WATER POLLUTION CONTROL RESEARCH SERIES T7010 EIP 05/71
Soluble Phosphorus in the
Activated Sludge Process
Part 1
Chemical-Biological Process Performance
U.S. ENVIRONMENTAL PROTECTION AGENCY
-------
WATER POLLUTION CONTROL RESEARCH SERIES
The Water Pollution Control Research Series describes
the results and progress in the control and abatement
of pollution in our Nation's waters. They provide a
central source of information on the research, develop-
ment, and demonstration activities in the Environmental
Protection Agency, through inhouse research and grants
and contracts with Federal, State, and local agencies,
research institutions, and industrial organizations.
Inquiries pertaining to Water Pollution Control Research
Reports should be directed to the Chief, Publications
Branch, Research Information Division, Research and
Monitoring, Environmental Protection Agency, Washington,
D. C. 20460.
-------
SOLUBLE PHOSPHORUS REMOVAL IN THE ACTIVATED SLUDGE PROCESS
PART I
CHEMICAL-BIOLOGICAL PROCESS PERFORMANCE
by
The Soap and Detergent Association
New York, N. Y. 10016
for the
Office of Research and Monitoring
ENVIRONMENTAL PROTECTION AGENCY
Project #17010 EIP
May 1971
-------
EPA Review Notice
This report has been reviewed by the Environmental
Protection Agency and approved for publication.
Approval does not signify that the contents neces-
sarily reflect the views and policies of the
Environmental Protection Agency, nor does mention
of trade names or commercial products constitute
endorsement or recommendation for use.
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $1.26
ii
-------
ABSTRACT
It was the objective of this research to develop and evaluate, at full
plant scale, the combined chemical-biological process of phosphorus
removal. The research was conducted in two major investigative phases
using the final aeration and settling tanks of that portion of the
Pennsylvania State University Wastewater Treatment Plant which treats
wastewater from the PSU campus.
The Phase I investigations indicated an Al/P (filt. ortho) weight ratio
of 2.25/1 was necessary to reduce the influent phosphorus of approxi-
mately 10 mg P/l to approximately 0.3 mg P/l in the filtered effluent.
Alum proved to be a more effective precipitant than sodium aluminate
in the moderately alkaline wastewater available for this study. The
best results were obtained when the chemical was added at or near the
effluent end of the aeration tank.
The total phosphorus concentration of the unfiltered effluent was
dependent upon effluent suspended solids levels as well as on effluent
soluble phosphorus concentrations.
Removal of organic matter was improved as a result of chemical addition
in the chemical-biological process.
Alum addition into the effluent channel from the aeration tank resulted
in an inhibition of nitrification apparently as a result of pH changes
which occurred. No such effect was noted with the use of sodium alumi-
nate.
The chemical-biological process produced approximately twice as much
weight of sludge as did the parallel control. Alum addition did not
reduce the sludge volume index (SVI) of the mixed liquor whereas sodium
aluminate addition resulted in significant decreases.
Costs for chemical precipitation of phosphorus in the chemical-biologi-
cal process to reduce the phosphorus from 10 mg P/l to 0.3 mg P/l are
estimated to vary from 7.3 C/1000 gal in a 1 MGD plant to 4.1 C/1000
gal in a 100 MGD plant. Total treatment costs are estimated to vary
from 39.6 C/1000 gal to 16.7 0/1000 gal for the 1 and 100 MGD plants
respectively.
This report was submitted in fulfillment of Project Number 17010 EIP
under the partial sponsorship of the Water Quality Office, Environ-
mental Protection Agency, and by the Soap and Detergent Association,
New York, New York.
iii
-------
TABLE OF CONTENTS
Page
LIST OF TABLES . ix
LIST OF FIGURES. . . xi
LIST OF SYMBOLS. xiii
CONCLUSIONS, . o 1
RECOMMENDATIONS FOR FURTHER RESEARCH 5
INTRODUCTION 7
GENERAL STATEMENT OF THE PROBLEM 7
CHEMICAL-BIOLOGICAL PROCESS CONSIDERATIONS 7
Objectives of the Present Investigation 8
EXPERIMENTAL APPROACH, 9
GENERAL PLAN OF RESEARCH . . . 9
PROCESS SCHEMESo ............. ... 9
ANALYTICAL TESTS AND PROCEDURES 11
PHASE I INVESTIGATIONS .......... 15
PROGRAM OF STUDY ........ 15
INFLUENT WASTE 15
PLANT OPERATIONAL PROCEDURES .......... 15
General. . . . . » . . . . . . . . . 15
Chemical Addition. ....... . 17
Sludge Wasting ............. 19
Mixed Liquor Solids Balance and Sludge Production.
Calculations ...................... 19
Reported Solids Age. ......... 19
ALUMINUM SULFATE STUDIES . 20
General Operation. 20
Influent Waste .......... 20
General Effluent Quality . 24
Phosphorus Removal , . 26
SODIUM ALUMINATE STUDIES ... 28
General Operation. ... ..... 28
Influent Waste .............. . . 32
General Effluent Quality 32
Phosphorus Removal ...... ... 35
-------
TABLE OF CONTENTS (continued)
Page
ACTIVATED SLUDGE STUDIES 35
Mixed Liquor 35
Sludge Production. 38
Mixed Liquor Phosphorus 39
GENERAL DISCUSSION 39
PHASE II INVESTIGATIONS 43
PROGRAM OF STUDY 43
General Operation ..... ......... 43
FLOWS AT OR BELOW DESIGN AVERAGE . 44
Influent Waste ........ 44
General Effluent Quality ..... 48
Phosphorus Removal ......... 48
FLOWS IN EXCESS OF DESIGN AVERAGE. ............ 49
Influent Waste ............. . 49
General Effluent Quality . . 55
Phosphorus Removal ............ 55
ALL FLOW DATA. ...................... 58
General. .................. 58
Influent Waste ................ 58
General Effluent Quality ........... 58
Phosphorus Removal . 67
ACTIVATED SLUDGE STUDIES ................. 68
Mixed Liquor . 68
Sludge Production. . 73
Mixed Liquor Phosphorus...... ........... 76
SPECIAL STUDIES. ............... 77
Polyelectrolyte Addition ........ 77
pH Studies .......... . 81
Identification of "Carry-Through" 88
Nitrification 88
Effluent Fertility ....... 90
vi
-------
TABLE OF CONTENTS (continued)
Page
CHEMICAL-BIOLOGICAL PROCESS DESIGN AND OPERATION 93
DESIGN CONSIDERATIONS 93
Preliminary Data 93
Chemical Handling and Feeding 93
Process Control 94
Hydraulic Loading on Clarifiers 94
Effluent Filtration 95
Solids Handling 95
OPERATIONAL CONSIDERATIONS 95
General 95
Process Control 95
Maintenance 96
COST ANALYSIS 99
GENERAL 99
CAPITAL COSTS 99
OPERATING COSTS 100
CHEMICAL COSTS 100
COST SUMMARY 102
GENERAL DISCUSSION 105
ACKNOWLEDGEMENTS 113
BIBLIOGRAPHY 115
vii
-------
LIST OF TABLES
Table Page
1 Summary Data from Aluminum Sulfate Studies Phase
I 21
2 Summary Data from Sodium Aluminate Studies Phase
I 29
3 Summary of Activated and Waste Sludge Data Phase
I 37
4 Comparative Waste Sludge Production Phase I .... 39
5 Comparison of Effluent Characteristics and Solids
Production Using Alum and Sodium Aluminate. ... 40
6 Summary Data from Phase II One Year Full Scale
Study for Flows Not Exceeding 1.040 MG Total
Daily Flow Per Tank . . 45
7 Summary Data from Phase II One Year Full Scale
Study for Flows Exceeding 1.041 MG Total Daily
Flow Per Tank 52
8 Summary Data from Phase II One Year Full Scale
Study for All Flow Data 60
9 Forms of Nitrogen in Mixed Liquor 67
10 Summary of Activated and Waste Sludge Data Phase
II. 72
11 Comparative Waste Sludge Production Phase II. ... 73
12 Effluent Phosphorus and Turbidity Relationships . . 83
13 Summary Data from Special Nitrification Study ... 89
14 Cost of Chemicals Delivered to State College,
Pennsylvania 102
15 Cost Summary for the Chemical-Biological Process to
Precipitate 97 Percent of the Influent Phosphorus
(1970 Dollars) 103
16 Estimated Total Treatment Costs (1970 Dollars). . . 104
17 Summary Phosphorus Removal Data Phase II June 15,
1970 to August 12, 1970 108
18 Comparison of Observed and Predicted Amounts of
Phosphorus in Waste Sludges 110
ix
-------
LIST OF FIGURES
Figure Page
1 Schematic Flow Diagram University Wastewater
Treatment Plant 10
2 Phosphorus Removal Process Simplified Flow Sheet . 16
3 Typical Variation in Flow and Phosphorus
Concentration , 18
4 BOD Removal as a Function of BOD Loading Phase I
Aluminum Sulfate Studies 25
5 Effluent Phosphorus Concentration as a Function
of Al/P (filt. ortho) Weight Ratio Phase I
Aluminum Sulfate Studies 27
6 BOD Removal as a Function of BOD Loading Phase I
Sodium Aluminate Studies 33
7 Effluent Phosphorus Concentration as a Function
of Al/P (filt. ortho) Weight Ratio Phase I
Sodium Aluminate Studies ........ 36
8 Statistical Distribution of Filtered Effluent
Total Phosphorus Tank 1 Flow £ 1.040 MGD .... 50
9 Statistical Distribution of Unfiltered Effluent
Total Phosphorus Tank 1 Flow £ 1.040 MGD .... 51
10 Daily Variation in Effluent Quality Tank 1 Flow
j> 1.041 MGD. 56
11 Statistical Distribution of Filtered Effluent
Total Phosphorus Tank 1 Flow >^ 1.041 MGD .... 57
12 Statistical Distribution of Unfiltered Effluent
Total Phosphorus Tank 1 Flow >^ 1.041 MGD .... 59
13 BOD Removal as a Function of BOD Loading Phase
II All Flow Data 64
14 Effluent BOD as a Function of Effluent Volatile
Solids Phase II All Flow Data. 66
15 Statistical Distribution of Filtered Effluent
Total Phosphorus Tank 1 All Flow Data 69
16 Insoluble Phosphorus as a Function of Effluent
Suspended Solids Phase II Tank 1 All Flow Data . 70
xi
-------
LIST OF FIGURES (continued)
Figure Page
17 Statistical Distribution of Unfiltered Effluent
Total Phosphorus Tank 1 All Flow Data 71
18 Mixed Liquor Settling Characteristics Phase II
Alum to Tank 1 74
19 Correlation of Activated Sludge Phosphate and
Calcium Contents 78
20 Correlation of Activated Sludge Phosphate Content
and Volatile Fraction 79
21 Comparative Settling With and Without Polymer
Addition to Tank 1 Mixed Liquor 80
22 Variation in Effluent Suspended Solids With and
Without Polymer Addition Tank 1 Flow >^ 1.041
MGD 82
23 Correlation of Effluent Phosphorus and Turbidity
with pH Adjustment 84
24 In situ pH Profile Alum Addition into Effluent
Channel from Aeration Tank 86
25 In situ pH Profile Alum Addition at Head End of
Aeration Tank 87
26 Chemical Costs as a Function of Filtered Effluent
Total Phosphorus Concentration . 101
27 Aluminum Dosage as a Function of Flow Phase II . . 109
xii
-------
LIST OF SYMBOLS
GENERAL
BOD - biochemical oxygen demand
BOD5 - five-day BOD
BOD5A ~ BOD5 aPPlied
BOD,-_ - BOD- removed
jK j
COD - chemical oxygen demand
IS - inorganic suspended solids
LF - activated sludge loading factor, wt BOD,-A/wt MLVS/t
DA
MLSS - mixed liquor total suspended solids
MLVS - mixed liquor volatile suspended solids
R - correlation coefficient
Sr - wt BODcl,/wt MLVS/t
_>K
SS - total suspended solids
SVI - sludge volume index
t - time
tg. - solids age
VS - volatile suspended solids
VS - VS produced
MATHEMATICAL SYMBOLS AND ABBREVIATIONS
< - less than
= - equals
> - greater than
log - common logarithm to base 10
xiii
-------
CONCLUSIONS
From the results of this study and other related work, it may be con-
cluded that:
1. A very high degree of removal of soluble phosphorus
can be obtained continuously in full plant scale opera-
tion using the chemical-biological process. The removal
efficiency is dependent primarily upon the Al/P ratio.
In the system studied, an Al/P (filt. ortho) ratio of
2.25/1 was necessary to achieve a filtered effluent
total phosphorus concentration of approximately 0.3
mg P/l when aluminum sulfate (alum) was used as the pre-
cipitating chemical.
2. Alum was a more efficient precipitant than was
sodium aluminate for the moderately alkaline wastewater
available for this study. An Al/P (filt. ortho) ratio
of 2.25/1 resulted in a filtered effluent total phos-
phorus concentration of approximately 0.5 mg P/l when
sodium aluminate was used.
3. The best results were obtained when the chemical
(alum or sodium aluminate) was added at the effluent
end of the aeration tank« No deterioration in effluent
quality was observed at very high Al/P (filt. ortho)
ratios when alum was added at the effluent end.
4. It was not feasible to adjust chemical feed rates
manually to compensate routinely and rapidly for
changes in influent flow and phosphorus concentration.
The effect of changing flows was much more significant
than that of changing phosphorus concentrations in the
system studied so that automatic pacing of chemical
feeders with flow would have improved significantly
the operating performance of the system
5. Removal of organic matter as measured by unfiltered
effluent BOD5 and COD was enhanced as a result of chemi-
cal addition.
6. Suspended solids losses from the chemical-biological
system were very heavy when flows were high and surface
settling rates in the final clarifier exceeded 800
gal/sq ft/day. At low flows, the effluent suspended
solids were significantly lower in the chemical-biologi-
cal system than in the parallel control.
7. Because of the demonstrated relationship between
effluent suspended solids and effluent insoluble phos-
phorus concentration, it is essential that effluent
filtration be provided where extremely high unfiltered
-------
effluent total phosphorus removals are required. Effluent
filtration may be advisable also in other situations as
a factor of safety to protect against unexpected solids
losses and resulting high phosphorus residuals.
8. Alum addition into the effluent channel from the
aeration tank resulted in an inhibition of nitrification
under the operating conditions experienced during the
study. It is thought that this inhibition was a pH
related phenomenon although this could not be shown con-
clusively from the observed data. No inhibition was ob-
served with the use of sodium aluminate which does not
depress the pH as alum does or when alum was added at
the influent end of the aeration tank with higher MLSS.
9. The study demonstrated that approximately twice as
many pounds of total solids were produced in the chemi-
cal-biological system as in the conventional activated
sludge control when alum was used as the precipitant.
The addition of alum did not decrease the SVI of the
mixed liquor whereas addition of sodium aluminate sig-
nificantly reduced the SVI. While these results reflect
conditions observed in the system studied, they are not
necessarily typical of results which would be obtained
in other systems because of the lower than normal SVI
observed in the activated sludge control.
10. Attempts to identify the "carry-through" which
occurs when the precipitant is added at the head end
of the aeration tank were not successful. It appears
to be a finely divided "aluminum phosphate" precipi-
tate which does not settle out in the clarifier. pH
has an effect on the amount of "carry-through" as does
the point of chemical addition but under the conditions
of this study it was impossible to separate these two
factors completely and to identify their respective
roles.
11. The removal of phosphorus by alum addition resulted
in lower fertility in the treated effluent compared to
the untreated effluent as measured by the Provisional
Algal Assay Procedure. These studies were very limited
in scope and should not be considered generally appli-
cable to field conditions.
12. Costs for phosphorus precipitation by the chemical-
biological process for 97% filtered effluent total phos-
phorus removal (based on 10 mg P/l in the influent
waste) vary from 7.3 0/1000 gal for a 1 MGD plant to
4.1
-------
units or the cost of filtration and sludge handling and
disposal. Inclusion of these costs raise the figures to
39.6 C/1000 gal (1 MGD) and 16.7 0/1000 gal (100 MGD)
based on the assumptions made.
-------
RECOMMENDATIONS FOR FURTHER RESEARCH
The observations and findings of this study suggest the following as
areas where further research would be beneficial:
1» The relative role of biological and chemical removal
of organic substrate should be identified so that pro-
cess design and operation parameters can be optimized.
20 The effect of alum addition on nitrification should
be identified conclusively so that process design and
operation can be modified to achieve desired degrees of
nitrification.
3. Additional investigations of the effect of both pH
adjustment and polymer addition as means of improving
effluent quality and optimizing chemical addition should
be made,,
4o Better definition of the role of phosphorus in con-
trolling eutrophication should be determined so that
realistic treatment requirements can be established.
-------
INTRODUCTION
GENERAL STATEMENT OF THE PROBLEM
The problems created by "cultural eutrophication" of our surface
waters now are recognized by the general public as well as by pro-
fessionals working in water pollution control. The aesthetic and
economic effects of the nuisance-level proliferation of aquatic
plants no longer can be ignored or merely talked about.
Nutritional limitation has been suggested as a means of controlling
eutrophication and phosphorus has been the nutrient most workers have
focused attention on as being most feasible to control. Domestic and
industrial wastewaters are, in many instances, the major sources of
phosphorus input to our surface waters. Since these represent point
discharges, they most easily are subject to control methods and
hence most attention has been directed towards development of phos-
phorus removal methods for treatment of wastewaters.
Nesbitt (24, 25) in his reviews of phosphorus removal included the
various work which has been done on the chemical precipitation of
phosphorus. Hall and Englebrecht (11) have also presented a compre-
hensive review of currently employed phosphorus removal technology.
Most workers seem to agree that some form of chemical precipitation
of phosphorus is the best way to achieve the high degrees of removal
being required. However, there is no general agreement as to which
of the several available precipitating chemicals or which of the many
process schemes is best. Nesbitt (24) concluded on the basis of the
figures presented in his report that chemical precipitation combined
with high rate activated sludge appeared to be the most economical of
the chemical precipitation processes.
CHEMICAL-BIOLOGICAL PROCESS CONSIDERATIONS
Several investigators have studied phosphorus removal by precipitation
with aluminim or iron salts within the aeration tanks of the activated
sludge process. The original work suggesting this concept of phos-
phorus removal was done by Stumm (41) and was further developed and
expanded by Tenney and Stumm (43). This basic work was then developed
into a complete process design utilizing a high rate activated sludge
system in a continuous flow bench scale study by Eberhardt and Nesbitt
(9). The above studies used alum precipitation and showed the chemi-
cal-biological process to be capable of a high degree of phosphorus
removal while at the" same time achieving removals of organic carbon
which are typical of high rate activated sludge systems. Earth and
Ettinger (2) in a 100 gpd pilot plant study were also able to demon-
strate excellent phosphorus removals using sodium aluminate as the
precipitant, Brenner (4, 5) has reviewed the chemical-biological
process and presents summary data from several studies of the process.
-------
In order to demonstrate the applicability of the chemical-biological
process for full plant scale use as well as to develop operating pro-
cedures and cost data for this process, it was proposed to study the
process in a prototype plant scale at the Pennsylvania State University
Wastewater Treatment Plant.
Objectives of the Present Investigation
The specific objectives of the study were:
1. To determine the relative merits of sodium aluminate and
aluminum sulfate (alum) as sources of aluminum ion for the
precipitation of phosphorus.
2. To determine the optimum precipitant dosage and point
of application in the system.
3. To determine the efficiency of the precipitating agent
in removing phosphorus and BOD over a one-year operating
period and to compare these results with those obtained
from a parallel control unit which was not to receive any
chemical addition.
4. To develop recommended design and operating procedures
and estimated costs for the process based on the system
studied.
-------
EXPERIMENTAL APPROACH
GENERAL PLAN OF RESEARCH
The research was conducted and is reported herein as two major inves-
tigative phases:
Chemical Selection and Process Optimization (Phase I)
Process Performance and Cost Analyses (Phase II)
The initial studies on chemical selection and process optimization
were designed to evaluate the relative effectiveness of two aluminum
compounds in precipitating phosphorus. Selection of the optimum
chemical dosage and point of application for each of the two chemi-
cals studied were also accomplished during this period.
The second phase of the project was designed to gather long term
operating data on process performance. These data were then used
as the basis for development of recommended operating procedures
and cost data for the chemical-biological process of phosphorus re-
moval .
PROCESS SCHEMES
The research was conducted using the final aeration and settling units
of that part of the Pennsylvania State University Wastewater Treat-
ment Plant which treats University wastewater and in the digestion
units serving the entire plant. It is the purpose of this section to
provide only a physical description of the treatment units. Pertinent
operating conditions will be reported elsewhere.
A simplified flow sheet of the complete treatment plant is shown in
Figure 1.
The final aeration tanks provide second stage biological treatment of
the University wastewater following first stage high rate trickling
filtration. Each of the two aeration tanks has a capacity of 15,600
cu ft which provides a theoretical detention time of 2.25 hours at the
original design flow of 2.0 million gallons per day (MGD) plus 25%
return sludge. Aeration is provided by rotary displacement variable
drive blowers using valved orifice diffusers on swing drop pipes.
Initial design air application rates were 0.56 cu ft/gal/day or 1,000
cu ft/day/lb BOD5 (42).
The final, settling tanks each have an area of 1,300 sq ft thus giving
a surface settling rate of 770 gal/sq ft/day at a design flow of 2.0
MGD. In 1965, a sludge flotation thickener unit was added to thicken
the waste activated sludge from the portion of the plant treating
Borough of State College wastewater. The underflow from the thickener
is discharged into the trickling filter effluent so that it receives
-------
FROM
BORO
BARMINUTOR
J
FROM
BORO
LEGEND
RAW WASTE OR SLUDGE
EFFLUENT
ft- SUPER
BARMINUTOR
SUPER \
STORAGE)
SLUDGE
TORAG
r*1 HEAD BOX
BORO\
(PRIMARY^
'AERATION1
X
BORO \
'SECONDARY)
I SLUDGE '
'THICKENER1
HIGH RATE
TRICKLING
FILTER
HIGH RATE
TRICKLING
FILTER
AERATION 1,1
FINAL
IGESTERJ SETTLING
TANKS
FINAL
AERATION
TANKS
FROM
UNIVERSITY
CHLORINE
CONTACT
TANK
L^S J77 ^s
^DIGESTED
SLUDGE
FINAL 4
EFFLUENT
Figure 1. Schematic Flow Diagram University Wastewater Treatment Plant
-------
treatment in the activated sludge units included in this study. This
additional flow reduces the design detention time in the aeration tanks
to 2.12 hours and increases the design surface settling rate to 825
gal/sq ft/day in the final settling tanks. The design air application
rates are reduced to 0.53 cu ft/gal/day or 925 cu ft/day/lb 8005 (42).
Recirculation of return sludge to Aeration Tank No. 1 was accomplished
using the existing sludge recirculation pumping station and a modified
piping scheme to permit return to Tank No. 1 only. The existing tele-
scopic drawoff valves in Final Settling Tank No. 2 were removed and
4 inch air lift pumps were installed in their place to permit recircu-
lation of sludge to Aeration Tank No. 2. Sludge from each of the tanks
was wasted by means of the existing waste sludge pumps. Return sludge
recirculation flows and waste sludge flows for each tank were individ-
ually metered by magnetic flow meters.
Feeding of chemicals was accomplished by metering the chemical through
a Wallace & Tiernan Model 747 diaphram pump with gravity flow through
1/2-inch insulated plastic lines to the point of chemical application.
The chemicals were purchased in liquid form and were stored in a 5,000
gallon vinyl lined wooden tank. The amount of chemical fed each day
was obtained from measured changes in the tank level. Pump calibration
curves were prepared to permit accurate feed rate adjustment with the
different chemicals used.
Typical properties of the commercial liquid chemicals used in the study
are:
Aluminum Sulfate Sodium Aluminate
Total Soluble A1203 8.3% 26.7%
Na203 -- 18.3%
Total Fe 0.2% 0.002%
Insoluble in water 0.3%
Viscosity, Centipoises
60°F 27 2100
90°F 15 260
Weight Ib/gal 60°F 11.2 13.1
ANALYTICAL TESTS AND PROCEDURES
Routine sampling of the influent to the aeration tanks (trickling
filter effluent) and the final effluent from each settling tank was
done on a 3-day per week schedule, usually Sunday, Tuesday and Thurs-
day, so the samples could be analyzed on Monday, Wednesday and Friday.
Samples were collected automatically over a 24-hour period by means of
SERCO Model NW-3-8 samplers. Grab samples of the mixed liquor from
each aeration tank were collected 5 days per week about 8:00 A.M.
11
-------
Composite samples of waste sludge from each tank were collected twice
weekly, usually on Monday and Wednesday.
A summary of the tests performed and analytical procedures used
throughout this investigation follows:
Alkalinity Alkalinity determinations were made in
accordance with Standard Methods (40) using a pH of 4.5
as the titration end point.
Aluminum Aluminum ion determinations were made using
the "Rapid Modified Eriochrome Cyanine R Method for the
Determination of Aluminum in Water" reported by Shull
and Guthan (36). Samples were filtered and diluted prior
to analysis.
Biochemical Oxygen Demand BOD was determined by the
method given in Standard Methods (40). Nitrification
was inhibited by means of allyl thiourea in all instances
except where noted.
Calcium and Magnesium Ca and total hardness were
measured by the EDTA titrimetric method of Standard
Methods (40) using Betz calcium and hardness indicators.
Mg was determined by difference from total hardness and
calcium results.
Calcium in Sludges Total calcium in sludge samples
was determined in accordance with the procedure reported
by Menar and Jenkins (23) except that Betz Calcium Indi-
cator was used in place of hydroxnapthol blue.
Chemical Oxygen Demand COD was measured by the pro-
cedure developed at the Sanitary Engineering Research
Laboratory (SERL) of the University of California (32).
Color Color determinations were made using a Hellige
Aqua Tester Model No. 611 with a 611-10 color disc.
Filtration Filtration for solids determinations and
where used prior to other analytical tests was performed
using 5.5 cm diameter Reeve Angel glass fiber filter
pads and Millipore filter flasks and adapters.
Nitrogen Ammonia and total Kjeldal nitrogen were
determined by the methods given in Standard Methods (40).
Oxidized nitrogen was determined by the Standard Methods
chromotropic acid method (40).
_p_H pH measurements were made using a Fisher Accumet pH
meter Model 210.
12
-------
Phosphorus The Stannous Chloride Method for Ortho-
phosphate in Standard Methods (40) was used for all of
the phosphorus determinations. All tests were conducted
in an incubator controlled at 20°C and ten minutes were
allowed for color development before readings were taken.
Readings were taken at a wave length of 600 my using a
Bausch and Lomb Spectronic 20 spectrophotometer with a
1-inch light path. The methods of sample preparation
varied as indicated below:
Total phosphorus determinations on mixed
liquor and waste sludge samples followed a
modified alkaline ash procedure used by
Eberhardt and Nesbitt (9). Total phos-
phorus determinations on influent and efflu-
ent samples were made using a modification
of the binary acid wet-ash procedure (46).
Total and orthophosphate were determined on
both filtered and unfiltered samples.
Sludge Volume Index SVI of the mixed liquor was
determined in accordance with Standard Methods (40).
Suspended and Volatile Solids All suspended and
volatile solids analyses were carried out using the
procedure reported by Eberhardt and Nesbitt (9).
Sulfate Sulfate analyses were conducted using the Tur-
bidimetric Method of Standard Methods (40). All samples
were filtered and diluted prior to analysis.
Turbidity All turbidity measurements were made with
a Hach Model 2100 Laboratory Turbidimeter.
13
-------
PHASE I INVESTIGATIONS
PROGRAM OF STUDY
The primary purpose of Phase I of the project was to compare the
relative merits of aluminum sulfate (alum) and sodium aluminate as
precipitating agents for use in removing phosphorus from domestic
wastewater. Selection of optimum points of chemical addition and
dosages were also accomplished during this period of study.
Addition of chemical to both aeration tanks was undertaken in
order to compare results obtained using various Al/P ratios and
points of addition in less time than would have been necessary if
one tank was maintained as a parallel control.
INFLUENT WASTE
The major portion of the raw waste influent to the part of the treat-
ment plant used in the study was derived from the Pennsylvania State
University campus. A portion of the flow originates from commercial
and residential buildings along College Avenue in the Borough of
State College. No significant industrial wastes are tributary to
the plant but since the major portion of the wastes originate on cam-
pus they could be characterized as more typical of institutional
wastes than residential domestic wastes. As can be seen from the flow
diagrams Figure 1 and 2, the wastes receive barminution, primary set-
tling and high rate trickling filtration prior to treatment in the
activated sludge units used in this study. Underflow from the flota-
tion thickener is directed to the activated sludge units under normal
operation as indicated previously. Data on wastewater characteristics
will be reported elsewhere.
Variations in student population during the different terms of the
school year together with the relatively long vacation periods when
students were nearly all absent from the campus created fluctuations
in flow and phosphorus concentration that were atypical of most munici-
pal installations. In addition, although the other portion of the
plant, intended to treat Borough flow, has a design capacity of 2.0
MGD, flows in excess of 1.7 MGD cause severe operating problems.
Therefore, all flows to this plant in excess of 1.7 MGD are diverted
to the portion of the plant used in the study. These extreme varia-
tions in flows together with flows well in excess of design capacity
(2.0 MGD) as a result of diversion from the other plant caused prob-
lems which will be discussed later.
PLANT OPERATIONAL PROCEDURES
General
Plant operational procedures followed those normally employed in con-
ventional activated sludge plants. Addition of the precipitating
15
-------
CHEMICAL
FEED
BUILDING
i
L CHEMICAL
f^FEED
LINES
>
TRICKLING^
FILTER
EFFLUENT
CHEMICAL
FEED -^
POINT
I
=-j-.l_
nr
FINAL
AERATIONT
NO. 1
LJ
rf
-
^ ''
TANKS
NO. 2
ii
jj^
FINAL
SETTLING
NO. 2
BLOWER
BLDG.
0
1 1
li=
TANKS
NO. 1
ii EFFLUENT
"TO CL2 /
L[] TANK /
f
SECONDARY
DIGESTER
\
SLUDGE
PUMPING i >
STATION WASTE
^
"1 SLUDGE
-J TO
PRIMARY
CLARIRER
Figure 2. Phosphorus Removal Process Simplified Flow Sheet
-------
chemical did not alter these procedures but did require some additional
operational considerations and time.
Chemical Addition
Since addition of chemical was solely for removal of phosphorus,
amounts of chemical to be added were calculated to achieve a desired
aluminum to phosphorus (Al/P) ratio. Phosphorus in the influent
waste can be measured in several different ways depending on the
particular analytical procedure used, i.e., filtered or unfiltered,
ortho or total, etc. It was decided for the purposes of this pro-
ject to use the filtered orthophosphate in the trickling filter
effluent (aerator influent) as the basis for chemical dosage calcu-
lations. This basis was selected since the particular test procedure
is simple, fairly rapid and is particularly well suited for use in a
typical wastewater treatment plant laboratory by plant personnel.
All reported Al/P ratios are weight ratios since they were easier to
use in day to day calculations thus minimizing errors.
Calculations for feed rate adjustment of chemical additions were based
on the average orthophosphate concentration in the filtered trickling
filter effluent, average wastewater flows and the desired Al/P ratio.
Various Al/P ratios were selected for testing in order to determine
the optimum ratio for each of the two chemicals tested and rates of
chemical addition were then calculated to achieve the desired ratios.
Actual Al/P ratios, for each day samples were collected, were later
calculated from the data using actual flow, filtered trickling filter
effluent orthophosphate concentration and the amount of aluminum fed.
Chemical feed rates were adjusted by treatment plant personnel three
or four times .daily (depending on flow patterns) in an attempt to
optimize chemical addition. Rate changes were based only on average
flow variations during the day. Variations in phosphorus concentra-
tions during the day were also observed but the magnitude of change
was not considered sufficient to incorporate into feed rate calcula-
tions during Phase I studies. Figure 3 shows typical variations in
flow and phosphorus concentration for periods when the University was
in session and a normal student population was on campus. Other
workers (10) have observed that phosphorus concentration vary much
more widely than those observed in this study. The relatively minor
variations observed herein are most likely explained on the basis of
the institutional nature of the raw waste and the habits of a student
population which are atypical of a residential-commercial domestic
waste flow pattern.
Because of daily variations in waste flow and influent phosphorus
concentrations from values used for calculation of feed rates, it
was impossible to achieve the desired Al/P ratio in all instances,
in fact, the deviation from the desired ratio was much greater than
had been anticipated. This will be considered again later when
results of Phase II are discussed. As a result of this variation,
17
-------
150
125
ui
§
iu 100
o
<
o:
UJ
I-
z
u
u
K
Ul
0.
50
25
rPHOSPHORUS
/ CONCENTRATION
DAILY
AVERAGE
j I
6A.M. 12 N
I
6 P.M.
TIME
12 M 6A.M.
Figure 3. Typical Variation in Flow and Phosphorus Concentration
18
-------
data presentation for Phase I studies is based on ranges of observed
Al/P ratios rather than on time periods.
Sludge Wasting
A schedule for wasting excess sludge from the aeration tanks was
established each day based on the mixed liquor suspended solids
analysis for that day. Sludge was wasted over periods of time
varying from 4 to 24 hours depending on the specific requirements
necessary to maintain the desired solids in the aerators.
An attempt was made to carry approximately the same level of volatile
solids (assumed to represent equal biomasses) in each aerator. It
should be noted that the volatile fraction of mixed liquor suspended
solids is much lower in the chemical-biological system than it is in
a conventional activated sludge system. Therefore, it is necessary
to carry a higher total mixed liquor suspended solids in the chemi-
cal-biological system than it is in conventional activated sludge if
equal volatile solids concentrations are to be carried in each system.
The volatile solids procedure which was used does not permit the
direct determination of organic solids but includes, in addition,
inorganic compounds which volatize at temperatures of up to 600°C.
Although some of the compounds which are formed during the precipita-
tion of phosphorus are aluminum-hydroxy-phosphates which undoubtably
volatilize, no attempt was made to correct the apparent volatile
solids data for purposes of plant control. Waste sludge volumes were
measured with magnetic meters as indicated earlier. Waste sludge
samples for phosphorus and solids analyses were also collected as
indicated previously.
Mixed Liquor Solids Balance and Sludge Production Calculations
For purposes of performing a system solids balance, the following
procedures were used: the weight of solids contained in the
aeration tanks was calculated from the measured solids concentration
and the design volume of the tanks. Effluent solids losses were com-
puted from measured solids concentrations and effluent volumes (deter-
mined from raw wastewater flow data). No correction was made for the
underflow from the flotation thickener or for solids in the clarifiers,
Waste sludge volumes and solids concentrations were used to calculate
the weight of sludge wasted.
Reported Solids Age
Solids age is defined consistently throughout this study as the
quotient of the total system volatile solids under aeration divided
by the volatile solids lost from the system and was calculated as
follows:
Solids, v MLVS (mg/1) x Aeration tank volume (1)
Age l SA; ~ Eff VS (mg/1) x Eff volume (1) + (VS in Waste) ± MLVS (mg)
Sludge (mg)
19
-------
Solids age data will be included with the other results presented
later.
ALUMINUM SULFATE STUDIES
General Operation
Addition of liquid alum to one aeration tank (Tank No. 2) was begun
on January 27, 1969 at a point two-thirds of the way along the tank
from the influent end and directly above the diffusers so the roll
of the tank would help to disperse the chemical. After an initial
decrease in settling tank effluent suspended solids from the treated
tank, observed for only a few hours after addition was started, these
solids increased substantially over those observed in effluent from
the untreated tank (Tank No. 1). In an attempt to remedy this solids
loss, the point of chemical addition was moved to a point in the
effluent channel carrying mixed liquor from the aeration tank to the
clarifier. After this change, effluent suspended solids from the
treated tank decreased and remained approximately the same as those
from the untreated tank.
Alum addition to Tank No. 1 was started on February 21, 1969. Dif-
ferent Al/P ratios were established for each of the two tanks in order
to obtain data which would permit evaluation of the optimum Al/P ratios
for this system. The data obtained during the alum runs and during the
brief period of no chemical addition are presented in Table 1. These
data are average values obtained from all observations wherein the
actual Al/P ratio observed fell into the ranges shown. Data from both
Tanks No. 1 and No. 2 are averaged together since the period prior to
chemical addition showed there was no significant difference in the
results obtained from each individually.
Influent Waste (Trickling Filter Effluent)
Data presented in Table 1, Part A, show the characteristics of the
influent waste to the aerators. As can be seen from the data, a con-
siderable variation in wastewater quality was observed. The relatively
low BOD was not unusual considering the removal that occurred during
high rate trickling filtration prior to activated sludge treatment.
Plant operating records show the BOD of the raw waste normally varies
between 300 and 600 mg/1 with an average of about 400 mg/1. The values
observed for nitrogen and phosphorus are typical of domestic waste-
water. The high percentage of orthophosphate (approximately 82%) is
evidently a result of the biological treatment in the trickling filter.
Eberhardt and Nesbitt (9) indicated that approximately 52% of the phos-
phate in the primary clarifier effluent from the University Treatment
Plant was in the ortho form. Only a trace of aluminum was observed in
the trickling filter effluent during the brief control period and the
sulfate concentration was sufficiently low so as not to create a prob-
lem.
20
-------
Table 1. Summary Data from Aluminum Sulfate Studies Phase 1
A. Trickling Filter Effluent Characteristics
Parameter
Suspended Solids
Total
Volatile
Unfiltered BOD5
Unfiltered COD
No
Chemical
Added
88±37
62±20
60±32
194+52
Chemical Added
Run Code
I
145±26
120+30
61±19
257±78
II
124±26
97±23
75115
216149
III
134+30
100131
87+20
244+77
IV
134+21
99H7
74124
227+47
V
138114
105116
68115
2491135
Phosphorus - P
Filtered
Ortho
Total
Unfiltered
Ortho
Total
Nitrogen - N
Ammonia
Oxidized
7.611.8 5.71.9 6.311.1 6.410.6 6.910.8 7.710.9
9.412.7 6.51.6 7.911.5 7.611.7 7.711.5 8.411.0
7.911.9 8.311.8 9.2+2.1 10.4+1.4 10.1+1.6 10.811.9
10.111.9 9.912.2 10.412.5 12.412.7 13.412.0 13.211.2
16.115.6 20.2+5.0 21.513.7 19.513.1 19.513.2 19.6+3.4
4.4+3.8 0.7+0.3 0.610.2 0.510.2 0.5+0.2 0.610.3
T. Kjeld.
Aluminum - AH
Sulfate - SO,
pH (units)
21.7 33.2
110.5 15.2
Tr. 0.101.14
2813 2613
7.351.30
32.5
16.9
0.151.07
2517
7.401.10
33.2
17.1
0. 151. 07
2414
7.401.10
31.5
14.5
0.151.07
2413
7.401.10
34.0
14.2
0.20
2511
7.451.20
Continued
Values shown are averages for all data included in run plus or minus
one standard deviation where more than one determination was made.
All values are mg/Jl except as noted.
21
-------
Table 1 (Continued)
B. General Performance and Effluent Quality
No
Chemical
Chemical Added
Run Code
Parameter
Suspended Solids
Total
Volatile
Unfiltered BOD5
mg/A
% Removal
Unfiltered COD
mg/A
% Removal
Nitrogen - N
Ammonia
Oxidized
T. Kjeld.
Aluminum - A£
Sulfate - SO,
pH
Chemical Dose
M2(S04)3'14 H
Added
31±16
20±9
18±8
66.6±9.8
94±45
57.5±15
7.8±4.6
12.917.8
14.6±5.8
Tr.
28±3
2°
I
16±4.5
15±2.6
613
88.8±6.4
43±17
83.0±7.8
19.3±4.7
1.611.6
23.1+6.4
Tr.
167124
6.70120
311+37
II
30114
22116
814
88.714.8
33117
83.119.9
20.313.5
1.010.3
26.316.1
0.151.10
139118
6.851.10
259161
III
19H1
1115
612
92.712.3
31115
86.3+9.0
17.513.0
1.5+1.1
21.0+3.8
0.231.12
131132
7.101.15
160135
IV
23H1
1418
612
89.315.9
36116
83.1+8.1
18.513.0
1.510.7
22.514.8
Tr.
77112
7.201.10
128120
V
25H2
14+7
712
88.913.6
54121
74.4118.1
19.513.1
1.710.9
25.4+6.3
0.20
73115
7.3010.10
108120
Continued
22
-------
Table 1 (Continued)
C. Final Effluent - Phosphorus Removal Performance
No
Chemical
Parameter Added I
Chemical Added
Run Code
II III IV V
Phosphorus - P
Filtered
Ortho
% Removal
Total
% Removal
Unfiltered
Ortho
% Removal
Total
% Removal
AA/Pfilt, ortho
Ratio
8.0±2.0 0.05±.04 0.161.16 0.28±.21 0.80±.34 1.57±.46
99.1±0.9 97.412.7 94.6+3.3 88.814.8 77.219.6
9.212.3 0.091.05 0.181.20 0.271.23 1.021.36 1.971.74
98.410.9 97.013,2 96.413.0 86.515.8 78.217.7
8.211.9 0.401.17 1.071.40 1.171.71 1.921.92 3.913.0
95.212.3 88.015.6 85.917,1 79.9111.2 70.919.9
9.7H.9 0.541.25 1.34+.62 1.481.76 2.4+1.2 4.714.2
94.1+2.7 86.5+7.2 87.216.7 82.619.4 78.217.6
4.73/1 3.01/1 2.04/1 1.55/1 1.17/1
+0.84 +0.23 10.19 10.10 10.11
Key:
No chemical added
period from December 2, 1968 to January 27, 1969,
no chemical addition
I - Alum addition with Al/Pfilt orthoRatio 1 3.50/1
II - Alum addition with 2.50/1 £ AJl/P £ 3.49/1
III - Alum addition with 1.75/1 £ A&/P _< 2.49/1
IV - Alum addition with 1.35/1 <. A£/P <_ 1.74/1
V - Alum addition with 0.95/1 <_ A&/P <_ 1.34/1
23
-------
General Effluent Quality
Data presented in Table 1, Part B, show that addition of alum defi-
nitely enhanced removal of BOD and COD over that observed during the
limited period of no chemical addition as indicated by the increase
in BOD removals from approximately 67% to 89 - 93% and of COD removals
from approximately 58% to 74 - 86% depending on the Al/P ratio em-
ployed. Allyl thiourea was used in all BOD determinations to inhibit
nitrification, hence, results shown reflect only carbonaceous demand.
The loading factors (LF) during the period in which alum addition was
used were as follows:
LF (Ib BOD5A/lb MLVS/day)
Average ± 1 std. dev. Range
Tank No. 1 0.517 ± 0.192 0.104 to 0.975
Tank No. 2 0.474 ± 0.194 0.119 to 0.932
From these results it can be seen the loadings on the activated sludge
units were generally in the normal range for so called conventional
activated sludge operation.
The BOD removal characteristics (Sr) for the same period were as fol-
lows:
Sr (Ib BOD5R/lb MLVS/day)
Average ± 1 std. dev. Range
Tank No. 1 0.462 ± 0.196 0.086 to 0.923
Tank No. 2 0.428 ± 0.200 0.087 to 0.983
The difference in the loading factors and removal rates between the two
tanks were not significant at the 5% level.
Figure 4 shows the relationship between BOD loading and BOD removal for
each tank during the alum studies. Regression analysis of the data
yielded the following equations for the lines of best fit:
Tank No. 1 Ib BOD5R = 0.94 Ib BOD - 0.02 Ib MLVS (1)
Tank No. 2 Ib BOD5R = 0.99 Ib BOD5A - 0.04 Ib MLVS (2)
The correlation coefficients were found to be highly significant sta-
tistically.
24
-------
1.000
0.800
g 0.600
UJ
0.4OO
0.200
a TANK No. 2
R 0.962
o TANK No. I
R 0.990
0.200 0.400 0.600 0.800 1.000
Ib. B009 APPLIED/Ib. MLVS
Figure 4. BOD Removal as a Function of BOD Loading Phase I Aluminum
Sulfate Studies
25
-------
Because hydraulic overloading of the final clarifiers during peak
flow periods during the day results in washout of solids from the
clarifiers, the effect of alum addition on suspended solids removal
was impossible to assess during Phase I. Indications were that im-
proved removal of suspended solids could be realized if hydraulic
overloading could be eliminated. Data presented later with the
results of Phase II support this observation.
The nitrogen data presented in Table 1, Part B, indicate that some
inhibition of nitrification occurred as a result of alum addition.
The results during the control period showed rather highly nitrified
effluents as could be expected from a two stage plant such as this.
However, since a major portion of the control period occurred during
a time when most of the students were away from the University and
consequently, the treatment plant was underloaded, it is impossible
to tell from these data how much of the observed change was due to
chemical addition and how much to change in organic loadings. Data
obtained during this period were too limited to permit comparison of
organic loading rates.
Aluminum values in the effluent remained quite low, never exceeding
0.2 mg/1 as an average value. Some "carry-through" of aluminum was
indicated by the increase in aluminum in the trickling filter effluent
after chemical addition was begun. Waste sludge from the aerators was
returned to the primary clarifiers and hence aluminum may have carried
through to the trickling filters.
Increased sulfate concentrations in the effluent were observed with
alum addition as was expected. The observed changes generally follow
the changes in alum dosage as indicated in Table 1, Part B. In no
instance did the effluent sulfate concentration reach the recommended
permissible limit for receiving waters intended for use as public
water supplies of 250 mg SO,/I but it did exceed the desirable criteria
of < 50 mg/1 (48). 4
The pH in the system varied from 6.7 at Al/P (filt. ortho) ratios over
3.50/1 to 7.3 at the lowest Al/P ratios. Later discussion of pH values
observed should be noted as it affects the results reported above.
Phosphorus Removal
Filtered phosphorus removals were excellent at high Al/P ratios as
shown in Table 1, Part C. Filtered effluent phosphorus concentrations
increase sharply at ratios less than 2/1. Figure 5 shows the effect
of varying Al/P ratios on the filtered effluent total phosphorus con-
centration. These results indicate an Al/P (filt. ortho) ratio (based
on filtered orthophosphate) of approximately 2.25/1 is necessary in
the particular system studied to achieve filtered effluent total phos-
phorus concentrations of approximately 0.3 mg P/l. Comparable values
for ratios based on unfiltered total phosphorus [Al/P (unfilt. total)]
and filtered total phosphorus [Al/P (filt. total)] are approximately
1.8/1 and 2.0/1 respectively.
26
-------
E
i 2.50
CO
a:
o
o
Q.
2.00
£ 1.50
o
1.00
UJ
o O.5O
ui
K
EFFP.-..79AL/Pfj|torfho+4.07
R 0.864 0.50£ AL/P < 2.25
EFF P « -0.04 AL/PSI4 ^ +0.34
fill, ortho
R » 0.277 2.26 < AL/P < 6.OO
1.00 2.00 3.00 4.00 5.OO
AL/PfMt. ortho ^.GHT RATIO
6.0O
Figure 5. Effluent Phosphorus Concentration as a Function of Al/P
Phase I Aluminum Sulfate Studies
Weight Ratio
' ortho
-------
Unfiltered effluent phosphorus removals were significantly lower at
all Al/P ratios indicating that considerable amounts of phosphorus
were escaping in the effluent suspended solids. Most of this was un-
doubtably precipitated phosphorus which was either entrapped in the
chemical-biological floe or chemical floe which was too small to
settle under the hydraulic conditions experienced.
On May 12, 1969 the points of alum addition were changed to the in-
fluent end of each aeration tank. The effluent data collected during
this period indicated some decrease in BOD and COD removals (81% and
63% removal respectively) when compared with the results obtained when
the chemical was added at the effluent end (Table 1). There was an
even more pronounced decrease in removal of unfiltered phosphorus (50-
60% removal). Filtered phosphorus removals were also somewhat lower
(82-88% removal). Insufficient data were available from this portion
of the study to permit comparison of results by the method used to
develop Table 1 and, therefore, were not included in Table 1. Visual
observation of effluents during this period showed them to be cloudy
or "milky" and certainly not as clear as when the alum was added to
the effluent end of the tanks. This same observation has been made
by other workers who have added the precipitant to the influent end
of the aerator (2, 52).
The Phase I alum studies were terminated on May 25, 1969.
SODIUM ALUMINATE STUDIES
General Operation
Addition of sodium aluminate to both aeration tanks was started on
May 28, 1969 following thorough cleaning of the chemical storage tank
and recalibration of the chemical feed pumps. Initially, addition of
aluminate was made into the effluent channels from the aeration tanks
at the same points used for alum addition. After a few days operation
it became visibly apparent that the effluent was more cloudy than had
been observed when alum was added at the same point. Since the amount
of aluminate which must be fed per day to achieve a given Al/P ratio
is only about one-third that required when feeding alum, it was decided
to move the points of chemical addition to the influent end of the
tanks to see if increased mixing time would improve effluent quality.
Several more days of operation under these conditions did not result
in a notable improvement in the effluent, in fact, it deteriorated
even further. The points of addition were then moved to a point two
feet from the effluent end of the aeration tanks directly over the
diffusers. Improvement in the visual appearance of the effluent was
noted within two days so it was decided to continue to operate with
chemical addition at this point.
Results obtained from the sodium aluminate runs are presented in Table
2. Data from the initial period prior to selection of the final point
of chemical addition were excluded since they were atypical of what can
be expected from normal operation with chemical addition at the proper
point.
28
-------
Table 2. Summary Data from Sodium Aluminate Studies Phase I
A. Trickling Filter Effluent Characteristics
Parameter
Phosphorus - P
Filtered
Ortho
Total
Unfiltered
Ortho
Total
Nitrogen - N
Ammonia
Oxidized
T. Kjeld.
Aluminum - A.S/3
Alkalinity -
IA
Run Code
IIA IIIA
IVA
VA
Suspended Solids
Total
Volatile
Unfiltered BOD5
Unfiltered COD
81±15
58±9
65±17
190±100
95±18
69±13
61±21
207+99
94±19
69±11
50±26
160±26
84±15
59±12
54±10
180±86
81±16
58±12
51±8
213±101
4.9±0.7 5.3±0.5 5.4±1.2 5.5+1.5 5.3±1.4
5.0±1.0 5.5±0.6 6.9+0.8 6.8±0.9 6.5±0.9
6.1±1.5 6.5±1.1 6.2±1.9 6.6±1.8 6
6.4+1.6 6.8±1.0 8.6±2.3 10.5±1.7 9.9±1.7
3.6±2.1 3.7±2.6 2.0±2.1 4.9±2.9 3.8±2.9
5.0±4.7 5.5±4.6 2.7±1.3 4.7±2.8 5.0±3.9
11.3±6.7 12.8±5.8 12.6±6.1 14.5±2.7 15.0±3.1
0.0
3 156+37
pH 7.20±.10 7.20±.05 7.15+.15 7.25±.05 7.25+.05
Continued
Values shown are averages for all data included in run plus or
minus one standard deviation where more than one determination
was made. All values are mg/& except as noted.
Results shown are for all data and are not separated on the basis
of A&/P ratios since too few data were available.
-------
Table 2 (Continued)
B. General Performance and Effluent Quality
Run Code
a
Parameter
Suspended Solids
Total
Volatile
IA
17110
11±5
IIA
23±10
11±6
IIIA
25±7
14±3
IVA
18±9
11±5
VA
20±3
12±3
Unfiltered BOD5
mg/H
% Removal
Unfiltered COD
mg/Jl
% Removal
Nitrogen
Ammonia
Oxidized
T. KJeld.
Aluminum - AJl
Alkalinity - CaC03b
PH
Chemical Dose
7±2 814
86.0±1.5 84.1±7
7±4 10±2 11±2
81.2±6.1 82.6±3.1 80.7±1.0
25±24 46±13 36±22 28±19 45122
84.5±14 76.9±10.2 78.7±10.4 79.4±6 78.1±10.6
0.5±0.6 0.2±0.3 0.210.5 0.3+0.4 0.2+0.3
9.316.3 12.5±4.1 13.716.5 13.319.9 13.1+4.1
2.912.6 6.015.2 4.514.7 1.612.4 1.912.2
0.410.1
10319
7.651.25 7.501.20 7.25+0.15 7.451.05 7.45+.20
77123 4619 4417 3015 22+3
Continued
30
-------
Table 2 (Continued)
C. Final Effluent - Phosphorus Removal Performance
Parameter3
Phosphorus - P
Filtered
Ortho
% Removal
Total
% Removal
Unfiltered
Ortho
% Removal
Total
% Removal
AJl/P-.^ , Ratio
tilt, ortho
IA
0.121.
97.4±1
0.12+.
97.5±1
0.321.
96.812
0.431.
93.615
10
.5
07
.1
31
.3
40
.9
6.23/1
12.37
0.
93
0.
93
0.
88
1.
80
2.
±0
HA
31±.
.9±4
34±.
.4+4
771.
.116
17±.
Run Code
IIIA
16
.2
17
.3
32
.7
31
.2±10.6
80/1
.46
0.51±.
92.614
0.571.
91.913
1.14+0
81.117
1.53+0
81.717
2.04/1
10.23
38
.8
25
.0
.56
.3
.61
.9
0.
87
0.
86
1.
78
2
79
1.
10
IVA
851.50
.518.0
89±.47
.8±8.0
40+. 53
.6+8.4
. 1±. 4
.1±5.7
47/1
.13
1.
82
1.
82
1.
65
3
68
1.
±0
VA
371.43
.517.1
371.45
.518.0
99±1.20
.1+25.9
.1+2.0
.6+19.3
19/1
.05
Run Code Key:
IA - Sodium aluminate addition with A&/P...... ., Ratio > 3.50/1
tilt, ortho
IIA - Sodium aluminate addition with 2.50/1 £ AJl/P >_ 3.49/1
IIIA - Sodium aluminate addition with 1.75/1 £ A&/P ^> 2.49/1
IVA - Sodium aluminate addition with 1.35/1 £ A&/P >_ 1.75/1
VA - Sodium aluminate addition with 0.95/1 £ Afc/P >^ 1.34/1
-------
Influent Waste (Trickling Filter Effluent)
Comparison of the characteristics of the aerator influent with those
observed during the alum studies (Tables 2A and 1A) show some dif-
ferences. It should be noted that the sodium aluminate runs were con-
ducted during the Summer term at the University when the student popu-
lation on campus is only about 40% as large as it is during the Fall,
Winter and Spring terms. This population reduction most likely
accounts for the lower values observed for all of the parameters
measured. The most noticable differences were those observed in the
relative amounts of nitrogen in the ammonia, oxidized and total
Kjeldal forms. The lower flows and higher temperatures apparently
resulted in appreciable oxidation of ammonia in the trickling filter.
It should be pointed out that considerable variability in results was
observed and relatively few data were obtained in each of the Al/F
ratio ranges shown.
General Effluent Quality
Data presented in Table 2, Part B, show that although BOD and COD
removals obtained during aluminate addition were higher than those
experienced during the period prior to alum addition, they were not
as high as those observed during alum addition. The loading factors
during the period in which sodium aluminate addition was used were as
follows:
LF (Ib BOiJ5A/lb MLVS/day)
Average ± 1 std. dev. Range
Tank No. 1 0.400 ± 0.161 0.230 to 0.708
Tank No. 2 0.408 ± 0.133 0.229 to 0.653
While these loadings are somewhat lower than those experienced during
the alum runs they were still in the normal range for conventional
activated sludge. The BOD removal characteristics for this same period
were as follows:
Sr (Ib BOD5R/lb MLVS/day)
Average ± 1 std. dev,
Tank No. 1 0.333 ± 0.146 0.172 to 0.632
Tank No. 2 0.358 ± 0.107 0.195 to 0.544
Figure 6 shows the relationship between BOD loading and BOD removal for
each tank during the sodium aluminate runs. Regression analysis of the
data yielded the following equations for the lines of best fit:
32
-------
I.OOO
w 0.800
£
o 0.600
u
0.400
UJ
n
O
O
CD
0.200
o TANK No. I
R- 0.994
TANK No. 2
R 0.992
_L
0.200 0.400 0.600 0.800
Ib. BOD9 APPLIED/lb. MLVS
1.000
Figure 6. BOD Removal as a Function of BOD Loading Phase I
Sodium Aluminate Studies
-------
Tank No. 1 Ib BOD5R = 0.90 Ib BOD5A - 0.03 Ib MLVS (3)
Tank No. 2 Ib BOD5R = 0.85 Ib BOD5A - 0.01 Ib MLVS (4)
The correlation coefficients were found to be highly significant sta-
tistically.
The differences between the constants for the BOD loading and removal
regression equations observed for the alum and sodium aluminate studies
were significant at the 1% level. It is impossible to tell from the
data whether the observed lower BOD removal rate and overall percentage
removal for the sodium aluminate runs were due to the lower applied
loading during the period or to the chemical used.
Although effluent suspended solids in the composite samples were gen-
erally lower than they were during alum addition, this reduction was
probably due more to the occurrence of lower flows during this period
than it was to the different chemical. Average flow through each tank
was approximately 0.98 MGD during the alum runs whereas it was only
about 0.66 MGD during the sodium aluminate runs.
Occasional visual inspection of settling tank effluents at various times
during the day during the alum runs had shown that when peak flows
occurred, periodic washout of solids resulted and at other times very
few solids were being lost. Similar inspections during the aluminate
runs when flows throughout the day were lower indicated a fairly uniform
loss of solids during the day so it is impossible from the data to de-
termine the relative effects of the two different chemicals on effluent
suspended solids.
The nitrogen data presented in Table 2, Part B, show that significant
nitrification occurred during chemical addition and hence no apparent
inhibition from chemical addition was evident. However, these data do
not permit a direct comparison with those obtained during the alum runs
since other conditions were not the same and no control data for the
aluminate runs are available. Temperatures and pH values were both
higher during the aluminate runs than during the original control period
and alum runs resulting in conditions which tend to optimize nitrifica-
tion. Additionally, the aluminate runs were conducted during the Summer
term at the University and because of a lower student population, the
BOD loading on the aeration tanks was somewhat lower which also tends
to favor nitrification. The data from this period are more nearly com-
parable to those obtained during the original period of no chemical
addition in this respect.
Aluminum values were generally higher than those observed during the
alum runs. Because of analytical problems with the aluminum test, it
is the writer's opinion that residual soluble aluminum values for both
the alum and aluminate runs were actually lower than those reported
herein. Investigations which were conducted as part of the digester
studies indicated some of the precipitated aluminum phosphate passes
through the filter and is dissolved during the aluminum test and hence
34
-------
is erroneously reported as "soluble" aluminum (26). This could explain
the higher aluminum values observed during the aluminate runs since
there was other evidence of greater "carry-through" of precipiated
aluminum phosphate as indicated by the phosphorus results. The higher
pH may also partially account for the higher Al*"1"* values observed
since the precipitation reaction is not as efficient at the higher pH.
Sulfate data were not collected during this part of the study since
sufficient background data on effluent sulfate concentrations were
obtained during the original period of no chemical addition and no
sulfate was being added to the system during the aluminate runs.
The effluent pH values generally ranged from 7.45 to 7.65 depending on
the Al/P ratio employed. There is no apparent explanation for the low
mean value of 7.25 experienced during Run Code III A. Individual values
ranged from 7.00 to 7.85 throughout the aluminate runs and means shown
were calculated from daily data correlated with Al/P ratio on a given
day.
Phosphorus Removal
As was experienced during the alum runs, filtered phosphorus removals
were very good at high Al/P ratios but they were not as good as those
obtained with alum. Unfiltered phosphorus removals compared even less
favorably with the alum results. This reduction apparently was due to
a "carry-through" of very small particles of precipitated "aluminum
phosphate" which did not occur with alum addition except when the alum
was added to the influent end of the aeration tank. This "carry-
through" was apparent visibly in the effluent as a cloudiness and on
no occasion was the effluent as clear during the aluminate runs as was
observed during the alum runs. Changing the point of chemical addition
did not completely eliminate this "carry-through" as it had with alum.
Figure 7 shows the effect of a varying Al/P ratios on the filtered
effluent total phosphorus concentration for aluminate additions. At
an Al/P (filt. ortho) ratio of 2.25/1, filtered effluent total phos-
phorus would be about 0.5 mg P/l which is about 70% higher than would
be expected using alum at the same Al/P ratio. The sodium aluminate
studies were terminated on August 12, 1969 thus completing Phase I of
the study.
ACTIVATED SLUDGE STUDIES
Mixed Liquor
Table 3 presents summary data from the activated sludge analyses which
were performed during Phase I. The volatile solids data presented have
not been corrected for apparent volatile solids production due to for-
mation of aluminum hydroxy-phosphate compounds which volatilize during
the standard volatile suspended solids analysis procedure. The sludge
volume index data for Tank No. 1 show there was no decrease in SVI as
a result of alum addition which was contrary to what was expected.
35
-------
cr*
6
i
(A
D
o:
o
o
I
o.
<
O
til
D
b.
u.
UJ
O
UJ
o:
2.50h
2.00
1.50
I.OOh
0.50
EFF P «-l.02 AL/P RATIO + 2.82
filt. ortho
R« 0.860 0.501 AL/P < 2.25
EFF P - -0.06 AL/P..14 .. RATIO -f 0.54
filt. ortho
R« 0.344 2.26 < AL/P <.6.00
1.00
2.00
3.00
AL/P
filt. ortho
4.00 5.00
WEIGHT RATIO
6.00
Figure 7. Effluent Phosphorus Concentration as a Function of Al/P... Weight Ratio Phase I
Sodium Aluminate Studies fc< ° °
-------
Table 3. Summary of Activated and Waste Sludge Data Phase I
Run
Code
Mixed Liquor
mg/A
Total Volatile
Waste Sludge
SVI
Mohlman
Ib/day
gal/day
Total
Volatile
Solids MLVS
Age P
Days %
pH
Temp
°C
C 1680+390 1260±280 68±13
A 24001570 1360±280 68±10
SA 1980+420 1010±210 26±6
A. Tank No. 1
12,300±5,540 504±120 321±84 1.33±.55 2.12±.74 7.10±.10
19,540+3,650 1443±573 775±326 1.29+.51 7.69+1.93 7.15+.25 17.8+2.4
12,59012,740 554+251 2511110 2.48+1.19 9.53+1.19 7.251.25 23.51.9
C 14901350 11901230 83+9
A 24501450 1390+250 60111
SA 2090+480 1050+210 2318
B. Tank No. 2
10,38012,350 1560+400
6,38012,270 639+298
8371219 1.371.54 6.76H.78 7.151.35 17.812.4
302+137 2.58+1.20 9.9213.30 7.30+.25 23.51.9
Run Code:
C - Period of no chemical addition.
A - Alum addition to both tanks.
SA - Sodium aluminate addition to both tanks.
Blanks indicate no data were collected during the period of no chemical addition.
-------
Initial settling rates were not measured but visual observations indi-
cated the chemical-biological sludge resulting from both alum and
aluminate addition settled much more rapidly than is usual for bio-
logical sludge. The data from the period of no chemical addition for
Tank No. 2 are too limited in number to permit any conclusions re-
garding the apparent reduction in SVI shown. The reduction in SVI
as a result of sodium aluminate addition was very dramatic. SVI's
during aluminate addition were approximately 30% as high as those ob-
served during alum addition. It should be noted that throughout the
Phase I operation, the SVI's for each of the tanks were lower than
those normally expected in a conventional activated sludge plant
treating domestic wastes. Earth and Ettinger (2) report sludge
density indexes (SDI) for pilot plant operation using conventional
activated sludge and sodium aluminate addition to the aerator of
0.5 and 1.3 respectively. Comparable values for SDI from Phase I
results reported above are 1.5 (period of no chemical addition), 1.6
(alum addition) and 4.1 (sodium aluminate addition).
The solids age during the period of no chemical addition and the period
of alum addition averaged 1.3 to 1.5 days which is much lower than that
normally employed for conventional activated sludge operation. The
solids age during the sodium aluminate runs was appreciably higher
(approximately 2.5 days) but still well below that employed in conven-
tional activated sludge operation. The higher solids age during the
aluminate studies was a result of the lower flows and loadings experi-
enced during this period.
Sludge Production
Sludge production is a very important consideration in wastewater
treatment because of the problems of solids handling and ultimate dis-
posal. Insufficient data were collected during the period of no
chemical addition prior to the alum runs to permit a significant
evaluation of the waste sludge production from the units under con-
ventional operation. The data presented in Table 3 indicate that a
significantly greater volume and weight of waste sludge was produced
from the use of alum than from the use of sodium aluminate. However,
if the data are correlated with flow (Table 4), these results indicate
that, although waste sludge volumes are essentially the same regardless
of which precipitant was used, the weight of total solids produced from
the use of sodium aluminate is only about 59% of that produced from the
use of alum and only about 40% as great for volatile solids. These
values may be atypical since organic loadings and influent phosphorus
concentrations were lower during the period of sodium aluminate addi-
tion than they were during the alum studies. This reduction would
also reduce the amount of solids produced, hence direct comparison
between the sludge production figures for the two precipitants is dif-
ficult.
38
-------
Table 4. Comparative Waste Sludge Production Phase I
Alum Sodium Aluminate
Tank No. 1 Tank No. 2 Tank No. 1 Tank No. 2
gal/day/M 20,000 10,640 19,160 10,420
Ib total solids/day/MG 1,470 1,600 837 965
Ib volatile solids/day/MG 793 957 380 457
Mixed Liquor Phosphorus
The percentage of phosphorus in the activated sludge during the period
of no chemical addition (2.12%) is typical of values which have been
reported in the literature. Jenkins and Menar (14) reported a weighted
mean value of 2.6% for the P content of activated sludge operating over
a range of substrate loadings. The percentage phosphorus increased
greatly as a result of chemical addition as shown by the data in Table
3. The observed higher percentage during the sodium aluminate runs
seems reasonable since the amount of volatile solids produced was much
less while the amount of phosphorus removed was nearly the same as with
alum addition.
GENERAL DISCUSSION
Table 5 is a summary comparison of results achieved from the use of
each of the two chemicals studied using data obtained with Al/P (filt.
ortho) ratios in the range of 1.75/1 to 2.50/1. These data show the
use of aluminum sulfate resulted in higher removals of BOD, COD and
phosphorus than were realized with sodium aluminate. Because of this,
alum was selected for use during Phase II of the project.
It should be noted that the pH values observed during both the alum and
the sodium aluminate studies were well above the optimum reported in
the literature for phosphorus removal using aluminum as the precipitant.
Therefore, lower effluent filtered phosphorus concentrations would be
achieved if the pH of the system were lowered in accordance with the
findings of Eberhardt and Nesbitt (9). This should reduce the Al/P
ratio required to attain a given effluent phosphorus concentration and
depending on the economics of pH adjustment, may result in lower total
chemical costs than those experienced without pH control. Full plant
scale evaluation of pH control was not a part of this study so this
question must be resolved at another time. However, some data on
additional pH observations made during Phase II will be discussed later
and will reflect on this.
It should not be concluded from the data shown that alum will be superior
to sodium aluminate with all wastewaters. The chemistry of the sodium
aluminate reaction favors low pH, low alkalinity systems which are not
39
-------
Table 5. Comparison of Effluent Characteristics and
Solids Production Using Alum and Sodium
Aluminate Phase Ia
Chemical
Parameter
BOD- - % Removal
COD - % Removal
Phosphorus - % Removal
Filtered
Ortho
Total
Unfiltered
Ortho
Total
Nitrogen - mgN/£
Ammonia
Oxidized
T. Kjeldal
PH
Aluminum - mgA/&
Sulfate - mgSO,/&
Alum Sodium Aluminate
92.7±2.3
86.3±9.0
94.6±3.3
96.4±3.0
85.9±7.1
87.2±6.7
17.5±3.0
1.5±1.1
21.0±3.8
7.10±.15
0.23±.12
131±32
81.2±6.1
78.7±10.4
92.6±4.8
91.9±3.0
81.1±7.3
81.7±7.9
0.2±0.5
13.7±6.5
4.5±4.7
7.25±.15
0.40 ±0.1
_
A1/Pfilt. orthoRatio
Weight 2.04±.19 2.04/1±.23
Waste Sludge Production0
gal/day/MG 20,000 19,160
Ib total solids/day/MG 1,470 837
Ib volatile solids/day/MG 793 380
Data for comparison are taken from Table 1, Code III, Table
2, Code IIA, and Table 6.
Chemical addition at effluent end of aeration tank only.
°Data from Tank 1 only.
40
-------
typical of Penn State University wastewater. Limited alkalinity data
which were collected during the latter part of the sodium aluminate
studies showed an average alkalinity in the trickling filter effluent
of 156 mg CaC03/l. Average pH values for the trickling filter effluent
were 7.4 for the alum runs and 7.2 for the aluminate runs, both well
above the desired pH for phosphorus precipitation with aluminum. Since
it is the pH after chemical addition which is important, it should be
noted that alum lowers the pH whereas sodium aluminate raises it. The
higher effluent pH's observed with use of sodium aluminate are most
likely the primary reason for the lower phosphorus removals experienced
with this chemical in this study.
Brenner (4) also discusses the relative advantages of using sodium
aluminate and alum when both chemicals were added at the head end of
the aeration tank. His observations that poorer fine floe capture
occurs as the pH approaches 6 is more likely due to conditions created
by adding the chemical at the head end of the aeration tank than it is
strictly to pH although pH is also important. Zenz and Pivnicka (52)
also observed a carry-over of fine floe described by them as "milky
white" and observed that the amount of floe increased with increasing
alum dosages which would decrease the pH. Since they also added the
precipitant at the head end of the tank, it is likely that this floe
carry-over could have been reduced or prevented by changing the point
of chemical addition to the effluent end of the tank. Recht and
Ghassemi (29) have shown that the reaction between aluminum and ortho-
phosphate is very rapid and may be instantaneous. Therefore, adequate
reaction time for phosphate removal should exist when adding the chemi-
cal into the effluent line from the aeration tank.
The reasons for the "carry-through": of phosphorus which was experienced
during the sodium aluminate studies and with alum when the chemical was
added at points other than at the effluent end of the aeration tank and
as noted by others are still not completely understood. It is believed
to be a function of: 1) pH, 2) solids age, and 3) point of chemical
addition. Since it does result in significantly higher effluent phos-
phorus concentrations than would otherwise be expected, additional study
during Phase II was undertaken in an attempt to resolve some of the
questions regarding this phenomenon. These observations will be reported
elsewhere.
As indicated earlier, in the chemical feeding system employed in this
study, liquid chemical was stored in a PVC lined wooden tank and diaphram
chemical feed pumps were used to meter the chemical. Flow to the point
of chemical addition was by gravity through 1/2-inch insulated plastic
lines. No problems with chemical handling were experienced during the
Phase I studies. Sodium aluminate is much more viscous than alum and
could cause handling difficulties during cold weather in some instances.
All of the aluminate runs were conducted during warm weather so it was
impossible to assess cold weather operation. The period of study using
alum began in January and no unusual problems were experienced during
cold weather operation.
41
-------
PHASE II INVESTIGATIONS
PROGRAM OF STUDY
Phase II of the project was conducted to obtain long term operating
data on the use of alum to precipitate phosphorus from domestic waste-
water as a result of the findings of Phase I studies. In addition,
several aspects of the operation which were not investigated during
Phase I were researched briefly during Phase II in an attempt to guide
future research into detailed areas of study not included herein or
yet generally understood.
The plan of operation and plant operational procedures for Phase II
studies were as outlined earlier for Phase I, Aluminum Sulfate Studies,
except as noted subsequently. In contrast to the procedure used during
Phase I, variations in the phosphorus concentration as well as flow
variations were incorporated into the chemical feed rate calculations
during Phase II. It should be noted again that the effect of the
phosphorus variation was very much less than that of flow variation
but that these results are probably atypical.
Sampling schedules and analyses of samples were as reported for Phase
I except that data on alkalinity, calcium, magnesium, and color were
collected routinely whereas data on aluminum and sulfate were collected
only at infrequent intervals.
General Operation
Phase II operation began on August 21, 1969 with the addition of liquid
aluminum sulfate (alum) into the effluent channel from Aeration Tank
No. 1 at the same point used for Phase I studies. Aeration Tank No. 2,
operating in parallel with Tank No. 1, was maintained as the untreated
control.
Operation over a one year period was accomplished in'order to furnish
data which would reflect operating capabilities of the chemical-bio-
logical process under varying operating conditions including those
resulting from changing loads, student populations, weather, and pro-
cess modification or adaptation as a result of chemical addition over
extended periods.
The Pennsylvania Sanitary Water Board (now Pennsylvania Environmental
Quality Board) has established water quality criteria for Spring Creek
(the receiving stream for the University Plant) which sets a limit of
0.4 mg P04/1 (0.13 mg P/l) total soluble phosphate (47). For the pur-
poses of the Phase II studies it was desired to produce an effluent
which contained not more than 0.3 mg P/l total phosphorus in a filtered
sample. Allowing for dilution, this then should meet the water quality
standards which have been established. Subsequent to the study, the
Board now uses the 0.4 mg P04/1 total soluble phosphate as an effluent
standard for plants discharging to the receiving stream.
43
-------
Daily flow variation during Phase II was extreme, varying from a low
of 0.33 MG through each of the two tanks to a high of 1.37 MG. Design
average flow is 1.04 MGD, based on a surface settling rate of 800 gal-
lons per day per square foot for the final settlers. Most of the low
flow days occurred during term breaks when students were absent from
campus although Sunday flows were often less than design average even
when students were on campus. The problem of flow variation was par-
ticularly severe during the Spring 1970 term at the University (March
30 - June 13, 1970). These extreme fluctuations in flow made evalu-
ation of the effects of the chemical-biological process on the various
parameters studied difficult and it was impossible to separate out
completely the effects of varying flow in analyzing the results ob-
tained. In an attempt to illustrate the effect of flow data presenta-
tion for Phase II operation has been broken down into three classifi-
cations: 1) data collected when daily flows did not exceed 1.040 MG,
2) data collected when daily flows exceeded 1.041 MG, and 3) all data
collected during Phase II regular operation. The daily flow was
greater than 1.040 MG 43.9% of the time during Phase II operation.
Both median and mean values have been presented so as to reflect bet-
ter the effect of varying flow and the resulting extreme values which
occurred. Considerable variance was observed in the data for most of
the parameters measured but it is believed the results which are pre-
sented herein do reflect the operating capabilities of the system.
Due consideration should be given in the interpretation of results to
the hydraulic overload that occurred on numerous occasions.
FLOWS AT OR BELOW DESIGN AVERAGE
Influent Waste (Trickling Filter Effluent)
Data presented in Table 6, Part A, show the characteristics of the in-
fluent waste to the aerators for those days on which flows did not
exceed the design average flow of 1.04 MGD. The mean flow for this
category was 0.74 MGD, well below design average. Peak rates of flow
did exceed design at different periods of the day on many occasions
but at these relatively low total flows they did not affect the results
observed significantly. The results are generally similar to those
observed during the aluminum sulfate studies of Phase I and are con-
sidered to be typic'l of what would be expected in a plant treating
domestic wastes. Note that approximately 90% of the unfiltered influ-
ent phosphorus is in the ortho form and that virtually 100% of the
filtered influent phosphorus was in the ortho form.
As expected, little oxidized nitrogen was found in the trickling filter
effluent on most occasions. The difference between the median and mean
values and the wide range of values observed suggest the flow and tem-
perature variations which occurred had more influence on nitrification
than on other processes involved.
44
-------
Table 6. Summary Data from Phase II One Year Full Scale
Study for Flows Not Exceeding 1.040 MG Total
Daily Flow Per Tank.
A. Trickling Filter Effluent Characteristics
Parameter3
Suspended Solids
Total
Volatile
Unfiltered BOD5
Unfiltered COD
Phosphorus - P
Filtered
Ortho
Total
Unfiltered
Ortho
Total
Nitrogen - N
Ammonia
Oxidized
T. Kjeld.
Alkalinity - CaC03
Calcium - Ca
Magnesium - Mg
pH - Units
Color - Units
Flow - MGD
Median Mean
98
74
72
152
6
6
9
10
13
0
21
.5
.5
.1
.1
.6
.8
.0
180
7
0
24
27
.60
50
.73
110
85
71
172
6
6
8
10
12
2
21
.7
.3
.9
.0
.8
.1
.0
168
7
0
25
24
.50
45
.74
Std.
Dev.
67
57
34
77
2.0
2.0
2.8
3.8
6.6
2.4
10.1
40
4
6
0.20
7
0.19
No. of
Range Observ.
38
21
15
32
2.6
3.1
2.9
3.6
1.8
0.1
5.8
85
18
11
6.90
30
0.33
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
448
396
211
436
14.2
13.6
21.2
21.6
22.7
9.3
47.5
230
34
31
7.85
50
1.03
71
70
57
63
70
68
70
67
56
49
55
61
49
48
63
32
77
Continued
45
-------
Table 6 (Continued)
B. General Performance and Effluent Quality
Tank lb
Parameter
Suspended Solids
Total
Volatile
Unfiltered BOD5
mg/£
% Removal
Unfiltered COD
mg/fc
% Removal
Nitrogen - N
Ammonia
Oxidized
T. Kjeld.
Alkalinity - CaC03
Calcium - Ca
Magnesium - Mg
pH - Units
Color - Units
Median
12
8
7
90.6
44
79.1
9.5
3.0
10.6
76
28
26
6.80
15
Mean
22
14
9
88.6
55
75.1
11.1
4.2
13.2
80
28
23
6.75
15
Std.
Dev.
37
19
6
5.8
47
15.6
6.4
2.8
7.8
42
3
7
0.50
7
Tank 2C
Median
22
19
12
82.1
58
69.8
7.6
5.0
7.8
120
26
26
7.30
30
Mean
26
20
13
81.4
68
66.7
7.7
6.6
9.7
120
24
23
7.35
30
Std.
Dev.
20
15
4
5.9
50
18.7
5.8
3.7
7.5
35
3
6
0,20
6
Continued
46
-------
Table 6 (Continued)
C. Final Effluent - Phosphorus Removal Performances
Tank
Parametera
Phosphorus - P
Filtered
Ortho
% Removal
Total
% Removal
Unfiltered
Ortho
% Removal
Total
% Removal
AA/Pfilt. orthoRati°
Median
0.
98
0.
98
0.
94
0.
92
10
.8
14
.1
52
.3
79
.8
2.62/1
lb
Mean
0.
97
0.
96
1.
92
1.
28
.5
36
.9
15
.1
41
91.1
2.97/1
Tank 2
Std.
Dev.
0.
3.
0.
4.
2.
6.
2.
6.
1.
46
0
56
5
32
8
62
9
20
Median
6
33
6
35
7
26
7
21
-
.9
.5
.7
.8
.4
.5
.5
.2
-
c
Mean
6.
31.
6.
33.
7.
26.
7.
22.
7
2
7
3
2
1
3
9
Std,
Dev.
1.55
16.1
1.89
14.5
1.87
17.1
2.08
12.7
All units are mg/& except as noted.
Alum addition to effluent end of tank.
°Control unit - no chemical addition.
47
-------
General Effluent Quality
The data presented in Table 6, Part B, permit comparison of the per-
formance of the chemical-biological system (Tank No. 1) with that of
the control unit (Tank No. 2). These data also suggest that even
though total daily flows were below design average, the flow rate
fluctuation throughout the day is sufficient to cause widely varying
effluent characteristics. While other factors can and do influence
treatment plant efficiency, it was felt that flow variation was the
most significant factor accounting for the major portion of the vari-
ability observed in plant performance during this study.
Comparison of the median values for suspended solids from each of the
two tanks indicate that considerably improved capture of suspended
solids can be expected from the chemical-biological system. However,
when the mean values are compared, no such conclusion can be reached.
It was noted earlier that the total solids which must be carried in
the chemical-biological system are much higher than those carried in
a conventional activated sludge system because of the large amount of
nonbiological solids present in the form of precipitated phosphorus.
This greater amount of solids which must be removed in the final set-
tling tanks makes hydraulic loading even more critical in this system
and probably accounts for the relatively small differences observed
between median and mean values for Tank No. 2 compared to those ob-
served for Tank No. 1.
The data also confirm the observation made during Phase I that alum
addition enhances removal of BOD and COD. The effluent BOD's from
the treated unit were only about 58% as high as those from the un-
treated unit if the median values are used for comparison. Compari-
son of mean values also shows a decided improvement in BOD removal
for the chemical-biological system but the difference is not as great
(69% as high). The same conclusions can be drawn from the COD data
although the noted improvement achieved by the chemical-biological
unit is not as great.
The nitrogen data in Table 6, Part B, also show a lower degree of nitri-
fication for the chemical-biological system similar to that which was
observed during the Phase I studies. The difference was not as great
as observed during Phase I which suggests again the importance of flow
and solids age as parameters affecting nitrification in activated sludge
units.
Phosphorus Removal
The effluent phosphorus data presented in Table 6, Part C, also reflect
the variability in results resulting from flow fluctuations even when
all flows are within design average values. As noted earlier, low flows
often occurred on Sundays even when a normal student population was on
campus. On these occasions, no attempt was made to reduce the alum feed
48
-------
rates from those normally used during the week. This resulted in an
overdose of alum in these instances which did not adversely affect
results but was uneconomical in terms of chemical usage.
Filtered effluent phosphorus removals were again excellent. The mean
values reported for both ortho and total phosphorus include data ob-
tained during periods when flows were changing and feed rate adjust-
ments were being made. On several occasions during these periods, the
effluent phosphorus increased to values in excess of 2 mg P/l before
the adjustments were completed. Figure 8 shows the statistical dis-
tribution of effluent total phosphorus concentrations for Tank No. 1.
The shape of the curve particularly at extreme values indicates the
distribution of data is not normal throughout the range of observations
and that results are affected by other factors such as very high or
very low Al/P ratios resulting from unpredicted changes in flow or in-
fluent phosphorus concentration. In any event, the overall percentage
removals of 98 to 99% for orthophosphate and 97 to 98% for total phos-
phorus show that with efficient solids separation either by sedimenta-
tion or filtration, the process is capable of a high degree of phos-
phorus removal at relatively high Al/P ratios, i.e., greater than 2/1.
Later discussion will attempt to identify those operating problems and
procedures which would optimize the economy and reliability of the
process.
The data on unfiltered effluent phosphorus concentrations also show
considerable variation for the reasons given above. Figure 9 presents
the statistical distribution of these data for Tank No. 1. These data
also are not normally distributed throughout the range of observation
so that statistical inferences from the data must consider this. No
attempt was made to normalize the data or to otherwise modify or uti-
lize sophisticated statistical techniques in analyzing the data. The
data on unfiltered effluent phosphorus demonstrate the importance of
effluent suspended solids capture in achieving low residual effluent
phosphorus values
FLOWS IN EXCESS OF DESIGN AVERAGE
Influent Waste (Trickling Filter Effluent)
'Part A, Table 7, presents data on the characteristics of the aerator
influent waste for those days on which flows exceeded the design
average flow of 1.04 MGD. The mean daily flow for this category was
1.23 MG. Peak rates of flow during the day would often go as h,igh as
1.6 MGD for short periods of time.
Data on the various parameters indicate heavier loadings on the plant
along with the increasing flow but no other significant differences
from earlier observations. These data show less variability than
those observed at the lower flows. However, the range of flow observed
was also much less in this category.
49
-------
1.0
Ul
o
o»
E
CO
Q.
to
O
0.1
Ul
U.
IL.
Ul
o
Ul
(T
Ul
TOTAL P(68 ANALYSES)
j i
0.01 O.I I 5 20 40 60 80 95 99
PERCENTAGE OF TIME OBSERVED VALUE < GRAPH VALUE
99.9 99.99
Figure 8. Statistical Distribution of Filtered Effluent Total Phosphorus Tank 1 Flow £ 1.041 MGD
-------
I
o
X
Q.
W
O
Q.
o 1.0-
UJ
Ul
o
UJ
oc
UJ
Z
O.I
TOTAL P(66 ANALYSES)
0.01 0.1 I 5 20 40 60 80 95 99
PERCENTAGE OF TIME OBSERVED VALUE < GRAPH VALUE
99.9 99.99
Figure 9. Statistical Distribution of Unfiltered Effluent Total Phosphorus Tank 1 Flow < 1.040 MGD
-------
Table 7. Summary Data from Phase II One Tear Full Scale
Study for Flows Exceeding 1.041 MG Total Daily
Flow Per Tank
A. Trickling Filter Effluent Characteristics
Parameter
Suspended Solids
Total
Volatile
Unfiltered BOD5
Unfiltered COD
Phosphorus - P
Filtered
Ortho
Total
Unfiltered
Ortho
Total
Nitrogen - N
Ammonia
Oxidized
T. Kjeld.
Alkalinity - CaC03
Calcium - Ca
Magnesium - Mg
pH - Units
Color - Units
Flow - MGD
Median Mean
112
93
91
216
7.4
6.9
9.8
10.8
18.6
0.6
28.0
198
24
22
7.70
50
1.23
117
98
95
223
7.5
6.9
9.8
10.5
18.7
0.7
28.3
199
25
23
7.65
50
1.22
Std.
Dev.
35
32
35
55
1.4
1.3
1.7
2.3
2.0
0.4
7.9
23
3
5
0.15
5
0.07
No. of
Range Observ.
63
48
30
122
5.2
4.3
6.2
5.6
11.9
0.0
14.2
160
20
14
7.20
1.05
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
275
241
222
342
13.9
9.5
13.4
15.7
23.9
2.0
49.0
256
32
35
8.05
1.37
61
60
54
53
59
49
58
53
55
49
55
53
39
38
59
41
64
Continued
52
-------
Table 7 (Continued)
B. General Performance and Effluent Quality
Tank 1L
Tank 2l
Parameter
a
Std. Std.
Median Mean Dev. Median Mean Dev.
Suspended Solids
Total
Volatile
Unfiltered BOD5
mg/fc
% Removal
Unfiltered COD
mg/A
% Removal
Nitrogen - N
Ammonia
Oxidized
T. Kjeld.
Alkalinity - CaC03
Calcium - Ca
Magnesium - Mg
pH - Units
Color - Units
39 105 127 30 30 13
27 62 69 23 23 10
14 22 29 16 15 4
85.8 68.3 48.1 83.3 81.2 9.3
68 124 123 71 75 35
76.2 58.6 42.0 68.0 64.9 16.3
17.4 17.1 2.6 13.1 12.7 3.6
1.5 1.8 1.0 3.5 4.5 2.1
21.0 23.1 7.9 16.5 17.0 5.5
120 123 21 160 158 22
30 29 3 26 26 3
19 22 6 25 24 6
7.10 7.15 0.25 7.50 7.45 0.15
20 25 10 30 30 5
Continued
53
-------
Table 7 (Continued)
C. Final Effluent - Phosphorus Removal Performance
Parameter
Tank 1
Tank 2
Std. Std.
Median Mean Dev. Median Mean Dev.
Phosphorus - P
Filtered
Ortho
% Removal
Total
% Removal
Unfiltered
Ortho
% Removal
Total
% Removal
0.54
94.7
0.65
94.1
2.9
69.4
2.9
73.7
0.80 0.93 7.0 7.2 0.8
9.1.5 10.5 30.5 28.0 16.1
0.80 0.82
92.0 8.9
6.2 6.5 1.2
35.2 33.5 12.4
7.0 8.9 7.9 8.2 1.7
25.8 95.0 22.8 20.1 18.6
7.7 11.5 8.1 8.2 2.1
39.2 68.7 22.8 19.0 15.5
Ai/P
filt. ortho
Ratio 2.28/1 2.18/1 0.61
All units are mg/£ except as noted.
Alum addition to effluent end of tank.
"Control unit - no chemical addition.
54
-------
General Effluent Quality
The data reported in Table 7, Part B, dramatically illustrate the
effect of flow on effluent quality. This effect is most apparent in
the comparison of suspended solids data and in the differences between
median and mean values for solids, BOD and COD. Figure 10 shows the
variation in effluent suspended solids, flow and unfiltered effluent
orthophosphate which can be considered typical of that which was ob-
served almost daily particularly during the Spring Term 1970.
The median values for effluent suspended solids show that although the
solids loss from Tank No. 1 was higher than that from Tank No. 2, it
was not greatly different. However, the significant point is the con-
tradiction of the results observed at the lower flows where comparison
of medians showed significant improvement in removal of suspended
solids as a result of chemical addition. Comparison of means reflects
the extremely high loss of solids from Tank No. 1 with increasing flows
whereas the loss of solids from Tank No. 2 did not increase signifi-
cantly over that observed at the lower flows.
The BOD and COD data for this category show a slight improvement in BOD
and COD removal for the treated system if median values are compared.
The mean values again reflect the influence of high suspended solids
which also increases the effluent BOD and COD and results in an appar-
ent advantage for the untreated system. These data also show that flow
has relatively less influence on the untreated system than on the
treated system when they are compared with the data observed for flows
below design average.
The higher flows and resulting lower solids age observed also effected
a reduction in nitrification in both units as can be seen in the com-
parison of data in Table 7 with that reported in Table 6. The lower
degree of nitrification observed for Tank No. 1 in the past also was
noted in these data.
Phosphorus Removal
The effluent phosphorus data presented in Part C, Table 7, show very
good removal of phosphorus on filtered effluent samples but the re-
movals are significantly less than those observed at the lower flows.
This is due largely to the lower Al/P ratios which resulted on those
occasions when flows were considerably in excess of those used for com-
puting chemical feed rates. It has been shown earlier in Phase I
results that effluent phosphorus concentration rises quite rapidly as
the Al/P (filt. ortho) ratio falls below 2/1. Figure 11 shows the
statistical distribution of the phosphorus data for Tank No. 1 for this
flow category.
The unfiltered effluent phosphorus removals were significantly higher
than those observed from Tank No. 2 but could in no way be considered
typical of chemical-biological process capabilities. The extremely
55
-------
(1570)
IO.O-T-
o
0
en
ON
-T-I3OO
SUSPENDED
SOLIDS
AVERAGE
FLOW
UJ
3-4°° 7-8°°
TIME OF SAMPLING
Figure 10. Daily Variation in Effluent Quality Tank 1 Flow >^ 1.041 MGD
-------
Ui
VJ
I
0)
=>
oe
a.
w
g
Du
U
U.
bJ
O
UI
1.041 MGD
-------
high loss of solids with resulting carry-over of precipitated phos-
phorus simply rendered the whole process ineffective without filtra-
tion of the effluent. Figure 12 shows the statistical distribution of
the unfiltered effluent phosphorus data for this category.
ALL FLOW DATA
General
Table 8 includes all data collected during Phase II operation begin-
ning on August 21, 1969 and terminating on August 20, 1970. During
this period, one occasion occurred where chemical addition was inter-
rupted for about twelve hours when heavy snows prevented needed deliv-
ery of alum. On at least one other occasion a plant operator failed
to make a scheduled feed pump rate change and underdosing occurred for
a period of about three'hours. These are the only known occasions
during the one year run when chemical addition was interrupted or re-
duce'd from scheduled feed rates except for brief periods of less than
one-hour duration to permit maintenance work on pumps and chemical
feed lines. These data reflect results obtained on these occasions
as well as those obtained during periods of flow adjustment when major
changes in student population were occurring. All data included in
Table 10 were used also in the data analyses for Tables 8 and 9.
Influent Waste (Trickling Filter Effluent)
As noted previously, the influent waste is characteristic of domestic
wastewater and presents no unusual problems in treatment. The mean
daily flow for all days on which data were collected was 0.96 MG
which is below the design average flow of 1.040 MG. Prior discussion
has pointed out the variability of flow experienced and the effect of
this variation on results obtained.
The influent wastewater has a moderately high alkalinity and approxi-
mately equal calcium and magnesium content.
It should be pointed out again that approximately 90% of the unfiltered
influent phosphorus and virtually 100% of the filtered influent phos-
phorus was in the ortho form. The phosphorus data for filtered samples
frequently showed higher values for orthophosphate than for total phos-
phorus on both influent and effluent samples. There was no apparent
explanation for this observation which has also been noted by others
(3).
General Effluent Quality
Because of the extremely wide variation in results obtained for most
of the parameters measured as a result of hydraulic overloading, the
median values reported offer a better basis for comparison than do
mean values. Therefore, unless otherwise noted the ensuing discussion
will involve comparison of reported median values.
58
-------
E
I
CO
3
K
i
Q.
CO
O
O.
10.0
(Ji
VD
bJ
_J
It.
la.
Ul
O
Itl
1C
bJ
1.01
0.01
TOTAL P(56 ANALYSES)
O.I I 5 20 40 60 80 95 99 99.9 99.99
PERCENTAGE OF TIME OBSERVED VALUE <. GRAPH VALUE
Figure 12. Statistical Distribution of Unfiltered Effluent Total Phosphorus Tank 1 Flow > 1.041 MGD
-------
Table 8. Summary Data from Phase II One Year Full Scale
Study for All Flow Data
A. Trickling Filter Effluent Characteristics
a
Parameter
Suspended Solids
Total
Volatile
Unfiltered BOD5
Unfiltered COD
Phosphorus - P
Filtered
Ortho
Total
Unfiltered
Ortho
Total
Nitrogen - N
Ammonia
Oxidized
T. Kjeld.
Alkalinity - CaC03
Calcium - Ca
Magnesium - Mg
pH - Units
Color - Units
Flow - MGD
Median Mean
110
85
81
193
7.0
6.5
9.5
10.8
17.6
0.8
26.3
190
24
25
7.60
50
0.98
114
91
83
195
7.1
6.6
9.3
10.3
15.7
1.4
24.7
182
25
24
7.60
45
0.96
Std.
Dev.
55
47
36
73
1.8
3.2
2.4
3.2
5.7
1.8
9.7
36
4
6
0.20
5
0.28
Range
38
21
15
32
2.6
3.1
2.9
3.6
1.8
0.0
5.8
85
18
11
6.90
30
0.33
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
to
448
396
222
436
14.2
13.6
21.2
21.6
23.9
9.3
49.0
256
34
35
8.05
50
1.37
No. of
Observ.
132
130
111
116
129
117
128
120
111
97
110
114
88
87
122
73
141
Continued
60
-------
Table 8 (Continued)
B. General Performance and Effluent Quality
Tank lb
Parameter
Suspended Solids
Total
Volatile
Unfiltered BOD5
mg/&
% Removal
Unfiltered COD
mg/£
% Removal
Nitrogen
Ammonia
Oxidized
T. Kjeld.
Alkalinity - CaC03
Calcium - Ca
Magnesium - Mg
pH - Units
Color - Units
Median
26
18
10
88.1
52
77.6
15.7
2.3
18.2
116
28
25
6.90
20
Mean
59
36
16
77.8
92
66.3
14.1
3.0
18.2
99
28
23
6.95
20
Std.
Dev.
98
55
22
36.6
101
33.3
5.7
2.4
9.2
40
3
6
0.45
10
Tank 2C
Median
25
22
15
83.0
64
69.2
11.5
4.0
14.3
147
26
25
7.40
30
Mean
28
22
14
81.3
72
65.7
10.5
5.5
13.5
139
25
23
7.40
30
Std.
Dev.
17
13
4
7.9
43
17.4
5.3
3.2
7.5
34
3
6
0.20
5
Continued
61
-------
Table 8 (Continued)
C. Final Effluent - Phosphorus Removal Performance
Tank lb
Parameter
Phosphorus - P
Filtered
Ortho
% Removal
Total
% Removal
Unfiltered
Ortho
% Removal
Total
% Removal
Median
0.
97
0.
97
1.
88
1.
88
26
.1
29
.0
30
.2
44
.1
Mean
0.
95
0.
94
3
66
4
67
51
.0
53
.7
.7
.2
.3
.8
Std.
Dev.
0
7
0
7
6
71
8
52
.76
.9
.69
.2
.8
.9
.5
.8
Tank
Median
7.
31.
6.
35.
7.
25.
7.
22.
0
2
5
5
5
4
9
0
2C
Mean
6
30
6
33
7
23
7
20
.9
.0
.6
.4
.7
.3
.7
.8
Std.
Dev.
1.3
16.1
1.6
13.3
1.8
18.0
2.1
14.3
A*/Pfilt. orthoRatio
Alum Dose
2.39/1 2.61/1 1.05
190
178
94
M
All units are mg/& except as noted.
Alum addition to effluent end of tank.
CControl unit - no chemical addition.
62
-------
The data on effluent suspended solids show there was virtually no
difference in suspended solids removal as a result of chemical pre-
cipitation of phosphorus within the activated sludge system. As
expected, the untreated unit has a significantly higher percentage
of volatile solids (88% compared to 69%).
The BOD data presented in Part B, Table 8, show a significant improve-
ment in BOD removal as a result of alum addition. The effluent BOD
from the chemical-biological system was only 67% as high as that from
the control system.
The loading factors (LF) for each of the units for the entire period
of study covered by Phase II were as follows:
LF (Ib BOD5A/lb MLVS/day)
Average ± std. dev. Range
Tank No. 1 0.567 ± 0.277 0.086 to 1.290
Tank No. 2 0.668 ± 0.363 0.099 to 1.920
These data show that organic loadings were generally in the normal
range for conventional activated sludge operation.
The BOD removal characteristics (Sr) for the same period were:
Sr (Ib BODclJ/lb MLVS/day)
jK
Average ± 1 std. dev. Range
Tank No. 1 0.527 ± 0.277 0.069 to 0.161
Tank No. 2 0.604 ± 0.304 0.070 to 1.274
Figure 13 shows the relationship between BOD loading and BOD removal
for each tank for all of the data collected during Phase II. Regres-
sion analyses of the data yielded the following equations for the lines
of best fit:
Tank No. 1 Ib BOD5R = 0.90 Ib BOD5A - 0.04 Ib MLVS (5)
Tank No. 2 Ib BOD5R = 0.87 Ib BOD5A - 0.03 Ib MLVS (6)
The correlation coefficients were found to be highly significant sta-
tistically. Although the differences between the constants for the
BOD loading and removal regression equations for the chemical-biological
63
-------
1.200
1.000 -
(0
^ 0.800
o
ui
0.600
O
o
09
0.400
0.200
« TANK No. I
R 0,953
a TANK No,2
R 0.984
0 0.200 0.400 0.600 0.800 1.000 1.200
Ib. BOD6 APPLIED/Ib.-MLVS
Figure 13. BOD Removal as a Function of BOD Loading Phase II All Flow
Data
-------
system and the control unit were small, they were statistically sig-
nificant at the 1% level.
Figure 14 shows the relationship between unfiltered BOD and effluent
solids for both units. A better correlation could have been achieved
if sufficient data had been available to compute the insoluble BOD and
these values used instead. Acknowledging this deficiency and the wide
variability of data, these curves and the resulting regression equa-
tions also indicate that lower effluent BOD can be expected from the
chemical-biological system. These curves also show the importance of
suspended solids removal if optimum BOD removal is to be achieved.
The nitrogen data show a significant difference in the degree of nitri-
fication which occurred in the chemical-biological system compared with
that of the control unit. Table 9 shows the average results of nitro-
gen analyses performed on three occasions on samples of mixed liquor
to see if the difference observed in effluent samples could be seen
in samples taken prior to alum addition. The same difference was
apparent indicating that alum addition does in some way reduce nitri-
fication under the conditions of operation experienced during these
studies.
As expected, the reduction in alkalinity in Tank No. 1 was much greater
than that observed in the control unit. Brenner (4) reports a reduc-
tion in alkalinity from 182 to 67 mg CaCOg as a result of alum addi-
tion. This is significantly greater than the reduction observed in
this study using'mean or median values but such large reductions were
observed on occasions where an overdose of alum occurred. Since
Brenner did not include data on alum dosages, direct comparison is
difficult.
Brenner (4) also reported in his analysis of chemical costs that "every
5 mg/1 of alum added (as A1+++) uses up 34 mg/1 alkalinity (as CaC03)."
The results of this study, using mean values, indicated that for the
particular system studied, 26 mg/1 of alkalinity were used up for each
5 mg Al/1 added in the form of alum which closely agrees with the
theoretical value of 27.8 mg CaC03/l.
Some increase in calcium in the effluent from Tank No. 1 was observed
(3 mg/1 from mean values) whereas no increase was noted in the effluent
from the control unit. Menar and Jenkins (23) have postulated the
release of calcium during activated sludge treatment. A possible ex-
planation for the observed increase in calcium in Tank No. 1 but not
in Tank No. 2 could be that the released calcium in Tank No. 2 precipi-
tated out as calcium phosphate whereas in Tank No. 1 the phosphorus had
already been precipitated with aluminum. This mechanism of phosphorus
removal, i.e., precipitation with calcium under certain conditions
within the aeration tank was the subject of an exhaustive investigation
by Menar and Jenkins.
Color removal was significantly higher as a result of alum addition.
On days where flows were low, the effluent from Tank No. 1 was crystal
65
-------
80r
70-
60h
o
O 50
til
It.
u.
Ill
DC
Ul
40
30
20 =
o TANK No.2
BOD9 0.19 VS + 10.4
R 0.472
o
e
o
TANK No.
BOD8 0.23VS + 6.0
R 0.876
24 48 72 96 120 144 168 192 216 240 264
EFFLUENT VOLATILE SOLIDS -
Figure 14. Effluent BOD as a Function of Effluent Volatile Solids Phase II All Flow Data
-------
Table 9. Forms of Nitrogen in Mixed Liquor
Nitrogen mg
Ammonia Oxidized Total Kjeldal
Tank No. 1 12.9 3.5 22.4
Tank No. 2 4.6 8.8 9.3
clear and sparkling whereas the effluent from Tank No. 2 had notice-
able color even though suspended solids were often quite low.
Only a limited amount of data on aluminum and sulfate concentrations
were collected during Phase II operation. The mean values for these
data are as follows:
Aluminum - mg Al/1 Sulfate - mg SO,/I
Trickling Filter Eff. 0.17 48
Tank No. 1 Eff. 0.28 134
Tank No. 2 Eff. 0.14 40
The aluminum results are quite similar to those observed for the Phase
I alum studies. The sulfate data for Phase I Run Code III (Table 1)
showed the sulfate concentration of the trickling filter effluent and
the final effluent during the period when no chemical was added to be
only about one-half that observed for Trickling Filter and Tank No. 2
effluent during Phase II operation. There is no apparent explanation
for this difference. The sulfate concentration of the alum treated
unit was virtually the same as observed during Phase I studies. The
sulfate data show that an average of 86 mg SO^/l was added as a result
of alum addition.
Each pound of aluminum added in the form of filter alum adds 5.33
pounds of sulfate ion to the water. If the observed increase in sul-
fate in the effluent from the treated unit is correlated with the mean
flow (0.96 MGD) and'mean alum dosage (178 mg/1), the observed increase
represents an addition of 5.35 pounds per pound of aluminum added
which verifies the accuracy of the sulfate data.
Phosphorus Removal
The effluent phosphorus data presented in Part C of Table 8 confirms
the observations which have already been made regarding phosphorus
removal. As was pointed out earlier, the reported results include all
67
-------
data collected throughout Phase II operation including those obtained
when alum feed was interrupted or when feed rate adjustments were
being made in response to changes in flow or phosphorus concentration.
The overall removal percentages for filtered phosphorus of 97 to 95%
depending on whether median or mean values are used for the chemical-
biological system indicate a high degree of removal can be obtained
over a long period of time. Figure 15 shows the statistical distri-
bution of the filtered effluent total phosphorus for all data obtained.
Figure 16 relates effluent suspended solids and effluent insoluble
phosphorus concentration and shows a high degree of correlation for
the data (R = 0.978). Although not plotted in Figure 16, regression
equations also were calculated for effluent insoluble phosphorus as
a function of effluent volatile and inorganic solids (R = 0.971 for
volatile solids and 0.957 for inorganic solids). The regression
equations for the lines of best fit are:
Eff P = 0.06SS - 0.61 (7)
Eff P = 0.11VS - 0.78 (8)
Eff P = 0.14IS - 0.19 (9)
This correlation further verifies the importance of effluent solids
in determining effluent total phosphorus concentrations. These data
show close agreement with the results obtained by Eberhardt and
Nesbitt (9), particularly for the relationship between insoluble
phosphorus and effluent inorganic solids. The comparable data from
their study shows approximately 9.5 mg P/l with 70 mg/1 of inorganic
solids whereas the results of this study would indicate approximately
9.7 mg P/l would be expected at this same effluent inorganic solids
concentration.
Figure 17 shows the statistical distribution of the unfiltered efflu-
ent total phosphorus for all data obtained.
ACTIVATED SLUDGE STUDIES
Mixed Liquor
Table 10 presents summary data from the activated sludge analyses per-
formed during Phase II. As was the case in Phase I, volatile solids
data were not corrected for apparent volatile solids production due to
volatilization of inorganics during the analysis procedure.
The sludge volume index data show there was no decrease in SVI as a
result of alum addition over those observed in the control unit.
Eberhardt and Nesbitt (9) observed a significant decrease in SVI as
a result of mineral addition into the high rate system used in their
study. The work by Zenz and Pivnicka (52) in a full scale conventional
activated sludge plant did not permit definite conclusions regarding
68
-------
I
(0
o
0.
o
X
Q.
o
Id
u.
Ul
o
UJ
1.0
O.I
TOTAL P (117 ANALYSES)
0.01 0.1 I 5 20 40 60 80 95 99
PERCENTAGE OF TIME OBSERVED VALUE 1 GRAPH VALUE
99.9 99.99
Figure 15. Statistical Distribution of Filtered Effluent Total Phosphorus Tank 1 All Flow Data
-------
2I.Or-
_i 18.0
to £ I5.O
o w
to C
x to I2.O
Q. (O
U.I3
O -I
O
9.0
iS6*
u. u.
3.O
o SINGLE DATA POINT
MEAN OF GROUP OF DATA POINTS
NUMBER « No. of Points Included
in meon
EFF P « 0.06SS-0.6I
R 0.978
36 72 108 144 180 216 252
EFFLUENT SUSPENDED SOLIDS - mg/J?
288 324
360
Figure 16. Insoluble Phosphorus as a Function of Effluent Suspended Solids Phase II Tank 1 All Flow Data
-------
CO
3
1C
o
Q.
CO
O
o
UJ
u.
IL
Id
O
u
tt
UJ
0.
TOTAL P(I23 ANALYSES)
0.01 O.I I 5 20 40 60 80 95 99 99.9 99.99
PERCENTAGE OF TIME OBSERVED VALUE < GRAPH VALUE
Figure 17. Statistical Distribution of Unfiltered Effluent Total Phosphorus Tank 1 All Flow Data
-------
Table 10. Summary of Activated and Waste Sludge Data Phase II
Parameter
Mixed Liquor Solids - mg/A
Total
Volatile
SVI (Mohlman)
SDIa
Solids Age - days
MLVS Phosphorus - %
pH
Temperature °C
Waste Sludge
Total - gal/day
- Ib/day
Volatile - Ib/day
- %
Tank
Mean
2500
1360
63
1.59
1.58
7.04
6.95
19.5
16,870
1,077
575
53.4
No. 1
Std. Dev.
480
310
11
0.76
3.01
0.30
3.5
4,300
436
237
Tank
Mean
1600
1150
63
1.59
2.09
3.67
7.00
19.5
5,670
513
368
71.7
No. 2
Std. Dev.
410
320
25
1.17
2.00
0.20
3.5
2,490
282
208
alculated from SVI: SDI
72
-------
the effect of alum addition on SVI. However, the mean SVI of 63 for
the control unit in the study reported herein is considerably lower
than normally expected for a conventional activated sludge plant
treating domestic wastes. The highest values recorded during the
study were 96 and 172 for Tank No. 1 and Tank No. 2 respectively.
The sludges from both tanks settled very readily as can be seen in
Figure 18. This settling rate is particularly significant since
even sludges with such excellent characteristics could not be handled
under the conditions of hydraulic overload experienced without causing
problems. As noted earlier, Tank No. 1 normally lost more solids in
the effluent than Tank No. 2 but this unit (Tank No. 1) had to handle
considerably more solids under the same hydraulic conditions.
The average solids age of the chemical-biological system was 1.58 days
which was significantly lower than the average of 2.09 days for the
control. This is probably most significant as it affects nitrifica-
tion since BOD and COD removals were generally better in the chemical-
biological system whereas nitrification was greater in the control.
Sludge Production
As indicated previously, solids handling is an important consideration
in any wastewater treatment scheme. Hence, the solids production from
the chemical-biological process is of prime concern. The data pre-
sented in Table 10 on waste sludge production were correlated with
flow (mean flow for Phase II) and are shown in Table 11.
Table 11. Comparative Waste Sludge Production Phase II
Ratio
Waste Sludge Tank 1 Tank 2 Chem-Biol./Control
Gal/day/MG 17,600 5,830 3.02
Ib total solids/day/MG 1,125 535 2.10
Ib volatile solids/day/MG 600 385 1,56
The results showing the weights of sludge wasted are more directly
comparable than are the volumes since the concentrations of the waste
sludges from each tank were significantly different. The results of
Phase II operation showed that for the system studied approximately
twice as many pounds of total solids and one and one-half times as
many pounds of volatile solids are produced in the chemical-biological
system as in the control.
Eberhardt and Nesbitt (9) developed the following equation to describe
the relationships between pounds BOD removed per day (6005^), solids
age in days (tg^) and volatile solids production in Ib per day (VSp).
73
-------
100
LJ
80-
< 60
o
2
o
O 40
K
2
U
U
oa
20
MLSS
SVI
Tonk No. I
2990
60
Tonk No.2
1980
56
Tank No.2
10 15 20
SETTLING TIME-Minutes
25
30
Figure 18. Mixed Liquor Settling Characteristics Phase II Alujn to Tank 1
-------
log (100 VSp/BOD5R) - -0.379 log t^ + 1.8761 (10)
Using this equation and mean values for BOD removal and solids age
observed in this study, the predicted volatile solids production for
each tank would be 452 Ib/day for Tank No. 1 and 395 Ib/day for Tank
No. 2. The predicted total solids production would be 847 Ib/day from
Tank No. 1 and 550 Ib/day from Tank No. 2 based on average volatile
solids content of the waste sludges.
The predicted and observed results show very close agreement for the
sludge production from Tank No. 2. Tank No. 1 results show much more
discrepancy, i.e., the predicted weight of sludge was only 78.6% of
the observed weight of sludge. As pointed out earlier, the volatile
solids test procedure used undoubtably reports results which are too
high because of the probable volatilization of aluminum hydroxy-
phosphate compounds. Eberhardt and Nesbitt (9) in referring to the
work by others reported weight losses of as high as 21.6% for ster-
rettite, an aluminum phosphate compound resulting from the reaction
between aluminum and dihydroxy phosphate. Recht and Ghassemi (29)
showed weight losses of 18.5% for the aluminum-orthophosphate precipi-
tates obtained under the conditions employed in their work. While the
results of these various studies are not directly comparable since con-
ditions under which the precipitates were formed differ, they do show
remarkably close agreement and suggest that approximately 20% of the
reported weight loss in the volatile solids procedure used herein may
have been due to volatilization of inorganic compounds. If the ob-
served results for the volatile solids produced in Tank No. 1 are cor-
rected on this basis, the corrected value of volatile solids would be
460 Ib/day (575 x 0.80) which is very close to the predicted value of
452 Ib/day. This then would suggest the actual inorganic solids pro-
duction in the chemical-biological system was 617 Ib/day (1077-460) or
0.43 Ib per Ib of alum (A12(S04)3 14 H20) added.
Another estimate of the inorganic solids produced from phosphorus pre-
cipitation with alum can be made by means of the following equations
if it is assumed that all of the aluminum ion reacts only with phos-
phate and hydroxide ions:
A12(S04)3 + 2P04 »»»>»» 2A1P04 + 3 S04 (11)
A12(S04)3 + 6H20 »»»)»» 2A1(OH)3 + 6H+ + 3S04 (12)
Using values reported for this study, the weight of inorganic solids
produced would be:
Flow = 0.96 MG Unfiltered Total P = 10.3 mg P/l
Alum dose = 178 mg/1 as A12(S04)3 14 H20
75
-------
For Eq. (11) mg/1 Alum » 1.0 x 10.3 x -^ = 113
For Eq. (12) mg/1 Alum - 178 - 113 = 65 (Excess)
Weight of solids produced:
From Eq. (11)
I 00
2 x 113 x 0.96 x 8.33 x ~~ = 370 Ib AlP04/day
From Eq. (12)
2 x 65 x 0.96 x 8.33 x ^||- = 136 Ib Al (OH)3/day
Total inorganic solids produced = 506 Ib/day
This method predicts only 0.36 Ib of inorganic solids per Ib of alum
added per day which is significantly less than the 0.43 Ib per Ib of
alum calculated above. This difference probably results from the
initial assumption above that aluminum reacts only with phosphate and
hydroxide ions. This is probably not true and the other reactions if
known could be incorporated into the calculation and could result in
better correlation of results.
Several samples of waste sludge were subjected to X-ray diffraction
analysis in an attempt to identify the precipitates which were formed
but all samples showed the precipitates to be amorphous, hence identi-
fication was impossible. Other workers have reported similar results
in their attempts to identify the compounds (29).
Mixed Liquor Phosphorus
The percentage of phosphorus in the mixed liquor from the chemical-
biological system was about twice as great as that observed in the
control system. Considerable variability was observed in this para-
meter in both systems. This variability was thought to be due in a
large part to the difficulty in analyzing sludge samples for total
phosphorus because of the high dilutions which must be made.
The phosphorus content of the control system of 3.67 percent was higher
than expected based on the data collected during the limited period of
no chemical addition during Phase I. It is also significantly higher
than the 2.62 percent reported by Jenkins and Menar for activated
sludge systems (14). Brenner (4) reported a phosphorus content in the
waste activated sludge of 3.4% on a volatile solids basis prior to
aluminum addition to the aerator. Earth and Ettinger (2) reported a
value of 3.0% on a volatile solids basis prior to aluminum addition to
their pilot plant.
76
-------
Menar and Jenkins (23) in their work on the mechanism of enhanced phos-
phate removal by activated sludge were able to correlate percentage of
calcium and percentage of phosphorus in the activated sludge (see
Figure 19). A very limited amount of data collected during the latter
part of Phase II are included in Figure 20, also taken from Menar and
Jenkins (23). These data further indicate that the observed percent-
age of phosphorus in the control system was not unusually high.
Perhaps the most interesting observation in the activated sludge data
was the difference in the filtered total phosphorus concentrations in
the mixed liquor from each of the two units. Mixed liquor samples
were collected ahead of the point of chemical addition so it was ex-
pected the filtered phosphorus values would be approximately equal in
each tank at the point of collection. The much lower values observed
in Tank No. 1 (1.5 mg P/l compared with 7.0 mg P/l for Tank No. 2)
indicate some precipitation or adsorbtion of phosphorus occurs as a
result of contacting the chemical-biological sludge with the influent
waste. However, continuous addition of alum is apparently necessary
to keep effluent phosphorus concentrations low since they rose very
rapidly when chemical addition was interrupted. Recht and Ghassemi
(29) reported that freshly precipitated aluminum hydroxides possess
little capacity to precipitate phosphates. The aging of the aluminum
hydroxy-phosphate precipitates in the chemical-biological system
during the time they are retained in the aerator may account for the
observed difference in filtered total phosphorus concentrations
between the two tanks.
SPECIAL STUDIES
Polyelectrolyte Addition
The heavy losses of solids from Tank No. 1 observed during periods of
peak flow were of continuing concern throughout the study. In an
attempt to apply corrective measures against this effluent degradation,
a series of polymer additions were made into the influent to Final
Settling Tank No. 1. The polymer selected for use on the basis of jar
tests was a moderately cationic flocculant (Nalco 673). A 0.5% solu-
tion of the flocculant was made up fresh daily and was diluted after
metering through a diaphram feed pump prior to application into the
settling tank influent channel.
Samples of mixed liquor taken from the influent channel prior to and
following polymer addition were allowed to settle in 1000 ml cylinders
for visual comparison of settling rates and effluent clarity. The
initial dosage of 0.25 mg/1 was sufficient to produce a somewhat faster
settling floe than was observed prior to polymer addition (see Figure
21). The floes appeared to have a more grainy texture after polymer
addition. The settling characteristics in the basin may have been
better than those shown in Figure 21 since the floe would have had some
additional mixing and time to build subsequent to the point of sampling.
The supernatant liquor in the cylinder was much clearer as a result of
77
-------
8
58
o^ 6
og 5
§u 2
oo
<0
+
i
12345
ACTIVATED SLUDGE PHOSPHATE
CONTENT, % P OF DRY SOLIDS
PILOT-PLANT EXPERIMENTS
STANDARD RATE
O LOW RATE
£ COMPARATIVE STUDY, VCSD PLANT
A COMPARATIVE STUDY, PILOT PLANT
OTHER DATA
D SERL ACTIVATED SLUDGE PLANT AND
GOLDEN GATE PARK ACTIVATED
SLUDGE PLANT
+ PSU CONTROL
Figure 19. Correlation of Activated Sludge Phosphate and Calcium
Contents [After Menar and Jenkin (23)]
78
-------
100
VO
85
CO
to
-------
too
UJ
O
>
o
a:
o
u.
o
UJ
o
(E
UJ
0.
80
20
Prior to polymer addition
/Af
After polymer addition
5 10 15 20 25
SETTLING TIME - minutes
30
Figure 21. Comparative Settling With and Without Polymer Addition to Tank 1 Mixed Liquor
-------
polymer addition although the sample taken prior to polymer addition
was more clear than normal activated sludge effluent as a result of
alum addition.
Visual inspection of the tank during periods of peak flows did not
show any noticeable benefit from polymer addition. Polymer dosages
were increased to 0.5 mg/1 without any significant reduction in solids
loss during peak flow periods. Because of limited storage capacity
for polymer, polymer addition was restricted to those hours of the
day when solids losses were heaviest, usually from 8:00 a.m. to 6:00
p.m. Although polymer addition did not significantly reduce the loss
of solids from Tank 1, the resulting settled effluent was more clear
and sparkling than normal. Figure 22 shows the typical variation in
effluent suspended solids from Tank 1 during periods of hydraulic
overload both with and without polymer addition. These data were
taken on two different days so the flow patterns and hence the pattern
of solids loss are not the same but similar. The data show conclusiv-
ely that, even with polymer addition, the effluent suspended solids
were too high for direct discharge to the receiving stream.
Since this was intended to be only a preliminary investigation of the
effects of polymers On the reduction of suspended solids, the investi-
gation was terminated after four days of operation because of the poor
results obtained.
pH Studies
A series of jar tests were conducted in an attempt to yield some pre-
liminary data on the effect of mixed liquor pH on the residual phos-
phorus concentrations and effluent clarity when alum was being added
at the- head end of the aeration tank. For the purposes of these
studies, samples of the mixed liquor from Tank No. 1 were brought in-
to the laboratory and six 1000 ml portions were set up on a jar testing
apparatus. The initial pH's of the samples were recorded and the pH's
of each of five samples were adjusted to the desired test pH with
addition of 0.1 N 1^504 or NaOH as required. The samples were mixed
during the pH adjustment step and were then allowed to settle for 30
minutes. Phosphorus, turbidity and pH determinations were made on
filtered and unfiltered supernatant samples from each jar. Table 12
shows the summary results from these studies.
As pointed out earlier, a haze or "carry-through" of what is thought
to be very finely divided "aluminum phosphate" occurs when the pre-
cipitant is added at the head end of the tank. The data in Table 12
suggest this is partially a function of pH since lower effluent insolu-
ble phosphorus and filterable turbidity results were obtained with a
reduction in pH (see Figure 23). The pH adjustments were made after
alum addition and mixing in the aeration tank so the fine precipitate
should have been present in the samples brought back to the laboratory.
Therefore, the data indicate that the fine precipitate was altered in
some manner during the pH adjustment step which resulted in lower in-
soluble phosphorus concentrations in the effluent with a reduction in
81
-------
isooh
oo
N)
E
I
30'
20
I
i
10
0
TYPICAL PATTERN
NO POLYMER ADDITION
I
\ n >r-TYPICAL PATTERN WITH
X y 0.5 mg/j? POLYMER ADDITION
I
I
I
7-8AM 11-12 N 3-4PM 7-8PM II-I2M 3-4AM
TIME OF SAMPLING
Figure 22. Variation in Effluent Suspended Solids With and Without Polymer Addition Tank 1 Flow >_ 1.041 MGD
-------
Table 12. Effluent pH Phosphorus and Turbidity Relationships
Initial Test pH
Parameter 5.30 5.80 6.30 6.80 7.30
Ortho P - mgP/fc3
Filtered 0.26 0.30 0.27 0.32 0.53
Unfiltered 1.31 2.02 6.08 5.37 7.59
Turbidity - Units3
Filtered 2.3 2.5 2.3 2.8 3.2
Unfiltered 10 16 25 30 34
pHa
Filtered 6.00 6.20 6.65 6.95 7.20
Unfiltered 6.05 6.10 6.45 6.70 7.00
ml of 0.1N H2S04 13.1 9.0 1.8
to reach test pH
ml of 0.1N NaOH 2.4 6.9
to reach test pH
Control
0.36
6.40
2.7
31
6.95
6.70
-
shown represent average values of analyses performed after
30 minutes settling following pH adjustment to test pH.
pH of control was 6.60.
83
-------
o»
E
l
en
D
o:
o
z
CL
S 8.0
o
UJ
cc
UJ
C 6.0
U»
«0
UJ
o
X
CD
*" 4O
QC ^
-------
pH. Additional study would be needed to determine if this has any
practical significance in terms of operating performance and proce-
dures ,
A series of in situ pH measurements were made in each of the two
aeration and settling tanks to determine actual pH profiles through
the tanks during normal operation. Figure 24 shows the pH profile
during normal operation with alum addition into the effluent channel
from the aeration tank as described earlier. These data show a very
severe depression of pH immediately following the alum addition (pH
6.65 reduced to 4.65) with an increase to pH 5.9 after mixing is com-
pleted. The pH stays about 5.9 through the final tank. This pH is
near optimum for phosphorus removal and is the result of the alum
addition alone. These results were significantly different from the
effluent pH's observed during regular collection of,routine data and
reported in Tables 6, 7 and 8 which showed values of 6.8 to 7.2 for
Tank No. 1 effluent depending on the flow category. Further investi-
gation showed the pH of the sample increases with time apparently as
a result of loss of entrained C02- The sample which showed a pH of
5.85 at the time of in situ measurements showed a pH of 7.10 twenty
hours later. The results of the jar tests reported in Table 12 also
show this increase in pH upon standing. Therefore, the composite
samples collected on a routine basis did not accurately reflect the
pH in the final tanks but rather some sort of equilibrium pH depending
on the length of time the sample had been standing. This same obser-
vation was made in the data from Tank No. 2 except the pH was essen-
tially the same throughout the aeration tank and clarifier. The
sample from Tank No. 2 which showed a pH of 6.95 at the time of in
situ measurements showed a, pH of 7.80 twenty hours later.
Figure 25 shows the pH profile data taken in a similar manner with
alum addition at the influent end of the tank. These data show no
measureable reduction in pH occurred as a result of alum addition
under these operating conditions. The vast buffering capacity of the
mixed liquor in the tank with the dispersion which occurred was able
to buffer the alum addition at this point. In contrast, the relatively
slow dispersion which occurred when alum was added into the effluent
channel was not sufficient to buffer the system.
Recht and Ghassemi (29) also showed in their work on precipitation of
phosphorus with aluminum that effluent residual turbidity is related
to pH. They showed a minimum residual turbidity at about,pH 5.5 with
significant increases as the pH rises above 6. This observation may
also help to explain the "carry-through" which occurred when alum was
added at the influent end of the aerator since the pH then was about
7 through the aerator and final clarifier whereas it was decreased to
about pH 6 through the final clarifier when alum was added into the
effluent channel.
85
-------
pH 6.70 6.65
DISTANCE
ALONG
TANK
1
0' 16'
6.65
1
32'
6.65 6.65
1
48' 64'
AERATION
| ALUM
4.65 5.55
1 1
4' 16'
5.90
1
32'
5.85 5.85 5.95
1
48' 64*
1
16'
5.85
1
32'
FINAL
5.85 5.85
1
48' 64'
TANK NO. I
00
pH 7.05 6.90 6.90 6.90 6.90
DISTANCE
ALONG
TANK
6.95 6.95 6.95 6.95
0'
1
16*
1 1
32' 48* 64'
AERATION
1
16'
1
32*
FINAL
1
46'
64'
TANK NO. 2
Figure 24. In situ pH Profile Alum Addition into Effluent Channel from Aeration Tank
-------
ALUM
PH
\
DISTANCE
ALONG
TANK
6.80 6.85
|
0' 16'
6.85 6.85
| |
32' 48'
6.85 6.95
64'
AERATION
1 1
16'
6.95
1
32'
6.95
1
48*
6.85 6.90
64'
|
16'
6.90
|
32'
FINAL
6.90
|
48'
6.90
64'
TANK NO. I
oo
pH 7.25 7.15 7.10 7.15 7.15
DISTANCE
ALONG
TANK
7.15 7.10 7.15 7.10
0'
1
16'
1
32
1
48 64'
AERATION
1
16
1
32'
FINAL
1
48'
64'
TANK NO. 2
Figure 25. In situ pH Profile Alum Addition at Head End of Aeration Tank
-------
Identification of "Carry-Through"
The attempts to characterize the "carry-through" which occurred when
alum was added at the influent end of the aeration tank were not suc-
cessful. A series of samples were analyzed by means of an atomic
absorption spectrophotometer but no significant differences were ob-
served between effluent samples obtained with alum addition at the
influent end. Preliminary particle size determinations with an opti-
cal microscope did not show any discernible difference between efflu-
ents even with use of 0.1 y pore size filter for purposes of particle
retention. This similarity would suggest that many of the particles
which make up the haze or "carry through" are extremely fine (pass
through a 0.1 y filter) and hence are not discernible except perhaps
by means of electron microscopy. Because of limitations of time and
the general lack of success in identifying or adequately characterizing
the "carry-through," it was decided to abandon such attempts for the
purposes of this project.
Nitrification
After the data collection for Phase II was terminated on August 21, it
was decided to increase the mixed liquor suspended solids in Tank No.
1 to see what effect this would have on nitrification. Chemical addi-
tion during this period was at the influent end of the aeration tank
for the purposes of a separate microbiology study and because of this
the pH of the system following alum addition was higher than that ob-
served during alum addition into the effluent channel as reported above.
The effluent from the chemical-biological system during this period was
more turbid than reported previously. However, on three occasions the
effluent was visibly less turbid than usual and on all three occasions
the pH of the mixed liquor sample was between 5.8 and 5.9 which was
significantly lower than the average pH for the period of 6.40. Table
13 presents summary data for this period (Aug. 24, 1970 to Nov. 11,
1970) based on a limited number of analyses (approximately 17- for each
parameter). The data on oxidized nitrogen show no significant dif-
ference between Tank No. 1 and 2 at the 5% level during this period
of operation. These data can be compared with the results reported
in Table 6, Part B, where the difference in oxidized nitrogen between
Tanks -No. 1 and 2 was significant at the 5% level and operation was
quite similar except for mixed liquor solids level and point of chemi-
cal addition. It would appear that either the higher solids or the
chemical addition into the influent end of the aeration tank or both
negated the previously exhibited inhibitory effect of alum addition or
nitrification. However, it is not possible from the data to determine
whether the higher mixed liquor suspended solids (and resulting higher
solids age) or the change in the point of alum addition was responsible
for the change observed. It is the writer's opinion that the lower pH
and greater pH shock which occurred with alum addition into the effluent
channel was primarily responsible for the apparent inhibition of nitri-
fication that was observed during Phases I and II. Therefore, while
the higher solids age experienced during this special study did undoubt-
ably increase the degree of nitrification which occurred, this same
88
-------
Table 13. Summary Data from Special Nitrification Study
Trickling
Filter Tank No. 1 Tank No. 2
Mean Std Dev Mean Std Dev Mean Std Dev
A. Mixed Liquor
MLSS - mg/H 3030 360 1500 200
MLVS - mg/£ ~ 1590 240 1050 160
SVI ~ -- 84 12 58 9
pH 6.40 0.40 7.00 0.10
B. Effluent
Oxidized Nitrogen
mgN/& 5.0 2.6 8.5 2.0 9.3 2.8
Phosphorus -
Filtered
Ortho 5.0 1.1 0.24 0.22 4.6 1.9
% Removal 96.8 38.7
Total 4.8 2.9 0.38 0.32 4.1 2.1
% Removal 94.9 45.4
Unfiltered
Ortho 6.3 1.9 1.58 1.17 5.6 2.1
% Removal 79.0 25.4
Total 7.5 3.6 1.97 1.36 5.8 3.7
% Removal 73.8 -- 22.7
C. Flow - MGD 0.75 0.16 --
. orthoRatl°
89
-------
observation would also be true for Tank No. 2. However, it probably
did not have a significant role in eliminating the inhibitory effect
of alum addition. Insufficient data were available to calculate the
respective solids ages for this period. Termination of project opera-
tion did not permit collection of comparative data with alum addition
into the effluent channel.
Comparison of the phosphorus data in Table 13 with similar data in
Table 6, Part C, illustrates the deterioration which occurs in efflu-
ent quality when alum addition is at the influent end of the aeration
tank. These data show that average removals of unfiltered effluent
total phosphorus decreased from 91.9% to 73.8% for operation under
what was otherwise very similar operating conditions. This difference
is most likely due to the "carry-through" which occurs with alum addi-
tion into the influent end of the aeration tank. The effect of fil-
tered effluent phosphorus concentrations is much less since much of
the mass of the "carry-through" is removed during the filtration step.
Effluent Fertility
Since the reason for removing phosphorus from wastewaters is to reduce
the fertility of the effluent, samples of trickling filter effluent
and effluents from the control unit and the chemical-biological system
were subjected to the Provisional Algal Assay Procedure (PAAP) (28) by
personnel from the FMC Corporation, Central Research Department, Prince-
ton, New Jersey (19). Selenastrum capricornutum was used as the test
organism and samples from Upper Spring Creek above the point of efflu-
ent discharge were used as the dilution medium for the tests.
Upper Spring Creek (USC) water alone showed significantly less growth
of test cells than did 0.3 PAAP cultures with additions of 10% Upper
Spring Creek water in one flask and 10% distilled water in another.
There was no apparent difference in growth potential between the 0.3
PAAP plus 10% USC and 0.3 PAAP plus 10% distilled water, USC water
with 1.0% addition of Tank No, 2 effluent showed about eight times
more growth of test cells than did the same dilution with Tank No. 1
effluent. Increasing the addition of effluent to 10% resulted in about
130 times more growth with Tank No. 2 effluent than was obtained with
Tank No. 1 effluent. Additional experiments showed that addition of
dibasic potassium phosphate (^HPO^) or sodium phosphate (Na2HP04) to
Tank No. 1 effluent to the same phosphorus concentration observed in
Tank No. 2 effluent resulted in virtually identical growth curves for
the two. The studies indicated that the lack of growth in the Tank
No. 1 effluent was not due to aluminum toxicity.
While very limited in scope, these preliminary studies did show a sig-
nificantly lower growth potential for effluents from Tank No. 1 when
compared with the untreated control. Further, spiking of Tank No. 1
effluent with phosphorus resulted in a restoration of fertility thus
substantiating to a degree the role of phosphorus in algal productivity
for these test conditions. Considerably more work must be done to
further define this role before the results can be projected to field
90
-------
conditions but the results reported above do give some indication that
conditions do exist where phosphorus removal from domestic wastewater
may result in lower productivity in certain waters.
Data collection for Phase II operation was terminated on August 20»
1970. Operation of the chemical-biological system and the control
unit were continued until December 1, 1970 to permit continued col-
lection of data for a satellite project on the microbiology of the
two systems and the special studies described above.
91
-------
CHEMICAL-BIOLOGICAL PROCESS DESIGN AND OPERATION
DESIGN CONSIDERATIONS
Preliminary Data
The chemical-biological process is particularly well adapted for use
in existing activated sludge wastewater treatment plants since it
makes maximum use of existing unit processes and structures thus mini-
mizing capital expenditures. The suggestions and recommendations con-
tained herein are applicable for new treatment facilities as well as
for modifications to existing plants. Some of the data on wastewater
flows and characteristics will, of course, have to be estimated for
new installations. Therefore, flexibility should be incorporated
into the design so that reasonable changes from predicted values can
be incorporated into plant operation without adversely affecting the
results obtained. The following parameters must be evaluated in order
to achieve an economical design that will provide the flexibility and
capability to meet the effluent requirements established for the plant:
1. Flow - design average with due consideration of peak
flow rates and daily, weekly and monthly variations.
2. Phosphorus concentration - variation in concentration
with time is important as are the relative amounts of
ortho and complex; soluble and insoluble phosphorus.
3. Alkalinity - higher alkalinity systems (125 mg CaC03/l
or above) tend to favor usage of alum. If alkalinity is
below 125 mg CaC03/l, sodium aluminate would probably be
the chemical of choice although alum plus lime should also
be considered
4, pH - related to alkalinity. If pH is 7.0 or above,
alum is preferred whereas low pH would favor usage of
sodium aluminate.
5. Sulfate - addition of appreciable amounts of sulfate
to wastewaters already high in sulfate concentration or
where effluents are to be discharged to stream used for
potable water sources may be undesirable. In this event,
sodium aluminate would probably be the chemical of choice.
Chemical Handling and Feeding
Liquid chemical handling and feeding systems are generally easier to
operate and maintain than are dry feed systems. However, transporta-
tion costs and inaccessibility to liquid chemical sources may dictate
use of dry chemicals in some instances. Chemical manufacturers should
93
-------
be consulted for detailed recommendations on chemical storage and un-
loading facilities. Provision should be included for measuring the
amount of chemical fed.
The point of Chemical addition should be located as near to the efflu-
ent end of the aeration tank as is practical. Because of the severe
pH shock which occurred when alum was added directly into the effluent
channel during this study, it is suggested that addition be made into
the aeration tank in order to take advantage of the greater buffering
capacity at that point. Some deterioration in effluent quality can be
expected as the point of addition is moved toward the influent end.
Excessive suction head on the chemical feed pumps should be avoided to
prevent siphoning of chemical through the feed pumps at high tank
levels. Pump manufacturers should be consulted for suction lift
characteristics of their pumps and maximum suction heads which can be
tolerated for the chemical involved. Exposed chemical feed lines and
storage tanks should be insulated for cold weather operation.
Process Control
It is highly desirable for the chemical feeders to be paced from in-
fluent flow meters so that chemical feed rate adjustment is more
responsive to flow changes. This is important both for economy of
operation but perhaps even more to avoid underdosing which can result
in intolerable increases in effluent phosphorus concentrations.
Since a proven reliable means for automated phosphorus analysis of
effluents which could be used in compound loop control of chemical
feed is not presently available, it is recommended that phosphorus
concentration variations be incorporated into the chemical dosage
calculations and the flow - feed rate schedule to compensate for the
variations as much as possible. It is suggested that a planned slight
overdose be practiced since this will help to ensure high quality
effluent where this is required. Where phosphorus removal requirements
are less stringent (90% removal or below), this is not nearly so criti-
cal as occasional underdoses do not have as serious an effect on the
overall results.
Hydraulic Loading on Clarifiers
Based on observations made during this project, it is recommended that
surface settling rates in final clarifiers should not exceed 600
gal/ft2/day for design average flow. Where unusually high peaks occur,
consideration should be given to even lower overflow rates. In some
instances it may be desirable to provide influent equalizing storage to
dampen out surges. This would be of assistance also in optimizing
chemical feed since many of the variations in flow and concentration
which otherwise occur would be eliminated.
94
-------
If hydraulic loads in existing treatment works are such that recom-
mended overflow rates are exceeded, consideration should give to in-
stallation of tube settlers or other devices to improve the efficiency
of the existing units. Severe cases of hydraulic overloading may re-
quire construction of new settling tanks or surge tanks to alleviate
the problem.
Effluent Filtration
It will usually be necessary to provide filtration of effluent in
order to meet very stringent phosphorus removal requirements. Filtra-
tion does provide a safety factor to protect the stream in the event
of heavy solids losses from the clarifiers. Various filter devices
are available and may be evaluated by the designer for the particular
application involved.
Solids Handling
No special solids handling equipment or requirements are necessary for
handling and disposal of the sludges resulting from chemical-biological
treatment. Flexibility in pumping units should be sufficient to handle
the greater weight of solids and volumes of sludges which result from
chemical-biological treatment in instances. Sludge weights approxi-
mately twice those obtained without chemical addition can be expected.
Dewatering and disposal of the chemical-biological sludges should not
present any unusual problems. These sludges are generally more easily
dewatered than are biological sludges and may be handled by ,any of the
normally employed unit processes. It is unlikely that recovery of the
precipitating chemical will offer any economic advantage except under
unusual circumstances.
OPERATIONAL CONSIDERATIONS
General
Operating experiences over the duration of the project resulted in the
following suggestions on plant operation in order to realize maximum
benefit from the chemical-biological process for phosphorus removal.
Process Control
It is recommended that filtered orthophosphate in the influent waste
be used as the basis for process control. The orthophosphate test pro-
cedute is rapid and relatively simple to perform compared to the pro-
cedure for total phosphate. The data presented earlier on the relation-
ship between Al/P (filt. ortho) ratio and effluent total phosphorus
concentration indicate the ratios which can be used as guides for dosage
95
-------
calculations. However, it is recommended that operating data from
each individual plant be reviewed regularly and adjustments in dosage
be made as needed to achieve the particular degree of removal required.
Different wastewaters will undoubtably require different Al/P ratios
to achieve a given effluent phosphorus concentration and these relation-
ships can only be developed from actual plant operating data.
Once the relationship between ortho and total phosphorus has been
determined for a given plant, the orthophosphate test can be used to
predict effluent total phosphorus values with sufficient accuracy for
control purposes.
Control of mixed liquor suspended solids must take into consideration
the much lower percentage of volatile solids in the chemical-biologi-
cal system. The mixed liquor volatile suspended solids should be
maintained at the necessary level to achieve the desired organic
loadings. Once the system has reached a balance, control can be based
on total suspended solids with regular checks on volatile solids so
that any changes can be incorporated into sludge wasting schedules.
Operators should check the amount of chemical actually fed each shift
to detect underfeed or overfeed as soon as possible. Unless the
chemical feed lines have a constant slope to the point of application,
they should be checked frequently for entrapped air and vented as
required.
Return of waste sludge from the chemical-biological system to the pri-
mary settling tanks during this study resulted in better concentration
of primary sludges in the withdrawal hoppers. This reduces the amount
of water which must be handled in the solids handling process and hence
can be of advantage where piping flexibility permits.
Maintenance
Some additional corrosion can be expected as a result of the lower pH's
which may occur in the chemical-biological system when alum is used.
Visual comparison in September 1970 of collection mechanisms in the
two final clarifiers used in the study revealed some additional cor-
rosion occurred as a result of alum addition. However, the treatment
plant foreman did not feel it was excessive for the length of time the
units had been in operation without being taken down for maintenance.
The most noticeable difference was in the color of the corrosion pro-
ducts. Those from Tank No. 1 were more rust colored whereas those
from Tank No. 2 were more black in color. This corrosion problem can
be taken care of by proper selection of protective coatings.
Less growth of slimes on weirs and tank walls was noted in Tank No. 1
than in Tank No. 2 during Phase II operation. This probably was due
to the lower pH and perhaps to a lesser degree the lower amount of
phosphorus in the effluent. Because of this, less routine cleaning
of weirs and walls in Tank No. 1 was required which resulted in some
savings in maintenance time.
96
-------
No unusual operating problems developed during the duration of the
study. Routine maintenance of equipment and structures is necessary
to achieve optimum process performance and to avoid major breakdowns.
97
-------
COST ANALYSIS
GENERAL
Process economics are as important as technical feasibility in the
final selection of unit processes for the solution of a particular
wastewater treatment problem. The chemical-biological process which
served as the basis for this study offers two significant economic
advantages over some of the other schemes which have been proposed
for removing phosphorus from domestic wastewater. First, for appli-
cation in existing wastewater treatment works, required capital ex-
penditures are minimal since no new treatment units normally would
be required. The addition of chemical storage and handling facili-
ties are the only plant additions necessary in most instances. In
some cases it may be necessary to modify waste sludge pumps and other
solids handling facilities in order to accommodate the increased
solids production which results from phosphorus removal. For new
facilities construction, no additional treatment units over and above
those required for activated sludge secondary treatment would be
required. Chemical storage and handling facilities, of course, would
be necessary also. Provision for handling the additional solids
could often be incorporated into the design for little or no addi-
tional cost. These considerations should make the chemical-biological
process an attractive alternative when considering the various phos-
phorus removal process schemes which are available.
CAPITAL COSTS
The major items requiring capital expenditures for the process include
chemical storage tanks, chemical feed lines and provision for auto-
matically pacing chemical feeders with flow. Housing for chemical
storage and feeding equipment also would be required in most instances.
The type of chemical storage is dependent upon the type of chemical
fed (liquid or dry) and the amount to be fed since this will influence
the method of shipment. The cost comparisons presented herein were
based on the use of liquid chemical for all three plant sizes shown
since liquid systems are generally less costly and wide distribution
of liquid alum producing facilities makes its use feasible in many
different geographical locations.
The selection of chemical feeding equipment to be used also is depen-
dent upon the form and amount of chemical to be fed. A diaphram feed
pump such as the B-I-F, Series 1700, was used as the basis for the
estimates for the 1 and 10 MGD applications and a volumetric feeder
such as the B-I-F Rotodip feeder was used for the 100 MGD plant.
Duplicate feeders were assumed to be provided in all instances and
all were equipped for automatic pacing from the influent flow meter.
Complete housing of equipment and housing or insulation of storage
tanks and feed lines was assumed for the purposes of this analysis.
99
-------
OPERATING COSTS
Some additional operator time is required for checking chemical
feeders and other routine operating duties in connection with phos-
phorus removal. Additional laboratory analyses would also be required
for process control but in most instances these would not involve
sufficient additional time to warrant additional cost allowances.
While this is generally true, it should be pointed out that additional
personnel costs can be significant in the smaller plant since the
chemical-biological process is relatively more complex than those
normally used for secondary treatment.
Additional power costs also would be incurred for chemical feed equip-
ment and handling of the increased solids which result from phosphorus
removal.
CHEMICAL COSTS
Alum is considerably less costly than sodium aluminate under present
pricing structures. Liquid alum is also cheaper than dry alum where
liquid alum is available within a reasonable haul distance. Figure
26 relates chemical cost to the effluent phosphorus concentration
based on the results of this study. The chemical costs used to develop
Figure 26 are bulk list prices F.O.B., the point of production, and are
based on the following:
Cents/lb as Cents/lb as
Alum or Aluminate Aj-f-H-
Alum
Liquid 1.13 25.7
Dry 2.89 31.8
Sodium Aluminate
Liquid 5.80 40.9
Chemical costs for any particular installation will be affected by
transportation costs. Since freight costs vary so widely, no allow-
ance for these costs were included in the overall cost analysis.
Table 14 presents cost data for delivery to the State College, Pennsyl-
vania area and gives some idea of the effect of transportation costs
where the delivery point is considered to be moderate to long distance
from the point of production.
Transportation costs in this instance would increase the liquid alum
costs presented in Figure 26 by approximately 35%. This is undoubt-
ably higher than would be experienced for many other geographical lo-
cations .
100
-------
2.0
E
i
-------
Table 14. Cost of Chemicals Delivered to State College, Pennsylvania
Cents/lb as Cents/lb as
Alum or Aluminate Al"*"*"*"
Alum
Dry
Bag
Truck (30,000 lb) 3.90 42.9
Car (80,000 lb) 3.61 39.7
Bulk
Truck (40,000 lb) 3.46 38.1
Car (40,000 lb) 3.71 40.8
Liquid
Tank truck 3.14 34.5
Tank car 3.33 36.6
Sodium Aluminate
Liquid
Tank truck 8.45 59.9
Tank car 8.30 58.5
COST SUMMARY
Table 15 summarizes the estimated total costs of chemical precipitation
of phosphorus using the chemical-biological process and based on the
findings of this study. The influent waste in each instance was
assumed to have 10 mg P/l unfiltered total phosphorus and 7 mg P/l
soluble orthophosphate. Alum was the chemical of choice and capital
cost estimates included only chemical storage and handling facilities.
The basis for amortization was the same as that used by Smith and
McMichael (38) in their recent work on costs of tertiary wastewater
treatment (4.5% - 25 year period, 6.744% of capital cost). All costs
are expressed in 1970 dollars based on the Environmental Protection
Agency, average 1970 National Index of 143.64 (15).
Brenner (4) has estimated phosphorus removals costs of 4.3, 3.3 and
2.9 c/1000 gal for 1, 10 and 100 MGD plants respectively with use of
the chemical-biological process. His costs were based on an assumed
90% removal of phosphorus. Data from this study would predict costs
of 6.1, 4.1 and 2.9 0/1000 gal for 90% removal for 1, 10 and 100 MGD
plants. The major differences between these two estimates occurs in
the estimated amortization and operating costs (exluding chemical costs).
102
-------
Table 15. Cost Summary for the Chemical-Biological Process
to Precipitate 97 Percent of the Influent Phos-
phorus (1970 Dollars)
Design Capacity
MGD
Costs
10
TOTAL ESTIMATED CAPITAL COST
Process Costs, cents/1000 gal.
100
Costs
Capital, dollars
Chemical Storage
Chemical Feeders
Housing and Insulation
Miscellaneous Plumbing
QTI A T7n ii-f TMnam t*
$ 2,000
1,800
5,200
3,000
$ 6,000
4,000
15,000
5,000
$65,000
6,000
15,000
8,000
$12,000 $30,000 $94,000
Amortization (25 yrs., 4.5%)
Alum (from Fig. 26 @ 97% Removal)b
Power
Operating & Maintenance Labor
Supervision & Payroll Overhead
Maintenance Materials
TOTAL ESTIMATED PROCESS COSTS,
0/1000 gal.
0.221
3.7
0.02
2.40
0.80
0.20
7.341
0.055
3.7
0.02
1.01
0.33
0.20
5.315
0.017
3.7
0.02
0.11
0.04
0.20
4.087
Duplicate units provided.
bLiquid alum was used for all three plant sizes and cost shown does
not include transportation costs.
°Taken as 30% of operating and maintenance labor (after Smith and
McMichael (38).
dTaken as 1/3 of maintenance and labor (after Smith and McMichael
(38).
103
-------
As stated earlier, the above costs only reflect the cost of precipi-
tating the phosphorus. These costs would be in addition to the cost
of activated sludge biological treatment. In addition, where phos-
phorus removal of more than 90% is required, it is recommended that
filtration of effluent be provided. This too would be an additional
cost. Table 16 summarizes the estimated costs for activated sludge
treatment, chemical-biological treatment for phosphorus removal and
multiple medium filtration of the effluent for the three plant sizes
shown. The costs for sludge handling shown in Table 16 include costs
for sludge thickening, digestion, elutriation, vacuum filtration and
incineration for solids resulting from primary settling and conven-
tional activated sludge. These costs do not reflect any adjustment
for the additional solids which are produced as a result of phosphorus
removal. Because of the wide variability in solids handling and dis-
posal methods used in individual plants, it was decided not to refine
sludge handling costs for this analysis. Cost refinements can be made
on an individual basis and used to adjust the cost data presented in
Tables 15 and 16.
Table 16. Estimated Total Treatment Costs (1970 Dollars)
Design Capacity, MGD
Costs - C/1000 gal^ ^^^ 1 10 100
Activated Sludge Treatment3 12.00 6.78 5.27
Chemical-Biological Phosphorus
Precipitation*5
Sludge Handling*
Multiple-Medium Filtration0
a
Chlorination
TOTAL TREATMENT COSTS
7.
13.
6.
0.
39.
34
07
24
95_
60
5.
7.
2.
0.
23.
32
48
70
74.
02
4
5
1
_0
16
.09
.46
.19
.66
^MHM
.67
3After Smith (37)
bFrom Table 15
CAfter Chen and Nesbitt (6)
104
-------
GENERAL DISCUSSION
Results of the various investigations conducted during this study were
discussed in detail in previous sections of this report. The general
discussion presented in this section is not intended to repeat the
earlier discussion but will focus on the general capabilities and
applicability of the chemical-biological process of phosphorus removal.
The Phase I studies showed aluminum sulfate (alum) was a better pre-
cipitant than sodium aluminate with the moderately alkaline wastewater
in this system. Uhler (45) in his pilot plant filtration study using
treated effluent from Tank No. 2 (during Phase I operation when chemi-
cal was being added to both units) showed that a dual medium filter
was capable of reducing the effluent total phosphorus concentrations
to approximately 0.1 mg P/l when alum was used as the precipitant. He
also showed that use of sodium aluminate resulted in filter effluent
phosphorus concentrations of approximately 0.3 mg P/l at comparable
Al/P ratios. The best effluent quality during this study was obtained
when the precipitating chemical was added at the effluent end of the
aeration tank. The exact point of addition is not considered to be
critical so long as adequate mixing is provided since the precipitating
reaction is virtually instantaneous. Excessive mixing such as occurs
when the chemical is added at the influent end of the aeration tank
results in the "carry-through" of material and a deterioration in
effluent quality. The problem of "carry-through" which has been re-
ported by other workers was not a problem in this study when the alum
was added to the effluent end of the aeration tank. Some "carry-
through" was noted with sodium aluminate even when it was added at
the effluent end. This probably accounts for the higher effluent
phosphorus observed by Uhler (45) in his filtration study and during
Phase I of this study when compared to the results obtained from use
of alum.
Overdosing of alum did not result in the adverse effects of higher
effluent turbidity and phosphorus which have been noted by others
(4, 52). Although Al/P ratios occasionally were as high as 6/1 or
more when unexpected low flows or low phosphorus or both occurred in
the influent waste, these high ratios actually reduced effluent tur-
bidity and phosphorus. The absence of increased turbidities at these
high Al/P ratios was probably due to the difference in the point of
chemical application since, as far as is known, chemical addition was
into the influent end of the aeration tank in the other studies
referred to above.
Solids loss from the chemical-biological system as a result of hydrau-
lic overload of the final clarifiers was a serious problem throughout
much of the study. This was not anticipated prior to commencing the
work and could not be eliminated during the course of the study. As
was discussed previously, suspended solids losses from the chemical-
biological system (Tank No. 1) were extremely heavy during periods of
excessive flow and at such times the effluent could not be considered
105
-------
acceptable for discharge to the receiving stream. This was contrasted
with periods of low flow when suspended solids were significantly lower
in the effluent from the treated system than they were in the control.
Density currents apparently were established in the final clarifiers,
at least during peak flow conditions. On several occasions, the sludge
blanket was observed to be within approximately 3 feet of the water
surface in Final Tank No. 1 moving toward the effluent end with an
appreciable velocity. As it approached the end of the clarifier, it
rose and went over the weirs of the effluent channels. The supernatant
liquid over the sludge blanket was extremely clear with very few sus-
pended solids. Solids losses from Final Tank No. 2 also increased
during periods of peak flow but never in the magnitude observed in Tank
No. 1. The sludge blanket in Tank No. 2 also never could be observed
in the same manner as that in Tank No. 1. This observable difference
in solids loss between the two tanks most likely is due to the greater
amount of solids which must be handled in the chemical-biological sys-
tem. Proper baffling of the clarifiers undoubtably would help to re-
duce this problem.
The results of the studies in both Phase I and Phase II demonstrated
the enhanced removal of 6005 (unfiltered effluent, carbonaceous demand
only) in the chemical-biological system. Eberhardt and Nesbitt (9)
also demonstrated this capability as have several other workers (2, 4).
The problems experienced with hydraulic overloading were also reflected
in the BODs data for effluent BOD's from the treated system often ex-
ceeded those of the control during periods of severe overloading. This
cannot be considered typical process performance and under normal oper-
ating conditions improved organic removal should be realized. The
study did not include investigations of oxygen uptake rates or other
parameters which could identify relative biological activity between
the two systems. Therefore, it was not possible to differentiate
between biological and chemical removal of organic matter in the
chemical-biological system. If it could be shown that a significant
portion of BOD removal is by chemical means, a reduction in aeration
tank capacity and air requirements could be made which would be reflec-
ted in lower costs for new facilities construction and operation.
Operating costs would also be reduced in existing plants due to the
lower air requirements which would help to offset chemical costs.
The data presented previously show that less nitrification occurred in
the chemical-biological system under normal operating conditions than
in the untreated control. Work still in progress on the microbiologi-
cal study may yield additional insight into this phenomenon.
Balakrishnan and Eckenfelder (1) have concluded that organic loading
and sludge age are the most critical parameters affecting nitrifica-
tion in the activated sludge process. They showed that the optimum
organic loading was about 0.3 Ib BOD/lb MLVS and a minimum sludge age
of 3-4 days was recommended. The reported results from the present
study show that normal operation did not fall within the limits recom-
mended above, hence nitrificatioti was not favored for either the con-
trol or the chemical-biological system.
106
-------
The phosphorus removal results for unfiltered effluent samples were
quite erratic because of the solids losses which occurred during
periods of hydraulic overload. Filtered effluent phosphorus removals
during Phase II were also somewhat erratic due largely to the diffi-
culty of responding to changes in flow and influent phosphorus concen-
trations in settling chemical feed rates. Automatic pacing of chemical
feed equipment from the flow meter would minimize this problem for most
installations. The process is capable of continuously providing very
high degrees of phosphorus removal under low flow conditions at rela-
tively high Al/P (filt. ortho) ratios. Table 17 shows the phosphorus
data for the period June 15, 1970 to August 12, 1970 when flows were
lower than normal due to the smaller student population on campus
during the summer term and flow variations were less severe. These
results show the average removal of unfiltered effluent total phos-
phorus was 91.9%. Table 17 also shows that average phosphorus removal
on filtered effluent was 98.9% for orthophosphate and 98.2% for total
phosphorus during the period reported. The average Al/P (filt. ortho)
ratio was higher than desired (2.71/1 instead of 2.25/1) since the
influent phosphorus was lower than anticipated. However, the only
adverse effect of this overdose was the increased chemical costs which
resulted.
The average Al/P (unfilt. total) ratio for Phase II operation was
1.6/1 based on average values for aluminum fed and unfiltered total
phosphorus for the trickling filter effluent. This is in the same
range reported by others (2, 4, 9) as necessary to achieve phosphorus
removals of approximately 95%. One of the difficulties in comparing
results of this study with those of others is they ordinarily do not
specify the phosphorus basis for the Al/P ratios reported. Most
workers have apparently used unfiltered total phosphorus in the influ-
ent waste as the basis for calculating the reported ratios but it is
not so stated in most instances.
Figure 27 shows the amount of Al fed per day in relation to flow
during Phase II operation. The desired dosage based on the average
filtered orthophosphate and using an Al/P (filt. ortho) ratio of
2.25/1 is also shown and illustrates the problems of overfeed and
underfeed which occurred due to inability to respond quickly to change
in flow. As indicated earlier, pacing of chemical feed from the in-
fluent flow meter would minimize this problem.
Attempts to make acceptable phosphorus mass balances were generally
unsatisfactory. This is thought to be due to the difficulty in ob-
taining reliable results for total phosphorus in the sludge samples.
This same observation was made by O'Shaughnessy (26) in his digester
studies. Since individual phosphorus mass balances yielded highly
variable results, the mean values for the amount of phosphorus ob-
served in the waste sludge from each aeration tank was compared with
the amount of phosphorus which was predicted to be in the waste sludge
based on observed phosphorus removal. These results are presented in
Table 18 and show reasonable agreement between observed and predicted
values.
107
-------
Table 17. Summary Phosphorus Removal Data Phase II June 15, 1970
to August 12, 1970a
Effluent
Phosphorus
Filtered
Ortho - mg P/£
% Removal
Total - mg P/&
% Removal
Unfiltered
Ortho - mg P/£
% Removal
Total - mg P/H
% Removal
Al/P (filt. ortho) Ratio
Flow - MGD
Tank
Mean
0.07
98.9
0.12
98.2
0.49
93.6
0.61
91.9
2.71/1
0.71
No. 1
Std. Dev.
0.05
0.9
0.08
1.2
0.27
2.8
0.27
2.7
0.57
0.05
Tank
Mean
5.2
30.9
5.7
25.6
5.9
27.1
6.3
19.8
0.71
No. 2
Std. Dev.
0.9
17.3
0.9
19.8
1.4
17.7
1.9
13.3
0.05
Sundays not included
5Based on unfiltered trickling filter effluent total phosphorus
(mean value = 7.4 mg P/5, ± 2.0)
108
-------
zoo
180
160
140
>.
e
v
£ 120
I
ui
(9
< 100
(0
O
o
X
3 8O
60
40
20
.*..
ACTUAL AL**+DOSAGE 109.1 FLOW
-1-28.2 R- 0.728
DESIRED AL*** DOSAGE
(2.25) (7.1) (8.33) FLOW
I Ib. AL*** - II Ib AI2(S04)3 14 H20
0.2 0.4 0.6 0.8 1.0 1.2
FLOW PER TANK -MGD
Figure 27. Aluminum Dosage as a Function of Flow Phase II
109
-------
Table 18. Comparison of Observed and Predicted Amounts of Phosphorus
in Waste Sludges
Tank No. 1
Tank No. 2
Obs.a Pred.b Obs/Pred Obs.a Pred.b Obs/Pred
Ib/day Ib/day % Ib/day Ib/day %
Phase I Alum
56.2 72.7 77.1
Phase I Sodium
Aluminate 31.6
Phase II Alum 50.2
Phase II Control
36.8 85.9
47.9 104.8
55.9 78.3 71.4
36.8 40.1 91.8
18.4 20.7 88.9
Predicted P
(Ib/day)
(8.33)
(Ib/gal)
(Ave. T.P. Unfilt. Inf. - Ave. T.P. Eff.) (Ave. Flow)
(mg/1) (MGD)
Observed P = (Unfilt. T.P. Waste Sludge)(Waste Sludge Flow)(8.33)
(Ib/day) (mg/1) (MGD) (Ib/gal)
Sludge production and sludge handling continues to be a major concern
with all phosphorus removal schemes. The results of this study have
shown that significantly more solids are produced and hence must be
removed from the chemical-biological system than from conventional
secondary treatment. Although the results of this study did not indi-
cate a reduction in SVI as a result of alum addition, this is consid-
ered to be atypical of what would be expected in most instances because
of the unusually low SVI observed in the control system. Use of sodium
aluminate did result in a significant reduction in SVI even in this
system. Work by 0'Shaughnessy (26) and others (2, 52) has shown that
the precipitated phosphorus is not released during anaerobic digestion.
Their work has also shown that waste sludges from the chemical-biologi-
cal system are more easily dewatered than are waste sludges front con-
ventional activated sludge treatment. Therefore, it does not appear
that sludges from the chemical-biological system would create any un-
usual operating problems. However, due consideration must be given to
the greater amount of solids which are produced.
No great differences in costs among the various process schemes using
chemical precipitation for phosphorus removal which have been reported
110
-------
by others and the chemical-biological process presented herein are
apparent. This similarity occurs because the chemical costs are
virtually the same regardless of which process scheme is selected
and they represent the major portion of phosphorus removal costs.
Costs are affected by the choice of chemical to be used and generally
the least costly chemical which will give the required degree of re-
moval should be chosen. Some cost advantage may be realized from use
of aluminum or ferric compounds as the precipitating agents over use
of lime since the dosages of the former are directly related to the
phosphorus concentration whereas those of the latter are not. There-
fore, any reduction in influent phosphorus would permit a reduction
in the amount of aluminum or iron used and a net savings in cost.
There does appear to be a definite cost advantage for the chemical-
biological system when compared to a tertiary treatment system. Most
of this advantage results from the lower capital costs for the chemi-
cal-biological process. Cost comparisons among the various processes
which have been proposed is difficult because all are on different
bases and none include all applicable costs.
Comparison of costs of the chemical-biological process with estimated
costs of various chemical-physical process schemes which have been
proposed (50, 53) also does not yield conclusive evidence in favor of
any given process. The results of this study have shown the chemical-
biological process is capable of effecting high degrees of removal of
phosphorus and organic material by use of well proven, easily operated
unit processes. Capital costs for use of the process in existing
activated sludge plants are very low and they would not add signifi-
cantly to capital costs for new facilities construction.
The problem of determining just what degree of phosphorus removal is
necessary to control eutrophication in a particular instance still
has not been resolved. Stream standards now being established show
widely varying requirements for phosphorus removal by specifying dif-
ferent levels for the various forms of phosphorus (i.e., filtered or
unfiltered, ortho or total) for different situations. This makes it
necessary to match unit process performance to a specific standard
which may well change in the near future as more results of research
become available. A distinct advantage of the chemical-biological
process is that varying phosphorus removal requirements can be met
easily by controlling the amount of precipitating chemical added while
still maintaining a high degree of organic removal.
Ill
-------
ACKNOWLEDGMENTS
This study was done and this report was prepared by David A. Long and
John B. Nesbitt of the Department of Civil Engineering of Pennsylvania
State University, University Park, Pennsylvania.
The valuable advice and assistance of the late Professor R. Rupert
Kountz is sincerely acknowledged.
Acknowledgment is also given to Messrs. Lloyd Niemann and Lawrence
Williams and to their operating staff at the University Wastewater
Treatment Plant for their assistance during the course of the inves-
tigation. The writers would also like to acknowledge the technical
assistance of Mrs. Francine Klein and Messrs. James C. 0"Shaughnessy,
David Smith and Michael Schimerlik.
This project was supported and financed in part by The Soap and
Detergent Association, New York, New York, and in part by a Research
and Development Grant from the Federal Water Pollution Control Adminis-
tration, Department of the Interior, pursuant to the Federal Water
Pollution Control Act. The assistance provided by the EPA Project
Officer, Dr. Robert L. Bunch, Advanced Waste Treatment Research
Laboratory, Cincinnati, Ohio, and the Phosphorus Committee of The
Soap and Detergent Association is acknowledged with sincere thanks.
113
-------
BIBLIOGRAPHY
1. Balakrishnan, S. and Eckenfelder, W. W., Nitrogen Relationships in
Biological Treatment Processes - I. Nitrification in the Acti-
vated Sludge Process, Water Research, Vol. 3, No, 1, January 1969,
pp. 73-81.
2. Barth, E. F. and Ettinger, M. B., Mineral Controlled Phosphorus
Removal in the Activated Sludge Process, Journal Water Pollution
Control Federation, Vol. 39, No. 8, August 1967, pp. 1362-1373.
3. Bell, G. R. , Libby, D. V. and Lordi, D. T., Phosphorus Removal
Using Chemical Coagulation and a Continuous Counter Current
Filtration Process, Water Pollution Control Research Series,
17010EDO 06/70, USDI, June 1970.
4. Brenner, R. C., Phosphorus Removal by Mineral Addition, Proceed-
ings of Conference on Nutrient Removal and Advanced Waste Treat-
ment, Federal Water Pollution Control Administration, Cincinnati,
Ohio, April 1969, pp. 1.9 - 1.18.
5. Brenner, R. C., Combined Biological-Chemical Treatment for Control
of Phosphorus, Advanced Waste Treatment and Water Reuse Symposium,
Dallas, Texas, January 1971.
6. Chen, J. and Nesbitt, J. B., Chemical Treatment, High-Rate Multi-
ple Medium Filtration of Raw Municipal Wastewater, The Pennsyl-
vania State University, Institute for Research on Land and Water
Resources, Research Publication No. 66, March 1971.
7. Dryden, F. D. and Stern, G., Renovated Waste Water Creates Recrea-
tional Lake, Environmental Science and Technology, Vol. 2, April
1968, pp. 268-278.
8. Eberhardt, W. A., Chemical Precipitation of Phosphorus Within a
High Rate Activated Sludge Process, Ph.D. Thesis, The Pennsyl-
vania State University, University Park, Pennsylvania, 1967.
9. Eberhardt, W. A. and Nesbitt, J. B., Chemical Precipitation of
Phosphorus in a High-Rate Activated Sludge System, Journal
Water Pollution Control Federation, Vol. 40, No. 7, July 1968,
pp. 1239-1267.
10. Finstein, M. S. and Hunter, J. V., Hydrolysis of Condensed Phos-
phates During Aerobic Biological Sewage Treatment, Water Research,
Vol. 1, No. 4, April 1967, pp. 247-254.
11. Hall, M. W. and Engelbrecht, R. S., Phosphorus Removal - past,
present, and future, Water and Wastes Engineering, Vol. 6, No. 8,
August 1969, pp. 51-53.
115
-------
12. Huber, L., Practical Experiences with Phosphate Elimination Plants,
Wasser U. Abwasser - Forschung, No. 3, 1968, pp. 99-101, (Trans-
lation from German).
13. Jaworski, N. A., Villa, 0., Jr. and Hetling, L. S., Nutrients in
the Potomac River Basin, Proceedings, 24th Industrial Wastes
Conference, May 1969, pp. 675-699.
14. Jenkins, D. and Menar, A. B., The Fate of Phosphorus in Sewage
Treatment Processes, I. Primary Sedimentation and Activated
Sludge, SERL Report 67-6, May 1967, Sanitary Engineering Research
Laboratory, University of California, Berkeley.
15. Journal Water Pollution Control, Vol. 43, No. 3, March 1971, p. 380.
16. Kuentzel, L. F., Bacteria, Carbon Dioxide, and Algal Blooms,
Journal Water Pollution Control Federation, Vol.. 41, No. 10,
October 1969, pp. 1737-1747.
17. Lee, G. F., Analytical Chemistry of Plant Nutrients, Water and
Sewage Works, Vol. 116, R. N., November 1969, pp. R-38 - R-43.
18. Levin, G. V. and Shapiro, J., Metabolic Uptake of Phosphorus by
Wastewater Organisms, Journal Water Pollution Control Federation,
Vol. 37, No. 6, June 1965, pp. 800-821.
19. Lindemann, E., Progress Report, Eutrophication, Penn State, Report
No. PCR 822, FMC Corporation Central Research Department, Prince-
ton, New Jersey, January 1971.
20. Little, Linda W., Phosphorus in Water and Wastewater, An Annotated
Selected Bibliography, UNC Wastewater Research Center Report No.
11, University of North Carolina, November 1970.
21. Mackenthun, K. M., Nitrogen and Phosphorus in Water: An Annotated
Selected Bibliography of Their Biological Effects, U. S. Depart-
ment of Health, Education, and Welfare, Public Health Service,
1965.
22. Mackenthun, K. M., The Phosphorus Problem, Journal American Water
Work Association, Vol. 60, September 1968, pp. 1047-1054.
23. Menar, A. B. and Jenkins, D., The Fate of Phosphorus in Sewage
Treatment Processes, Part II, Mechanism of Enhanced Phosphate
Removal by Activated Sludge, SERL Report 68-6, August 1968,
Sanitary Engineering Re&earch Laboratory, University of Cali-
fornia, Berkeley.
24. Nesbitt, J. B., Removal of Phosphorus from Municipal Sewage Plant
Effluents, Engineering Research Bulletin B-93, The Pennsylvania
State University, February 1966.
116
-------
25. Nesbitt, J. B., Phosphorus Removal - The State of the Art, Journal
Water Pollution Control Federation, Vol. 41, No. 5, Part 1, May
1969, pp. 701-713.
26. O'Shaughnessy, J. C., Evaluation of the Anaerobic Digestion of
Sludges Containing Aluminum Phosphate Precipitates, M. S. Thesis,
The Pennsylvania State University, University Park, Pennsylvania,
1970.
27- Popel, J., Removal of Phosphate from Sewage by Coagulation with
Ferric Iron and Aluminum, Water Research, Vol. 2, No. 1, January
1968, p. 95.
28. Provisional Algal Assay Procedure, Joint Induetry/Government Task
Force on Eutrophication, New York, February 1969.
29. Recht, H. W. and Ghassemi, M., Kinetics and Mechanism of Precipita-
tion and Nature of the Precipitate Obtained in Phosphate Removal
from Wastewater Using Aluminum (III) and Iron (III) Salts, Water
Pollution Control Research Series, 17010EKI04/70, U. S. Gov.
Printing Office.
30. Rudolfs, W., Phosphates in Sewage and Sludge Treatment - I.
Quantities of Phosphates, Sewage Works Journal, Vol. 19, No. 1,
January 1947, pp. 43-47-
31. Rx for Ailing Lakes - A Low Phosphorus Diet, Environmental Science
& Technology, Vol. 3, No. 12, December 1969, pp. 1243-1245.
32. Sanitary Engineering Research Laboratory, University of California,
News Quarterly, 10:5, July 1960.
33. Sawyer, C. N., Some New Aspects of Phosphates in Relation to Lake
Fertilization, Sewage and Industrial Wastes, Vol. 24, No. 6, June
1952, pp. 768-776.
34. Sawyer, C. N., Causes, Effects, and Control of Aquatic Growths,
Journal Water Pollution Control Federation, Vol. 34, No. 3,
March 1962, pp. 279-288.
35. Scalf, M. R., Pfeffer, F. M., Lively, L. D0, Witherow, J. L. and
Priesing, C. P., Phosphate Removal at Baltimore, Maryland, ASCE,
Journal of the Sanitary Engineering Division, Vol. 95, No. SA5,
October 1969, pp. 817-827.
36. Shull, K. E. and Guthan, G. R., Rapid Eriochrome Cyaninc R Method
for Determination of Aluminum in Water, Journal American Water
Works Association, Vol. 59, No. 11, November 1967, pp. 1456-1468.
37. Smith, Robert, Preliminary Design and Simulation of Conventional
Wastewater Renovation Systems Using the Digital Computer, Water
Pollution Control Research Series, WP-20-9, Federal Water Pollu-
tion Control Administration, Cincinnati, Ohio, March 1968.
117
-------
38. Smith, R. and McMichael, W. F., Cost and Performance Estimates for
Tertiary Wastewater Treating Processes, Robert A. Taft Research
Center, FWQA, Report No. TWRC-9, June 1969.
39. Sources of Nitrogen and Phosphorus in Water Supplies, Task Group
Report, Journal American Water Works Association, Vol. 59, No. 3,
March 1967, pp. 344-366.
40. Standard Methods for the Examination of Water and Wastewater, 13th
Ed., American Public Health Association, New York, 1971.
41. Stum, W., Discussion, International Conference on Water Pollution
Research, London, 1962, Advances in Water Pollution Research,
Vol. 2, Pergamon Press, Oxford, 1964, pp. 216-229.
42. Summary Report, Sewage Treatment Plant Sludge Thickener Additions,
The Pennsylvania State University, University Park, Pennsylvania,
prepared by Gilbert Association, Inc., Reading, Pennsylvania,
January 1965.
43. Tenney, M. W. and Stumm, W., Chemical Flocculation of Microorganisms
in Biological Waste Treatment, Journal Water Pollution Control
Federation, Vol. 37, No. 10, October 1965, pp. 1370-1388.
44. Thomas, E. A., Phosphat - Elimination in der Belebscllammanlage
von Mannedorf und Phosphat - Fixation in Seeund Klarschlamm,
Vierteljahrsch. Naturforsch. Fes. Zurich (Switzerland) Vol. 110,
pp. 419, 1965.
45. Uhler, K. A., Fixed Bed Filtration of Secondary Sewage Effluent
for Phosphorus Removal After Chemical Treatment With Aluminum,
M. S. Thesis, The Pennsylvania State University, University Park,
Pennsylvania, December 1969.
46. Vacker, D., Connell, C. H. and Wells, W. N., Phosphate Removal
Through Municipal Wastewater Treatment of San Antonio, Texas,
Journal Water Pollution Control Federation, Vol. 39, No. 5,
May 1967, pp.-750-771.
47. Water Quality Criteria, Article 301, Sanitary Water Board, Rules
and Regulations, Commonwealth of Pennsylvania.
48. Water Quality Criteria, Report of the National Technical Advisory
Committee to the Secretary of the Interior, Federal Water Pol-
lution Control Administration, April 1, 1968, p. 20.
49. Weaver, P. J., Phosphates in Surface Waters and Detergents, Journal
Water Pollution Control Federation, Vol. 41, No. 9, September
1969, pp. 1647-1653.
50. Weber, W. J., Hopkins, C. B. and Bloom, R. J., Physiocochemical
Treatment of Wastewater, Journal Water Pollution Control Federa-
tion, Vol. 42, No. 1, January 1970, pp. 83-99.
118
-------
51. Weidner, R. B., Christiansen, A. G., Weibel, S. R. and Robeck, G.
G., Rural Runoff as a Factor in Stream Pollution, Journal Water
Pollution Control Federation, Vol. 41, No. 3, March 1969, pp.
377-384.
52. Zenz, David R. and Piunicka, J. R., Effective Phosphorus Removal
by the Addition of Alum to the Activated Sludge Process, Pro-
ceedings of the 24th Annual Purdue Industrial Wastes Conference,
May 6, 7, 8, 1969, Part 1, pp. 273-301.
53. Zuckerman, M. M. and Molof, A. H., High Quality Reuse Water by
Chemical - Physical Wastewater Treatment, Journal Water Pollu-
tion Control Federation, Vol. 42, No. 3, Part 1, March 1970,
pp. 437-456.
MJ.S. GOVERNMENT PRINTING OFFICE: J972-484-483/98
119
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
1. Report No.
3. Accession No.
w
4. Title SOLUBLE PHOSPHORUS REMOVAL IN THE ACTIVATED SLUDGE
PROCESS - PART I - CHEMICAL-BIOLOGICAL PROCESS
PERFORMANCE,
7. Author(s) Longj D. A>> Nesbitt, J. B., and
Kountz, R. R.
9. Organization
The Pennsylvania State University
University Park, Pennsylvania 16802
12. Sponsoring Organization
15. Supplementary Notes
5. Report Date
6,
8, Performing Organization
Report No.
10. Project No.
11. Contract I'Grant No.
17010 EIP
13. Type of Report and
Period Covered
16. Abstract
It was the objective of this research to develop and evaluate, at full plant scale,
the combined chemical-biological process of phosphorus removal. Phase I inves-
tigations indicated an Al/P (filt. ortho) weight ratio of 2.25/1 was necessary to
reduce the influent phosphorus of approximately 10 mg P/l to approximately 0.3 mg
P/l in the filtered effluent. Alum proved to be a more effective precipitant than
sodium aluminate in the moderately alkaline wastewater available for this study.
The best results were obtained when the chemical was added at or near the effluent
end of the aeration tank. Total phosphorus concentration of the unfiltered effluent
was dependent upon effluent suspended solids levels as well as on effluent soluble
phosphorus concentrations. Removal of organic matter was improved as a result of
chemical addition in the chemical-biological process. The chemical-biological
process produced approximately twice as much weight of sludge as did the parallel
control. Alum addition did not reduce the sludge volume index (SVI) of the mixed
liquor whereas sodium aluminate addition resulted in significant decreases. Costs
for chemical precipitation of phosphorus in the chemical-biological process are
estimated to vary from 7.3 cents/1000 gal in a 1 MGD plant to 4.1 cents/1000 gal
in a 100 MGD plant for 97% removal.
17a. Descriptors
#Phosphorus, ^Chemical precipitation, ^Activated sludge, *Aluminum, *Efficiencies,
^Nitrification, Wastewater treatment, Effluents
17b. Identifiers
^Operations, #Alum, #Aluminate, *Chemical removal, Waste sludge, Dosages, Costs
17c. COWRR Field & Group
18. Availability
19. Security Class,
(Report)
20. Security Class.
(Page)
Abstractor
D. A. Long
21. No. of
Pages
22. Price
Send To:
WATER RESOURCES SCIENTIFIC INFORMATION CENTER
U.S. DEPARTMENT OF THE INTERIOR
WASHINGTON, D. C. 20Z4Q
lnstitutionve.pt. of Civil Engrg., Penn State University
WRSIC 102 (REV. JUNE 1971)
SPO 913.261
------- |