of Sludges from Train
No. 2 (Alum) 95
36 Volumes and Characteristics of Sludges from Train
No. I CNo Alum) 96
37 Sludge Production Summary for 1/25/72 to 8/27/72 97
38 Conditions in Primary Digester During Alum Treatment
Investigation 98
-------
ACKNOWLEDGMENTS
Numerous individuals were involved in the various phases of the project de-
scribed in this report. The help and encouragement from the project officers,
Dr. Robert Bunch and Mr. Richard Brenner of the U.S. Environmental Protection
Agency, who suggested the full-scale alum studies, are gratefully acknowledged.
Appreciation is also expressed to Dr. James C. Lamb III, who directed the pro-
ject in its first year, and to Drs. A. Energin Eralp and Donald E. Francisco,
who served as Research Associates during some phases of the contract.
The research work on the project was initiated in July 1969 under Contract No.
14-12-505. The contract, scheduled to terminate on June 6, 1972, was extended
until October 6, 1972 to allow full-scale alum addition studies initiated in
January 1972 to be completed. That portion of the research work covered in
Contract No. 14-12-505 excluding the alum addition studies was reported in
METHODS FOR IMPROVEMENT OF TRICKLING FILTER PLANT PERFORMANCE - PART I -
MECHANICAL AND BIOLOGICAL OPTIMA, James C. Brown, Linda W. Little, Donald E.
Francisco, and James C. Lamb III, August 1973. (19).
Assistance of graduate students in this research is gratefully acknowledged.
Graduate assistants included George Budd, A. T. Rolan, Thomas Bates, William W.
Sun, Enrique J. LaMotta, Martin Strauss, Robert Hanson, Ronald Sims, and
Ronald Benton. The analytical laboratory staff, responsible for the numerous
analyses conducted, included William C. Walker, UNC Wastewater Research Center
Laboratory Supervisor; James E. Hayes, Cornelia Jones, William James, E. Patrick
Jessup, Robert Moore, and Bruce DiCintio of the UNC Wastewater Research Center
Laboratory; and Tony Owen and Susan Rappaport of the UNC Limnology Research
Laboratory.
The cooperation and support of Cities Service Company in the studies on ferric
chloride and ferric suIfate addition are appreciated.
The invaluable assistance of George Burns, research mechanic, and John Street,
plant operator, is acknowledged, as is the help rendered by Ernest Rogers and
Robert Parrish.
Special recognition is due Delores E. Plummer, who was responsible for the
typing and retyping of reports and publications emanating from the project.
XI
-------
SECTION
INTRODUCTI ON
Over 3,500 secondary municipal wastewater treatment plants in the United States
use trickling filters as the biological units. Many of these are high-rate
installations, characterized by relatively heavy rates of wastewater applica-
tion with recirculation of treated effluent to dilute influent before appli-
cation to the filter media. Removal of phosphorus is seldom greater than 10
percent, and usually little or no phosphorus removal occurs if digester super-
natant is discharged back through the plant.
Contract No. 14-12-505 had as its overall research objective the development
of information to help design engineers and operating personnel select among
alternatives for improving performance of trickling filter plants. This re-
port describes that portion of the contract research directed toward improve-
ment of phosphorus removal and concommitant enhancement of general plant
performance with the addition of phosphorus precipitation minerals to plant
flow units.
The general approach was based on laboratory, pilot-, and full-scale experi-
mental investigations. These investigations were conducted at the Mason Farm
Sewage Treatment Plant in Chapel Hill, North Carolina, operated for the Town
by personnel of the Department of Environmental Sciences and Engineering at
the University of North Carolina. The most recent plant enlargement (1968)
included modifications to provide unusual flexibility in full-scale operation,
as well as facilities for laboratory and pilot studies. The plant is designed
to permit operation as two separate parallel identical trickling filter trains
between which the influent flow can be divided in any desired proportion,
with capability for independent control of recirculation and other aspects of
operation in each. These unusual features allowed simultaneous full-scale
investigation of: (I) the effect of alum dosage on phosphorus removal, (2)
the effect of a I urn addition on overall plant performance, including BOD and
suspended solids removals, and (3) the effect of hydraulic loading on a I urn
precipitation.
Preliminary phosphorus removal investigations were conducted prior to the full-
scale studies. These investigations included a review of the literature perti-
nent to phosphorus removaI, characterization of the quality and quantity of
plant flows, jar testing with several coagulants (lime, alum, and iron salts)
and coagulant aids, and pilot studies to determine the effect of the point of
alum addition on phosphorus removal in a high-rate trickling filter system.
-------
After the preliminary studies and following consultation with the Project
Officer, full-scale plant studies were conducted with alum. Liquid alum was
added to the effluent from the trickling filter of one train only immediately
prior to its introduction into that train's final settling tank. Liquid a I urn
was not added to the other trickling filter train which served as an ideal
control system, particularly during those experimental phases when plant in-
fluent flow was equally divided between the two trains. Plant flows were
measured and determinations made of the constituents in the raw wastewater,
the primary and final effluents, and the various sludge and recycle streams.
In addition, digester performance and sludge production were closely monitored.
-------
SECTION I I
CONCLUSIONS
Investigations at the Mason Farm Wastewater Treatment Plant, Chapel Hill, North
Carolina, a typical high-rate trickling filter installation, demonstrated marked
improvement of phosphorus removal and concommitant enhancement of overall plant
performance with addition of phosphorus-precipitating minerals. The experi-
mental program included laboratory, pilot-, and full-scale studies. From the
results of these studies, the following conclusions were reached:
I. Characteristics of Plant Influent: The concentration of phosphorus and
other constituents in the plant influent varied diurnally, daily, monthly,
and seasonally. Diurnal variations were the most pronounced. For example,
considering diurnal variations in flow and orthophosphate concentration,
orthophosphate diurnal loading ranged from 9 to 112 kg/day (19 to 246 Ib/
day) as P, or from 16 to 204 percent of the daily average. The relative
proportion of inorganic phosphorus to total phosphorus increased during
treatment, physical removal of organic phosphorus in the primary clarifier
being responsible for much of this change. Diurnal variations in influent
quality were largely dampened by recirculation of trickling filter effluent
through the primary clarifiers so that the quality of the filter effluent
was more consistent.
2. Jar Tests: With the relatively soft water indigenous to the Chapel Hill
area, removal of total inorganic phosphorus to a residual of less than
I mg/l as P required 300-400 mg/I of lime Cas Ca(OH)2J with a corresponding
pH of II or greater. At pH levels substantially less than II, lime reacted
with soluble phosphorus to form a finely divided insoluble material, much
of which did not settle out during the jar test procedure, but was retained
on either a 0.45 y or a 0.22 y membrane filter. Addition of an appropriate
cationic polyelectrolyte enhanced coagulation and settling of finely divided
insoluble phosphorus and permitted total inorganic phosphorus removals to
residuals less than 1.0 mg/l at a pH of 9.3 to 9.5 and a lime dose of about
80 mg/l. Addition of fluoride had no effect on lime precipitation of phos-
phorus.
Jar test a I urn dosages required for 97-98 percent removal of total phosphorus
vary with the degree of prior treatment, with a dose of approximately 200
mg/l [as AI^SO.), • 18 H2OH required for plant influent and 150 mg/l for
secondary effluent. However, analysis of the results on the basis of Al:TP
indicate that a ratio of 1.5-1.6 (wt/wt) is required to effect the same
removal, regardless of the point in the treatment sequence. Coagulant aids
enhanced phosphorus removal to varying degrees.
-------
In jar tests with iron salts, Fed I I) requirements for removal of phos-
phorus, solids, and turbidity were lower with ferric chloride than with
ferric sulfate, but ferric sulfate produced less sludge and a stronger floe.
With iron salts alone, some iron remained in the supernatant. Experiments
on trickling filter effluent with ferric sulfate and alkalis (lime and
sodium hydroxide) indicated that this iron "leakage" could be eliminated
by control of pH to above 7. Lime was the more effective alkali. With a
trickling filter effluent total phosphorus concentration of 9 mg/l as P,
approximately 5C mg/l of iron and 90 mg/l of Ca(OH)2 were required to
obtain iron residuals of less than I mg/l.
3. Pi lot Plant Studies: Increased phosphorus removals were observed in trick-
ling filter pilot plant studies with increasing alum dosages up to 200 mg/l
as Al (SO ), • 18 hLO (the highest dose tested). Alum addition to trick-
ling fiIter influent or to trickling filter effluent above the takeoff
point of recirculation was less effective than addition to primary clari-
fier influent or to trickling filter effluent below the takeoff point of
recirculation. Overall removals of BOD,-, total organic carbon, and phos-
phorus were essentially the same when alum was dosed to primary clarifier
influent or to trickling filter effluent below the takeoff point of recir-
culation.
4. Full-Scale Studies: Following jar tests and pilot plant studies, full-
scale were initiated with alum addition. Liquid alum was added to the in-
fluent of the final clarifier of one of two parallel trains at the Chapel
Hill, North Carolina Wastewater Treatment Plant during the period from
January 25 through October 5, 1972. The other train was not dosed and
served as a control when influent flow was equally divided between the two
trains. The Chapel Hill plant is a typical high-rate trickling filter
facility. The two parallel trains consist of identical clarifier, trick-
ling filter, and final clarifier units. Single primary and secondary
anaerobic digesters treat the sludge from the two trains. The objective
of the experimental program was to explore the potential for achieving
efficient phosphorus removal and generally enhancing overall plant perfor-
mance with alum addition to this type of municipal wastewater treatment
plant.
During most of the experimental program, influent flow was divided equally
between the two plant trains. Alum was applied on a three-step per day
basis to approximately match the phosphorus loading entering the final
clarifier of the dosed train. Alum dosage varied from 143 to 245 mg/l,
and the Al:influent TP (mole) varied from 1.0 to 2.7 during the eighteen
experimental periods of the investigation. During periods of equal flow
division (15 of the 18 periods), influent flow to each train averaged
5,299 m3/day (1.40 mgd), recirculation flow around the filters 7,040 m3/day
(1.86 mgd), plant influent total phosphorus 11.9 mg/l, plant influent BOD,-
170 mg/l, plant influent total suspended solids 244 mg/l, and plant influ-
ent total organic carbon 180 mg/l. Corresponding final effluent concentra-
tions for the alum dosed train for the same 15 periods averaged 2.4 mg/l
total phosphorus (79 percent removal), 16 mg/l BOD5 (91 percent removal),
30 mg/l total suspended solids (88 percent removal), and 24 mg/| total
organic carbon (87 percent removal). These values are contrasted to the
-------
average final effluent concentrations of the non-dosed control train for
the 15 periods; total phosphorus 9.7 mg/l (18 percent removal), BOD5 39
mg/l (84 percent removal), total suspended solids 54 mg/l (78 percent re-
moval), and total organic carbon 54 mg/l (70 percent removal). During
five of the 15 equal flow division periods, small amounts of different
polyelectrolytes were added to the dosed train final clarifier in conjunc-
tion with alum. No improvement in performance was noted with the combina-
tion dosing over alum addition alone. The surface overflow rate of the
dosed final clarifier, more than either a I urn concentration dosage or Al:
influent TP, was the most significant operating parameter affecting solids
capture and overall removal efficiencies.
Sludge production for the two plant trains during the alum treatment in-
vestigation is summarized below:
Vo 1 ume
Total
Sol ids
Volati
Sol ids
, Pumped, ga 1 /mi 1 gal
Sol ids, %
Pumped, 1 b TS/mi 1 ga 1
le Fraction, %
Pumped, 1 b VS/mi 1 gal
Train No. 2
(Alum added)
6,275
3.57
1,894
67
1,265
Train No. 1
(Alum not
added)
4,748
3.85
1,483
76
1,120
Conversions:
I gal/mi I gal = 0.001 S,/m~
I Ib/miI gal = 0.12 g/m3
The mixture of alum and conventional primary and secondary sludges to the
primary digester resulted in a gradual decrease in the buffering capacity
of the system. Primary digester alkalinity decreased from a normal level
of over 2500 mg/l to about 1500 mg/l. Digester pH was less stable and
tended to drift downward. It was necessary to add lime to correct this
condition on one occasion. The primary digester functioned satisfactorily
during the experimental program provided the pH was maintained above 6.8.
The most serious problem encountered was the failure of the secondary di-
gester to produce a concentrated sludge for centrifugation and the inabil-
ity to secure a reasonable quality supernatant for return to the plant head
works. Secondary digester underflow (centrifuge feed) concentration de-
creased from a normal value of 6 to 7 percent solids to an average of 3.8
-------
percent solids. The total suspended solids concentration of the superna-
tant increased from a normal level of less than 1000 mg/l to over 10,000
mg/l and remained high during the entire a I urn treatment investigation.
The high solids content of the supernatant caused serious problems with
the supernatant return system.
The digested sludge produced during the investigation was satisfactorily
dewatered in a Bird Solid-Bowl Centrifuge and on sand drying beds. Because
of reduced solids concentration of the centrifuge feed, however, it was
necessary to operate the machine for longer periods of time.
On the basis of an alum [AI2(SO ), • 18 hLO] dose of 175 mg/l, the chemical
cost of alum treatment was $41 per million gallons of wastewater treated
at a unit price of $58 per ton of equivalent dry alum.
-------
SECTION I I I
RECOMMENDATIONS
I. A flow paced alum feeding system should be used for application of a I urn
in wastewater treatment plants.
2. The correct a I urn dosage must be determined for each wastewater. On the
basis of full-scale Chapel Hill results, the alum dosage in mg/I is more
significant than Al:TP (mole). However, the significant correlation of
AI:TP noted in jar tests needs further study.
3. When alum is applied ahead of final clarifiers at trickling filter plants,
the average surface overflow rate on the clarifiers should be limited to
500 gpd/ft2.
4. In upgrading existing plants by application of alum ahead of final clarifi-
cation, consideration should be given to the inclusion of fine solids re-
moval facilities following clarification. Settling ponds or granular
media filters may be appropriate. Additional research is needed to deter-
mine the most cost-effective fine solids removal system to follow final
clari f ication.
5. Alum sludge results in decreased alkalinity and a lower than normal pH in
anaerobic digesters. It is recommended that permanent equipment to facili-
tate the addition of lime to the digesters be included in plant upgrading
programs where a I urn treatment is to be used.
6. Because of the problem encountered in settling digested sludge in the sec-
ondary digester at Chapel Hill during the a I urn addition study, it is recom-
mended that separate facilities be considered for stabilizing and disposing
of a I urn humus sludge withdrawn from final clarifiers. Further research
designed to determine feasible and cost-effective methods for separate
handling of this sludge should be conducted. If separate handling facili-
ties are incorporated in plants utilizing a I urn addition, the need for per-
manent equipment to feed lime to digesters, as recommended in No. 5 above,
becomes marg i naI.
7. Additional research at full-scale should be conducted to investigate the
potential of ferric sulfate precipitation at slightly alkaline pH. Iron
precipitation of phosphorus followed by adjustment of the pH to ~7.5 with
lime appears to achieve high phosphorus removals with minimal iron leakage,
while producing an effluent of near neutral pH.
-------
Full-scale studies should be performed to compare the relative effective-
ness of a I urn addition ahead of the primary clarifier, to trickling effluent
ahead of the recircuiation takeoff point, and to trickling filter effluent.
after the recircuIation takeoff point.
-------
SECTION IV
PRELIMINARY PHOSPHORUS REMOVAL INVESTIGATIONS
Preliminary phosphorus removal investigations were conducted prior to full-scale
plant studies. These investigations included:
I. Review of the literature pertinent to phosphorus removal (I)
2. Determination of influent phosphorus levels and the diurnal, daily, monthly,
and seasonal variations
3. Determination of average phosphorus concentrations at various points in the
plant flow scheme
4. Jar tests for:
a. Comparison of the effectiveness of lime, iron salts, and aluminum salts
in removing phosphorus from wastewater
b. Comparison of phosphorus removal efficiencies in wastewater after vari-
ous degrees of treatment (influent, primary effluent, trickling filter
effluent, secondary effluent)
c. Determination of the probable levels of a I urn addition required in pilot-
and full-scale studies
d. Examination of the effectiveness of several coagulant aids in removing
phosphorus
5. Pilot studies to determine the effect of the point of a I urn addition on phos-
phorus removal in a trickling filter system.
CHARACTERIZATION OF PLANT INFLUENT
Throughout the study period, routine analyses were made on composited samples of
plant influent. Table I summarizes selected characteristics of the plant influ-
ent during the period from September 1969 to February 1972. During a portion of
the study period, both total inorganic phosphorus (TIP) and total phosphorus
(TP) concentrations were determined. The resulting ratios of TIP to TP are
shown in Table 2.
In 1970-71, TP and TIP concentrations were determined at various points in the
flow scheme on both plant trains; average monthly values are shown in Table 3.
As indicated in this table, little or no phosphorus removal was achieved in
either the primary or secondary treatment processes.
Diurnal and weekday variations in plant influent orthophosphate (OP) concentra-
tion and loading were evaluated in a special study conducted in July 1969 (2).
In order to cover a representative week, samples of the wastewater influent were
-------
TABLE I
CHARACTERISTICS Of PLANT INFLUENT*
September 1969 - February 1972
(Monthly Averages)
Month
9/69
10/69
1 1/69
12/69
1/70
2/70
3/70
4/70
5/70
6/70
7/70
8/70
9/70
10/70
11/70
12/70
1/71
2/71
3/71
4/71
5/71
6/71
7/71
8/71
9/71
10/71
11/71
12/71
1/72
2/72
Ave
Max
Mi n
Tot.
BOD 5
mg/l
167
176
153
170
193
182
159
165
142
1 17
126
176
141
136
143
150
128
134
134
136
159
156
136
134
188
140
170
183
170
161
154
193
126
TSS
mg/l
238
262
186
159
170
185
162
150
189
169
146
187
159
198
195
175
156
156
163
189
172
195
168
146
136
120
167
187
186
162
174
262
120
TOC
mg/l
140
125
1 13
107
139
124
116
109
100
1 17
1 16
1 12
1 14
1 10
1 16
130
1 1 1
120
I 1 1
123
148
132
121
130
142
1 14
154
148
146
138
124
154
100
Kje
-N
ing/
42.
38.
43.
43.
29.
36.
37.
37.
33.
33.
36.
32.
26.
37.
35.
35.
28.
28.
28.
31 .
30
30
25
28
30
28
31
43
25
Id.
1
3
1
4
8
8
6
7
2
5
6
6
2
3
1
5
8
8
2
8
2
0
2
6
1
.6
.5
.2
.8
.6
NH
mg/
29.
27.
28.
27,
19.
20.
21.
23.
23.
21 .
22.
22.
22.
26.
22.
23.
22.
20.
20.
26.
25
20
19
24
20
21
23
29
19
1
1
0
4
3
5
0 .
9
7
0
7
2
2
3
3
8
2
0
6
2
5
5
5
2
7
.3
,7
, 1
.1
.2
N03
~N
mg/l
0. 10
0.08
0.05
0.05
0.30
0.30
0.28
0.30
0.26
0.19
0.23
0. 16
0. 12
0. 14
0.20
0. 14
0. 18
0.10
0. 10
0. 10
0.20
0.17
0.20
0.50
0.26
0.13
0. 19
0.50
0.05
TP
mg/l
as P
14.0
8.8
.
. — -
0.8
1.5
0.6
0.3
1.2
I.I
0.9
1.6
1.5
8.8
8.6
9.0
9.7
10.8
9,7
9.4
10.9
9,1
10.4
14.0
8.8
TIP
mg/l
as P
10. !
7.4
-• —
-, •
9.3
8.3
8.5
8,2
7.9
8,8
8.6
7.9
9.2
8.5
6.5
7.5
6.4
6.0
5.8
6.8
5.8
8. 1
7,0
6.7
7.4
7.0
7,7
10. 1
5.9
Alk,
mg/l
MBAS as
mg/l pH CaC03
3.85
3.57
3,34
3,43
2.91
2 . 39
2.75
2,76
2.67
3.04
3.06
2.57
2.60
3. 10
7.3
7.4 141
7.0 133
7.1 131
7.2 152
7.2 150
7.0
7,0
3.00 7.2 141
3,85 7.4 152
2.39 7.0 131
*Based on analytical methods described in Table 5.
10
-------
TABLE 2
RATIO OF TOTAL INORGANIC TO TOTAL PHOSPHORUS
CONCENTRATION IN PLANT INFLUENT
Date
5/70
6/70
7/70
8/70
9/70
10/70
1 1/70
12/70
1/71
2/71
3/71
4/71
7/71
8/71
9/71
1/72
2/72
3/72
4/72
5/72
6/72
No. of
Days
Samp led
1 1
12
6
9
5
3
2
7
1 i
12
5
2
3
7
6
7
16
21
14
16
16
Ave
0.75
0.78
0.81
0.76
0.83
0.72
0.60
0.61
0.58
0.68
0.68
0.77
0.75
0.73
0.72
0.70
0.77
0.73
0.72
0.76
0.79
TIP:TP
Max
0.82
0.85
0.92
0.83
0.89
0.83
0.60
0.70
0.73
0.78
0.85
0.78
0.78
0.80
0.76
0.90
0.88
1 .00
0.86
0.94
1 . 19
Mi n
0.65
0.76
0.62
0.62
0.74
0.61
0.59
0.52
0.48
0.53
0.45
0.75
0.71
0.55
0.68
0.56
0.57
0.61
0.62
0.65
0.61
Average of monthly averages 0.72
Range of monthly averages 0.58-0.81
NOTE: One-, two-, and three-day composites were analyzed in 1970.
AM analyses in 1971 and 1972 were performed on one-day
compos ites.
collected on seven different days. However, to (I) collect samples during dry
weather flow only and (2) avoid overloading the analytical laboratory, no at-
tempt was made to sample on consecutive days.
On each of the seven days, sample collection started at 0001 hr and ended at
2400 hr. The 24-hr interval was subdivided into 12 two-hr sampling periods,
which started and ended on even hours, during which separate composite samples
-------
TABLE 3
PHOSPHORUS CONCENTRATION IN PLANT FLOWS
Mason Farm Treatment Plant, April 1970-March
1971
Month
Apr
May
June
July
Aug
Sept
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
June
July
Aug
Sept
Oct
Nov
Dec
Jan
Feb
Mar
Inf luent
1 1,48
10.8!
11.46
10.57
10.33
1 1.23
1 1.07
10.90
1 1.64
11.53
8.81
8.62
8. 18
7.92
8.85
8.61
7.88
9. 17
8.49
6.50
7.46
6.42
5.95
5.84
TOTAL PHOSPHORUS
P- 1 P-2 F- 1
1 1.10
1 1.23
10.34
10.30
10.73
11.16
10.35
1 1.36
1 1.41
9.02
8.82
8.98
9.59
9.55
9.20
8.53
9.37
9.01
7.55
8.15
7.48
6.85
6.56
1 1.56
1 1 .05
11.47
10.84
10.30
11.15
1 1.77
9.35
11.49
1 1.70
9.09
9.12
INORGANIC
9.02
8.92
9.79
9. 10
8.71
9.59
8.90
6.65
8. 19
7.29
6.87
6.50
1.39
1.26
1.17
0.21
1.07
1.66
0.65
1.35
1.04
9.06
8.84
PHOSPHORUS
9.62
9.69
9.89
9.26
8.51
10.10
9.66
8.35
8.39
7.64
7.10
6.60
F-2
1 1.00
1 1.51
1 1.41
12.03
10.30
11.15
12.40
10.40
11.14
1 1 .74
9.16
9.02
9.14
9.13
9.82
9.46
8.87
9.05
9.46
6.85
7.95
7.70
7.24
6.42
S-l
1 1.07
1 1.06
1 1.17
10. 1 1
ro.se
1 1.13
10.20
11.05
1 1.07
8.92
8.58
9.32
9.72
9.71
8.40
8.69
9.77
9.21
7.25
8.50
7.72
7.08
6.40
S-2
1 1.00
10.69
1 1.23
1 1.01
10.08
10.95
1 1.40
10.25
1 1.52
1 1.50
8.76
8.60
9.08
8.94
9.82
9.16
9.01
9.37
8.79
6.95
8.24
7.64
7.1 1
6.32
IP
100
np i
May
June
July
Aug
Sept
Oct
Nov
Dec
Jan
Feb
Mar
Legend :
/ r. *.->
73.26
77.22
81,46
76,33
81 .70
76.69
59.63
64.09
55,68
67.53
67.75
P - Primary Clarif ier
S - Secondary Clarifle
86.39
85.04
88.97
82.82
87.32
80.73
72.95
71.74
65.60
75.94
74.37
Effluent; F
)r Effluent;
/O.UJ-
80.72
85.35
83.95
84.56
86.01
75.62
71.12
71.28
62.31
75.58
71.27
- Trlckl
1 and 2
83.10
85.07 79.32
87.83 86.06
82.90 78.64
83.35 86.17
91.24 81.16
82.85 76.29
78.40 65.86
73.92 71.36
69.20 65.59
78.37 79.04
74.66 7J.I8
Ing Fi Iter Effluent
- Systems Numbers
— -^* — —
87.84
87.79
75.20
85.95
89.96
82.75
71.08
76.92
69.74
79.37
74.59
82.54
83.63
.87.44
83.20
89.38
85.57
77.10
67.80
71.53
66.46
81.16
73.49
12
-------
of influent were collected. The even hour periods were selected after prelim-
inary study of previous influent flow records indicated relatively less change
in flow during these periods as opposed to those 2-hr periods starting and end-
ing on odd hours.
The composite sample collected during each two-hr period was composed of eight
120-ml portions collected at 15-min intervals. The first portion was collected
15 min after the start of the sampling period, the last at the end of the two-
hr period. Samples were not composited according to flow since previous flow
records showed relatively little change in flow magnitude during each period
with the exception of the periods from 0001-0200 hr and 0800-1000 hr. The un-
filtered samples were preserved with mercuric chloride and refrigerated until
ana Iyzed for OP.
Flow data were obtained from the main-plant records. Influent flow is measured
in a stilling well connected to a Parshall flume, and the signal transmitted to
a strip chart recorder.
Results of the above intensive seven-day study are presented in Table 4. Week-
day variations in OP loading were relatively small, ranging from 42 to 54 kg/day
(92 to 118 Ib/day) as P. Diurnal variations in concentration in contrast were
large, ranging from 36.4 percent (0600-0800 hr) to 151 percent (1200-1400 hr)
of the daily average. When flow was considered and OP loading calculated,
diurnal variations were even more pronounced, ranging from 9 to 112 kg/day (19
to 246 Ib/day) as P, or from 16 to 204 percent of the daily average.
JAR TESTS
Materials and Methods
In a typical jar test, the sample was subjected to rapid mixing on a magnetic
stirrer for 30 sec, then transferred to a gang stirrer (Eberbach or Phipps &
Bird) for 30 min of flocculation mixing at 30 rpm. After completion of mixing,
the sample was allowed to settle quiescently for 30 min before collection of the
supernatant for analysis.
Test jars were 1.5 liter battery jars or I qt Kerr wide-mouth jars. Unless
otherwise stated, test chemical solutions were freshly prepared from reagent or
purified grade chemicals. Distilled water was employed in preparation of stock
solutions.
Turbidity was measured with a Hach Laboratory Turbidimeter Model 2100 (Hach
Chemical Co., Ames, Iowa). pH was measured with a Leeds & Northrup pH Meter
Series 7400-A2 or a Fisher Accumet pH Meter Model 310. Sludge volumes were
measured utilizing plastic or glass I liter Imhoff cones, allowing 30 min settl-
ing time. Other analyses were performed as indicated in Table 5.
13
-------
TABLE 4
DIURNAL AND WEEKDAY VARIATIONS IN INFLUENT ORTHOPHOSPHATE CONCENTRATION AND LOADING
Orthophosphate,
T ime
0001-0200
0200-0400
0400-0600
0600-0800
0800-1000
1000-1200
1200-1400
1400-1600
1600-1800
1800-2000
2000-2200
2200-2400
Ave
% of Ave
Load ing
Ib/day as P
Sun
8.3
6.2
4.0
2.9
2.4
7.6
1 1.9
10.6
10.9
9.7
8.7
8.4
7.6
94.2
92
Mon
6.6
5.5
4.7
3.1
4.7
1 1.4
12.9
10.8
10.2
1 1.4
9.6
9.1
8.3
103.4
133
Tue
8.0
5.8
4.6
3.2
4.6
1 1.7
12.2
12.0
10.6
9.2
9.8
10.8
8.5
106.0
132
Wed
6.8
4.4
3.1
2.5
4.1
8.5
10.6
9.5
9.6
7.4
8.7
8.0
6.9
86. 1
I 13
Thu
6.6
4.8
3.4
2.7
9.6
4.1
1 1.9
10.5
9.9
9.5
9.8
10. 1
7J
96.1
123
mg/l as P
Fri
8.5
6.0
4.2
3. 1
4.7
10.9
12.0
1 1.4
1 I.I
10. 1
9.1
10. 1
8.4
105.0
128
Sat
8.3
6.7
4.9
3.0
4.0
10.4
13.4
13.6
11.7
10. 1
9.5
9.6
8.8
1 10.0
118
Ave
7.6
5.6
4.1
2.9
4.9
9.2
12. 1
1 1.2
10.6
9.6
9.3
9.4
8.0
% of Ave
94.2
69.9
51.3
36.4
60.5
1 14.6
150.6
139.0
131.2
119.5
1 15.6
1 17.2
100.0
Load ing
Ib/day
as P
94
44
24
19
81
198
246
225
190
166
155
152
120
% of ave
78.4
36.3
20.0
16.0
67.2
164.5
204.4
187.2
158.4
138.2
129.2
126.4
100.0
Note: I Ib/day = 0.4536 kg/day
-------
TABLE 5
ANALYTICAL PROCEDURES
PARAMETER
METHOD
SOURCE*
Alkalinity, Total Cas
Biochemical Oxygen Demand
(BOD, 5 day, 20 °C)
Carbon - Inorganic
Organic (TOO
Chemical Oxygen Demand (COD)
Chloride (CD
Dissolved Oxygen
Methylene Blue Active Substances
(MBAS)
Metals, Total
Di ssolved
Nitrogen, A.nmonia (NH^ -N)
Nitrogen, Kjeldahl, Total
(Kje!d-N)
Nitrogen, Nitrate (NO,~-N)
Nitrogen, Nitrite (N02 -N)
pH
Total Phosphorus (TP)
Total Inorganic Phosphorus (TIP)
Soluble Phosphorus (SP)
Orthophosphate Phosphorus (OP)
Electrometric Titration - pH 4.5 1
YSI DO Analyzer (probe method) 2
(modified blank depletion)
Dow-Beckman Carbonaceous Analyzer I
Model No. 915 (Dual Channel)
Dichromate reflux - 0.25 N 2
Mercuric Nitrate Titration 2
Winkler Azide or YSI DO Analyzer 2
(probe method)
Methylene Blue 2
Perkin-Elmer Model 303 Atomic I
Absorption Unit
Filtration through 0.45 y membrane
fi Iter
Technicon AutoAnalyzer - Sodium 1
Phenol ate
Technicon AutoAnalyzer - Digestion I
+ Phenol ate
Technicon AutoAnalyzer - Hydrazine I
Reduction
Technicon AutoAnalyzer - I
Diazoti zation
EIectrometric 2
Persulfate Digestion + Technicon I
AutoAnalyzer Automated Stannous
Chloride
Automated (single reagent) Hydra- I
zine SuI fate Reduction Modifica-
tion *
Filtration through 0.45 u membrane I
fiIter
Automated Stannous Chloride Method I
15
-------
TABLE 5 (continued)
ANALYTICAL PROCEDURES
PARAMETER
METHOD
SOURCE*
Sol ids, Total CTS)
Solids, Total Volatile (TVS)
Solids, Total Suspended (TSS)
Solids, Volatile Suspended (VSS)
Sol ids, Settleable
Solids, Total Suspended (after
sett I ing)
Solids, Volatile Suspended
(after sett I ing)
Solids, Mixed Liquor
Suspended (MLSS)
Turbidity (JTU)
Volatile Acids
Gravimetric, 103 °C (Method 224 A) 2
Gravimetric, 550 °C (Method 224 B) 2
Gooch Crucible Filtration, 103 °C 2
(Method 224 C)
Gooch Crucible Filtration, 103 °C 2
Gravimetric, 550 °C (Method 224 D)
Volume (Method 224 F) 2
Method 224 C, on supernatant prepared 2
by Method 224 F
Method 224 D, on supernatant prepared 2
by Method 224 F
Known volume of sample is centri-
fuged and solids removed are
dried and weighed
Hach Model 2100 Turbidimeter
Distillation Method (tentative)
UNC Waste-
water Re-
search Center
method
Hach manual
3
*Tota I Inorganic Phosphorus (Automated Method). The unfiltered sample is treated
by mild acid hydrolysis (2.5 N H2S04 at 90 °C), followed by orthophosphate
determination. Ammonium molybdate reacts with phosphorus in an acid medium to
form a phospho-molybdate complex. This complex is reduced to an intensely blue-
colored complex by hydrazine sulfate. The color is proportional to the phosphorus
concentration. The result includes dissolved and suspended orthophosphates and
acid-hydrolyzable phosphates originally present in the sample.
**!FWPCA. 1969. FWPCA Methods for Chemical Analysis of Water and Wastes, U.S.
Department of Interior, Federal Water Pollution Control Administration. Analy-
tical Quality Control Laboratory, Cincinnati, Ohio.
2APHA, AWWA, WPCF. 1965. Standard Methods for the Examination of Water and
Wasteuater, 12th edition. American Public Health Association, Inc., New York,
New York.
llth edition, I960.
16
-------
Lime Precipitation
Investigations of lime precipitation of phosphorus were conducted to examine
(I) the effects of lime dosage on pH, phosphorus residual, and turbidity and
C2) the degree of incremental phosphorus removal achieved with filtration,
coagulant aid addition, and fluoride addition in conjunction separately with
Iime add ition.
Typical results of jar tests to examine the relationship of lime dosage to pH,
phosphorus residual, and turbidity are shown in Figures I and 2 and Table 6.
These results are consistent with those commonly reported (3). The lime dosage
required to raise the pH from 9.5 to 11.5 was approximately three times that re-
quired to raise the pH from 7.5 to 9.5. As indicated in Figure I, for Chapel
Hill wastewater, which has an alkalinity of about 140 mg/l (as CaC03), 75-125
mg/l of lime [as Ca(OH)2H increased the pH to 9.5, whereas an additional 225-
375 mg/l were required to raise the pH from 9.5 to 11.5. To obtain TIP residuals
£ 1 mg/I required a pH of II or higher as illustrated in Figure 2. Residual
TIP and TP concentrations for 24 jar tests are summarized in Table 6. Table 6
also shows residual turbidity for each test, and, as indicated, initial incre-
ments of lime increased turbidity, but further lime addition produced a sharp
decrease in turbidity.
Observations of turbidity in jar tests indicated the presence of large amounts
of finely divided suspended solids at pH levels below II. Filtration studies
were conducted in conjunction with lime addition to determine if significant
amounts of phosphorus were associated with these solids. Membrane filters
(Miilipore Filter Corporation) of 0.22 y or 0.45 y pore size were employed.
The test filters were preceded by a prefilter and a 0.8 y filter. Comparison
of Figures 3 and 4 reveal that with both pore sizes, filtration drastically re-
duced the residual TIP concentration above a calcium hydroxide dose of about 50
mg/l and a pH level of about 9. In general, filtration through either membrane
pore size was capable of reducing TIP to I mg/l or less at a calcium hydroxide
dose of 100 mg/l and a pH level of 9.5. This lime dose is considerably less
than that required to effect the same TIP removal by jar settling alone. This
work indicated that at lower lime doses and pH levels, significant quantities
of phosphorus are insolubiI ized but remain in suspension.
A variety of coagulant aids were tested to determine if they would enhance set-
tleability of the finely divided particulate phosphate produced at low lime doses.
Anionic, cationic, and nonionic polyelectrolytes were evaluated in preliminary
experiments. The best results were obtained with cationic polye Iectrolytes such
as Cat-Floe (Calgon Corporation). Cat-Floe then was investigated more extensive-
ly. Using plant effluent and a lime dose sufficient to raise the pH to 9.3
[80 mg/l of i ime as CaCOH^H, Cat-Floe was evaluated in doses up to 20 mg/l.
Data from this study are summarized in Table 7 and Figure 5. With lime alone,
TIP was reduced from 8.9 to 5.3 mg/l; with lime and Cat-Floe (3 mg/l), to 0.8
mg/l. Such experiments indicate that with addition of small amounts of cationic
polyelectrolyte, about 90 percent removals of phosphorus can be achieved at a
pH of 9.5 or less, which would require much less lime, reduce or eliminate the
need for further pH adjustment, and reduce the amount of sludge formed.
17
-------
! ' I ' I ! T
®- Trickling Filter Effluent
O Plant Effluent
12 h
0 SOO 200 300 400 500 600
Calcium Hydroxide Dose (mg/l)
FIGURE I. RELATIONSHIP BETWEEN pH AND CALCIUM HYDROXIDE
DOSE.
-------
c, 15
O
_ f Plant Effluent (3 Runs)
*~x
en
15
en
c
10
a)
o
c.
o
o
1-
0
Trickling Filter Effluent
8
9 10
pH
12
a.
15
CJ>
E
c
O
10
c
a>
o
CL
0
Plant Influent
9 10
pH
12
FIGURE 2.
RELATIONSHIP BETWEEN pH AND RESIDUAL TOTAL INORGANIC PHOSPHATE IN VARIOUS PLANT SAMPLES.
-------
TABLE 6
JAR TEST RESULTS FOR LIME ADDITION TO TRICKLING
FILTER EFFLUENT
Jar Test
Number
Initial
Cond ition
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
15
16
17
18
19
20
21
22
23
24
Lime Dose
mg/l as
Ca(OH)2
0
0
25
32.5
40
52.5
65
80
90
105
120
135
160
200
280
400
480
800
10
12.5
15
17.5
20
20
pH
7.6
7.3
7.3
8.7
8.9
9. 1
9.35
9.5
9.8
9.95
0.2
0.4
0.6
1.0
1.2
1 .5
1.7
1.8
2. 1
8.4
8.2
8.4
8.5
8.65
8.7
Turbidity
JTU
30
13
14
16
18
20
22
21
20
20
15
16
15
13
15
8
0
0
0
15
13
13
14
16
15
TIP
mg/l as P
8.0
8.0
8.0
7.3
6.8
6.8
6.5
6.2
5.5
5.2
4.2
4.2
4.0
3.4
3.0
1. 1
O.I
<0. 1
<0. 1
7.9
7.7
7.6
7.4
7.4
7.3
TP
mg/l as P
8.7
8. 1
8.3
7.9
7.7
7.4
7.5
7.2
6.0
5.8
4.9
4.8
4.4
3.8
3.6
2.3
0.3
0. 1
8.4
8.6
8.7
8.6
8.7
8.7
20
-------
Q.
O Unfiltered
D Filtered
Q.
y>
o
o>
E 10.0
c
o
D
i_
•«—
C
o
o
o
CL
0 !00 200 300 400 500
)aicium Hydroxide Dose(mg/l)
O Unfiltered
D Filtered
FIGURE 3. EFFECT OF LABORATORY FILTRATION ON RESIDUAL TOTAL INORGANIC PHOSPHORUS AS A FUNCTION
OF CALCIUM HYDROXIDE DOSE AND pH LEVEL (0.45 y FILTER).
-------
NJ
• -Unfiltered
o-Filtered
10.0
0.01
0 100 200 300 400 500
Calcium Hydroxide Dose (mg/l)
-Unfiltered
O-Filtered
I!
9 10
PH
FIGURE 4. EFFECT OF LABORATORY FILTRATION ON RESIDUAL TOTAL INORGANIC PHOSPHORUS AS A FUNCTION
OF CALCIUM HYDROXIDE DOSE AND pH LEVEL (0.22 y FILTER).
12
-------
TABLE 7
EFFECT OF CAT IONIC POLYELECTROLYTE ON PHOSPHORUS REMOVAL FROM
PLANT EFFLUENT AT CONSTANT LIME DOSE, pH 9,3
Jar Test
Number
untreated
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
Lime Dose
mg/ 1 as
CaCOH)2
80
80
80
80
80
80
80
80
80
80
80
80
80
80
pH
7.0
9.3
9.2
9.3
9.3
9.3
9.3
9.3
9.3
9.2
9.3
9.3
9.25
9.3
9.3
Polyelect.
Dose
mg/l
_
0
0
1
2
3
4
5
6
7
9
1 1
15
17
20
Turbidity
JTU
14
16
16
10
5
2
2
2
1
2
2
5
13
13
10
TIP
mg/l as P
8.9
5.3
5.2
3.5
1 .9
0.8
0.8
0.9
0.9
1 . 1
1.0
1 .0
1 .8
2.8
1 .2
TP
mg/l as P
10.3
5.9
6.5
4.3
2.5
0.8
0.8
0.9
0.9
1 .2
1.2
1 .4
2.9
3.4
3.3
Fluorapatite [Ca|Q(P04)6 F2U is a highly insoluble form of calcium phosphate
(4) and conversion of soluble phosphate to this form should produce very low
phosphorus residuals. Therefore, the effect of f I uoride addition on lime pre-
cipitation of phosphorus using secondary effluent was evaluated. One set of
samples dosed only with calcium hydroxide served as controls, while another
set was dosed in addition with 10 mg/l of sodium fluoride. At the conclusion
of the tests, samples were analyzed before and after filtration through a 0.22
y filter. Results of this investigation are shown in Figure 6. Under the
conditions employed, fluoride addition had little or no effect on the precipi-
tation of phosphorus.
Alum Precipitation
When AIUI I) or Fed II) is added to a solution containing orthophoshate, the
precipitation of the metal phosphate is extremely rapid, reaching completion
in less than a second (5). With alum, phosphate precipitation is favored over
precipitation of the hydroxide (6). Aluminum hydroxide is less efficient as a
phosphate precipitant (5).
23
-------
- io.o
C7>
—• E
z> —
2.1
r I
T> =
. _
-------
to
O
c
O
10.0 —
c
4)
O
C
O
O
I I
Fluoride Added
O No Fluoride Added
PH
FIGURE 6.
EFFECT OF SODIUM FLUORIDE ADDITION ON THE
RELATIONSHIP BETWEEN pH AND RESIDUAL
TOTAL INORGANIC PHOSPHORUS DURING LIME
PREC I P I TAT I ON .
25
-------
With wastewaters, the precipitate formed after a I urn addition appears to be an
amorphous compound with a composition intermediate between aluminum phosphate and
aluminum hydroxide 16).. Also, in practice an AI:P mole ratio in excess of the
stoichiometric requirements for aluminum phosphate formation is necessary for
satisfactory phosphorus removaI. Alum requirements as determined experiment-
ally with various types of wastewater have led to the hypothesis that there
is"an 'alum demand' which must be satisfied before soluble phosphorus can be
effectively removed'1 (.71. This alum demand varies for different wastewaters,
due to their different chemical and physical characteristics.
The interactions of aluminum and condensed phosphates have not been extensively
investigated. Studies by Recht and Ghassemi (5) indicate that the pH range
for highly efficient precipitation of condensed phosphates with a I urn is very
narrow, between 4.5 and 6.5 with an optimum of 5.5.
Other common wastewater components are not considered to affect directly the
precipitation of phosphate with aluminum (6). Nevertheless, investigators
working with a I urn treatment of wastewater commonly encounter variations in
alum requirements which cannot be readily explained by simple analysis of the
usual chemical and physical parameters. Yuan and Hsu (8) have presented evi-
dence that sulfate, kaolinite, montmoriIlonite, and fluoride can affect phos-
phate precipitation with aluminum, and that the type and concentration of these
and other components can influence the optimum pH for phosphorus removal from
wastewater.
In the jar tests performed in this study, such variations in requirements were
observed. Therefore, numerous runs were performed with a variety of a I urn
dosages and samples from several points in the process train. Alum stock solu-
tions were prepared from reagent grade aluminum sulfate CA^tSC^^ • 18 h^CG.
Results from these tests were plotted, and the range of suitable a I urn dosages
and AI:P ratios were defined.
In the tests described, unless otherwise stated, pH adjustments were not made
before or after a I urn addition. In practice, in water of moderate to low alka-
linity, addition of a I urn often lowers the pH sufficiently for good precipita-
tion and if not, the addition of excess a I urn is probably more practical than
maintaining and operating a second chemical dosing system. Further, limited
tests with pH adjustment indicated that adjustment to pH 6.0, the pH of minimum
orthophosphate solubility, frequently led to gasification and flotation of
sludge, possibly due to C02 release. Slightly higher pH levels (6.5-7.0) con-
sequently gave better overall phosphorus removal.
Treated samples were analyzed for phosphorus (total and/or total inorganic),
turbidity, and pH. From these studies, results were compiled and polynomial
regression models constructed indicating the degree of total phosphorus removal
as a function of alum dosage, the ratio of applied Al(lll) to initial total
inorganic phosphorus in the respective' samples, and the ratio of applied Al(lll)
to initial total phosphorus in the respective samples. Models were based on
the following number of observations: influent, II; primary effluent, 12;
trickling filter effluent, 19; and secondary effluent, II. In most cases, both
second and third degree models were derived. The higher the degree, the more
26
-------
closely the function follows the observed points. The best models based On
an analysis of predictability and statistical significance are given in Tabl
8, The models themselves are illustrated {n Figures 7 to 10.
TABLE 8
POLYNOMIAL REGRESSION MODELS OF CHOICE FOR PHOSPHORUS REMOVAL WITH
ALUM ADDITION
Parameter
Stream
Model Degree of Choice
Alum dosage
Al :TP
Al:TIP
Influent
Primary effluent
Trickling filter effluent
Secondary effluent
Primary effluent
Trickling filter effluent
Secondary effluent
Primary effluent
Trickling filter effluent
Secondary effluent
2
2
2
3
2
3
3
2
3
3
Based on the models, a I urn dosages necessary for 97-98 percent removal of total
phosphorus are shown in Table 9.
TABLE 9
ALUM DOSAGES REQUIRED FOR 97-98 PERCENT REMOVAL OF TOTAL PHOSPHORUS
Stream
Alum [AI2(S04)3 • 18 H20] Required, mg/l
2nd Degree Model
3rd Degree Model
InfIuent
Primary Effluent
Trickling Filter Effluent
Secondary Effluent
200*
180*
187*
161
175
170
150*
*Model of Choice
27
-------
100
90
80
s*
15 70
>
o
£
0)
cc
2 60
o
Q.
in
o
£ 50
o
^
\—
40
30
20
10
0
1 ' ' ^ ' 0*i
• A 00
• 00
• 00
• oo
OO A
• 0
A • o
r oo ~
• 0
• 0
oo
• 0
0
•0
•0
o
•
00
O fl&
o
- •
0 •
o
0 •
o
0 •
o •
0
_ o •
0 •
o •
o
o •
£
0 • LEGEND
0
O *^ A - Experimental Data
•* ^ O " ntted Values for 2 Degrees
O • £ - Fitted Values for 3 Degrees
A O ••
100 120 140 160 180 200
AI2(S04 )3 -I8H20 (mg/l)
FIGURE 7, POLYNOMIAL REGRESSION MODELS FOR PHOSPHORUS REMOVAL FROM
INFLUENT AS A FUNCTION OF ALUM DOSAGE,
28
-------
N)
100
90
60
ro
g 60
"
1
I 50
5
I
jf
2
O
•" 30
20
| £
o
1 I
— oooo &
9' IBHjO (mg/l)
1 I
h,-. - - n.ww in~» A •-
CD?
A oo «*^» A o
A ot»» A o
OM O
A M
0
f
•
•
•0
•0
«o A
0
•3
£90
•o
•o
o
0
0
o
m
• A
o
o
o
o
<# (C)
* • i i '
0.73 1.19 1.65 2.10
Al: TP (wl /wl|
Graph A - As a Function of Alum Dosage
Graph B - As a Function of AhTIP
Graph C - As a Function of Al:TP
A- Experimental Data
O- Fitted Values for 2 Degrees
•- Fitted Values for 3 Degrees
FIGURE 8. POLYNOMIAL REGRESSION MODELS FOR PHOSPHORUS REMOVAL FROM PRIMARY EFFLUENT WITH ALUM ADDITION,
-------
100
90
80
70
60
50
40
o 30
20
10
a • caooooo
• coo
• CO
co A
•GO
•o
A
o»
00 • A
o • A
o
A o •
o •
o •
o • A
0 •
o •
o •
0 •
o •
o •
o •
0 •
0 *
o •*•
0 •»
0
o
-
A (A)
,,\ l i I ._ J
A A co o •-
A •!"* A ° *A
** • o •• o
• ° •*• «• o
• o A
• 0 0
o o
A*P°
• o
0
• A
o
A
A
«
o
o
o
a»
o
°«
A (B)
A
i i i i
AA.^*AA "***^ °°o •"
A Z^° *»— w*
o ••••••••• o
• A o
• 0
o
A ^
*°
•3
A
4^
o
• A
o
o
•
o
•
O
o
o
0 *A (0
A A
i i i i i i i
100 125 ISO 179
AI2
-------
100
90
80
„ 70
^
""*
o
o 60
£
-------
These results indicate there is a general trend to lower alum requirements
with progressively greater degrees of treatment, due primarily to decreasing
initial TP concentrations a.s, the degree of prior treatment increases,
Based solely on an analysis of AI:T|P vglues, it appears that of the streams
tested, the secondary effluent is the most efficient point for alum applica-
tion. Phosphorus reductions of 97-98 percent were achieved at an AI;TIP sub-
stantially lower than that required for trickling filter effluent or secondary
effluent as summarized in Table 10. However, when the tests were analyzed in
terms of A1:TP values, Table II shows that differences due to the point of
alum application disappeared.
TABLE 10
RATIO OF ALUMINUM TO TOTAL INORGANIC PHOSPHORUS REQUIRED FOR 97-98
PERCENT REMOVAL OF TOTAL PHOSPHORUS
1
Stream
Primary Effluent
Trickling Filter Effluent
Secondary Effluent
nitial TIP,
mg/l
5.7-7.0
3.9-8.5
6.8-8.5
Al :TIP,
2nd Degree Model
2.30*
2.46
1 .84
wt/wt
3rd Degree Model
>2.45
2.22*
1.84*
*Model of Choice
TABLE I I
RATIO OF ALUMINUM TO TOTAL PHOSPHORUS REQUIRED FOR 97-98 PERCENT REMOVAL
OF TOTAL PHOSPHORUS
Stream
Primary Effluent
Trickling Filter Effluent
Secondary Effluent
Initial TP,
mg/l
7.7-1 1.9
5.0-12.9
8.8-1 1 .0
Al :TP, wt/wt
2nd Degree Model 3rd
1.51*
1 .60
1 .55
Degree Model
1.65
1.63*
1 .60*
l of Choice
32
-------
As indicated in these tables, the AI:TP values for the waste streams analyzed
most uniformly predicted the amount of alum required to achieve essentially
complete total phosphorus removal, irrespective of the degree of prior treat-
ment. While this statement may seem somewhat obvious, the literature (.3, 10)
reveals many instances where alum dosage is chosen on the basis of the type
of waste stream, AhTIP values or Ahsoluble phosphorus values. Despite the
fact that the total phosphorus determination requires digestion of a sample
prior to phosphorus measurement, it appears that this determination should be
the parameter of choice to monitor. Although there may be some disagreement
over the relative significance of organic phosphorus and inorganic phosphorus
in eutrophication, the ultimate availability of most organic phosphorus com-
pounds for algal growth would appear to dictate control measurements based on
total phosphorus content (II, 12, 13, 14, 15, 16, and 17).
Alum Precipitation in Conjunction wrth Coagulant Aid Addition
Jar tests were conducted with a I urn and a variety of coagulant aids to deter-
mine if significant phosphorus removal could be achieved with lower alum
doses. Tests were conducted with grab samples, in most cases on a "one-shot"
basis. Coagulant aids were prepared according to manufacturers' recommenda-
tions. Phosphorus removals were calculated on the basis of the phosphorus
remaining in the test jar supernatant compared to that remaining in supernatant
of a control jar to which nothing was added.
As shown in Table 12, with trickling filter effluent and an a I urn dose of 150
mg/l, phosphorus and turbidity removals were markedly enhanced by the addition
of low concentrations of Cat-Floe or Magnifloc. Calgon WT-3000 somewhat im-
proved phosphorus removal at an a I urn dose of 150 mg/l, especially with secon-
dary effluent. However, improvement was not marked and was substantially less
effective than was increasing the a I urn dose to 200 mg/l as indicated in Table
13. To a small extent, Natron Floe Aid (0.5 mg/l) improved phosphorus re-
moval from secondary effluent at an alum dose of 150 mg/l. Similar results
were obtained by increasing the a I urn dose alone from 150 to 175 mg/l as summa-
rized in Table 14.
Iron Free i p itati on
Iron compounds effectively precipitate phosphate from wastewater. Ferric and
ferrous iron compounds are being utilized currently in several large scale
installations (3). However, a number of problems remain unsolved in regard to
optimum utilization of iron compounds. Recent studies recommend further re-
search in the following areas:
(I) Evaluation of the effect of ionic constituents on the efficiency
of phosphate removal (5),
(2) Characterization of the sludge produced (water content, compact-
ability, dewaterabiIity)(3, 5, 9), and
(3) Prevention of leakage of iron into plant effluent by use of
polyelectrolytes or other means (3).
33
-------
TABLE 12
EFFECT OF CAT-FLOC AND MAGNIFLOC ON ALUM PRECIPITATION
OF PHOSPHORUS FROM TRICKLING FILTER EFFLUENT
Sample
Control
AI2(S04
)
+ Cat-F
t Cat-
t Magn
t Magn
F
i
i
3
1
•18
oc,
loc,
f
f
loc,
loc,
H20, 1
3 mg/l
3 mg/l
3 mg/
3 mg/
50 mg/
(Run
(Run
1 (Run
1 (.Run
1
1)
2)
1)
2)
mg/l
1 1.4
5.5
0.9
0.8
1.4
1.2
TP
% Remova
51.8
92,1
93.0
87.7
89.5
TIP
1 mg/l
8.5
3.9
0,8
0.7
I.I
1.0
% Remova 1
54. 1
90.6
91 .8
87. 1
88.2
Turbid ity
JTU
20
14
2
2
4
4
TABLE 13
EFFECT OF CALGON WT-3000 ON ALUM PRECIPITATION OF PHOSPHORUS
FROM TRICKLING FILTER EFFLUENT AND SECONDARY EFFLUENT
Sample
Trickling Filter Effluent
Control
+ WT-3000, 0.2 mg/l
Alum, 150 mg/l
+ WT-3000, 0. 1 mg/l
t WT-3000, 0.2 mg/l
*Alum, 200 mg/l
Secondary Effluent
Control
+ WT-3000, 0.2 mg/l
Alum, 150 mg/l
+ WT-3000, 0. 1 mg/l
+ WT-3000, 0.2 mg/l
*Alum, 200 mg/l
mg/l
10.5
10.3
3.0
2.4
2.2
0.4
1 1.0
1 1.3
9.2
6.3
4.0
1.3
TP
% Remova 1
1.9
71.4
77. 1
79.0
96.2
16.4
42.7
63.6
88.2
Turbid ity
JTU
19
20
10
7
7
2
29
28
24
18
1 1
6
*Alum
as
18 hLO
34
-------
TABLE 14
EFFECT OF NATRON FLOG AID ON ALUM PRECIPITATION OF PHOSPHORUS FROM SECONDARY EFFLUENT
TP
Sample
Control
*Alum, 100
+ Natron,
*Alum, 125
*Alum, 150
+ Natron,
*Alum, 175
*Alurn, 200
mg/l
0.5 mg/l
mg/l
mg/l
0.5 mg/l
mg/l
mg/l
mg/l
8.8
6.7
8.7
3.3
I.I
0.5
0.4
0.3
% Remova 1
23.9
I.I
62.5
87.5
94.3
95.4
96.6
TIP
mg/l
7.7
5.7
6.6
2.7
0.6
0,3
0.3
0.2
% Remova 1
26.0
14.3
64.9
92.2
96.1
96.1
97.4
TOC
mg/l
37
27
33
18
14
14
12
10
Kjeldahl-N
% Remova 1 mg/ 1
27
II
51
62
62
68
73
41.0
39.0
25.0
37.5
34.5
23.5
31.0
25.5
% Remova 1
4.9
39.0
8.5
15.8
42.7
24.4
37.8
NJ-U-N
mg/T"
20.5
20.0
22.0
20.5
20.5
21.5
20.5
20.5
Turbidity
JTU
9
13
18
7
4
2
2
2
*AI2
-------
Jar tests were conducted with iron using iron equivalent doses of 10, 20, 30,
40, and 50 mg/l. Two series of tes.ts were conducted, one with. pH control and
one without.
Iron was supplied by Cities Service Company rn the form of production chemi-
cals. Ferric suIfate was supplied in dry form as Ferri-Floc, with 21.8 per-
cent water soluble iron. Ferric chloride was supplied in liquid form as 31
percent ferric chloride (II percent trivalent iron). Stock solutions of each
chemical were prepared immediately before testing.
In each set of tests, a control jar was included. This jar was not dosed with
iron, but was in all other ways treated like the test samples. The use of
controls enabled distinction of the effectiveness of iron addition from the
effectiveness of simple slow mixing and settling.
Tests were conducted on plant influent (after screening and grit removal) and
trickling filter effluent immediately upon collection. Analyses were perform-
ed on the samples prior to chemical application and on the supernatant in the
jars following chemical addition, flocculation, and settling. The number and
type of analyses varied according to the type of experiment. Parameters mea-
sured included total phosphorus, soluble phosphorus (0.45u filter procedure),
BODc;, total organic carbon, total suspended solids, total solids, pH, turbidity,
total iron, soluble iron (0.45y filter procedure), total aluminum, and soluble
aluminum (0.45y filter procedure).
Based on results of the above tests as partially summarized in Figures II to
14, the following conclusions can be drawn: (I) Under the conditions tested,
ferric chloride was effective at lower doses than was ferric suIfate for re-
moval of phosphorus, total suspended solids, and turbidity; (2) Sludge volume
production was less with ferric sulfate and the sludge appeared stronger than
that produced with ferric chloride; and (3) Some iron remained in solution
with both FeCU and Fe2(S04)3, as evidenced by a faint brown color.
Four additional sets of experiments were conducted with Ferri-Floc (ferric
sulfate) to determine the effectiveness of pH control in and the effect of the
sequence of chemical dosage on removing phosphorus. In these tests, trickling
filter effluent was utilized. Chemical solutions employed in pH control were
prepared from laboratory chemicals. Lime [CatOH^D was prepared from purified-
grade power, sodium hydroxide (NaOH) and hydrochloric acid (HCI) from reagent
grade liquids. Controls were utilized as previously described in this section.
First, experiments were conducted to determine whether lime or sodium hydroxide
was more effective for pH control. Tests were conducted over the nominal pH
range of 6 to 9. As indicated in Tables 15 to 17, phosphorus removals and con-
trol of supernatant iron were consistently better when lime was used for pH
control. This difference in performance of lime and sodium hydroxide was
especially marked at the lower iron doses.
Second, experiments were performed to determine the optimum pH for removal of
phosphorus and elimination of effluent iron. Table 18 shows that with an Fe
(III) dose of 50 mg/l, phosphorus removal was essentially the same over the
36
-------
TABLE 17
pHOSPHORUS REMOVAL WITH FERRIC SULFATE: COMPARISON OF NaOH AND Ca(OH)2 FOR pH CONTROL AT pH 8
Supernatant
Test Sample
Run 1 :
Control
Fe Dose oH
mg/l Before After
0 7.05
Base Req'd
mg/l
pH Turbidity
JTU
7.05 21
Fe(Tot)
mg/l
0.8
TOC
mg/l
33
TP
mg/l
9.4
Adjustment with NaOH
I
I I
I I I
IV
V
VI
VI I
Adjustment with Ca(OH)2
I I
I I I
IV
V
VI
VI I
0
20
40
60
80
100
120
0
20
40
60
80
100
120
,05
.20
,60
,15
,50
, 15
3.05
7.00
6.20
6.05
5.45
3.60
3. 15
3.02
8. 10
7.95
8.00
8. 10
8.20
8.00
8. 10
8.20
8.30
8.10
8.20
8. 10
8. 15
8.20
26
68
74
120
168
200
262
7.60
.00
.95
.55
,60
,65
,60
7.70
23
21
18
8
6
3
2
20
14
8
4
I
4
I
0.5
0'.6
0.8
3.8
3.4
0.6
0.4
0.4
0.2
32
28
22
20
17
46
13
34
23
49
15
13
10.7
5.7
3.4
1.0
0.7
0.5
0.6
9. I
3.2
I .-3
0.6
0.8
0.6
I . I
Run 2:
Control
Adjustment with NaCH
7.3
7.0
28
37
8.6
1
i 1
1 1 1
IV
V
VI
VI 1
Adjustment with Ca(OH>2
1
II
1 1 1
IV
V
VI
VI 1
0
20
40
60
80
100
120
0
20
40
60
80
100
120
7.3
6.2
5.6
4.7
3.3
3.2
3.1
7.3
6.2
6.0
4.8
3.4
3.2
3.1
8.2
8.2
8.0
8.0
8.0
8. 1
8. 1
8.3
a. 4
8.2
8.2
8.0
8.6
8.2
*
*
*
*
*
*
*
20
50
78
114
146
194
266
7.8
7.8
7.6
7.6
7.6
7.7
7.6
8.0
8,2
7.8
7.4
7.6
8.1
7.7
28
19
17
8
12
5
5
26
12
8
8
12
8
4
0.8
6.0
5.0
1 .6
0.9
1.0
1 .0
0.7
3.1
1.2
0.5
0.5
0.6
0.4
45
24
18
15
15
14
13
43
24
16
14
15
15
14
9.1
3.7
I.I
0.4
0.2
0.2
0.2
8.6
1.5
0.4
0.2
0.2
0.2
O.I
'Amount of NaOH not measured.
-------
nominal pH range of interest CpH 6 to 8), However, as shown in Table 18
and Figure 15, iron capture wa,s enhanced by increasing pH to levels greater
than 7- Comparison of soluble and total iron concentrations in Figure 15
indicates that removal of the insoluble iron portion was primarily responsible
for the enhanced iron capture noted at the higher pH levels. It appears pro-
bable that the so-called "iron leakage" observed in practice is due to the
escape of colloidal fron particles which fail to settle under the usual opera-
ting condftions. These jar tests suggest that concomitant addition of lime
will improve coagulation and settling of colloidal iron.
Third, experiments were conducted to determine if any effect could be ascribed
to the order of addition of iron and lime. An Fell I I) dosage of 50 mg/l was
chosen. To one jar, iron was added, followed by addition of lime to the de-
sired pH, To a similar jar, the same amount of lime was added prior to addi-
tion of the iron. Results of these tests are summarized in Table 19. Little
difference was noted in phosphorus and iron removal. However, it was observed
that when iron addition preceded lime addition, the floe was more compact and
settled more rapidly. Incidentally, this experiment again confirmed the
effectiveness of pH control for improving iron capture. With iron alone (pH 6),
total iron remaining in the supernatant was 6 mg/l, contrasted with about 2
mg/I at a pH of 8.
And finally, a fourth set of additonal jar tests compared the effectiveness of
alum (production liquid, 4.4 percent Al(lll), Allied Chemical Company) and
ferric sulfate (Ferri-Floc) in removing phosphorus at different pH levels.
Iron was dosed to provide an FerTP of 3:1 (molar basis); aluminum, to give
ratios of 1.5:1 or 1.8:1 (molar basis). At these dosages, iron compared very
favorably with aluminum for phosphorus removal as summarized in Table 20.
While iron "leakage" was substantial with the use of iron alone, when pH was
controlled to 7 or above with lime, iron carryover was markedly reduced. It
also appeared that with the wastewater sample used in this series of tests,
the Fe(lll) dose for effective phosphorus removal could have been reduced to
40 mg/l or less.
From the four sets of additional jar test experiments just described, the fol-
lowing conclusions were reached regarding chemical treatment of trickling
filter effluent with ferric sulfate: (I) Iron leakage can be minimized by con-
trol of pH to 7 or above, (2) Lime is more effective than sodium hydroxide for
pH control, (3) The optimum pH for iron-lime precipitation is approximately
7.5, (4) The sequential order of addition of iron and lime has little effect
on phosphorus removal efficiency, (5) Using lime for pH control, ferric sulfate
is superior to a I urn for phosphorus removal, and (6) With trickling filter
effluent from the Town of Chapel Hill Treatment Plant, approximately 90 mg/l
of Ca(OH)2 are required to raise pH to above 7 after dosage with 50 mg/l of
Fed I I) (229 mg/l of Ferri-Floc).
PILOT PLANT STUDIES OF PHOSPHORUS REMOVAL WITH ALUM
Two trickling filter pilot plants were constructed during 1966 prior to the
initiation of the work reported here. Two additional trickling filter pilot
plants were constructed during the course of this study. The pilot plants
44
-------
TABLE 15
PHOSPHORUS REMOVAL WITH FERRIC SULFATE: COMPARISON OF NaOH AND Ca(OH)7 FOR pH CONTROL AT
pH 6 AND pH 9
Supernatant
Test Samp 1 e
Control
Control + acid
Adjustment with NaOH
i
i
1 1
1 1 1
IV
V
VI
VII
Adjustment with Ca(OH)2
1
II
III
IV
V
VI
VII
Fe Dose
mg/l
0
0
0
40
40
60
60
80
80
0
40
40
60
60
80
80
pH
Before
7.00
~>.oo
6.95
5.72
5.80
4.70
4.70
3.25
3.20
6,95
5,85
5,80
4.60
4,56
3,25
3. 15
After
6.05
9,35
6.00
8.92
5,92
9,12
6,81
9,20
9, 18
6,35
9,50
6,00
9,05
6,20
8,90
Base Req'd
mg/l
*
*
*
*
#
#
*•
100
30
170
52
156
120
200
pH
6.85
6.45
9.00
6,20
8.70
6.20
8,85
6.60
9.00
8,95
6,60
9,15
6,55
8,80
6,65
8.45
Turbi-
dity
JTU
17
18
27
8
15
7
8
2
3
28
7
10
2
4
1
3
- Fe
(tot)
mg/l
0.5
0.6
0.8
4,4
6.5
3.2
3.3
0,7
1. 1
0,4
3,0
0,5
1,0
0,9
0,4
0.4
TOC
mg/l
48
48
56
21
28
20
26
14
21
38
21
20
18
20
16
19
TP
mg/l
10.6
10.3
1 1 .1
1.6
2.3
0.6
1 .1
0,3
0.9
5.6
0.9
0.5
0.3
0.5
0.3
0.4
*Amount of NaOH not measured
-------
TABLE 16
PHOSPHORUS REMOVAL WITH FERRIC SULFATE: COMPARISON OF NaOH AND Ca(OH)2 FOR pH CONTROL AT pH 7
Supernatant
Test Samp 1 e
Control
1 ron +
1 ron +
1 ron +
1 ron +
1 ron +
1 ron +
NaOH
CaCOH)2
NaOH
Ca(OH)9
z
NaOH
Ca(OH)2
Fe Dose
mg/l
0
40
40
60
60
80
80
pH
Before
6.95
5.90
5.90
4.80
4.55
3.40
3.40
After
6.95
7.00
7.00
7.05
6.90
6.92
Base Req'd pH Turbidity Fe(tot)
mg/l
7.
* 6.
58 6.
* 6.
96 6.
* 6.
128 6.
JTU mg/!
95
70
90
85
95
80
90
15 0
6 3
3 1
2 0
2 0
1 0
-------
75
r>
H
25
120
o>
1 3
. 90
o
(S)
•o
-------
75
50
253
- 120
o>
E
90
60
-30
-------
1
0
8
e
a.
4
2
0
c
_ 6
E
- 4
CL
2
n
> O~ Ferric Chloride
^ ^ 0- Ferric Sulfate
O
o •
— —
0
•
0
>
— —
0
0
$ 8
0 10 20 30 40 50 60
Fe (mg/l)
FIGURE II. PHOSPHORUS REMOVAL FROM PLANT INFLUENT
WITH FERRIC CHLORIDE AND FERRIC SULFATE
ADDITION.
37
-------
10
8
o<
E
6
0>
14
Q.
h-
o
o
- Ferric Chloride
- Ferric Sulfate
o
. 8 9
o
J _ ,
10 20 30 40 50 60
Fe (mg/l)
FIGURE 12. PHOSPHORUS REMOVAL FROM TRICKLING FILTER
EFFLUENT WITH FERRIC CHLORIDE AND FERRIC
SULFATE ADDITION.
38
-------
TABLE 18
EFFECT OF pH ON PHOSPHORUS AND COLLOIDAL IRON CAPTURE WITH FERRIC SULFATE AND LIME
Supernatant
Test Sample
Sample as Col lected
Control
Control
Control
Ca(OH)7 only
Ca(OH)^ only
Ca(OH)" only
Ca(OH)^ only
1 ron on 1 y
i ron + Ca(OH)?
1 ron + Ca(OH)^
Iron + Ca(OH)^
Iron + Ca(OH)"
Iron + Ca(OH),
Iron + Ca(OH)^
Iron + CaCOH)^
Fe Dose
mg/l
0
0
0
0
0
0 x
0
0
50
50
50
50
50
50
50
50
Before
7.00
7.00
7.00
7.05
7. 10
7.20
7.05
7.00
5.70
5.78
5.86
5.86
5.86
5.86
5.70
pH
After
6.52
6.70
7.00
7.25
7.60
7.75
8. 10
5.80
6.50
6.72
7.00
7.28
7.52
7.78
8.02
Cat OH), Req'd
mg/l
—
—
5
1 1
14
29
—
47
51
78
81
90
102
1 19
PH
7.2
6.6
6.6
6.9
7.2
7.4
7.6
8.1
5.9
6.5
6.7
7.0
7.2
7.4
7.5
7.8
Turbidity
JTU
50
32
31
32
32
32
32
32
6
7
6
6
6
5
6
5
Fe(Tot)
mg/l
1.4
0.7
0.7
0.6
0.7
0.7
0.7
0.7
6.0
3.0
1.6
2.0
1 .2
1.0
I.I
I.I
Fe(Sol)
(mg/l)
0.3
0.3
0.6
<0.l
0.3
0.3
0.3
0.3
2.7
0.4
<0. 1
0.3
0.4
0.3
0.3
0.3
TOC
mg/l
107
78
77
79
78
77
80
79
27
28
29
29
28
27
28
29
TSS
mg/l
100
37
27
30
53
53
53
86
10
12
7
6
5
5
7
7
TP
mg/l
13.9
1 1.4
10.7
10.7
10.5
10.3
10.5
9.9
0.6
1 . 1
1.0
0.6
0.9
1.3
1 . 1
0.9
P(Sol)
mg/l
6. 1
8.4
7.6
8.8
6.6
6.7
6.2
6.9
0.2
0.2
0.2
0. 1
0.6
0.2
0.9
0.5
-------
o> ,
J
J 2
o •
en
0
_ 8
-x
a>
£ 6
2 4
|2
0
O
.08-
- 50 mg/1 Iron Added
- No Iron Added, pH
Control Only
o
o
, . y't0*?^^
0" 5.5 6.0 6.5 7.0 7.5 8.0
PH
FIGURE 15. EFFECT OF pH ON Fe CAPTURE DURING IRON
ADDITION FOR PHOSPHORUS REMOVAL.
46
-------
TABLE 19
EFFECT OF ORDER OF ADDITION OF IRON AND LIME ON PHOSPHORUS REMOVAL
AND COLLOIDAL IRON CAPTURE
Test Sample
Samp le as Co 1 1 ected
pH 7 Control
Fe, Lime
Lime, Fe
pH 8 Control
Fe, Lime
Lime, Fe
Fe
Dose
mg/l
0
0
50
50
0
50
50
Supernatant
pH
7.4
7.2
7.2
6.9
8.2
7.7
7.2
Turb id i ty
JTU
67
56
10
14
56
9
10
Fe(Tot)
mg/l
2.5
1 .4
3.0
4.0
1,0
2. 1
2.0
FeCSo'l )
mg/l
0. 1
0.2
0. 1
0. 1
<0.l
0. 1
0. 1
TOC
mg/l
130
96
25
32
96
26
32
TSS
mg/l
246
59
15
36
1 17
14
21
TP
mg/l
16.6
12,8
1 .3
1.5
12.4
1 .0
1 .0
PCSol )
mg/l
6.6
7.4
0.6
0.2
5.9
0.2
0.4
ron only
50
6.0
6.0
0.5
28
I .3
0. I
-------
TABLE 20
COMPARISON OF FERRIC SULFATE AND ALUM FOR PHOSPHORUS REMOVAL
CD
Supernatant
Test Sample
Contro 1
1 ron on 1 y
Iron + Ca(OH>2
Iron + Ca(OH)2
Iron + Ca(OH>2
A 1 urn on 1 y
Alum, +Ca(OH>2
Alum, +Ca(OH)2
Alum, +Ca(OH)2
A 1 urn on 1 y
Alum, 4Ca(OH)2
Metal
Dose
mg/l
50
50
50
50
12
12
12
12
14
14
Molar
Ratio
Met: TP
3:1
3:1
3:1
3:1
1.5:1
1.5:1
1.5:1
1.5:1
1.8:1
1.8:1
pH
Before
6.9
5.6
5.6
5.6
5.6
6.4
6.4
6.5
6.5
6.5
6.5
After
7.0
7.6
8.4
7.0
7.5
8.1
7.6
PH
7.2
6.2
7.0
7.5
8.4
6.8
7.2
7.4
7.9
6.6
7.5
Turbi-
dity
JTU
32
6
5
3
4
II
13
17
12
7
5
Metal (Tot) Metal (Sol)
mg/l mg/l
Fe Al Fe
1.0 2.0 0.6
4.5 2.9
0.7 O.I
0.4 O.I
0.6 <0. 1
3.0
3.4
4.8
3.2
1.3
1.4
Al
0.3
0.3
0.3
0.3
0.4
0.9
0.2
TOC
mg/l
79
15
26
27
28
33
30
34
37
24
29
TSS
mg/l
58
13
10
7
9
30
28
44
25
17
15
TS
mg/l
313
321
385
392
363
286
333
340
352
248
303
TP
mg/l
9.0
0.3
0.2
0.2
0.2
2.6
2.4
3.5
2.3
, I.I
0.9
P(Sol)
mg/l
6.1
0.4
0.2
O.I
O.I
0.3
0.3
0.6
0.7
0.3
0.4
mg/T
53
12
13
10
28
23
19
22
20
17
17
-------
were designed to treat raw Chapel Hill wastewater which had passed through the
main plant bar rack, a degritting chamber, and a fine bar rack to remove stringy
solids which would tend to clog the small pumps and pipes in the pilot plants.
Influent to the pilot plants was delivered through a 2-in (5-cm) plastic pipe
at a flow rate substantially in excess of pilot plant requirements. Excess
flow was wasted and the desired amount of influent delivered to the operating
pilot units by means of a variable speed pump with DC motors regulated by
silicon controlled rectifiers. Flow to each of the pilot units was proportioned
with the use of an overhead rotating distributor discharging into a circular
distribution box with four equal radial sectors. Flow was by gravity from the
distribution box to the primary settling tank of each pilot plant.
Each pilot plant unit consisted of a primary settling tank, a trickling filter,
and a final settling tank. Recirculation was provided around the filters
through the primary settling tanks, using a 2:1 recirculation ratio. A general
flow diagram of a single pilot plant unit for single-stage filtration operation
is shown in Figure 16.
The sizes of the settling tanks and filters were selected to provide detention
times and, in the case of the filters, a hydraulic loading about the same as
experienced in the main plant at a flow rate of 11,355 m^/day (3 mgd). As the
pilot settling tanks were not as deep as the main plant units, the surface
overflow rates for the pilot units were substantially less than those in the
main plant. All settling tanks were equipped for hydrostatic sludge removal.
The pilot trickling filters were designed to operate under conditions similar
to those of the main plant filters. A filter diameter of 1.2 m (4 ft) was
selected, this being considered reasonably safe for minimizing wall effects.
Conventional clay tile filter underdrains were used. Filter media depth was
also fixed at 1.2 m (4 ft). Inner and outer walls of the filters consisted of
two vertical concentric sections of Armco steel pipe, 1.8 m (6 ft) long and
122 cm (48 in) and 137 cm (54 in) in diameter respectively. The annular space
between the inner and outer pipes provided insulation to reduce heat loss dur-
ing cold weather operation. Filter media was granite stone selected to pass
a 9 cm (3.5 in) screen with less than 75 percent passing a 6.4 cm (2.5 in) screen
Design conditions for the various plant units are given in Table 21 below. On
the basis of an influent BOD^ of 180 mg/l and 35 percent removal in the primary
settling tanks, the organic loading on the filters approximated 1500 Ib BOD5/
acre-ft/day.
TABLE 21
DESIGN CONDITIONS FOR TRICKLING FILTER PILOT UNITS
Primary Settling Tank
Trickl ing Fi Iter
Final Settl ing Tank
Flow
gpm
3.6*
3.6*
1 .2
Detention
Time, hrs
1 .8
2.0
Overflow Rate
qpd/ff2
470
436
Hydrau 1 ic Load
mgad
18.0
i ng
49
-------
Recycle
Plant Screening
and
Degrifting
Fine Screen
Speed Control
Primary Settled Wastewater and
Unsettled Trickling Filter Recycle
Pump
To Waste'
Pilot
Plant
Influent
Recycle
Rotary)
Distributor
Trickling
Filter
Filter Effluent
r
>
i
">
Finn
Sett
Tan*
Pilot Plant Effluent
ffl «-Valve
T Sludge
FIGURE 16. FLOW DIAGRAM OF TRICKLING FILTER PILOT PLANT
FOR SINGLE-STAGE FILTRATION.
50
-------
Chemicals for phosphorus precipitation may be added at one or more of several
points in a treatment sequence (3, 6, 9). Prior to this study, the relative
effectiveness of different points of addition in a conventional trickling
filter plant had not been evaluated simultaneously. To perform such a study,
all four trickling filter pilot plant systems had to be employed. This was
justifiable since previously all four pilot plant systems were found to provide
comparable performance under identical loading conditions.
Three studies with alum doses of 100, 150, and 200 mg/l were conducted utiliz-
ing the above four units. The experimental design in each study was as follows:
System TF- I Control - no a I urn addition
System TF-2 Alum addition to cone at which influent entered
primary clarifier
System TF-3 Alum addition to dosing chamber of trickling filter
System TF-4 Alum addition to cone (recircu lation funnel) at
which filter effluent passed to the secondary clari-
fier, just above point of recircu lation (simulating
sp I it-add ition)
Alum addition was accomplished by pumping a concentrated stock alum solution
at a constant rate to the desired input point with a manifolded Harvard pump.
The design of the pilot units assured turbulence and thorough mixing at the
various points of addition.
Alum CA lotSO^^ ' 18 H203 doses were calculated on the basis of influent flow
to the units, without consideration of the recycle flow. Therefore, in each
instance the actual a I urn concentration was 1/3 of the stated feed concentration.
In the following discussion, the alum concentration will be referred to in
terms of the nominal dose, i.e., the dose calculated on the basis of influent
f low on I y.
The results of these pilot studies are summarized in Tables 22 to 24. Data
collected on days in which unusual operational difficulties were encountered
are not included in these summaries.
Analysis of Tables 22 to 24 indicates that alum addition to the primary clari-
fier is more effective than addition to the trickling filter and as good or
better than split-addition to the primary and secondary clarifiers.
In a follow-up pilot-scale study, two of the four trickling filter systems were
employed with a constant alum dosage of 200 mg/l. One system (TF-4) was modi-
fied to permit a I urn addition into the secondary clarifier below the point of
recycle withdrawal; the point of a I urn addition in the other system (TF-3) was
the primary clarifier. The results of this follow-up study are shown in Table
25.
The two systems exhibited little difference in overall TSS removal efficiency,
77 percent in System TF-3 compared to 75 percent in System TF-4. Examination
of the data indicates a dramatic decrease in TSS concentration in the effluents
from the individual process units to which a I urn was added, but little overall
difference in the final effluents. Average TSS solids removal of 77 and 78
51
-------
TABLE 22
COMPARISON OF ALUM ADDITION (100 mg/l) TO PRIMARY CLARIFIER INFLUENT, TRICKLING FILTER INFLUENT, AND
SECONDARY CLARIFIER INFLUENT AHEAD OF RECIRCULATION
BOO,- (Tot)
j
TOC
TSS
VSS
TS
TIP
TP
Ave, mg/l
Max, mg/l
Mi n, mg/l
% Rernova 1
Avo, mg/l
Max, mg/l
Win, mg/l
% Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
« Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
$ Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
J Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
2 Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
% Remova 1
Control - No Alum Addition
Inf P-l F-l S-l
134 43 35 24
144 65 68 30
114 26 52 20
68 74 82
114 54 46 32
122 73 65 40
103 46 34 18
53 60 72
170 56 57 30
273 87 86 81
122 37 36 12
67 66 82
146 46 49 25
239 63 78 63
104 32 31 9
68 66 83
454 — ~ 291
494 — ~ 367
398 — — 260
36
7.5 7.7 7.5 7.3
9.7 9.6 9.0 9.1
6.4 6.4 2.4 2.0
0 03
9.7 8.7 9.1 8.1
10.7 10.3 10.0 10.4
7.7 7.3 7.9 5.1
10 6 16
Al (S04> • 18 H20 Added
to Pri . Clar. Inf.
P-2 F-2 S-2
98 44 35
171 53 43
45 38 29
27 67 74
86 49 36
218 72 45
38 39 28
25 57 68
171 94 44
528 148 77
69 56 25
0 45 74
125 63 33
388 92 66
61 41 18
14 57 77
309
413
131
32
5.3 5.6 4.0
8.9 12.9 6.2
1.7 2.0 2.1
29 25 47
9.1 7.9 5.5
14.6 15.2 7.6
3.7 3.1 2.2
6 19 43
Al (S04)3 • 18 H20 Added
to Tr. Fi It. Inf.
P-3 F-3 S-3
99 53 41
183 63 63
41 42 30
26 60 69
87 52 48
226 74 66
44 38 32
24 54 58
137 99 57
477 201 110
49 57 16
19 42 66
109 69 39
365 153 70
36 49 16
25 53 73
347
410
314
24
6.6 5.4 5.0
11.4 8.5 8.0
3.9 2.3 2.2
12 28 33
8.8 7.6. 7,3
15.5 11.2 9.6
5.2 3.9 4.7
9 22 25
Al (SO.) • 18 H20 Added
to Sec. Clar. Inf.
P-4 F-4 S-4
42 32 40
45 33 53
38 31 13
69 76 70
93 61 47
84 122 69
55 41 34
18 46 59
152 108 64
340 192 131
67 67 23
II 36 62
100 78 44
277 123 96
55 53 19
32 47 70
346
379
302
24
6.7 5.3 5.2
9.7 11.9 7.8
3.6 2.2 2.1
II 29 31
10. 1 7.7 6.9
15.1 15.2 9.6
7.1 4.5 2.9
0 21 29
Ul
Legend: P - Primary Clarifier Effluent; F- Trickling Filter Effluent
S - Secondary Clarifier Effluent; I, 2, 3, and 4 - System Numbers
-------
TABLE 22 (continued)
COMPARISON OF ALUM ADDITION (100 mg/l) TO PRIMARY CLARIFIER INFLUENT, TRICKLING FILTER INFLUENT,
AND SECONDARY CLARIFIER INFLUENT AHEAD OF RFC IRCULATION
N02-N
N03-N
NH/-N
Kjeld-N
pH
Al
Ave, mg/l
Max, mg/l
Min, mg/l
Ave, mg/l
Max, mg/l
Min, mg/ 1
Ave, mg/l
Max, mg/l
Min, mg/l
Ave, mg/l
Max, mg/l
Min, mg/l
Ave, mg/l
Max, mg/l
Min, mg/l
Ave, mg/l
Max, mg/l
Min, mg/l
Control - No Alum Addition
Inf. P-l F-l S-l
0.05 0.17 0.12 0.16
0.17 0.48 0.26 0.38
>O.OI 0.07 0.07 0.08
O.I 0.4 0.8 0.7
0.3 1.8 2.7 2.1
>0. 1 O.I 0.3 O.I
19.3 17.4 18.1 17.0
31.0 31.0 34.0 36.0
13.0 II. 0 12.5 9.5
28.7 23.7 23.5 18.9
34.0 31.5 27.5 26.5
22.5 17.5 19.5 15.5
7.1 7.2 7.3 7.3
7.3 7.4 7.4 7.5
6.8 6.8 7.1 7.0
0.6 0.4 0.4 0.8
1.3 1.0 0.9 5.2
O.I 0. 1 0. 1 0. 1
Al (SO ) • 18 HO Added
to Prl . Clar. Inf.
P-2 F-2 S-2
0.09 0.15 0.14
0.17 0.34 0.30
0.02 0.08 0.05
0.4 1.2 0.8
1.6 3.4 2.1
>0.l O.I O.I
17.5 18.7 17.5
32.5 29.5 26.0
10.0 9.5 9.0
26.1 21.5 21.5
37.0 32.5 30.5
20.0 15.5 14.0
7.0 7.2 7.2
7.2 7.3 7.4
6.8 6.9 7.0
7.9 6.0 4.5
12.0 8.6 6.4
3.3 0.3 3.2
Al (SO ) • 18 HO Added
to Tr. Fi It. Inf.
P-3 F-3 S-3
0.06 0.12 0.17
0.22 0.21 0.35
0.01 0.06 0.09
O.I 0.4 0.4
0.3 1.0 0.7
>0.l >0.l O.I
19.3 19.3 19.7
32.5 29.0 37.0
14.0 12.5 13.0
27.2 24.3 24.3
33.0 35.0 30.5
22.0 21.0 21.0
7.1 7.1 7.2
7.2 7.2 7.3
6.8 6.8 6.9
5.0 5.9 5.2
9.5 10. 1 8.0
3.2 4.6 2.0
Al (SO,), • 18 H_0 Added
243 2.
to Sec. Clar. Inf.
P-4 F-4 S-4
0.04 0.06 0.07
0. 1 1 0.15 0. 16
0.02 0.01 0.03
0.6 0.7 0.6
4.3 5.5 5.3
0. 1 0. 1 O.I
20.3 19.8 20.0
29.5 35.0 .37.0
13.5 13.5 13.0
28.2 27.8 23.8
34.0 34.5 30.0
23.5 21.5 18.0
7.0 7.1 7.1
7.2 7.2 7.3
6.8 6.8 6.7
5.3 5.6 5.9
8.0 8.0 7.0
0.8 0.9 3.5
Legend: P - Primary Clarltier Effluent; F- Trickling Filter Effluent
S - Secondary Clarifier Effluent; I. 2, 3, and 4 - System Numbers
-------
TABLE 23
COMPARISON OF ALUM ADDITION (150 mg/ I ) TO PRIMARY CLARIFIER INFLUENT, TRICKLING FILTER INFLUENT,
AND SECONDARY CLARIFIER INFLUENT AHEAD OF REC IRCULATION TAKEOFF POINT
BOD (Tot)
TOC
TSS
VSS
TS
TIP
TP
Ave, rng/l
Max, mg/l
Min, mg/l
% Removal
Ave, mg/l
Max, mg/l
Win, mg/ 1
% Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
% Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
$ Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
? Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
Ave, mg/l
Max, mg/l
Min, mg/ 1
? Remova 1
Control - No Alum Addition
Inf P-l F-l S-l
116 61 39 58
120 118 50 138
114 23 35 24
47 64 50
122 82 45 52
152 185 75 114
80 32 17 25
33 63 57
146 102 40 43
200 265 87 154
100 36 25 II
30 73 70
113 75 32 38
158 154 66 138
71 26 16 5
34 72 64
433 — — 327
487 — — 463
389 — — 262
24
6.9 7.4 7.4 7.1
9.0 9.6 8.8 10.0
4.9 5.2 6.0 4.9
9.2 9.3 8.7 8.9
12.0 13.3 11.8 13.2
7.3 7.4 5.5 6.9
6 3
AI2(S04>3 ' 18 H20 Added
to Pri . Clar. Inf.
P-2 F-2 S-2
49 26 33
120 34 48
29 20 10
58 78 72
77 34 30
235 52 39
24 12 13
37 72 75
129 71 46
523 130 84
47 30 19
12 51 68
87 48 28
362 7 1 56
35 16 14
23 58 75
325
372
284
24
6.1 4.6 4.2
17.9 8.7 6.1
1.5 1.7 2.3
6.9 7.3 5.6
16.8 12.5 9.4
3.6 4.2 3.1
25 21 39
AI2(S04)3 ' 18 H20 Added
to Tr. Fi It. Inf.
P-3 F-3 S-3
69 49 30
120 89 45
44 24 23
41 58 74
78 52 42
216 64 72
22 40 28
36 57 66
128 97 68
387 139 158
39 71 36
12 34 53
•95 62 50
302 75 123
34 54 29
84 45 56
362
524
287
16
5.7 5.4 4.8
9.8 6.5 6.5
2.5 2.4 2.9
i 7 97 on
7.8 10.3 '7.2
12.2 15.8 16.2
4.2 7.5 4.5
15 21
Al (S04), ' 18 H20 Added
to Sec. Clar. Inf.
P-4 F-4 S-4
82 46 35
118 60 56
50 34 18
29 60 70
87 47 31
221 73 56
27 8 9
29 61 75
1 62 97 69
335 154 205
64 65 27
34 53
123 65 45
249 96 145
31 39 20
42 60
359
502
289
17
5.7 4.6 4.1
9.3 7.1 5.7
2.3 1.8 2.4
17 T^ ^Q
7.8 7.0 5.4
11.4 10.0 9.0
4.0 3.8 3.9
15 14 14
Legend: P - Primary Clarifier Effluent; F- Trickling Filter Effluent
S - Secondary Clarifier Effluent; I, 2, 3, and 4 - System Numbers
-------
TABLE 23 (continued)
COMPARISON OF ALUM ADDITION (150 mg/l) TO PRIMARY CLARIFIER INFLUENT, TRICKLING FILTER INFLUENT,
AND SECONDARY CLARIFIER INFLUENT AHEAD OF RECIRCULATION TAKEOFF POINT
NO--N
z
N03-N
+
NH4 -N
Kjeld-N
pH
Al
Ave, mg/l
Max, mg/l
Min, mg/ 1
Ave, mg/ 1
Max, mg/l
Min, mg/l
Ave, mg/l
Max, mg/l
Min, mg/l
Ave, mg/l
Max, mg/l
Min, mg/l
Ave, mg/l
Max, mg/l
Min
Ave, mg/ 1
Max, mg/l
Min, mg/ 1
Control - No Alum Addition
Inf P-l F-l S-l
0.06 0.17 0.14 0.16
0.30 0.34 0.28 0.28
0.01 0.03 0.09 0.04
0.5 0.4 0.7 0.7
2.3 0.7 1.2 1.6
O.I O.I 0.3 0.3
21.3 19.6 18.5 18.3
33.0 35.5 32.0 34.5
13.5 13.5 13.5 12.0
24.7 23.7 20.4 20.7
39.5 42.0 34.5 36.5
15.0 15.0 14.5 14.5
7.0 7.1 7.2 7.1
7.5 7.5 7.7 7.7
6.6 6.8 6.8 6.9
0.9 0.7 0.5 1.2
2.5 1.4 0.8 7.5
O.I O.I 0. 1 O.I
AI2
-------
TABLE 24
COMPARISON OF ALUM ADDITION (200 mg/I) TO PRIMARY CLARIFIER INFLUENT, TRICKLING FILTER INFLUENT, AND
SECONDARY CLARIFIER INFLUENT AHEAD OF RECIRCULATION TAKEOFF POINT
BOD5(Tot)
TOC
TSS
VSS
TS
TIP
TP
Ave, mg/l
Max, mg/l
Min, mg/l
% Remove 1
Ave, mg/|
Max, mg/l
Min, mg/l
% Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
% Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
£ Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
$ Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
$ Remova 1
Ave, mg/l
Max, rng/l
Min, mg/l
$ Remova !
Control - No Alum Addition
Inf P-l F-l S-l
163 73 32 18
174 135 39 19
156 39 25 16
55 80 89
112 65 38 39
152 160 47 67
55 29 28 27
29 66 65
IIS 61 29 35
171 163 44 180
72 24 15 5
48 79 70
100 51 23 29
154 128 33 179
31 II 12 2
49 77 71
447 — _. 313
533 — — 397
296 — — 247
30
6.7 6.6 6.7 6.7
8.9 7.8 8.7 8. i
2.8 3.8 4.3 4.3
200
9.7 8.9 9.0 8.7
14.4 1 1 .7 14.6 1 1 .4
4.0 4.7 5.7 5.4
8710
AI2(S04>3 • 18 H20 Added
to Pri . Clar. Inf.
P-2 F-2 S-2
51 26 II
77 50 18
23 6 2
69 84 93
51 27 18
85 38 29
16 10 10
54 76 84
91 43 20
276 94 34
II 20 4
23 64 83
69 33 15
196 58 25
8172
31 67 85
325
467
239
27
4.0 2.6 2.4
7.0 6.9 6.3
<0.5 <0.5 <0.5
40 61 64
5.6 3.8 3.1
10.0 10.5 9.0
<0.5 <0.5 <0.5
42 61 68
AI2(S04>3 ' 18 H20 Added
to Tr. FI It. Inf.
P-3 F-3 S-3
87 35 31
89 43 41
84 24 II
47 78 81
54 38 34
88 58 70
31 15 II
52 66 70
72 77 54
128 157 166
24 32 26
42 35 54
54 55 39
92 138 104
19 23 15
46 45 61
352
480
299
21
4.6 4.6 3.9
8.3 11.5 8.0
1.0 0.7 <0.5
31 31 42
7.1 7.6 6.5
16.3 21.0 19.0
1.9 2.6 <0.5
27 22 33
AI2
-------
TABLE 24 (continued)
COMPARISON OF ALUM ADDITION (200 mg/l) TO PRIMARY CLARIFIER INFLUENT, TRICKLING FILTER INFLUENT, AND
SECONDARY CLARIFIER INFLUENT AHEAD OF RECIRCULATION TAKEOFF POINT
NO-N
NO,-N
NH +-N
Kjeld-N
pH
Al
Ave, mg/l
Max, mg/l
Min, rng / 1
Ave, mg/l
Max, mg/l
Min, mg/l
Ave, mg/l
Max, mg/ 1
Min, mg/l
Ave, mg/l
Max, mg/l
Min, rng/ 1
? Remova 1
Ave, mg/l
Max, mg/l
Min, mg/l
Ave, mg/ 1
Max, mg/ 1
Win, mg/l
Control - No Alum Addition
Inf. P-l F-l S-l
0.03 — — 0.14
0.05 — — 0.22
3 ' 18 H20 Added
to Pri . Ciar. Inf.
P-2 F--2 S-2
0.38
1.20
-------
TABLE 25
COMPARISON OF ALUM ADDITION (200 mg/l) TO PRIMARY CLARIFIER INFLUENT AND SECONDARY INFLUENT
AFTER RECIRCULATION TAKEOFF POINT
Without
Al urn
2/13/72
Ave
With
Alum
Add i t ion
2/22/12
2/26/72
2/29/72
3/5/72
3/7/72
3/9/72
Ave
138
123
*
195
164
169
194
285
160
193
TSS,
46 56
72 61
52 45
33 57
44 75
27 51
H7 94
126 123
61 69
mg/l
27
28
52
29
38
12
25
34
42
30 89
46 92
152 84
138 132
146 36
106 148
70 254
96 281
136 163
14
26
22
25
49
40
45
38
47
BOD,-, mq/l
204 71 62 39 77 159 50
126 27 29 29 114 68 15
150 36 42 27 78 112 23
204 39 38 23 84 117 21
160 34 36 26 92 99 20
TOC, mg/l
93
43 46
38 46
65 29
93
152
125
MO
139
193
151
147
43 46
35 35
32 38
41 43
28 32
42 42
56 49
39 40
38 46
32 127
21 112
37 58
21 107
28 86
28 101
31 105
65 29
85 23
99 26
115 30
102 28
121 26
141 25
107 28
Total System
TP Remova 1 , %
0 8
30 26
16 12
15 15
70 76
77 75
67 68
85 69
72 73
77 80
72 70
Ul
CXi
Legend: P - Primary Clarifier Effluent; F - Trickling Filter Effluent
S - Secondary Clarifier Effluent; 3 and 4 - System Numbers
* On System 3, 200 mg/l AI2(SO ) -18 H20 added to primary clarifier influent.
On System 4, 200 mg/l AL^CSOJ-,-I 8 hLO added to secondary clarifier influent
after reelrculation takeoff point.
-------
percent were achieved in the week just prior to the alum study. Therefore,
it must be concluded that under the conditions of this pilot study, TSS removals
were not enhanced by a I urn precipitation.
Visual observation, however, indicated that the character of effluent TSS during
alum addition differed from that when no alum was added. Alum floe was visible
during alum addition and probably contributed to the effluent TSS.
TOC removal was slightly better (80 percent) in System TF-4 than in System TF-3
(73 percent). Likewise, 6005 removal was slightly better in System TF-4 (87
percent, compared to 83 percent). Again, examination of the data shows a drama-
tic decrease of 6005 and TOC in the effluents of the individual process units
to which alum was added, but little overall difference in total plant efficiency.
Insufficient data were available from the time period immediately prior to this
study to make a valid comparison of BOD^ and TOC removals with and without a I urn
add it ion.
Phosphorus removal efficiency of the two systems was about the same, 72 percent
in System TF-3 compared to 70 percent in System TF-4.
From these pilot-scale investigations, the following conclusions were reached:
(I) Alum addition substantially improved phosphorus removal,
(2) Increased alum dosages up to 200 mg/l (the highest dose tested)
resulted in increased phosphorus removal,
(3) Alum addition to trickling filter influent or to trickling
filter effluent above the takeoff point of recirculation was
less effective for phosphorus removal than was addition to
primary clarifier influent or to influent to the secondary
clarifier below the takeoff point of recirculation, and
(4) Overall removals of 8005, TOC, and phosphorus were essentially
the same when alum was dosed to primary clarifier influent or
to influent to the secondary clarifier below the takeoff point
of recirculation.
'. 59
-------
SECTION V
ALUM ADDITION TO CHAPEL HILL MAIN PLANT
On the basis of laboratory and pilot-scale results, the addition of metal salts
to trickling filter effluent appeared to offer a technically feasible alter-
native of upgrading overall trickling filter plant performance. This was con-
firmed by a preliminary report (18) of full-scale treatment with liquid alum
at Richardson, Texas. The Richardson results indicated that the addition of
alum to trickling filter effluent just prior to final clarification was effect-
ive in removing phosphorus and enhancing general plant performance. An ini-
tial trial adding alum ahead of primary clarification at Richardson resulted
in apparent overloading and impending failure of the plant's unheated anaerobic
digesters, necessitating termination of that trial. Split addition of a I urn
with 20 percent to the primary clarifier and 80 percent to the final clarifier
produced very efficient phosphorus removal. However, when the split was alter-
ed with 30 percent to the primary clarifier and 70 percent to the final clari-
fier, early but unmistakable signs of digester problems again developed. In
a final trial, alum addition ahead of the final clarifier provided good phos-
phorus removal with no indications of impending digester upset. This point of
a I urn application was adopted for extended study at Richardson.
The addition of a I urn to pilot trickling filter plants at Chapel Hill, as de-
scribed previously in this report, indicated that approximately equal results
could be obtained with the addition of alum to either primary or final clari-
fiers. However, primarily because of the preliminary work at Richardson,
Texas, it was decided to apply alum in the Chapel Hill main plant at a point
just ahead of final settling in one of the plant's two trains. No a I urn was
added to the other train. Therefore, when the plant flow was equally divided,
overall performance of the two trains could be compared on the basis of phos-
phorus removal or any other parameter of interest.
There were a number of reasons for attempting to confirm the promising results
obtained at Richardson, Texas at the plant in Chapel Hill. The principal con-
siderations in this regard were the following:
I. The Richardson plant is a standard-rate trickling filter facility,
whereas the Chapel Hill plant is a high-rate facility.
2. Except for secondary sludge return to the primary settling tanks,
no recirculation is provided at Richardson. At Chapel Hill,
trickling filter effluent is recirculated through the primary
sett I ing tanks.
60
-------
3. Only one final settling tank exists at the Richardson plant,
and it is, therefore, impossible to simultaneously compare
the results of normal operation with chemical treatment. On
the other hand, it is possible to operate the two trains of
the Chapel Hill plant as if they are two separate facilities.
With this flexibility, many direct comparisons are possible
between normal operation and operation with chemical addition.
4. The single final settling tank at Richardson was quite large
relative to normal design practice. The surface loading on this
tank at the design flow of 5,678 m3/day (1.5 mgd) is 16.7 m3/day/m2
(410 gpd/ft2). The average surface loading on the Chapel Hill
final tanks is about 36.7 m3/day/m (900 gpd/ft2), which is
more typical of final tanks at^many high-rate trickling filter
pI ants.
5. At Richardson, secondary sludge is returned to three primary
settling units which consist of two-story tanks; the upper levels
are clarifiers and the bottom sections are unheated digesters.
Since the combined primary and secondary sludges drop directl-y
through the primary clarifiers into the digesters below, it is
impossible to measure the actual quantity of sludge produced.
At Chapel Hill, all sludge, both primary and returned secondary,
is accumulated in the two primary settling tanks and pumped
independently from each train to one separate anaerobic digestion
facility. Accordingly, it is possible to measure the quantity
of sludge produced on each side of the plant. The quantity of
sludge resulting from chemical treatment is one of the most impor-
tant factors in designing solids treatment and disposal systems.
THE CHAPEL HILL TREATMENT PLANT
The wastewater treatment plant for Chapel Hill is a conventional high-rate
trickling filter installation treating predominantly domestic wastewater.
There is substantially no industrial or other unusual contribution, except for
the hospital and laboratories of the University of North Carolina. Figure (7
is a partial flow sheet for the plant, and Table 26 summarizes characteristics
and design parameters of major plant units.
Incoming wastewater passes through a mechanically cleaned bar screen, with a
manual unit serving as a backup in case of failure. Subsequently, the flow is
metered and grit removed in a detritor. Design of the grit removal effluent
structure allows splitting of flow into any desired proportions for diversion
to the two identical treatment trains.
Based on a total plant infIuentf low of M.355 m3/dav (3.0 mgd) equaI Iv divided be-
tween two trains, and a recycle ratio of 2:1, me Zl.3-m (70-ft) primary claritiers
provide 1.8 hr detention time at an overflow rate of 48 m^/day/m^ (1,180 gpd/
f t ). Each trickling filter is 36.6 m (120 ft) in diameter with a stone depth
of 1.3 m (4.25 ft), providing a "design" loading of about 0.56 kg/day/m3 (35
61
-------
RAW WASTEWATER
SUPERNATANT
DIGESTERS
o-
SIJUD6E_AND_
SCUM
CENTRIFUGE
CAKE TO
LANDFILL
SLUDGE
DRYING
BEDS
CENTRATE
X
MANUAL AND MECH.
BAR SCREENS
>< PARSHALL FLUME
GRIT REMOVAL
FLOW SPLITTER
PRIMARY
SETTLING
TRICKLING
FILTER
WET
WELL
FINAL
SETTLING
EFFLUENT
FIGURE 17. PARTIAL FLOW SHEET FOR CHAPEL HILL WASTEWATER TREATMENT PLANT
-------
TABLE 26
CHARACTERISTICS OF AND DESlbN PARAMETERS FOR UNITS IN
CHAPEL HILL WASTEWATER TREATMENT PLANT
CURRENT AVERAGE FLOW:
Approximately 3.0 mgd
SCREENS:
a) One automatic, mechanically-cleaned
b) One manually-cleaned (standby)
GRIT REMOVAL:
One mechanically-cleaned detritor
PRIMARY SETTLING (Two units):
a) Diameter 70 ft
b) Water depth 12 ft
c) Detention time 1.8 hr (@ 2:1 Recycle)
d) Overflow rate 1,180 gpd/ft2 (@ 2:1 Recycle)
e) Mechanical sludge and scum removal
TRICKLING FILTERS (Two units):
a) Diameter 120 ft
b) Stone depth - 4.25 ft
c) Rotary distributors
d) BODj loading about 35 Ib/day/IOOO ft (assuming 1/3 removal during primary
treatment)
e) Hydraulic loading 17 mgad (@ 2:1 Recycle)
FINAL SETTLING (Two units):
a) Diameter 45 ft
b) Water depth = 10 ft
c) Detention time 1.9 hr
d) Overflow rate = 960 gpd/ft2
e) Mechanical sludge removal
RECIRCULATION PUMPS:
In each battery, one 1.5 mgd and two 3.0 mgd units
ANAEROBIC DIGESTERS:
a) One 75 ft diameter, mechanically mixed, heated, floating cover (primary
d igester)
b) One 50 ft diameter, mechanically mixed, heated, floating cover (secondary
d igester)
c) One 50 ft diameter, no mixing, floating cover (not now in operation)
d) Heat exchanger (digester gas or propane) for units in operation now,
including pumps, control valves, and interconnecting piping
SLUDGE DEWATERING:
a) 18 sand drying beds, 25 ft x 50 ft, uncovered
b) One 18-in solid-bowl, 15-17 gpm, Bird centrifuge
Convers ions:
I . I in 2.54 cm
2. I ft 0.3048 m
3. I gpd/ft2 0.04074 m3/day/m2
4. I lb/day/1,000 ft3 0.01602 kg/day/m3
5. I gpm 0.06308 I/sec
6. I mgad 9.553 m3/day/ha
63
-------
Ib BOD^/day/IOOO ft )(assuming one-third BOD5 removal in the primary clari-
fiers) at a hydraulic loading of approximately 159,000 m3/day/ha (17 mgad).
Trickling filter effluent passes through a wet well from which any or all of
three pumps take recycle at rates up to 22,710 m^/day (6.0 mgd) in each train.
Net plant flow (no recycle) passes to 13.7 m (45 ft) diameter final clarifiers,
providing 1.9 hr 'detention time with an overflow rate of 39.1 nvVday/rr/ (960
gpd/ft2) at 5,678 m3/day (1.5 mgd) through each train.
Normal plant operation is based on recycle to each primary clarifier influent;
however, a connection has been provided to permit recycling directly around
each trickling filter. Series or stage operation of the filters is not possi-
ble. Typically, the plant operates with the two trains in parallel as shown
in Figure 17, in effect providing two separate treatment facilities. The in-
fluent wastewater flow can be divided between these as desired for operation
at different loadings. Recycle in each train may be adjusted independently.
Sludge from each final settling tank is returned to its respective primary
clarifier where it resettles in combination with primary sludge. Sludge and
scum are pumped from the primary clarifiers to a 22.9 m (75 ft) diameter first-
stage or primary anaerobic digester equipped with floating cover and mixer.
A 15.2 m (50 ft) diameter digester, with infrequently used mechanical mixer,
serves as a second-stage digester. Supernatant from the secondary digester is
slowly decanted during periods of low plant flow and is returned to the plant
headworks. Gas produced in the process is utilized for heating the sludge
digesters, and the excess is flared.
Digested sludge is usually dewatered with the plant's centrifuge. This machine
is backstopped by 18 uncovered sand drying beds. The centrifuge also may be
used for dewatering undigested sludge if unusual circumstances require reduc-
tion of loading on the sludge digesters.
PREPARATORY WORK FOR ALUM TREATMENT
As it was decided to add a I urn just ahead of one final clarifier, it was neces-
sary to examine the diurnal characteristics of the trickling filter effluent
entering this tank. Accordingly, a short term sampling program was conducted
during December 1971 and January 1972. Around-the-clock samples were obtained
of final settling tank influent. The parameter measured was total phosphorus
as this was thought to be most significant in terms of a I urn requirements.
During this sampling period, filter reelrculation ratios were held at about
2.0. The recirculation almost completely suppresses the diurnal variation in
concentration of normal plant influent wastewater constituents, e.g., BOD, TSS,
etc. This suppression of variation in concentration was also found in the
case of total phosphorus. In fact, no discernable pattern of diurnal varia-
tion in total phosphorus concentration was observed in final clarifier influ-
ent. The loading rate (mass/time) of total phosphorus to the final clarifier
was found to be approximately proportional to flow. The typical total phos-
phorus diurnal loading pattern as determined by the sampling program is shown
by the solid curve in Figure 18.
64
-------
200
150
o
T3
100
50
Iiiiiiiiiir^
TP LOADING
— HIGH AND LOW PUMP
.__ j
9
A.M.
I
I
LOW PUMP
I ill I I
5 7
P.M.
300
250
200
150
100
50
-
u
o
c/>
£
m
CL
Q
I
3 5
A.M.
FIGURE 18. TYPICAL DIURNAL VARIATION IN TOTAL PHOSPHORUS LOADING TO FINAL
CLARIFIER AND CORRESPONDING AI DOSAGE PATTERN.
-------
With the budget remaining at the time the plant-scale a I urn work was started,
it was not possible to purchase a flow-paced alum feeding system. However,
the pumping capacity of two available Lapp chemical pumps was adjustable and
an arrangement was devised using simple electrical timing devices so that the
pumps could be sequenced in and out of service at selected times of the day.
The initial pumping capacities of the two pumps were adjusted to 970 ml/min
and 388 ml/min (15.35 gph and 6.14 gph). Using liquid alum with an AKIN)
content of 4.4 percent, by weight, these pumping rates are equivalent to 82.0
kg (180.7 Ib) and 32.8 kg (72.3 Ib) of AKIN) per day, respectively. With
both pumps operating together the resulting Aid I I) feed rate was 114.8 kg/
day (253 Ib/day). These three feed rates were fitted to the phosphorus load-
ing curve as illustrated by the dashed lines in Figure 18. The daily total
phosphorus loading shown by the solid curve was equal to 55.8 kg/day (123 Ib/
day). The total daily dosage of Al(lll) represented by the dashed lines was
93 kg/day (205 Ib/day). Figure 18 represents a daily ratio of Al:TP on a
weight basis of 1.67; on a molar basis the ratio was 1.92. Daily alum dosage
could be varied by adjusting the feed rate of each pump. When this was done,
both pumps were changed by the same percentage.
FACILITIES FOR ALUM TREATMENT AT CHAPEL HILL
The program of alum treatment was undertaken during the last few months of a
three-year contract to explore methods of improving the operation of trickling
filter treatment plants. When the a I urn treatment program was initiated, little
money was available for new equipment. It proved to be possible, however, to
locate and install equipment for handling, storing, and feeding liquid alum
with minimum capital cost.
Liquid a I urn was stored in a 25 m (6,600 gal) plastic-1ined swimming pool. As
liquid a I urn weighs about 1,318 kg/m^ (II Ib/gal), or 32 percent more than water,
the pool was never filled above the 19 nrp (5000 gal) level. Concern about pos-
sible problems with the pool proved unjustified; no difficulties were experi-
enced. The a I urn tank was set on a sand base and enclosed in a simple unheated
plywood structure. This structure also enclosed the a I urn pumps and the pump
control system. Although the ambient temperature fell as low as 10 °F during
night time periods in February 1972, no crystallization problems were encount-
ered.
Alum was fed with two positive displacement chemical feed pumps programmed to
operate as previously described. Liquid alum was drawn up over the side of the
storage tank with the pumps setting close to the tank, about 0.3 m (I ft)
above the tank floor. Separate discharge lines from the two pumps fed into a
3.8 cm (I 1/2 in) polyethylene pipe, and the liquid a I urn flowed by gravity,
undiluted with water, a distance of about 18 m (60 ft) where it was discharged,
in free fall, into the entrance box of the final settling tank. An elevation
view of the a I urn feed system is shown in Figure 19.
When a I urn is introduced into water containing soluble phosphorus, the reaction
resulting in the final formation of AI(OH)3 may reduce the amount of soluble
Al(lll) available for the precipitation of phosphorus. On the other hand, the
driving forces of the phosphate precipitation reaction indicate that if the
66
-------
ALUM PUMPS
D-n
^WW ^^^X,™™^
LIQUID ALUM TANK
SECONDARY SLUDGE ^_
TO PRIMARY TANKS
TR.FILT.
EFF.
I HP MIXER
FINAL SETTLING TANK
FIGURE 19. ELEVATION VIEW OF ALUM FEEDING SYSTEM.
-------
soluble AI(I I I) is rapidly and complete Iy dispersed in the water, phosphate
will be precipitated before any significant amount of aluminum reacts with
the hydroxide alkalinity. It is important, therefore, to introduce the alum
at a point where the turbulence is high in the entire flowing stream. A 0.75-
kW (l-hp) mixer was installed in the entrance box to the final clarifier. Al-
though a considerable amount of natural turbulence existed in this box due to
the entrance of water into the drop pipe leading to the clarifier, the mixer
provided additonal turbulence to insure very rapid dispersion of the alum.
After initial mixing of the alum with the waste stream in the entrance box, the
wastewater flowed down a 61 cm (24 in) drop pipe and then horizontally under
the settling tank to the vertical riser pipe in the center of the tank. Waste-
water discharged into the tank through ports in the top of the vertical pipe
inside a 76 cm (30 in) deep annular skirt, and then flowed in a radial direction
from the bottom of the annular skirt to the effluent weirs located on the peri-
meter of the tank.
Flocculation of precipitated aluminum-phosphorus compounds and AKOhD-j complexes
along with the entrained organic solids occurred in the settling tank feed pipes,
in the annular section enclosed by the skirt, and in the settling tank itself.
It is logical to assume that more effective flocculation could be obtained in an
efficiently designed flocculation chamber. On the other hand, the simple ad
hoc system described above functioned satisfactorily, i.e., well-formed floe
particles were observed in the settling tank most of time when the a I urn dosage
was above I 50 mg/l.
Settled secondary sludge from the alum treatment train was returned continuous-
ly to its primary settling tank by the secondary sludge pump at a rate of 13.2
I/sec (210 gpm). The combined primary and secondary sludges were pumped from
the primary settling tank to the primary digester twice each day, at 7 a.m.
and 5 p.m. Secondary sludge from the train not treated with alum (i.e., the
control train) was returned continuously to its primary settling tank at a
rate of 12.0 I/sec (190 gpm). Combined primary and secondary sludges from the
control train were separately pumped to the common primary digester. It was,
therefore, possible to separately measure the flow rate of and sample the
sludge pumped from each primary settling tank.
Sludges from both trains were digested anaerobicaIly in the single 22.9 m (75 ft)
diameter primary digester. This digester was mixed daily and was maintained
at a temperature of 95 °F. Sludge from the primary digester was transferred
daily to the 15.2 m (50 ft) diameter secondary digester. The secondary diges-
ter was neither heated or mixed; its function was to thicken the digested
primary sludge prior to final dewatering and to allow the separation of a re-
latively clear supernatant for return to the head end of the treatment plant.
Settled sludge from the secondary digester could be dewatered in either the
plant's solid-bowl centrifuge or on the sand drying beds. The centrifuge was
utilized during the alum treatment investigation to the greatest extent possi-
ble.
68
-------
SAMPLING AND ANALYSIS
During the a I urn treatment study, daily flow-weighted composite samples were
obtained from all main-flow streams and the secondary sludge returns on both
trains of the treatment plant using an automatic sampling system. The sampl-
ing points included in the automatic system are listed below:
Influent wastewater
Primary effluent
Trickling filter effluent
Final effluent
Secondary sludge return
The automatic sampling system was necessary as the Chapel Hill plant is not
manned from 6 p.m. to 7 a.m. All waste streams sampled automatically were
piped to a set of standpipes at a central location. Waste streams flowed
through the standpipes continuously at a velocity sufficient to prevent the
accumulation of solids. Each sample delivery pipe was equipped for back flush-
ing. Wastewater flowing through the standpipes was continuously wasted to the
recirculation well on one side of the plant. Plant influent and primary
effluents were delivered to the standpipes by gravity through 3.8 cm (I 1/2
in) piping. Trickling filter effluents and secondary sludge returns were
delivered through 2.5 cm (I in) piping under the discharge pressure of the
filter recirculation pumps and secondary sludge return pumps. Final effluents
were delivered by means of small submersible centrifugal pumps mounted adja-
cent to the final tank weirs. Samples of the individual waste streams flowing
through the standpipes were obtained with a rotary multitube peristaltic pump.
Each pump tube of the peristaltic pump was connected to a particular standpipe.
The peristaltic pump was actuated by a program timer so that the interval be-
tween sampling was approximately inversely proportional to plant flow. During
the first 30 sec of sample pump operation, the lines to the sample containers
were flushed to waste. Sample portions were obtained during the final second
of pump operation. The flushing-sampling system was controlled by separate
adjustable timers. Each final effluent sample was accumulated in a separate
container stored in a cold chest and maintained at a temperature of 2-4 °C.
The normal sampling day began and ended at 8 a.m. The daily flow-weighted
composite samples were collected Sunday through Thursday during each week.
In addition to the above, grab samples were also obtained on a regular basis
from the following points in the plant:
Sludge pumped to digester - both trains
Digested sludge in primary digester
Digested sludge in secondary digester (sludge feed to centrifuge)
Centrifuged sludge
Centrate from centrifuge
Supernatant from secondary digester
A summary of sampling points and analyses conducted is given in Table 27. The
methods for chemical analyses used in this study are standard procedures listed
previously in Table 5.
69
-------
TABLE 27
POINTS OF SAMPLING AND ANALYSES CONDUCTED FOR MAIN-PLANT ALUM ADDITION STUDIES
BOD5 (Tot)
TOC
TSS
VSS
TS
VS
TP
TIP
Sol. P
NH/-N
Kje!d-N
NO,-N
Turbid ity
PH
Raw
W. W.
A
A
A
B
-
-
A
A
-
B
B
B
A
A
Primary
Eff.
1 & 2
A
A
A
B
-
-
A
A
-
B
B
B
A
A
Tr.
Fi it.
1 & 2
A
A
A
B
-
-
A
A
-
B
B
B
A
A
Sec.
Eff.
1 & 2
A
A
A
B
-
-
A
A
A
B
B
B
A
A
Sec.
S 1 udge
Return
1 & 2
B
A
A
B
-
-
A
-
-
-
-
-
-
-
Si udge
to
Digester
-
-
-
-
A
A
-
-
-
-
-
-
-
-
Digester
Sup.
-
B
B
B
-
-
B
B
-
B
B
-
-
-
Digested
SI udge
-
-
-
-
B
B
-
-
-
-
-
-
-
-
Centri-
f uged
S 1 udge
-
-
-
-
B
B
-
-
-
-
-
-
-
-
Centrate
B
B
B
B
-
-
B
-
-
-
-
-
-
-
Key: A = 5 days/week; B = I or 2 days/week; - = analyses not performed
NOTE: Alkalinity, pH and volatile acid measurements were run on primary digester sludge, I or 2 days/week
-------
DESCRIPTION OF EXPERIMENTAL PROGRAM AND PERFORMANCE RESULTS
The objectives of the plant-scale experimental work with a I urn were to determine
the effectiveness of a I urn application in a final settling tank at a typical
high-rate trickling filter plant as related to phosphorus removal and general
enhancement of plant performance and to determine the sludge quantities result-
ing from the alum treatment.
During the first phase of the investigation, experiments were conducted at
several alum dosage rates and with three different dosage program schedules.
The objective during this phase was to determine which overall dosage rate
and daily application schedule would be most effective. This phase extended
from January 25 through May 25, 1972. During this entire period the flow was
divided equally between the dosed and control trains.
The second phase of the program consisted of diverting lesser fractions of the
total plant flow to the alum dosed train while maintaining an effective dosage
program as determined in the first phase. The total amount of alum applied
was, of course, reduced in proportion to flow. The objective of this phase was
to determine the effect of hydraulic loading on final settling tank performance.
This phase extended from June 3 through August 27, 1972.
During the final phase of the investigation, experiments were conducted using
two polye Iectrolytes along with alum to determine if polye Iectrolyte addition
would benefit the process. This phase extended from August 28 through October
5, 1972.
A summary of the experimental periods throughout the entire alum treatment in-
vestigation is given in Table 28. The a I urn dosage programs are shown in Table
29. Following are brief statements concerning the rationale for the 18 experi-
mental periods into which the three general phases of the investigation were
divided. Performance results for the 18 experimental periods for both the
dosed and undosed trains are summarized in Tables 30 to 33.
Experimental Periods
Period I. The Aid II) dosage was set at 93 kg/day (205 Ib/day). Dosage Pro-
gram A as shown in Table 28 was followed. Flow to the a I urn train averaged
5,413 m3/day (1.43 mgd). The average alum [AI2(S04)3 • 14 H20] concentration
dosage was 195 mg/l. Based on influent or raw wastewater total phosphorus,
AI:TP (mole) averaged 2.2. Effluent total phosphorus averaged 1.6 mg/l for
the alum dosed train and 7.3 mg/l for the undosed train. Influent total phos-
phorus averaged 9.1 mg/l. Removals of BOD5, TSS, and TOC were substantially
higher on the train receiving alum treatment.
Period 2. The weight dosage of AI(lll) was reduced to 90 percent of that of each
time interval in Period I. No change was made in the timing of the dosage program
Flow to the alum train averaged 5,564 m^/day (1.47 mgd). Alum concentration
dosage averaged 171 mg/l, and the mean Al:influent TP (mole) was 1.8. Effluent
total phosphorus for the alum train averaged 2.0 mg/l. As will be shown later,
it is probable that the higher effluent total phosphorus was due to the slight
increase in flow rate through the final settling tank.
71
-------
TABLE 28
EXPERIMENTAL PERIODS FOR FULL-SCALE PHOSPHORUS REMOVAL STUDIES
Experimenta 1
Period No.
1
2
3
4
5
6
7
8
9
10
II
12
13
14
15
16
17
18
Dates
1/25-2/13/72
2/15-2/25
2/26-3/8
3/9-3/19
3/20-3/26
3A27-4/6
4/7-4/20
4/26-5/25
6/3-6/19
6/24-7 /I 1
7/17-8/2
8/21-8/27
8/28-8/30
9/6-9/1 1
9/12-9/17
9/18-9/21
9/25-9/28
9/30-10/3
10/4-10/5
Dosage
Program*
A
B
C
D
E
A
F
F
G
H
1
F
F
F
F
J
J
G
Percent Flow
to Alum Train
50
50
50
50
50
50
50
50
33
20
40
50
50
50
50
50
50
50
Po-lyelectrolyte
-
-
-
-
-
-
-
-
-
-
-
-
Nalco 610
(0. 1 mg/l)
-
Nalco 610
(0. 1 mg/l)
Nalco 677
(1.0 mg/l)
Nalco 677
(0.3 mg/l)
Nalco 677
(0.3 mg/l)
*Refer to Table 29.
72
-------
TABLE 29
ALUM DOSAGE PROGRAMS
Program
Time, hr
AI (I I I) Dosage Rate
Ib/day
A
0000-0300
0300-0700
0700-1000
1000-2400
181
72
181
253
B
Same as A
A x 0.90
0000-0300
0300-0700
0700-1000
1000-2400
Same as A
Same as A
2330-0300
0300-0700
0700-0930
0930-2330
Same as F
Same as F
Same as F
Same as F
163
0
163
228
A x 0.85
A x 0.80
181
72
181
253
F x 0.67
F x 0.40
F x 0.80
F x I.10
Conversion: I Ib/day =0.4536 kg/day
73
-------
TABLE 30
MAIN PLANT PHOSPHORUS REMOVAL
Period
No.
Dates
1 I./25/72-2/ 13/72
2 2/15
3 2/26
4 3/9
5 3/20
6 3/27
7 4/7
8 4/26
9 6/3
10 6/24
II 7/i7
12 8/21
13 8/28
14 9/6
15 9/12
16 9/18
17 (9/25
17 19/30
18 10/4
-2/25
-3/8
-3/19
-3/26
-4/6
-4/20
-5/25
-6/19
-7/11
-0/2
-8/27
-8/30
-9/1 1
-9/17
-9/21
-9/28
-10/3
- 1 0/5"
Inf. TP
mg/l
9. 1
9.4
I 1 .0
1 1 .9
14.3
13.0
13. 1
11.8
13.9
1 1 .3
13.7
1 1.6
1 1.7
10.9
11.5
12.3
14. 1
12.2
Train
Flow Alum
mgd mg/l
1.43 195
1.47 171
i.36 165
1.19 192
1.51 143
1.51 183
1.54 177
1.40 198
0.73 250
0.45 254
0.90 241
1.07 242
1.41 191
1.18 245
1.35 202
1.47 202
1.42 212
1 . 62 113
No. 2 (Alum)
Al :TP
[mole]
2.2
1.8
1.7
1.7
1.0
1.5
1.4
1.7
1.9
2.2
1.9
2.2
1.7
2.7
1 .7
1 .7
1.6
0.95
Eff
TP
mg/l
1.6(1)
2.0
2.0(2)
1 .3
4.6
2.2(3)
2.3
1.5
0.8
1.2
1.2
1 .8
2.3
3.0
2.1
2.2
2.6
4.4
Eff.
Sol. P
mg/l
0.9
0.2
0.2
0.25
0.3
0.2
0.2
0.3
0.3
0.5
0.7
0.8
0.4
0.7
0.4
0.4
0.5
1.0
% Rem ' 1
TP
82
79
82
69
68
83
82
87
94
89
91
84
80
72
82
32
32
64
Train
Flow
mgd
1.43
.47
.36
.19
.51
.51
.54
.40
.46
.78
.35
.07
.41
. 18
1.35
1.47
1.42
1.62
No. 1 (No Alum)
Eff. Eff.
TP Sol. P
mg/l mg/l
9.1 3.8
8.4 4.6
8.8 5.9
9.1 6.2
10.0 5.5
10.5 6.6
10.4 6.5
8.3 6.0
10.9 7.5
9.1 7.2
10.3 7.5
10.5 8.5
9.6 8.2
10.5 7.3
9.8 7.4
10.6 7.5
9.7 7.7
% Rem' 1
TP
0
1 1
20
23
30
19
21
30
22
19
25
10
18
4
15
14
—
21
Remarks
(1) excludes eff. observation
for 1/29/72: 6.7
(2) excludes eff. observation
for 2/28/72: 6.9
(3) excludes eff. observation
for 4/10/72: 1 1 .7
+0.1 mg/l Mai co 610 to Train No. 2
t O.I mg/l Nalco 610 to Train No. 2
t 1.0 mg/l Nalco 677 to Train No. 2
+0.3 mg/l Nalco 677 to Train No. 2
t 0.3 mg/l Nalco 677 to Train No. 2
•-J
Conversion: I mgd = 3,785
-------
TABLE.3 I
MAIN PLANT BOD5 REMOVAL
Period
No. Da
tes
1 1/25/72-2/13/72
2 2/15
3 2/26
4 3/9
5 3/20
6 3/27
7 4/7
8 4/26
9 6/3
10 6/24
il 7/17
12 6/21
13 8/28
14 9/6
15 9/12
16 9/18
17 {9/30
18 10/4
-2/25
-3/6
-3/19
-3/26
-4/6
-4/20
-5/25
-6/19
-7/1 1
-8/2
-3/27
-8/30
-9/1 1
-9/17
-9/21
-9/28
-10/3
-10/5
Inf. 30D5
mg/l
174
153
159
169
149
173
190
160
223
173
223
176
178
164
160
176
177
196
Flow
fngd
1 .43
1 .47
1 .36
1. 19
1.51
1.51
1 .54
1 .40
0.73
0.45
0.90
1.07
1.41
1. 18
1 .35
1 .47
1 .42
1 .62
Tra in No.
Eff. BOD
mg/l
19(4)
16
16
13
17
14
15
1 1
9
9
1 1
17
19
14
17
17
21
18
2 (Alum)
Eff. Sol.
BOD,-
mg/1
-
T
-
-
-
-
5
3
5
5
5
9
6
7
9
1 1
12
i Rem' 1
BOD5
89
90
90
92
89
92
92
93
96
95
95
91
89
91
89
90
88
91
Flow
mg/l
1 .43
.47
.36
. 19
.51
.51
.54
.40
.46
.78
.35
.07
.41
. 18
.35
.47
.42
1 .62
Train No.
Eff. B005
mg/l
39
52
44
43
47
41
52
35
51
36
52
35
24
27
33
50
41
28
1 (No A 1 urn)
Eff. Sol.
BOD5 %
mg/l
_
_
-
-
-
-
-
8
14
16
13
9
12
13
1 1
1 1
17
14
Rem1 1
BOD5
76
66
72
74
70
76
72
78
77
79
76
81
86
83
79
.73
77
86
Remarks
(4) excludes eff. observation
for 1/25/72: 63
+ O.I mg/l Nalco 610 to Train No. 2
+0.1 mg/l Nalco 610 to Train No. 2
+ 1.0 mg/l Nalco 677 to Train No. 2
+ 0.3 mg/l Nalco 677 to Train No. 2
+0.3 mg/l Nalco 677 to Train No. 2
-J
Ul
Conversion: I mgd = 3,785 m^/day
-------
TABLE 32
MAIN PLANT TOTAL SUSPENDED SOLIDS REMOVAL
Per iod
No.
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
15
16
17
18
Dates
1/25/72-2/13/72
2/15
2/26
3/9
3/20
3/27
4/7
4/26
6/3
6/24
7/17
8/21
8/28
9/6
9/12
9/18
fl/25
V30
10/4
-2/25
-3/8
-3/19
-3/26
-4/6
-4/20
-5/25
-6/19
-7/11
-8/2
-8/27
-8/30
-9/1 1
-9/17
-9/21
-9/28
-10/3
-10/5
Inf. TSS
mg/l
167
175
200
227
258
245
319
267
310
229
291
240
323
230
253
230
268
261
Train
Flow
mgd
.43
.47
.36
. 19
.51
.51
.54
1.40
0.73
0.45
0.90
1.07
1.41
1.18
1.35
1.47
1.42
1.62
No. 1 (Alum)
Eff. TSS %
mg/l
29
29
30
27
45
38
49
25
20
19
15
22
30
28
24
25
23
32
Rem1 1
TSS
83
82
84
87
81
83
84
90
96
92
94
90
90
88
90
89
91
88
Train
Flow
mgd
.43
.47
.36
. 19
.51
.51
.54
.40
.46
.78
.35
.07
.41
.18
.35
.47
1.42
1 .62
No. 2 (No
Eff. TSS
mg/l
63
57
55
58
69
77
74
56
72
43
56
43
37
38
51
51
53
33
Alum)
t Ren'l
TSS
63
65
71
74
69
67
76
78
77
81
81
82
88
84
79
78
60
87
Remarks
+0.1 mg/l Ma Ico 610 to Train No. 2
+ 0'. 1 mg/l Nalco 610 to Train No. 2
+1.0 mg/l Nalco 677 to Train No. 2
+0.3 mg/l Nalco 677 to Train No. 2
+ 0.3 mg/l Nalco 677 to Train No. 2
o\
Conversion: I mgd = 3,785 m-Vday
-------
TABLE 33
MAIN PLANT TOTAL ORGANIC CARBON REMOVAL
Period
No.
Date
1 1/25/72-2/13/72
2 2/15
3 2/26
4 3/9
5 3/20
6 3/27
7 4/7
8 4/26
9 6/3
10 6/24
II 7/17
12 8/21
13 8/28
14 9/6
15 9/12
16 9/18
17 {9/25
19/30
18 10/4
-2/25
-3/8
-3/19
-3/26
-4/6
-4/20
-5/25
-6/19
-7/1!
-8/2
-8/27,
-8/30
-9/1 1
-9/17
-9/21
-9/28
-10/3
-10/5
Inf.
TOC
mg/l
151
147
154
162
220
195
250
162
193
151
198
193
166
174
180
182
192
149
F low
mgd
1.43
1.47
1.36
1. 19
1.51
1.51
1.54
1.40
0.73
0.45
0.90
1 .07
1.41
1. 18
1.35
1.47
1 .42
1.62
Train No.
Eff .
TOC
mg/l
22
28
25
22
36
37
32
17
15
12
13
19
19
21
23
23
22
20
2 (Alum)
Eff. Sol.
TOC
mg/l
15
20
17
20
25
26
21
12
15
9
12
15
12
14
16
13
13
12
% Ram' \
TOC
84
81
82
88
84
79
87
89
92
92
93
90
89
88
87
87
88
86
Flow
mgd
.43
.47
.36
.19
.51
.51
.54
1.40
1.46
1.78
1.35
1.07
1.41
1.18
1.35
1.47
1.42
1.62
Train No.
Eff.
TOC
mg/l
59
51
56
61
77
78
82
44
44
42
55
45
38
41
42
48
52
29
1 (No Alum)
Eff. Sol.
TOC
mg/l
38
36
22
31
36
44
37
19
31
21
23
29
24
23
22
22
22
22
t Rem' 1
TOC
58
65
62
66
65
58
67
72
62
72
73
76
77
76
76
73
73-
80
Remarks
+ O.I mg/l Nalco 610 to Train No. 2
+ O.I mg/l Nalco 610 to Train No. 2
+ 1.0 mg/l Nalco 677 to Train No. 2
t 0.3 tng/| Nalco 677 to Train No. 2
+0.3 mg/l Nalco 677 to Train No. 2
Conversion: I mgd = 3,785 mVday
-------
Period 3. The weight rates of Al(lll) application for the different time in-
tervals were maintained the same as in Period 2 with the exception of the low
flow interval from 0300 to 0700 hr. During this interval, both a I urn feed pumps
were shut off. It was hypothesized that elimination of a I urn dosage during the
hours of low flow would make little difference in terms of daily phosphorus
removal or overall plant performance. The data in Tables 30 to 33 indicate
this hypothesis was correct. However, during the early morning hours, a rather
high degree of turbidity was noted in the final settling tank. This dosage
method was, therefore, discontinued. Alum concentration dosage and Al:influent
TP (mole) during this period averaged 165 mg/l and 1.7, respectively. Flow to
the alum train averaged 5,148 m3/day (1.36 mgd). This lower flow may account
for relatively small changes in plant efficiency at the decreased alum dosage.
Period 4. The weight dosage of Al(lll) applied was maintained at 85 percent
of that applied in each time interval during Period I. Dosage between 0300
and 0700 hr was resumed. Flow entering the a I urn train during this period was
less than usual, 4,504 m-Vday (1.19 mgd), because of University holidays. Alum
concentration dosage averaged 192 mg/l and the mean Al:influent TP (mole) was
1.7. Average removals of phosphorus, 6005, TSS, and TOC were higher than in
any of the previous periods. This increase in efficiency can probably be attri-
buted to improved flocculation and solids capture in the final clarifier at the
reduced flow.
Period 5. During Period 5, the Al(lll) weight dosage during each time interval
was reduced to 80 percent of that applied in Period I. The flow to the a I urn
train during this period averaged 1.51 mgd, the highest to this point in the
experimental program. The high flow and a higher influent total phosphorus
concentration, along with the reduced alum dosage, resulted in a mean Al:
influent TP (mole) of 1.0. Removals of phosphorus, TSS, and TOC were reduced,
particularly phosphorus. It is interesting to note, however, that the soluble
phosphorus in the effluent was about the same as in earlier periods. This
seems to indicate that phosphorus was precipitated but that too little Al(lll)
remained for coagulation and formation of good settleable floe. A well-formed
floe is obviously necessary to entrain the precipitated aluminum phosphate and
other nonsettleable solids if the process is to function effectively.
Period 6. On the basis of the results of the first five periods, the Al(lll)
dosage pattern was set the same as during the first period, 93 kg/day (205 Ib/
day). The dosage program was also maintained on the same time schedule as in
Period I. Effluent phosphorus during this period averaged 2.2 mg/l, as com-
pared with 1.6 mg/l during the first period. On the other hand, the alum train
average influent was 5,715 m^/day (1.51 mgd) as compared with 5,413 mVday (1.43
mgd) during the first period. Although the mean Al:influent TP (mole) was 1.5,
it seems probable, as will be shown later, that the reduced removal of phosphorus
was primarily a result of the high flow through the final settling tank.
Period 7. During earlier periods, a higher than normal turbidity was observed
in the final effluent from 0900 to 1000 hr. It was thought that this was due
to insufficient alum dosage, relative to flow and phosphorus loading, as is
apparent in Figure 18. Therefore, the time when the highest a I urn dosage was
applied was advanced 30 minutes, i.e., the high and low pumps were operated
78
-------
from 0930 to 2330 hr. The total daily A I CM I) dosage was maintained at 93 kg/
day (.205 Ib/day).
8- During all earlier periods when the a I urn treatment process was func-
tioning efficiently, well-formed floe particles were observed escaping over the
effluent weirs of the final settling tank. It was thought that floe capture
might be improved if the escaping floe were entrained in a sludge blanket in
the settling tank. It was hypothesized that a deeper annular skirt in the cen-
ter of the settling tank might enhance the formation of a sludge blanket and
help prevent short circuiting. Accordingly, during the period from April 21
to April 24, 1972, the skirt was extended from 72 cm (30 in) to 198 cm (78 in)
below the settling tank water surface. After installation of the skirt exten-
sion, the AKIN) weight dosage program was continued as in Period 7. The
mean Al: influent TP (mole) during this period was 1.7, and alum train influent
flow was 5,299 m3/day (1.40 mgd). Effluent phosphorus averaged 1.5 mg/l and
removals of BOD5, TSS, and TOC were higher than during any previous period.
It is difficult to prove that the skirt extension enhanced the process, but it
certainly did no harm.
Period 9. Beginning with this period and continuing through Period II, lesser
fractions than 50 percent of the influent flow were diverted to the plant train
receiving a I urn treatment. The objective was to investigate the effect of final
settling tank hydraulic loading on process performance. AI(MI) weight dosages
were maintained at 67 percent of those utilized for the four time intervals in
Periods 7 and 8. During Period 9, 33 percent of the total plant influent flow
was directed to the a I urn dosed train. The daily dosage of AIM I I) was 62 kg
(137 Ib). The average flow to the alum train was 2,763 m^/day (0.73 mgd).
The Al: influent TP (mole) averaged 1.9, and a I urn concentration dosage was 250
mg/l. The alum dosage was high as the flow was lower than anticipated.
Effluent phosphorus averaged 0.8 mg/l and removals of BOD^ and TSS were the
highest experienced in the entire program. One might speculate that the excel-
lent removal was a result of a high alum dosage. Later analysis indicates that
final tank hydraulic loading was the most significant factor.
Period 10. During this period, only 20 percent, 1,703 m3/day (0.45 mgd), of
the total plant flow was directed to the a I urn treatment train, the lowest expe-
rienced during the entire program. The Al(lll) weight dosage was accordingly
reduced to 40 percent of that applied during Periods 7 and 8, maintaining the
same dosage time schedule. Effluent BOD5, TSS, and TOC were similar to those
in Period 9. Total phosphorus concentration in the effluent was higher than
expected (1.2 mg/l). Examination of the individual daily data does not reveal
any factor to account for this apparent anomaly.
Period II. Forty (40) percent of plant flow was directed to the alum train dur-
ing this period, and the Al(lll) weight dosage schedule was held at 80 percent
of the rates used in Periods 7 and 8. Alum train influent flow averaged 3,407
m-Vday (0.90 mgd). The mean Al: influent TP (mole) was 1.9, and the alum con-
centration dosage was 241 mg/l. Excellent removals of 6005, TSS, and TOC were
obtained. Before assuming that the high alum dosage accounted for the high
removals, the results during Period 4 should be reviewed. Effluent phosphorus
averaged 1.3 mg/l during Period 4 even though the a I urn concentration dosage was
79
-------
less than 200 mg/I. The influent flow during Period 4 was only 4,504 m /day
(1.19 mgd).
Period 12. During this period, the division of flow between the two trains of
the plant was equal, Flow averaged 4,050 m3/day (1,07 mgd) to each train. Al
(III) dosage was maintained at 93 kg/day (205 Ib/day). Effluent total phos-
phorus averaged 1.8 mg/l, higher than expected based on earlier results.
Period 13. Flow division was continued at 50 percent to each train. The dos-
age of AI(I I I) was continued at 93 kg/day (205 Ib/day), and the application
of Nalco 610 was initiated. Nalco 610 is a high molecular weight cat ionic poly-
electrolyte designed for use in sludge dewatering and wastewater clarification.
The polyelectrolyte was applied with two small positive displacement feed pumps
programmed to operate from the same timing system as the alum pumps. Poly-
electrolyte dosage was maintained at an average of 0.I mg/l, selected on the
basis of laboratory jar tests. The test run during this period was quite short
(3 days) and, although no improvement in effluent quality was apparent, the
results should not be considered conclusive.
Period 14. Difficulties with the preparation of polyelectrolyte solution re-
sulted in the temporary cessation of polyelectrolyte addition. While awaiting
the delivery of polyelectrolyte dispersion equipment, the dosage of Al(lll)
was continued at 93 kg/day (205 Ib/day). Phosphorus removal during this period
was disappointing.
Period 15. The addition of Nalco 610 polyelectrolyte was resumed during this
period. Al(lll) dosage was maintained at 93 kg/day (205 Ib/day)- Polyelectro-
lyte addition was again set at O.I mg/l. No significant improvement in perfor-
mance was observed.
Period 16. During this period, Nalco 677 was added at the rate of 1.0 mg/l.
Nalco 677 is a liquid anionic polyelectrolyte designed to enhance the removal
of suspended solids in municipal or industrial wastewater treatment applications.
The dosage of Al(lll) was increased to 102 kg/day (225 Ib/day), or 10 percent
more than the daily rate employed in Periods 12 through 15. No significant
improvement in performance was noted.
Period 17. Dosage of Nalco 677 was continued but was reduced to 0.3 mg/l follow-
ing consultation with a company representative. AIM I I) dosage was maintained
at 102 kg/day (225 Ib/day). Some difficulties were experienced with lumps of
undissolved polyelectrolyte clogging the feed pumps. In addition, it was found
that one check valve on an alum feed pump was sticking. Performance results
during this period were disappointing.
Period 18. Aid I I) dosage was reduced to 67 percent, 62 kg/day (137 Ib/day),
of that used in Periods 12 through 15, while Nalco 677 continued to be fed at
O.I mg/l. It was hoped that the polyelectrolyte would improve the settling
characteristics of the floe to the extent that substantially lower alum dosages
would be effective. Unfortunately, this test coincided with a period of high
plant flow. The resulting Al:influent TP (mole) ratio was less than 1.0 and
phosphorus removal was seriously reduced.
80
-------
DISCUSSION OF RESULTS
The overall Improvement in performance of the a I urn dosed train versus the un-
dosed train is obvious in Tables 30 to 33. Total phosphorus removal averaged
82 percent on the alum train and 18 percent on the undosed train during the
entire set of experimental periods. Other average treatment efficiencies were
as follows: BOD5 removal, 91 percent versus 77 percent; total suspended solids
removal, 88 percent versus 77 percent; total organic carbon removal, 87 percent
versus 69 percent.
Examination of average results indicates that the single most important variable
affecting the removal of phosphorus, BOD5, TSS, and TOC on the alum treatment
train was flow rate. This effect was visually evident during the entire experi-
mental program, i.e., when the flow through the final settling tank was less
than 4,542 nwday (1.2 mgd), the tank effluent was quite clear and only a few
small floe particles could be observed passing over the effluent weir. At
higher flow rates, larger floe particles could be seen and the effluent turbi-
dity was higher. The a I urn treatment process used at Chapel Hill functioned
well at alum dosages between 175 and 250 mg/l and an Al:influent TP (mole) be-
tween 1.4 and 2.2. Within these ranges, the final tank hydraulic loading ap-
pears to be the principal factor controlling process efficiency.
It is also apparent that when the Al:influent TP (mole) approaches 1.0, phos-
phorus is precipitated but not effectively settled, as illustrated by the results
during Period 5. During this period, effluent total phosphorus averaged 4.5
mg/l, the highest during the experimental program; however, effluent soluble
phosphorus averaged only 0.3 mg/l.
The effect of final settling tank hydraulic loading or surface overflow rate
on process efficiency can be seen in Figures 20 to 22. The data plotted on
these graphs do not include the results of Period 5 when the Al:influent TP
(mote) was 1.0, nor those of Periods 13 through 19 during which polyelectrolytes
were used in addition to alum. Plots of Al:influent TP (mole) versus mean re-
movals for the experimental period yielded no apparent trend. An upward trend
in removal of phosphorus and other constituents was noted with increasing a I urn
dosage, but this can largely be explained by the fact that at any particular
alum weight feed rate, the alum concentration was inversely proportional to flow.
It is interesting to note that when the flow to the final settling tank in which
alum was being applied was less than 50 percent of plant flow (Periods 9, 10,
and II), the percentage removal of total phosphorus, 8005, total organic carbon,
and total suspended solids averaged higher than during any other period.
It is also significant that the levels of soluble phosphorus, soluble BOD5, and
soluble organic carbon were substantially lower than the total values during all
experimental periods. The amount of these soluble constituents remaining after
treatment gives an indication of the effluent quality which might be obtained
with the use of effective fine solids removal facilities following secondary
clari f ication.
As mentioned earlier, settled sludge from the final clarifiers was returned to
the primary clarifiers where it resettled along with raw wastewater solids.
The return of alum sludge had a significant effect on the performance of the
81
-------
CO
M
FINAL CLARIFIER SURFACE OVERFLOW RATE (gpd/ft.2)
300 400 500 600 700 800 900 1000 MOO
100
g^
_l
tr
o.
90
80
70
0.4 0.6 0.8 1.0 1.2
FLOW (mgd)
1.4
1.6
2.4
2.0
1.5
1.0
1.8
m
c
m
H
TJ
«o
FIGURE 20. TOTAL PHOSPHORUS REMOVAL AS A FUNCTION OF FLOW AND FINAL CLARIFIER
OVERFLOW RATE FOR EXPERIMENTAL PERIODS I THROUGH 4 AND 6 THROUGH 12.
-------
100
Co
o
10
Q
O
CD
85
FINAL CLARIFIER SURFACE OVERFLOW RATE (gpd/ft.2)
300 400 500 600 700 800 900 1000 1100
0.4 0.6 0.8 1.0 1.2
FLOW (mgd)
1.4
1.6
20
15
m
~n
m
CD
O
o
10
1.8
FIGURE 21. BOD,- REMOVAL AS A FUNCTION OF FLOW AND FINAL CLARIFIER OVERFLOW
RATE FOR EXPERIMENTAL PERIODS I THROUGH 4 AND 6 THROUGH 12.
-------
FINAL CLARIFIER SURFACE OVERFLOW RATE (gpd/ft.2)
300 400 500 600 700 800 900 *IOOO 1100
-o
00
§
o
2
LU
o:
(/)
CO
m
~n
~n
r~
m
H
(/)
cn
I*
iQ
80
1.0
FLOW
1.8
FIGURE 22. TOTAL SUSPENDED SOLIDS REMOVAL AS A FUNCTION OF FLOW AND FINAL
CLARIFIER OVERFLOW RATE FOR EXPERIMENTAL PERIODS I THROUGH 4
AND 6 THROUGH 12.
-------
primary clarifier to which it was returned, Table 34 summarizes primary effluent
characteristics on the plant trains during the period from 1/27/72 to 8/30/72.
TABLE 34
QUALITY OF PRIMARY EFFLUENT FROM 1/27/72 to 8/30/72
Exp.
Period
1
2
3
4
5
6
7
8
9
10
1 1
12
13
Trai n
BOD5
mg/l
64
73
65
59
63
63
65
50
58
30
51
52
45
No. 1
TSS
mg/l
59
72
59
53
70
88
77
93
76
27
43
50
62
(Alum)
TP
mg/l
5.3
7. 1
6.9
5.8
7.5
8.4
7.0
6. 1
5.3
4.7
5.4
6.8
7.9
Train
BOD 5
mg/l
76
82
78
82
70
78
81
64
93
64
93
80
65
No. 2
TSS
mg/l
71
80
71
84
95
123
103
65
49
68
79
64
92
(No Alum)
TP
mg/l
8.3
8.5
9.3
10.3
1 1.0
1 1.5
10.7
9.1
10.9
9. 1
10.8
10.4
9.9
As Table 34 indicates, the removal of BOD5 and total suspended solids was signi-
ficantly higher in the primary clarifier to which the a I urn sludge was returned.
The reduction in influent total phosphorus was even greater. Influent total
phosphorus was reduced from an average concentration of 11.6 mg/l to 10.0 mg/l,
85
-------
or 14 percent, in the primary clarifier which received no alum sludge. The pri-
mary clarifier receiving return a I urn sludge reduced the influent total phosphorus
to an average concentration of 6.5 mg/l (a 44 percent reduction). Obviously the
return of alum sludge enhanced primary clarifier settling efficiency. The im-
proved phosphorus removal was probably due to a combination of enhanced settling
efficiency and adsorption of phosphorus on the floe structure.
Although it was possible to control the fraction of plant influent entering
either side of the plant, the actual amount of flow entering e'ither train on any
particular day or at any particular time could not be controlled. Both flow
and influent phosphorus concentration varied from day to day. No means were
available, however, to automatically vary alum feed rate with flow or influent
phosphorus concentration. Consequently, the daily dosage of a I urn in mg/l and
the daily Al:influent TP (mole) varied during any given experimental period.
Because of these variations, a number of statistical procedures were used treat-
ing each day's data as a separate experiment in an attempt to develop some mean-
ingful correlations.
Data Sorting Procedures
A computerized data sorting method was the first technique used in analyses of
daily data. Because flow appeared to be the most important single factor affect-
ing performance, flow was included in each sorting. Five flow intervals were
used:
1,438-2,971 m3/day (0.38-0.78 mgd)
2,972-4,334 m3/day (0.79-I.14 mgd)
4,335-5,129 m^/day (I.15-1.35 mgd)
5,130-5,886 rrrVday (I.36-I.55 mgd)
5,887-7,002 m3/day (I.56-1.85 mgd)
In the first sorting, the data were separated into three levels of total phos-
phorus removal: 70-78 percent, 79-87 percent, and 88-96 percent. The printout
listed alum dosage in mg/l, Al:influent TP (mole), influent total phosphorus
concentration, total suspended solids in final settling tank influent, flow,
percent removal of total suspended solids, and percent removal of total phos-
phorus.
The pertinent questions and observations derived from the first sorting
follow:
Is the percent removal of total phosphorus or total suspended solids in-
fluenced by flow?
86
-------
A: Yes, removal decreased as flow increased.
Q: Is the percent removal of total phosphorus or total suspended solids influ-
enced by alum concentration dosage or Al:influent TP (mole)?
A: Increased flow resulted in decreased alum concentration dosage and Al:influ-
ent TP (mole). The data did, however, include some days when the a I urn con-
centration dosage and Al:influent TP (mole) were high even though the flow
was high. On such days, relatively poor removals were obtained, indicating
that the important variable was flow rather than a I urn concentration dosage
or Al:influent TP (mole).
Q: Does the concentration of total suspended solids in the influent to the
final settling tank (trickling filter effluent) have an effect on removal
of total phosphorus or total suspended solids? (The basis for this question
is that a high concentration of colloidal matter in settling tank influent
may require more alum for destabi I ization, thereby affecting precipitation
of phosphorus or removal of suspended solids).
A: The amount of total suspended solids in final settling tank influent seemed
to have little effect on the removal of either total suspended solids or
total phosphorus.
In the second sorting, data were separated into the previously listed flow ranges
and also into four alum concentration dosage categories: 130-165 mg/l, 166-210
mg/l, 211-260 mg/l, and 261-310 mg/l. The printout listed the same variables
as the first sorting. The pertinent question and observation derived from the
second sorting is as follows:
Q: Considering the two variables flow and alum concentration dosage, which one
produces the most significant effect on removal of total phosphorus and
total suspended solids?
A: Decreased flows and increased alum concentration dosages appeared to result
in increased removal of total phosphorus. In general, as flow increased,
a I urn concentration dosage decreased. On the basis of this sorting, it was
impossible to determine the separate effects of a I urn concentration dosage
and flow.
The third sorting listed data by flow categories and also by AI:infIuent TP
(mole) ranges as follow: 1.2-1.7, 1.8-2.4, and 2.5-3.1. The pertinent question
and observation emanating from the third sorting are given below:
Q: Is the removal of total phosphorus and total suspended solids affected more
by flow or by Al:influent TP (mole)?
A: Increasing the Al:influent TP (mole) above 1.4 results in little if any
observable improvement in total phosphorus and total suspended solids re-
movaI.
The fourth sorting simply listed flow categories along with percent total phos-
phorus removal and the pH of the final effluent. The pertinent questions and
observations were as follows:
87
-------
Q: Is there a relationship between flow and total phosphorus removal?
A: Yes, as shown in Figure 23.
Q: Is pH affected by flow?
A: Yes, low flows are normally coincident with high alum concentration dosages
and consequently with low pH values.
Scattergrams of the data used in the sorting analysis are shown in Figures 24
to 26. The general trend of increasing total phosphorus removal with increas-
ing alum concentration dosage and decreasing flow is apparent. The lack of
any trend in total phosphorus removal with the Al:influent TP (mole) can also
be seen. Because the amount of alum fed each day during any experimental period
was fixed, the a I urn concentration depended on flow, decreasing as flow increased.
Statistically, therefore, it is difficult to independently determine the effect
of flow and a I urn concentration dosage.
A plot of percent total phosphorus removal versus a I urn concentration dosage is
shown in Figure 27. The odd point on this chart is the average of phosphorus
removals and a I urn concentration dosage on II days when both flow and a I urn con-
centration were high: flow > 5,299 rrvVday (1.40 mgd) and a I urn dosage > 190 mg/l.
This point does not correspond with the curve in Figure 27. It does, however,
correspond well with the curve in Figure 23. This indicates that above some
minimum alum concentration dosage, flow is the more significant factor affect-
ing total phosphorus removal. If sufficient project funds had been available
to install an alum feeding system automatically paced by plant flow, the alum
feed concentration could have been maintained at any desired level and more
definitive results might have been obtained relative to the effect of this
variable.
Regression Analysis
Attempts were made to fit various types of regression equations to the daily
plant data. In one set, the dependent variable was percent total phosphorus
remaining in the plant effluent. Independent variables were various combina-
tions of flow, alum concentration dosage, Al:influent TP (mole), influent total
phosphorus concentration, and trickling filter effluent total phosphorus con-
centration. In another set, the dependent variable was plant effluent total
phosphorus concentration. Both linear and log-linear equations were fitted to
the data, but in all cases, the multiple R^ value was less than 0.35. Because
of the very low correlations obtained, the various regression equations are not
reported here. It is believed the performance data for the 18 experimental
periods are summarized in Tables 30 to 33, along with the general results of
the sorting analyses, present the best view of the alum treatment process as
operated at Chapel Hill.
Sludge Production
The incremental sludge production which results from the addition of alum or
any other coagulant to a wastewater treatment plant is an essential item of data
to both the designer and the plant operator. The effect of the combined organic-
88
-------
00
95
90
§
o
CL
h-
85
80
75
FINAL CLARIFIER SURFACE OVERFLOW RATE (gpd/ft.2)
100 200 300 400 500 600 700 800 900 1000
I
0.5
1.0
FLOW (mgd)
T
T
1.5
FIGURE 23. EFFECT OF FLOW ON TOTAL PHOSPHORUS REMOVAL FROM SORTING ANALYSIS.
-------
O
100
90
80
70
UJ
or
60
50
40
30
20
0 o
opooo c
O OOOOQO O
O _n OOOD O
gCP o ooo
o omoo oo o
fcoooDCCD
000 0
O
o
o
o o
oo
o o
o
o
O
o
O
o
o
100 150 200
250 300 350 400 450 500 550
ALUM DOSAGE (mg/l)
FIGURE 24. SCATTERGRAM FOR ALUM CONCENTRATION DOSAGE VERSUS PERCENT TOTAL
PHOSPHORUS REMOVAL.
-------
100
90|-o
o o
OO
o o
80
70
o 60
LU
50
40
30
20
o « o
80°
°2>o°o Q
o~ o> & o
o o _ o
o 08 oo o
00 ° q>
o o ^
o o
00 O
o o o
o oo
_ o
O O
o o o
I
I
1.0
2'° MOLE RATIO 3'°
4.0
FIGURE 25. SCATTERGRAM FOR AL:INFLUENT TP (MOLE) VERSUS PERCENT TOTAL
PHOSPHORUS REMOVAL.
-------
ho
100
90
80
70
o 60
Ld
o:
Q_
I- 50
40
30
20
I
o
CO
•
o o
o
o
o
o
I I I I I
o
Q> o
X) O O
COO COD O O
00 O 00 CO O O
O O CO O O O O
O O OCD
I
8
o
o
o o
oo o
o
o
o
OCD CO
o o o
OJSD coo o Q ^
oo
o
o
CD
00 O
CD
CO
I
I
0.0 .20 .40 .60
.80 1.00 1.20
FLOW (mgd)
I
1.40 1.60 1.80 2.00
FIGURE 26. SCATTERGRAM FOR FLOW VERSUS PERCENT TOTAL PHOSPHORUS REMOVAL.
-------
95
90
§
o
LJ
Q_
I-
85
80
75
n f Dosage > 190 mg/1
I Flow> 5,299 m3/day (1.40 mgd)
125
150
175 200 225
ALUM DOSAGE (mg/l)
250
275
FIGURE 27. EFFECT OF ALUM CONCENTRATION DOSAGE ON TOTAL PHOSPHORUS REMOVAL FROM
SORTING ANALYSIS.
-------
chemical sludges on solids treatment and disposal systems is another import-
ant consideration.
At the Chapel Hill Plant, the secondary sludge settling in the final clarifi-
ers is returned to the primary clarifiers where it resettles and combines with
the raw sludge. Separate secondary sludge return systems are used on each
train, i.e., there is no mixing of the sludges from the two trains until they
combine in the anaerobic digester. The same positive displacement pump is used
to remove sludge from the two primary clarifiers, but at different times per-
mitting the individual sludges to be sampled at the pump discharge.
Sludge was pumped to the digester from each primary clarifier twice a day,
at 7 a.m. and 5 p.m. During the early periods of the investigation, three
samples were collected during each pumping period to determine the rate of
decrease in sludge solids concentration with pumping time. On the average,
the solids concentration of the sludge decreased linearly with time during
pumping. Sludge samples were subsequently taken at the middle of each pump-
ing period during the remainder of the investigation.
Soon after the introduction of alum to the No. 2 train of the plant, the sludge
from No. 2 primary settling tank was observed to be of heavier consistency at
the end of a normal period of pumping. Therefore, the pumping time from the
No. 2 train was increased until the sludge being pumped at the end of the pump-
ing period was of approximately the same consistency as that observed at the
end of the pumping from the No. I train. The sludge volumes actually pumped
from the two plant trains during each experimental period of this investiga-
tion, along with other pertinent sludge characteristics, are shown in Tables
35 and 36.
The data presented in Tables 35 and 36 indicate a decrease in sludge solids
concentration on both trains beginning in Period 12. Various factors which
might explain this decrease, e.g., sampling methods, analytical procedures,
etc., have been investigated. However, no reasonable explanation for this
apparent decrease has been found. The calculated grams (Ib) of total solids
or grams Clb) of volatile solids per million gallons of wastewater treated
averaged about the same for both trains during the last seven weeks of the
37-week experimental program. As a I urn dosages and general levels of phos-
phorus and suspended solids removals on the No. 2 train did not decrease
significantly during the last seven weeks, the sludge solids data for this
period are suspect. Because of this, the summary of information presented in
Table 37 does not include sludge solids data from the last seven weeks of the
experimental program.
Table 37 indicates that on a unit treated flow basis, the total volume of
sludge removed from the alum dosed train was approximately 32 percent greater,
the pounds of total solids removed were about 28 percent greater, and the
pounds of volatile solids removal were about 13 percent greater than from the
undosed train. The decrease in percent volatile solids- in the sludge from the
alum dosed train was to be expected as the precipitated chemical sludge con-
tributed non-volatile material.
94
-------
TABLE 35
VOLUMES AND CHARACTERISTICS OF SLUDGES FROM TRAIN NO. 2 (ALUM)
Period
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
15
16
17 {
18
Dates
1/25-2/13/72
2/15-2/25
2/26-3/8
3/9 -3/19
3/20-3/26
3/27-4/6
4/7 -4/20
4/26-5/25
6/3 -6/19
6/24-7/1 1
7/17-8/2
8/21-8/27
8/28-8/30
9/6 -9/1 1
9/12-9/17
9/18-9/21
9/25-9/28
9/30-10/3
10/4-10/5
Total
Sol ids
%
4,4
3.1
4.0
4.0
4.8
3.9
3.8
3.8
2.7
2.5
2.3
2.6
2.6
2.2
2.3
2.3
-2.6
3.5
Volati le
Fraction
%
66
72
68
67
66
65
67
66
62
65
65
67
63
65
66
59
64
65
Vo 1 ume
Pumped
gpd
8,250
8,950
8,708
8,500
8,721
8,400
8,210
9,1 13
8,014
5,243
6,955
6,943
8,250
7,516
8,250
8,250
8,250
8,800
Vb 1 ume
Pumped
gal
mil ga 1
5,770
6,100
6,300
7,150
5,770
5,570
5,330
6,510
1 1,000*
1 1,630*
7,750
6,500
5,850
6,380
6,1 10
5,610
5,800
5,430
Sol ids
Pumped
Ib TS
mil ga 1
2, 120
1,580
2,100
2,380
2,300
1,810
1,690
2,060
2,470*
2,420*
1,490
1,410
1,260
1, 170
1,170
1,080
1,260
1,590
Sol ids
Pumped
Ib VS
mil ga 1
1,400
1, 140
1,430
1,590
1,520
1,170
1, 130
1,360
1,530*
1,500*
970
940
790
760
770
640
810
1,030
*Excessive pumping - not included in average amounts shown in Table 37.
Conversions; I gpd = 3.785 £/day; I gal/mi I gal = 0.001 &/m3; I Ib/mil gal
0.12 g/m3
95
-------
TABLE 36
VOLUMES AND CHARACTERISTICS OF SLUDGES FROM TRAIN NO. I (NO ALUM)
Period
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
15
16
17 {
18
Dates
1/25-2/13/72
2/15-2/25
2/26-3/8
3/9 -3/19
3/20-3/26
3/27-4/6
4/7 -4/20
4/26-5/25
6/3 -6/19
6/24-7 /I 1
7/17-8/2
8/21-8/27
8/28-8/30
9/6 -9/1 1
9/12-9/17
9/18-9/21
9/25-9/28
9/30-10/3
10/4-10/5
Total
Sol ids
%
4.3
2.7
4.4
4.0
3.6
4.2
4.8
3.8
4. 1
3. 1
3.3
2.7
2.7
2.6
2.9
2.5
3. 1
2.8
Volati le
Fraction
%
69
85
77
79
78
78
78
77
71
78
66
75
74
71
67
73
66
70
Vol ume
Pumped
gpd
7,150
6,250
6,416
6,500
6,521
6,600
6,600
6,830
6,521
7,761
6,988
6,1 18
6,600
6,966
6,966
6,600
6,700
6,700
Vol ume
Pumped
gal
mil ga 1
5,000
4,250
4,730
5,450
4,310
4,370
4,280
4,880
4,460
4,320
5, 170
5,760
4,680
5,900
5, 160
4,500
4,720
4,130
Sol ids
Pumped
Ib TS
mil ga 1
1,800
960
1,740
1,820
1,300
1,530
1,720
1,540
1,530
1, 120
1,430
1,300
1,050
1,280
1,250
940
1,220
970
Sol ids
Pumped
Ib VS
mil gal
1,240
815
1,340
1,440
1,010
1,190
1,340
1,190
1,090
875
950
970
775
910
840
690
805
680
Convers ions:
I gpd = 3.785 £/day, I gal/mil gal = 0.001
I Ib/mi I gal = 0.12 g/rrH
96
-------
TABLE 37
SLUDGE PRODUCTION SUMMARY FOR 1/25/72 TO 8/27/72
Vo 1 ume
Total
Sol ids
Volati
Sol ids
Pumped, ga I/mi 1 gal
Sol ids, %
Pumped, 1 b TS/mi 1 gal
le Fraction, %
Pumped, 1 b VS/mi 1 ga 1
Train No. 2
(Alum)
6,275
3.57
1,894
67
1,265
Trai n
(No A
4,748
3
1 ,483
76
1,120
No. 1
1 urn)
.85
Conversions:
I gal/mi I gaI = 0.001
I Ib/miI gal - 0.12 g/m3
SIudge Digestion
As described, the Chapel Hill plant employs anaerobic digestion as the first
step in sludge disposal. Sludge is treated in a 22.9 m (75 ft) diameter pri-
mary digester with a 6.I m (20 ft) water depth. Digested sludge is thickened
and supernatant is separated in a 15.2 m (50 ft) diameter secondary digester.
The primary digester is heated and maintained at about 95 °F and is equipped
with a centrally located draft tube mixer which is seldom used. Mixing is
normally accomplished with a 18.6 kW (25 hp) centrifugal pump which takes
suction from the bottom of the digester and discharges into the center of the
tank. This pump is operated for about one hour each day. The secondary di-
gester is neither heated nor mixed.
The detention time in the primary digester, based on normal plant operation
without alum addition, is about 48 days at a volumetric loading of 4.6 £ of
sludge per m3 (4600 gal sludge per mil gal) of wastewater treated. The normal
volatile solids loading is about 0.62 kg/day/m3 (0.039 Ib/day/ft3) of primary
digester. Because of the relatively long digester detention time and the low
volatile solids loading, no operating difficulties were anticipated with the
digestion process during the alum addition program.
A summary
treatment
of conditions
investigation
in the primary digester during the course of the alum
is given in Table 38.
97
-------
TABLE 38
CONDITIONS IN PRIMARY DIGESTER DURING ALUM TREATMENT INVESTIGATION
Period
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
15
16
17
18
Volati le Acids
mg/l
230
101
1 10
103
94
94
101
82
94
936
132
141
158
51
73
123
292
384
Al ka 1 in ity pH
mg/l as CaC03
2,069
2,516
2,736
2,817
2,850
2,873
3,043
2,779
2,166
1,857
2,350
1,588
1,531
1,641
1,806
1,449
1,517
1,384
7.0
7. 1
7. 1
7.1
7.3
6.9*
7.1
7.1
7.0
6.8**
7.1
6.9
6.9
6.9
6.9
6.9
6,8
6.8
* Sludge feed transferred to secondary digester.
** Lime added to primary digester.
98
-------
During the first two months of operation with a I urn addition, no difficulties
were experienced with the primary digester. All parameters remained within a
range typical for digestion without alum addition, Late in March 1972 (Peri-
od 6), the pH dropped from above 7 to 6.9, and the methane content of the di-
gester gas decreased from a normal level of about 68 percent to slightly less
than 60 percent. It was suspected this might be due to inadequate mixing.
Feed sludge was diverted to the 15.2 m (.50 ft) secondary digester for several
days, and the primary digester was given extra mixing. The primary digester
recovered rapidly, and normal operation was resumed.
A similar condition developed in July 1972 (Period 10). In this instance,
diversion of feed sludge to the secondary digester did not result in recovery
of the primary digester. It was necessary to add 680 kg (1500 Ib) of hydrated
lime to the primary digester to raise the pH. As before, the methane content
of the digester gas had decreased to about 60 percent and, prior to the addi-
tion of lime, the volatile acids increased from a normal level of 100 mg/I to
over 900 mg/l. The addition of lime corrected the low pH, and the volatile
acids concentration quickly returned to a typical low value. During the
final two and one-half months of the alum treatment investigation (Periods II
through 18), the alkalinity in the primary digester was about 1,000 mg/l less
than normal (1,500 versus 2,500 mg/l), the pH tended to drift downward (7.1
to 6.8), volatile acids were gradually increasing (132 to 384 mg/l), and the
methane content of the gas was less than usual.
The decreased detention time and the increased volatile solids loading might
be suspected as causes of the difficulties with the primary digester. How-
ever, even with the additional sludge resulting from a I urn addition to one
train, the detention time in the primary digester was decreased by only eight
days to 40 days and the average volatile solids loading was increased by only
0.10 kg/day/m3 (0.006 Ib/day/ft3) to 0.72 kg/day/m3 (0.045 Ib/day/ft3). Both
factors were still considered conservative for normal operation. Sulfide
toxicity (via suIfate reduction) was also suspected, but the addition of 200
mg/l of alum to one plant train would result in a digester sulfide concentra-
tion of only about 16 mg/l, well below the concentration typically considered
toxic. Whatever the cause of the problem with the primary digester, recovery
was rapid after a I urn treatment was stopped in October 1972. Within two weeks,
the pH increased to a value above 7, the alkalinity was back up above 2,000
mg/l, and volatile acids decreased to about 100 mg/l. On the basis of these
results, it has been decided to provide convenient facilities for the addition
of lime to the primary digester when permanent alum treatment facilities for
dosing to both trains are constructed at the Chapel Hill plant.
In spite of the difficulty in maintaining a normal pH in the primary digester,
the digestion process itself continued to produce a normal reduction in vola-
tile solids throughout the entire a I urn treatment study. The volatile content
of the mixture of digester influent sludges from the two plant trains averaged
70 percent. The average volatile fraction of centrifuged sludge was 50 per-
cent. Assuming no change in fixed solids during digestion, these results in-
dicate an average of 57 percent reduction in volatile solids in the digestion
process. In fact, the volatile solids fraction in the centrifuged sludge
averaged 47 percent during the final seven weeks of the investigation while
the primary digester pH and alkalinity were trending down and the volatile
acids concentration was rising.
99
-------
Another problem was encountered severaI weeks after the start of a I urn treat-
ment. Supernatant is decanted from the secondary digester continuously dur-
ing the night hours. The total suspended solids concentration of the super-
natant is generally less than 1,000 mg/I, About the end of February 1972
(Period 3), supernatant solids increased rapidly and remained high (8,000 to
20,000 mg/l) during the remainder of the experimental program. At the same
time, the percent solids in the centrifuge feed, which is drawn from the bottom
of the secondary digester, decreased from a normal range of 6 to 7 percent to
an average of 3.8 percent. The sludge in the secondary digester did not
thicken in a normal manner, nor was it possible to obtain a reasonable quality
supernatant after the first month of alum treatment. The problem with heavy
supernatant was particularly serious as the supernatant return line tended to
clog quite rapidly after backfIushing. On several occasions, it was necessary
to discharge heavy supernatant to the undosed No, I train through a 15 cm (6
in) line at high rates, causing temporary upsets in the operation and perfor-
mance of that train.
Because of the problems encountered in handling the mixture of alum humus and
conventional primary sludges in the primary and secondary digesters, it has
been recommended that a separate system for stabilization and disposal of alum
humus sludge be included in any permanent alum treatment installation at Chapel
Hi I I.
Sludge Dewatering
Two methods of sludge dewatering are available at the Chapel Hill plant. The
method utilized routinely is the plant's solid-bowl centrifuge. The centri-
fuge is backstopped by 18 uncovered sand drying beds.
No difficulty was encountered in dewatering digested sludge with the centri-
fuge during the a I urn treatment program. It was, however, necessary to increase
the operating time of the centrifuge because of the reduced solids concentra-
tion in the secondary digester underflow (the centrifuge feed). During normal
plant operation without a I urn addition, the feed to the centrifuge averages be-
tween 6 and 7 percent solids and the digested solids can be dewatered with
approximately 23 hours of centrifuge operation each week. Due to the increased
volume of solids fed to the digesters and poor thickening in the secondary
digester during the alum addition experiments, the volume of wet digested
sludge to be dewatered increased about 70 percent, necessitating longer periods
of time for centrifuge operation. Prolonged periods of centrifuge operation
were not always possible as only one sludge hauling truck was available and
the plant is not manned around the clock. It was frequently necessary 'to use
the sand drying beds to accommodate the extra volume of sludge. >
If alum had been added to both trains of the Chapel Hill' Plant, the volume of
digested sludge to be dewatered would have increased to approximately 2.4 t
the conventional operation norm. Dewatering this volume of sludge would be
impossible at the Chapel Hill plant without around the clock operation or the
installation of additional dewatering equipment.
100
-------
No problems were encountered in drying the digested sludge on the sand beds.
The beds were filled to a depth, of about 3Q cm (12 in) and dried to a fork-
able consistency .in periods ranging from three to six weeks during the late
spring, summer, and early- fall. The sludge dried with a whitish-gray surface
crust which was apparently due to the various compounds of aluminum and phos~
phorus.
Chemical Costs
content, by weight, of about 4.4 percent was used
The chemical cost was $54.80 per equivalent ton of
Liquid alum with an Aid I I)
during this investigation.
dry filter a I urn [A^CSO^ • 14.3 H2CfL On the basis of a dosage of 175 mg/l
of dry alum, the chemical cost per million gallons of wastewater treated was
$41. Capital costs for permanent chemical storage and feeding equipment
along with the requirement for additional maintenance and operator attention
will add to the total cost. Additional sludge handling facilities, if needed,
will also add to the total cost.
IQI
-------
SECTION VI
REFERENCES
I. Little, L. W., Phosphorus in Water and Wastewater: An Annotated Selected
Bibliography. UNC Wastewater Research Center Report No. II, Depart-
ment of Environmental Sciences and Engineering, UNC-Chapel Hill (1970).
2. Hanson, R. L., Walker, W. C,, and Brown, J, C., Variations in Character-
istics of Wastewater Influent at the Mason Farm Wastewater Treatment
Plant, Chapel Hill^ North Carolina. UNC Wastewater Research Center
Report No. 13, Department of Environmental Sciences and Engineering,
UNC- Chapel Hi I I (1970).
3. Process Design Manual for Phosphorus Removal. U. S. EPA Office of Tech-
nology Transfer, Washington, D. C. (October 1971).
4. Van Wazer, J. R., Phosphorus and Its Compounds, Vol. 1: Chemistry. Inter-
science Publishers/ Inc., New York, N. Y. (1958),
5. Recht, H. L., and Ghassemi, M., Kinetics and Mechanism of Precipitation
and Nature of the Precipitate Obtained in Phosphate Removal from Waste-
water Using Aluminum (III) and Iron (III) Salts. Water Pollution
Research Series No. 17010 EKI 04/70. U. S. Dept, of the Interior,
Federal Water Quality Administration (1970).
6. Jenkins, D., Ferguson, J. F., and Menar, A. B., "Chemical Processes for
Phosphate Removal." Water Research, 5, 369-389 (1971).
7. Bell, 6. R., Libby, D. V., and Lordi, D. T., Phosphorus Removal Using
Chemical Coagulation and a. Continuous Counter current Filtration Process,
Water Pollution Research Series 17010 EDO 06/70, U. S. Dept. of the
Interior, Federal Water Quality Administration (1970).
8. Yuan, W. L., and Hsu, P. H., "Effect of Foreign Components on the Preci-
pitation of Phosphate by Aluminum," Presented at the 5th International
Water Pollution Research Conference, San Francisco, Calif. (1970).
9. Wuhrmann, K., "Objectives, Technology, and Results of Nitrogen and Phos-
phorus Removal Processes," pp. 21-48. J[n_ E. F. Gloyna and W, W. Ecken-
felder, Jr. (ed.), Advances in Water Quality Improvement, University
of Texas Press, Austin (1968).
10. Eckenfelder, W. W., "Development of Tertiary Treatment Methods for Waste
Water Renovation." Water Pollution Control 1969, 584-591 (1969),
102
-------
II. Foree, E. G., Jewell, W, J., and McCarty, P. L., "The Extent of Nitrogen
and Phosphorus Regeneration from Decomposing Algae." Presented at
the 5th International Water Pollution Research Conference, San
Francisco, Cal if. (1970),
12, Abbott, W., "Nutrient Studies in HyperfertiI ized Estuarine Ecosystems
I. Phosphorus Studies," pp. 729-739. In S, H. Jenkins (ed.) Advances
in Water Pollution Research, Pergamon P?ess, Oxford (1969).
13. Grill, E. V., and Richards, F. A., "Nutrient Regeneration from Phytoplank-
ton Decomposing in Seawater." Journal of Marine Research, 22, 51-69
(1964).
14. Johannes, R. E., "Nutrient Regeneration in Lakes and Oceans," pp. 203-
213. In M, R. Droop and E, J. F. Wood (eds.), Advances in Microbiology
of th~Sea, Academic Press, New York (1968).
15. Kerr, P. C,, Paris, D. F., and Brockway, D. L., The Interrelation of
Carbon and Phosphorus in Regulating Eeterotrophic and Autotrophic
Populations in an Aquatic Ecosystem, U. S. Department of Interior,
FWQA, Southeast Water Laboratory, National Pollutants Fate- Research
Program, Athens, Ga., 53 pp. (1970).
16. Golterman, H. L., "Mineralization of Algae Under Sterile Conditions or by
Bacterial Breakdown." Verhandlungen, International Vereinigung fur
Theoretische und Angewandte Lirmologie, 15., 544-548 (1964).
17. Kuenzler, E. J., "Dissolved Organic Phosphorus Excretion by Marine Phyto-
plankton." Journal of Phycology, 6, 7-13 (1970).
18. Laughlin, J. E., "Modification of a Trickling Filter Plant to Allow
Chemical Precipitation," Advanced Waste Treatment and Water Reuse
Symposium, Dallas, Texas (January 1971).
19. Brown, J. C., Little, L. W., Francisco, D. E., and Lamb, J. C., Methods
for Improvement of Trickling Filter Plant Performance. Part 1,
Mechanical and Biological Optima. U. S. Environmental Protection
Agency, EPA-670/2-73-047a (August 1973).
103
-------
APPENDIX A
ABSTRACT OF PUBLICATION RESULTING
FROM PROJECT
PHOSPHORUS IN WATER AND WASTEWATER - AN ANNOTATED SELECTED BIBLIOGRAPHY.
University of North Carolina, Chapel Hill. Wastewater Research Center.
Linda W. Little. Wastewater Research Center Report No, II, November, 1970.
118 pp. FWQA Contract No. 14-15-505.
This volume comprises a selected annotated bibliography pertinent to sources
of phosphorus in water and wastewater, effects of phosphorus on aquatic sys-
tems, behavior of phosphorus in soils and waters, phosphorus analysis, and
removal of phosphorus from wastewater. Contains 281 entries.
104
-------
APPENDIX B
ABSTRACT OF THESES RESULTING
FROM PROJECT
ROBERT L. HANSON. Variations in Characteristics of Wastewater Influent at
the Mason Farm Wastewater Treatment Plant, Chapel Hill, North
Carolina. (Under the direction of JAMES C. LAMB III). Master's
thesis submitted to the University of North Carolina in partial
fulfillment of the requirement for the Master of Science in Sanitary
Engineering, 1970.
Twelve composited samples of domestic Wastewater influent were collected over
two-hour intervals on each of seven different days, Sunday through Saturday,
in July 1969 so diurnal variations in flow and constituent concentrations and
loadings could be observed. The samples were analyzed for BOD, COD, TOC,
nitrogen, phosphorus, MBAS, and specific solids and metal constituents. In-
fluent flow was found to vary from 39 to 144 percent of average with the
maximum flow between 1000-1200 hours and the minimum flow between 0400-0600
hours. The wastewater constituents showed a wide range of concentrations and
loadings. Generally, the maximum concentrations and loadings occurred between
1000-1400 hours and the minimum values between 0600-0800 hours. The ratio
of maximum to minimum concentrations for the constituents varied from 4-12 to
one, while the same ratio for loadings varied from 10-40 to one.
GEORGE C. BUDD, JR. Laboratory Studies of Phosphate Removal by Addition of
Lime to Wastewater. (Under the direction of JAMES C. LAMB III). Master's
thesis submitted to the University of North Carolina in partial fulfill-
ment of the requirement for the Master of Science in Environmental
Engineering, I 97 I.
A study of phenomena associated with lime precipitation of phosphate in domes-
tic wastewater, with observations based on the results of "jar tests" per-
formed at the Chapel Hill Wastewater Treatment Plant. Effects of filtration,
coagulation aid addition, fluoride addition, and hydroxyapatite addition were
determined. On the basis of the results, it was concluded that additional
phosphate reduction could be effected by filtration and coagulant aid addition.
Neither fluoride addition nor hydroxyapatite addition had significant effects
on phosphate removal.
105
-------
THOMAS BATES. Phosphorus Removal from Trickling Filter Effluent Using Ferric
Sulfate and Lime Precipitation, (Under the direction of LINDA W.
LITTLE). Master Is thesis submitted to the University of North. Carolina
in partial fulfillment of the requirement for the Master of Science in
Public Health, May, 1973,
At the laboratory scale, phosphorus removal from trickling filter effluent was
achieved by ferric sulfate-lime treatment. Four series of experiments were
conducted to investigate the following parameters:
I. Comparison of the effectiveness of lime CCa(OH)2Il with sodium hydroxide
(NaOH) for pH control when used with iron (III) for phosphorus removal
and control of residual iron.
2. Determination of the optimum pH for removal of phosphorus and elimination
of effluent iron.
3, Effect, if any, of the order of addition of iron (III) and lime.
4. Comparison of the effectiveness of alum precipitation with iron-lime pre-
c i pitation.
Based on the results obtained, the following conclusions were reached:
I. Lime proved more effective than sodium hydroxide when used with iron (III)
for phosphorus removal and minimizing iron residual in the supernatant.
2. Optimum pH for iron-lime precipitation is approximately 7.5; iron (III)
leakage can be minimized by pH control to above pH 7-
3. Iron (111) added prior to lime addition produces a floe that is more com-
pact and settles more rapidly.
4. Using pH control with lime, iron (III) compares favorably with aluminum
(III) for phosphorus removal.
MARTIN STRAUSS. Effect of Colloidal Surface Area on the Removal of Phosphorus
by Aluminum. (Under the direction of JAMES C. BROWN). Master's thesis
submitted to the University of North Carolina in partial fulfillment of
the requirement for the Master of Science in Environmental Engineering,
1973.
Effluents from trickling filters bear a substantial amount of colloidal parti-
culates. AluminumdI I) added to such effluents for the purpose of phosphorus
precipitation also coagulates colloids. Coagulation proceeds primarily as chemi-
cal sorption of a aluminum species onto colloidal surfaces. Colloidal surface
area concentration is used as a parameter of a IuminumU I I) demand and its inter-
ference with phosphate removal is investigated in artificial systems containing
phosphate and silica colloids.
Experiments showed that the range of particulate surface area in a tricklinq
filter effluent is 50-100 m2/|. RemOvaI of 10 mg/l of phosphate as P is dras-
tically ,mpa,red ,f colloldal si I lea Is added to the phosphaVe solution at
representative sur ace area concentrations. A Iinear relationship exists be-
tween alummum(lll) dosage required to remove phosphorus and initia? coMoida I
106
-------
surface area concentration. The effect of colloids on required aluminumUII)
dosage is. less pronounced at pH 6.0 than at pH 7.0. Additional aluminumUII)
requirements for removal of phosphorus, caused by colloidal silica surface area
are due to competition of silica colloids with colloidal phosphate precipi-
tates rather than with solution phase phosphate,
Inferences with respect to plant scale phosphorus removal are made. An experi-
mental procedure feasible to investigate the effect of different size parti-
culates on removal of phosphorus by aluminumUM) in trickling filter efflu-
ents is proposed.
107
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-77-OI2
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
METHODS FOR IMPROVEMENT OF TRICKLING FILTER PLANT
PERFORMANCE - PART II - CHEMICAL ADDITION
5. REPORT DATE
January I977 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
James C. Brown
Linda W. Little
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
University of North Carolina
Department of Environmental Sciences and Engineering
School of Pub Iic Health
Chapel Hill, North Carolina 275 I 4
10. PROGRAM ELEMENT NO.
IBC6I I
11. CONTRACT/GRANT NO.
Contract #14-12-505
Project #11010 DGA
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory - Gin., OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final . I97I - 1972
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Studies undertaken on Part I of this project were previously
reported in EPA-670/2-73-047a (PB-224 715), "Methods for Improvement of Trickling
Filter Plant Performance - Part I - Mechanical and Biological Optima." August 1975.
IB.ABSTRACT An experimental program to explore potential methods for removing phosphoru:
and generally enhancing trickling filter plant performance was conducted at the Mason
Farm Wastewater Treatment Plant, Chapel Hill, North Carolina. Preliminary investi-
gations included jar testing with several coagulants and coagulant aids and pilot
studies to determine the effect of the point of a I urn addition on phosphorus removal.
Follow-up fuI I-seale stud ies utiIi zed the Chape I Hill h igh-rate trick I, i ng f iIter
plant which consists of two parallel identical main-stream trains. From January 25
through October 6, 1972, alum was added to the influent of one final clarifier. Alum
dosage and influent flow rates to the dosed train were varied and phosphorus removal,
general plant performance, sludge production, and sludge digestion performance were
mon itored.
Alum addition effectively removed phosphorus and enhanced overall plant perform-
ance. Optimization of a I urn precipitation will require a flow-paced a I urn feed system,
restriction of average dry weather final settling tank surface loadings to 20.4 m3/
day/m (500 gpd/ft2), and inclusion of tertiary fine solids removal facilities.
Alum sludge decreased the alkalinity and pH in the primary anaerobic digester
and led to liquid/solids separation problems in the secondary digester. Separate
facilities may be necessary for handling alum-humus sludge from the final settling
tank.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
*Trickl
Sewage treatment, *Trickling filtration,
larificat ion, *Chemical removal (sewage
treatment), Coagulation, Upgrading,
^Aluminum sulfate, Sludge digestion
b.IDENTIFIERS/OPEN ENDED TERMS
;. COS AT I Field/Group
^Phosphorus removal,
*AI urn precipitation,
3B
18. DISTRIBUTION STATEMENT
Re I ease to Pub I ic
19. SECURITY CLASS (This Report)'
UncI ass i f ied
21. NO. OF PAGES
I 20
20. SECURITY CLASS (This page)
UncI ass i f ied
22. PRICE
EPA Form 2220-1 (9-73)
108
£ U.S. GOVERNMENT PKINIINO UH-IU: I 9//--757-056/5562 Region No. 5-11
------- |