EPA-670/2-73-060
August1973
Environmental Protection Technology Series
nhancing Trickling Filter Plant
Performance By Chemical Precipitation
fice of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and
Monitoring, Environmental Protection Aqency, have
been grouped into five series. These five broad
categories were established to facilitate further
development and application of environmental
technology. Elimination of traditional grouping
was consciously planned to foster technology
transfer and a maximum interface in related
fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
**. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL
PROTECTION TECHNOLOGY series. This series
describes research performed to develop and
demonstrate instrumentation, equipment and
methodology to repair or prevent environmental
degradation from point and non-point sources of
pollution. This work provides the new or improved
technology required for the control and treatment
of pollution sources to meet environmental quality
standards.
-------
EPA-670/2-73-060
August 1973
ENHANCING TRICKLING FILTER PLANT PERFORMANCE
BY CHEMICAL PRECIPITATION
by
Robert E. Derrington
David H. Stevens
Jaimes E. Laughlin
For the: City of Richardson, Texas
Richardson, Texas 75080
Grant No. S800685
Project 11010 EGL
Project Officer
Richard C. Brenner
U. S. Environmental Protection Agency
National Environmental Research Center
Cincinnati, Ohio 45268
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D. C. 20460
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20102 - Price $1.45
-------
EPA REVIEW NOTICE
This report has been reviewed by the Office of Research and
Development, EPA, and approved for publication. Approval does
not signify that the contents necessarily reflect the views
and policies of the Environmental Protection Agency, nor does
mention of trade names or commercial products constitute
endorsement or recommendation for use.
ii
-------
ABSTRACT
Two years of plant-scale studies indicated metal addition was an
effective effluent polishing technique at this conventional waste-
water treatment plant. Effluent phosphorus (P), five-day BOD and
suspended solids were reduced to 0.5, 5, and 7 mg/1 respectively.
Aluminum sulfate was more effective than ferric chloride. Alum
addition ahead of the final clarifier proved the best arrangement.
An optimum mole ratio (metal/phosphorus) of 1.6 developed; tnis
ratio shows moles of aluminum fed per mole of incoming total phos-
phorus. Chemical costs, of which one-third was for transportation,
were 5 cents per 1,000 gallons of flow treated, or 36 cents per
pound of phosphorus removed when in the 96 percent removal range.
Chemical addition doubled the volume of digested sludge but de-
watering on sand beds took half as long as previous conventional
operations. During this demonstration the treatment system re-
ceived some 1.6 MGD of typical domestic discharge, essentially
its design loading. Hydraulic loading on clarifiers was minimized
by drastic reduction of recirculation flows.
This report was submitted by the City of Richardson, Texas (P. O.
Box 309, Zip Code 75080) in fulfillment of Grant Number 11010 EGL,
Project Number S800685 under the partial sponsorship of the U. S.
Environmental Protection Agency.
111
-------
CONTENTS
SECTION PAGE
I CONCLUSIONS 1
II RECOMMENDATIONS 3
III INTRODUCTION 5
Project Development and Schedule 5
Fundamentals of Metal Salt Addition 6
IV DESCRIPTION OF TREATMENT FACILITIES 9
Facilities Prior to Project 9
Modifications to Treatment Units 11
Improvements in Flow Control and Sampling 16
Miscellaneous Improvements 17
Costs of Modified Facilities 18
Pilot Tertiary Treatment Units 19
V PLANT LOADING AND CONVENTIONAL PERFORMANCE 21
VI ALUM TRIAL 27
Chemical Feed Preceding Final Clarifier 31
Chemical Feed Preceding Primary Clarifiers 33
Split Feeding to Primary and Final Clarifiers 36
General Observations 37
VII IRON TRIAL 39
General Observations 48
VIII EXTENDED ALUM RUN 51
Effect of Supernatant Treatment on Performance 54
-------
CONTENTS (Continued)
SECTION PAGE
Effect of Low Wastewater Temperature
on Performance 55
Effect of Peak Flow Rates on Performance 56
Effect of Varying Metal to Phosphorus Mole
Ratio on Performance 58
Chemical Costs 59
Effluent Sulfate Levels 60
Effluent Alkalinity Levels 61
Sludge Production 62
IX PILOT-SCALE FILTRATION AND CARBON ADSORPTION 65
X DISCUSSION 71
Overall Results 71
Equipment and Facilities 73
Clarifier Performance 74
Sludge Production, Digestion, and Drying 75
Supernatant Treatment 77
Pilot Adsorption and Filtration 78
Drainage from Sand Drying Beds 79
Things That Did Not Work 80
Unanswered Questions 82
Costs 83
XI ACKNOWLEDGMENTS 85
XII REFERENCES 87
XIII APPENDICES
89
VI
-------
FIGURES
NO. pAGE
1 Twenty-Four Month Schedule of Data Accumulation 6
2 Treatment Plant Prior to Modification 9
3 Original Batch-Type Supernatant Treatment System 10
4 Plant Modified for Chemical Addition 12
5 Conversion of Junction Box to Flash Mix Unit 13
6 Improved Continuous Supernatant Treatment System 15
7 Recirculation Sampler 17
8 Pilot Units for Filtration and Adsorption 20
9 Typical Variation in Incoming Phosphorus Concentration 22
10 Typical Variation in Incoming Phosphorus Load 23
11 Results of Alum Treatment of Digester Supernatant 24
12 Clarifier Detention Periods at Various Flows 26
13 Alum Trial Activities, Fall 1970 27
14 Jar Tests of Phosphorus Removal with Alum 28
15 Plant Performance with Too Few Alum Feed Adjustments 29
16 Plant Performance after Adjustment of Alum Feed Rates 31
17 Results of Alum Feed Preceding Final Clarifier 32
18 Effluent Phosphorus Profile When Feeding Alum Ahead
of Primary Clarifiers 33
19 Results of Alum Feed Ahead of Primary Clarifiers 34
20 Effluent Phosphorus Profile During Split Feed of Alum 36
21 Iron Trial Schedule, Winter 1971 39
22 Jar Tests of Phosphorus Removal with Ferric Chloride 40
via
-------
FIGURES (Continued)
NO. PAGE
23 Phosphorus Removal Trends when Feeding Ferric Iron
to Final and Primary Clarifiers 41
24 Results when Feeding Iron Prior to Final Clarifier 42
25 Results when Feeding Iron to Primary Clarifiers 43
26 Phosphorus and Iron Leakage when Feeding Ahead
of Primary Clarifiers 44
27 Phosphorus and Iron Leakage when Feeding Ahead
of Final Clarifier 45
28 Summary of Iron Leakage Data 46
29 Reduction of Iron Leakage with Polymer 47
30 Reduction of Iron Leakage through Treatment Units 48
31 Extended Alum Run Activities, April 1971 - March 1972 51
32 Overall Performance During Extended Alum Run 52
33 Typical Daily Performance During Optimized Control 53
34 Effect of Untreated Supernatant on Plant Performance 54
35 Plant Performance Problems from Infiltration and
Untreated Supernatant 55
36 Effect of Plant Flow on Effluent Phosphorus 57
37 Relation of Effluent Phosphorus to Flow Through
the Final Clarifier 58
38 Relationship of Mole Ratio to Effluent Phosphorus 59
39 Cost of Chemical Injected for Various Levels of
Effluent Phosphorus 60
40 Relationship of Effluent Sulfate to Effluent
Phosphorus 61
viii
-------
NO
FIGURES (Continued)
41 Relationship of Effluent Alkalinity to Effluent
Phosphorus 62
42 Schedule of Pilot Tests 65
43 Performance of Multi-Media Filter and Carbon Columns 66
44 COD Levels in Wastewater Passing Through 14-Ft
Carbon Column 68
45 Service Time Until COD Breakthrough at Various
Depths of Carbon 69
IX
-------
TABLES
NO
Data on Treatment Units 1:L
Duration of Flash Mix and Flocculation 14
Character of Inflow and Conventional Effluent 21
Loads on Biological Units and Clarifiers 25
Plant Performance During Extended Alum Run 53
Typical Low Temperature Performance 56
-------
SECTION I
CONCLUSIONS
Based on two years of mineral addition to the 1.6 MGD City of Richardson,
Texas, single-stage standard-rate trickling filter plant:
1. Characteristics of the wastewater were typical of
domestic sewage.
2. The plant was operating at near design load during the
project. Hydraulic loadings on clarifiers were kept
low by reducing recirculation to a minimum. Daily
peak flows were minimized by control of main pumps.
3. Addition of liquid alum ahead of the final clarifier was
the easiest and most effective means of chemical addition.
Mole ratios (Al/P) of 1.6/1.0 yielded consistent effluent
concentrations (mg/1) of 0.5 for phosphorus, 5 for BOD5
and 7 for suspended solids. This level of treatment
represents a significant upgrading of overall performance
compared to the capabilities of the plant when operated in
a conventional mode without chemical addition.
4. Chemical costs were 5 cents per thousand gallons flow, or
36 cents per pound of phosphorus removed, with phosphorus
removal at the 96 percent level.
5. Higher aluminum dosages, as much as 50 percent more than
cited in 3 above, produced only slightly better results.
6. Addition of liquid alum ahead of the primary clarifiers
was not as effective an approach as alum addition to the
final clarifier and caused solids handling problems in
the digesters. Split feed (simultaneously to primary and
final clarifiers) was an effective approach but was not
worth the extra effort required.
7. Liquid ferric chloride addition did not work as well as
alum addition. Discrete iron colloids escaped the treat-
ment system under all iron addition approaches tried.
Effluent quality did not match alum treatment even at iron/
phosphorus mole ratios of 2/1.
8. Polymers aided in the capture of discrete colloids but
were not needed when alum addition was performed properly.
-------
9. To prevent phosphorus breakthrough/ the rate of metal
addition had to be matched with the incoming phos-
phorus load. This required changing pump settings four
times per day.
10. Separate supernatant treatment was necessary for optimum
plant performances. A single continuous alum treatment
system reduced pollutants in supernatant to levels
below those found in raw wastewater.
11. Residual pollutant levels in alum clarified trickling
filter effluent were reduced by 50 percent or more after
passage through a tertiary pilot-scale high-rate multi-
media filter unit. When pilot carbon columns were added
to the tertiary sequence, pollutants were reduced to near
trace levels.
12. Chemical treatment doubled the volume of anaerobically
digested sludge to be handled; however, the digested
sludge could be dried on and removed from sand beds in
half the usual time.
-------
SECTION II
RECOMMENDATIONS
The personnel involved in this project have pooled their observations
and experiences to offer suggestions to others considering similar
operations.
To those who wish/ outside of chemical addition, to improve overall
plant performance:
1. Treat supernatant from anaerobic digesters before
returning it to the plant inflow. Such treatment
can involve a simple continuous system which yields
a very treatable return liquor. Cost should be less
than one quarter cent per thousand gallons of plant
flow.
2. When sand beds are "stripped" of dry sludge, take
time to maintain them. Fluff or scarify the sur-
face of the beds and let the sand dry. Add coarse
sand as required to offset attrition. These simple
measures can cut drying time by one-third and pro-
duce bed underflow of secondary effluent quality.
3. Require operators to become members of the laboratory
team. Encourage them to observe gross features such
as turbidity, color, and smell. Also, have them
make in situ tests for temperature, pH, dissolved
oxygen and settable solids. More to the point,
teach operators to respond to changes in these
parameters rather than report them for historical
record only.
To those who wish to enhance effluent quality by chemical addition at
a standard-rate trickling filter plant:
1. Provide continuous around-the-clock operation at the
plant.
2. Equalize flow through the plant by every available
means until it is as near constant as possible. This
does not infer recirculation is equalization; the
aim is to equalize untreated inflow.
3. Reduce hydraulic loading on final clarifiers to a
minimum. Five hundred gallons per day per square
foot (based on surface overflow) is a realistic
and important goal.
-------
4. Reduce debilitating effects of strong local currents in
all clarifiers by providing velocity dissipation of
inflow. Use chemical floe as a tracer to indicate inflow
baffling needs and progress.
5. Take all possible means to increase low energy floccula-
tion and settling times in clarifiers. This infers
increasing clarifier design depth in new plants, and
possible addition of skirts or baffles in the feed
wells of existing units.
6. Monitor and characterize incoming phosphorus levels.
Adjust chemical feed rate to match changes in the rate of
incoming phosphorus. Monitor the effluent to evaluate
phosphorus removal effectiveness.
7. Pipe chemical feed facilities to provide several possible
feed points. Try all of these/ and combinations of several,
to determine the best chemical dosing regimen for a
particular plant.
8. Insure flash mix operations are truly high energy complete
dispersion operations. G-values should exceed 500 for
metal dispersal, but should be reduced to 25 to 100 for
polymer mixing.
9. Do not draw digested sludge too deeply into drying beds
(10 or 11 inches is typical maximum) or chemical sludge
will compress, blind the bed off, and result in long
drying times.
10. In judging the effectiveness of chemical addition,
consider changes in suspended solids, turbidity, bio-
chemical or chemical oxygen demand, total organic carbon,
and coliforms as well as phosphorus.
-------
SECTION III
INTRODUCTION
The overriding purpose of this study was to demonstrate that controlled
in-plant chemical addition could substantially upgrade overall perfor-
mance of a full-scale standard-rate trickling filter plant. One major
objective was operation of the plant to reduce phosphorus concentration
(as P) to a level of one mg/1 or less. Other goals included reduction
of BODn and suspended solids to concentrations of 15 mg/1 or less. It
was felt that improving treatment in a conventional plant to these levels,
consistently and economically, would serve as a valuable test case and
could thus make a contribution towards enhancing performance of thousands
of other trickling filter plants in use today.
Facilities for this plant-scale study were provided, to the greatest
extent possible, through modification of existing treatment units.
Existing facilities were adapted to new roles rather than installing
major new units alongside them. The scope of this report includes a
description of modifications undertaken, a description of chemical feed
equipment and dosing options, a summary of pertinent results, and dis-
cussion of those results.
The only tertiary treatment evaluation undertaken in this project
involved pilot-scale units for multi-media filtration and carbon absorp-
tion. Operation of these pilot-plant units was considered a secondary
study, and is presented in that perspective in this report.
PROJECT DEVELOPMENT AND SCHEDULE
In the middle 1960's, operators of the Richardson, Texas treatment plant
began to add chemicals, of several types and in various ways, in attempts
to improve plant performance. Initial results were encouraging and, in
1966, a letter was written to the precursor of the United States Environ-
mental Protection Agency inquiring whether there was merit in an expanded
plant-scale investigation. A favorable response from the Agency led to
further developments and an application for support of the present study
was submitted in April, 1969. A research and development grant was
awarded to the City in June, 1969.
It took a full year to complete detailed plans, order and receive
equipment, and complete all phases of plant construction. However, it
was possible to make operational baseline studies during the latter
portion of that first year (Figure 1).
-------
I97O
A I M| J | J[A S | O INI
1971
A|M|J|J|A|S[0]N|D|J|F|M
1972
PLANT
BASELINE
RUN
ALUM
TRIAL
IRON
TRIAL
EXTENDED ALUM RUN, INCLUDING A
VARIETY OF OPERATING CONDITIONS
PLANT SCALE MINERAL ADDITION
PIUDT SCALE HLTRATION-ADSORBTION
RUN
RUN 2
RUN 3
FIGURE 1. TWENTY-FOUR MONTH SCHEDULE OF DATA ACCUMULATION.
Trial operations with liquid alum were carried out during the last
quarter of 1970. A trial with ferric chloride was made during the
first quarter of 1971. Finally, after alum was selected as the more
promising chemical additive, an extended alum run was made during the
next eleven months through March, 1972.
FUNDAMENTALS OF METAL SALT ADDITION
These introductory comments are not intended to detail an extensive
literature search nor do they offer an extended theoretical considera-
tion of the chemistry involved. Instead, several fundamentals dictated
the physical arrangement and mode of operation and, in retrospect, seem
to be important insights into assembling a chemical addition system for
a conventional secondary treatment plant. Those fundamentals are dis-
cussed below.
Chemical addition has two major overall functions: (1) precipitation of
phosphorus and (2) removal of the greatest possible amount of colloids.
It was assumed that phosphorus precipitation was brought about solely
by metal salts. Conversely, coagulation or destabilization of colloids
was considered a function of both metals and polymers. Finally, floc-
culation of destabilized colloids was also presumed a function of both
metals and polymers.
The first function mentioned, precipitation of phosphorus, involves some
obscure reactions. Recht and Ghassemi (1) have undertaken explora-
tions into this field and offer numerous references to other investiga-
tions. Reaction products appear to be a variety of metal phosphates and
-------
related species, according to a study committee of the American Water
Works Association (2) and Theis, et al (3). An important factor in
kinetics has been demonstrated: the precipitation reaction is essential-
ly complete in less than one second. Addition of acid metal salts will
depress pH, but this would rarely interfere with phosphorus precipita-
tion. Since polymers are not involved in this reaction; their addition
should be deferred until this phase is completed.
The second function, coagulation of all types of colloids in the water
phase, has received more intensive study over recent years. Again, the
American Water Works Association has recently published an extensive
committee report on this subject (4). It is now widely accepted that
metal coagulation is a very rapid reaction, taking place in less than
one second. The resulting metal colloids appear to be extremely compli-
cated and probably involve a series of related polymers as described by
Bilinski and Tyree (5) . Polymers also induce coagulation but their
reaction rates are on the order of seconds to minutes in duration. The
very process of biological treatment of wastewater apparently creates a
separate variety of natural polymers; their role is obscure at present
according to Dean (6) and Busch and Stumm (7) . In any event, the function
of coagulation in this project was considered to be destabilization of
discrete colloids so that they might be flocculated and separated from
the wastewater.
The final function of the chemicals added was promotion of flocculation
or progressive agglomeration of the colloids into solids which could be
physically separated and handled. Flocculation is certainly the most
visible and probably the best understood of the reactions mentioned here.
The process of flocculation was visualized and physically provided for
via the same general approach used in present water treatment technology.
When taken together, the fundamentals reviewed here fairly well dictated
the physical facilities which would be required. For each point of chemi-
cal addition, flash mixing of metal salts was provided for a period of a
few seconds. Following that, high energy flocculation was established for a
period of one to five minutes followed finally by low energy flocculation for
approximately five to twenty minutes. Polymers were added at a point some
two minutes into the high energy flocculation phase. In all cases, the
facility requirements were modest and were largely inherent in the existing
treatment units.
More specifically, the only precalculated addition of chemical reaction
(precipitation, coagulation, or flocculation) equipment was for the purpose
of flash mixing. Available hydraulic conditions were used to promote a.
reasonable degree of flocculation. Although this was the least expensive
approach, it also entailed the greatest risk of operational difficulty. How-
ever, this approach may very well be the one that would most likely be taken
at other plants.
Finally, the operators involved in this project were those who, except for
normal personnel turnover, worked at the plant both before and since
-------
the time of the study. In terms of experience (5 year average) and
training they probably are near the norm for the operating profession.
The question of whether they could meet the challenge of understanding
chemical precipitation and applying this understanding to control of the
equipment was a very real part of the effort reported here.
-------
SECTION IV
DESCRIPTION OF TREATMENT FACILITIES
Treatment facilities at the City of Richardson, Texas consist of a con-
ventional secondary system designed to handle domestic wastewater. The
first phase was built in 1953; the addition of a parallel unit in 1961
extended plant capacity to 1.6 MGD. Modifications for chemical addition
in 1969 were minor and did not materially alter the basic facilities or
flow pattern. Tertiary pilot plant units were added beside the final
clarifier in 1971.
FACILITIES PRIOR TO PROJECT
The plant is a typical standard-rate single-stage trickling filter system.
Plant facilities are shown in Figure 2.
• FINAL SLUDGE 8 RECIRCUlATION
INFLOW
PRIMARY
CLARIFtERS
a
DIGESTERS
(UNHEATED)
EFFLUENT
FINAL
CLARIFIER
1
1
DRYING
1
|
1 1
BEDS
1 1
FIGURE 2. TREATMENT PLANT PRIOR TO MODIFICATION.
-------
A mechanical bar screen precedes a wet well serving four raw sewage
pumps which lift the wastewater into a flow splitter box. Proportional
weirs there divide flow between three clarifier-digesters. Primary
effluent is combined in a splitter box, then divided and sent to two
standard-rate rock filters. Filter effluent is combined and carried to
the final clarifier. Chlorination and settling occur at the same time
in that clarifier. A mixture of settled trickling filter humus and recir-
culated effluent are drawn from the bottom of the final clarifier and
returned to the head of the plant, the amount of recirculation usually
being regulated by a level control system in the raw sewage wet well.
Sludge is digested in the lower compartment of each primary clarifier-
digester. No heat is provided (gas is wasted through a burner) and mixing
consists of gentle stirring by a 3 rph mechanism revolving on the same
shaft as the clarifier rakes above.
Digested sludge is dried on sand beds. Filtrate collected in the under-
drains flows back to the head of the plant. Prior to the grant project,
digester supernatant was drawn and batch-treated before return to the
head of the plant as shown in Figure 3.
DRYING
BED
SLUDGE
MAIN
WET WELL
f UNTREATED
(SUPERNATANT
PORTABLE
COMPRESSOR
FIGURE 3. ORIGINAL BATCH-TYPE SUPERNATANT TREATMENT SYSTEM.
(Equipment was modified during project as shown in Figure 6.)
10
-------
Three 500-gallon filL-and-draw tanks received raw supernatant. Approximately
250 mg/1 of alum was added followed by 20 minutes of air agitation to yield
a finished liquor which separated into sludge (which went to drying beds)
and treated supernatant with strength comparable to raw sewage.
Detailed plant data are condensed into Table 1.
TABLE 1
DATA ON TREATMENT UNITS
Prim. Clar. No. 1
2
3
All Prim. Clar.
Final Clarifier
Filter No. 1
2
Filters Combined
Digester No. 1
2
3
Digesters Combined
Diam
(Ft)
40
40
40
70
84
120
40
40
40
Depth
(Ft)
8
10
10
6.5
6.5
14.3<2)
14.3<2>
14.3(2)
Sludge Drying Beds 12,000 Square Feet
(1)
Area
(Sq Ft)
1257
1257
1257
Volume
(Cu Ft) (Gal)
3771
3848
5542 U)
11310(1)
16852(1)
1257
1257
1257
10,054
12,570
12,570
35,194
23,088
36,000
73,500
109,500
13,000
13,000
13,000
39,000
75,200
94,000
94,000
263,200
173,000
97,000
97,000
97,000
291,000
(2)
Area in acres: 0.127, 0.260 and 0.387, respectively
14.3 Effective, 18.0 SWD, 15.8 Clear @ Center
Before the present study began, existing facilities were brought to
their best mechanical efficiency. All three digesters were drained and
cleaned. Flow meters were recalibrated.
MODIFICATIONS TO TREATMENT UNITS
In this project, chemical treatment was intended as an adjunct to the
physical and biological treatment already provided. Further, it was
11
-------
intended that required modifications be as simple as possible and that
existing facilities be fully utilized.
Early in the project two coagulants were selected for operational trials;
(1) aluminum in the form of liquid alum and (2) iron as liquid ferric
chloride. Both were available in bulk from commercial firms at haul
distances of about 250 miles. Both were similar enough in character
to permit use of common storage and feeding hardware. Use of polymers
was also projected based on the assumption they would be worthwhile in
improving settling characteristics of solids involved.
At: this point the arrangement of the plant was modified to permit
addition of these chemicals in the main wet well and just ahead of the
final clarifier as indicated in Figure 4.
FINAL SLUDGE 9 RECIRCULATION
EFFLUENT
\
DRYING
1
1
1 1
BEDS
1 !
; 4. FJjANT MODIFIED FOR CHEMICAL ADDITION.
One 6000-gallon fiberglass tank was installed for storing liquid coagu-
lant in a central location. This capacity was sufficient to receive
tanktruck lots of any chemical considered. Two chemical feed pumps
were installed beside the tank and piped to deliver to either the head
12
-------
or effluent ends of the plant, or both simultaneously. Both pumps
included variable feed controls covering their 0-110 gph discharge
range. Pump controls also were equipped with automatic-manual capability.
Wetted parts were selected of materials resistent to alum, ferric chloride,
sodium aluminate, and pickle liquor.
Two 1200-gallon fiberglass polymer storage tanks were provided, one near
the plant influent sewer and the other near the final clarifier. Both
had feed pumps similar to the pair at the coagulant tank. Both polymer
stock tanks were fitted with eductor assemblies for dissolving polymer,
and 3-hp mixers for blending fresh batches of polyelectrolytes.
The junction box preceding the final clarifier was modified to provide
flash mixing of coagulant. The change involved baffling off a section
and installing a 3-hp mixer to promote rapid dispersal of the metal
salts injected. The arrangement is shown in Figure 5.
FIGURE 5. CONVERSION OF JUNCTION BOX TO FLASH MIX UNIT.
(Physical changes were simple, but carefully planned.)
13
-------
The mixer delivered a measured 2.2 water hp, making the approximate
velocity gradient, or G-value, equal to 650/second. At the average
flow of 1.6 MGD, detention was some 50 seconds, so Gt equaled 32,000.
Water confined in front of the baffle was considered in the mixing zone.
After flowing over the baffle, water in transit through the remainder of
the junction box, through the clarifier inlet pipe, or in the clarifier
centerwell was considered in a high energy flocculation zone. Observa-
tion of floe-laden water just outside the centerwell allowed an estimate
of volume undergoing low energy flocculation. Volumes involved were:
Flash Mix 1,000 gal
High Energy Flocculation 4,000
Low Energy Flocculation 20,000
TOTAL VOLUME 25,000 gal
Table 2 shows nominal detention times in the above zones under different
rates of flow.
TABLE 2
DURATION OF FLASH MIX AND FLOCCULATION
Flow Rate Coagulation Flocculation Time (Minutes)
MGD GPM (Minutes) High Energy Low Energy Total
1 700 1.42 5.71 28.6 34.3
1.5 1,050 0.95 3.81 19.1 22.9
2 1,400 0.71 2.86 14.3 17.2
2.5 1,750 0.57 2.28 11.4 13.7
3 2,100 0.48 1.91 9.5 11.4
Kinetic energy of turbulent flow entering the wet well was used for
flash mixing coagulants added to raw sewage. Chemicals were injected
at a manhole to initiate the mixing process in a ten-foot length of
steeply descending sewer carrying plant inflow to the wet well. Dis-
persal was completed in a confined receiving zone in the wet well.
After a brief (and indeterminant) stay in the wet well, incoming flow
was pumped into the splitter box preceding the primary clarifiers. Deten-
tion time was short and energy levels were fairly high from pumps to
14
-------
primary clarifiers. This was followed by flocculation in the centerwell
area of the clarifiers. An estimate of the size of the flocculation
zones in this section of the plant was not possible.
In summary, coagulant dispersal and flocculation in raw sewage took
place at ill-defined energy levels and reaction periods. The arrange-
ment was probably not as effective as that provided after the trickling
filters.
At both injection points in the plant, polymer injection facilities
delivered into high energy flocculation zones. Polymer stock solution
water was metered, then mixed with 20 gpm carriage water and jetted
into a hydraulic regime where there was sufficient turbulence to promote
dispersal. A two-minute lag time was intended between injection of
coagulant and addition of polymers.
In 1971, continuous supernatant treatment evolved from the original
batch system. As shown in Figure 6, one of the three 500-gallon tanks
was piped to serve as a flow-through chemical addition chamber with air
agitation. The remaining two tanks were rigged to serve as settling
vessels, working in series as shown, or parallel. With inflow diverted
upwards at midpoint, both settling tanks offered half their 500-gallon
volume to the 30 gpm flow, yielding some eight minutes detention in
each tank. This arrangement reduced required operator attendance and
produced a high quality treated supernatant.
WATER „ WET
* WELL
BEDS
UNTREATED
(SUPERNATANT
FIGURE 6. IMPROVED CONTINUOUS SUPERNATANT TREATMENT SYSTEM.
(Total of 30 minutes of treatment included chemical treat-
ment plus two-stage settling. Operation became largely
unattended.)
-------
IMPROVEMENTS IN FLOW CONTROL AND SAMPLING
One primary clarifier was found to suffer poor inlet hydraulics due to
the piping arrangement into its centerwell skirt. A combination splitter-
deflector was fabricated and installed to redirect inflow and served to
effectively dissipate velocity in the centerwell.
There were some inaccuracies in control of flow in the splitter box
preceding the trickling filters. Previously, this had been regulated
by manual adjustment of sluice gates. Proportional weirs were fabri-
cated and installed in the box, insuring an accurate division to the
two different size filters at all rates of flow.
Recirculation flow (which included settled trickling filter sludge) had
not been sampled and analyzed previously. The flow was drawn from the
bottom of the final clarifier through a gravity line to the raw sewage
wet well. A vault housed a flow meter and an air-operated throttling
valve at the midpoint of the line. Facilities for automatic sampling
were installed in that vault to function as shown in Figure 7.
A continuous sample flow was withdrawn and split between a constant
head shunt and a sample shunt which normally diverted to drain. When
the flow meter generated a signal indicating flow in the recirculation
line, that same signal energized a solenoid diverting sample flow to a
receiving can. The amount of sample caught was proportional to the amount
of recirculation flow.
The throttling valve on the plant recirculation line had, for years,
been controlled by water level in the wet well. Recirculation occurred
on a demand basis, making up the deficit between the selected pumping
rate and plant inflow. For reasons discussed later it became necessary
to sharply reduce this flow. An electric timer was wired into the valve
control circuit in a manner which allowed it to override other signals.
This timer was eventually set to trigger a 25-second flushing flow every
20 minutes; this pattern established a 70,000 gpd recirculation rate.
Underdrain facilities beneath the sludge drying beds collected filtrate
from wet sludge. Underflow from seven beds drained to either of two
filtrate manholes, and from there to the wet well. The manholes were
partially dammed and a sump pump placed in each. A standard water meter
was installed in the discharge line of each pump to record bed drainage.
Two magnetic flow meters were added to the supernatant treatment system,
one measuring raw flow coming in, the other measuring treated superna-
tant returning to the head of the plant. The difference between their
cumulative readings gave the volume of precipitated sludge drawn to
drying beds.
Three gas meters, one for each digester, were installed to permit accurate
measurement of gas generated during sludge digestion.
16
-------
n
CONTINOUS
FLOW AT
CONSTANT
HEAD
DRA1N
j~ TIME-PULSE
1
Ik RECIRCULATION *-
FLOW
METER
3
SIGNAL TO RECORDER
FIGURE 7. RECIRCULATION SAMPLER.
(Although a mechanical success, irregularities in quality of
recirculation flow made this system a practical failure.)
MISCELLANEOUS IMPROVEMENTS
A small manually adjusted chlorinator had served the plant adequately
for some years. At the time of the project it was replaced by an _
automatic 2000 ppd unit with compound-loop automatic controls. This
changeover was undertaken more to improve overall operation than to
modify the plant for chemical addition. After trial operations, the
wastewater flow sensing leg was disconnected in the compound-loop
control as chlorine flow in this installation could be adequately
controlled by utilizing only the automatic residual analyzer.
17
-------
A recording pH meter and a dissolved oxygen meter were installed to
monitor plant influent and final effluent, respectively. Both supplied
information valuable to the demonstration project, but neither would be
necessary when modifying a plant to add chemical precipitation.
Ten small pipelines, mostly PVC, were installed at various locations
around the plant. These delivered coagulant, polymer dilution water,
diluted polymer, rinse water, and sample flows to designated receiving
points.
A duplex strainer was added to the plant water system. This served to
remove occasional large solids from treated effluent used for chlorine
water supply and other similar needs. Also, a larger pressure tank and
a new compressor were installed to upgrade the plant water system.
COSTS OF MODIFIED FACILITIES
Part of the Richardson treatment plant was built in 1953 at a total cost
of $75,000, exclusive of land. In 1961, treatment facilities were
enlarged to the present arrangement at an additional cost of $250,000,
excepting land.
A new laboratory building was added in 1969 at a total cost of $33,000.
This facility is considered an integral part of the treatment system,
but perhaps one-third of its cost went to extra space for the demonstra-
tion project.
The laboratory was outfitted at a cost of $11,000. Furniture accounted
for $2,000 of the total, and the $9,000 balance went for equipment and
supplies. Some of the equipment including a $3,000 zeta meter, a $600
recording pH meter, a $700 recording dissolved oxygen meter, and some
$700 worth of special glassware and chemicals would not have been
required except for the demonstration project. The remaining $4,000
in laboratory facilities would have been spent just to support chemical
coagulation in the treatment plant; major items included $800 for an
advanced type jar test apparatus, an $800 analytical balance, and a
$500 spectrophotometer.
If the laboratory had been built solely to support chemical coagulation
in the treatment plant, it would have cost about $21,000 for the building
and $6,000 for furniture, equipment and apparatus.
Modifications to the treatment plant exclusive of the laboratory building
and its equipment came to a total cost of $53,000 distributed as follows:
18
-------
Materials $35,000
Labor 7,000
Supervision 3,500
Design and Misc. 7,500
TOTAL $53,000
Construction labor and supervision were provided by city personnel, and
figures cited include a factor for overhead. The high ratio of materials
to labor relates to such expensive equipment as a $6,800 automatic
chlorinator, two magnetic flow meters at $4,200 total, three fiberglass
chemical tanks at $5,000 total, two polymer mixers at $2,200 total, and
four chemical feed pumps at $6,800 total. All these items total $25,000
which is considerably more than it cost to install them. The $53,000
total is reduced to $38,000 when discounts for improvements related
only to the demonstration project and not essential for chemical addition
are made.
In summary, costs of all improvements related to chemical addition were
about $65,000 in 1970 dollars.
PILOT TERTIARY TREATMENT UNITS
In mid-1971, pilot treatment studies were added as an adjunct to the
plant-scale investigation. Facilities were provided to direct treated
effluent to a multi-media granular filter and a carbon absorption
system shown in Figure 8.
The filter consisted of 30 inches of mixed media housed in a vertical
tube with a cross sectional area of 0.11 square feet. Media consisted
of selected fractions of hard coal, sand, and garnet. Appurtenances
included headloss gage, backwash piping, surface wash nozzle, and other
items needed for a complete filtration system.
Four carbon columns were assembled in series. Each vertical column was
60 inches high and 0.0825 square feet in cross sections. The system
was loaded with a total of 28 pounds of granular 8 x 35 carbon derived
from lignite coal. Sampling points were provided for system influent
and effluent and also between each column.
All pilot units were closed and operated under pressure. A supply
pump delivered several gpm to the system. This was trimmed (by wasting)
to 5 gpm/sq ft through the filter and carbon columns.
19
-------
SECONDARY
EFFLUENT
ANTHRACITE-
SILICA-
GARNET-
2.5'
3.5'
PRODUCT
WATER .
MULTI-MEDIA ACTIVATED CARBON ADSORPTION COLUMNS
FILTER (DOWNFLOW SHOWN)
FIGURE 8. PILOT UNITS FOR FILTRATION AND ADSORPTION.
(Direction of flow in carbon columns was optional.)
Pilot units were provided without payment of rental. They were sheltered
in a skid mounted metal building. The building, electrical service,
and piping came to a total cash outlay of about $1500.
20
-------
SECTION V
PLANT LOADING AND CONVENTIONAL PERFORMANCE
Richardson's wastewater treatment plant has historically operated at a
higher degree of efficiency than is normally expected for the type of
facility involved. This is partly because the system has been attended
full time for a period of several years by qualified personnel who have
given considerable attention to cleanliness, daily upkeep and maintenance.
Data were kept, throughout the course of this investigation, to develop
broad overall values describing the character of the incoming wastewater.
Quality of effluent discharged during conventional treatment (i.e.,
without chemical addition) was documented during baseline periods. During
baseline periods the plant was operated without chlorination and without
treatment of supernatant in order to produce an effluent characteristic
of the main treatment units themselves.
Performance data were analyzed statistically to yield "averages" or a
reliable measure of central tendency. In several cases, zero end con-
straint strongly affected families of data; in such instances the geomet-
ric mean was a better "average" than the arithmetic mean. Finally, it
should be recognized that the reported values cover a broad period of
time. Climatic extremes have evened out to provide a general overview,
but the resulting values might not occur on any specific given day of
operation. The results of these baseline studies are shown in Table 3.
TABLE 3
CHARACTER OF INFLOW AND CONVENTIONAL EFFLUENT
(mg/1 unless noted)
Influent Effluent
Flow (MGD) 1.5 1.5
Suspended Solids *155 *15
BOD5 166 *20
Phosphorus (P) *11 8
Total Kjeldahl Nitrogen (N) 24 12
Iron *0.82 0.29
Aluminum *0.25 0.10
Alkalinity *180 175
*Geometric means, all other values are arithmetic means.
21
-------
Detailed studies were also made of the rate of incoming phosphorus in
the raw sewage. A typical diurnal variation in concentration (P in mg/1)
is shown in Figure 9. The relatively constant concentration of phosphorus
entering the plant on the day shown in Figure 9 could be highly mislead-
ing. When incoming phosphorus is converted to pounds per day, by com-
bining concentration and rate of flow, a considerably more dramatic vari-
ation in phosphorus loading occurs (Figure 10). Additional studies of
phosphorus removal during conventional treatment indicated a fairly con-
sistent reduction of approximately 25% of the incoming phosphorus load.
18
c
_c
ST
3
-------
220
PHOSPHOROUS LOAD
IN PLANT INFLUENT
MID-
NIGHT
FIGURE 10. TYPICAL VARIATION IN INCOMING PHOSPHORUS LOAD.
(When reexpressed as absolute load, the rate of incoming
phosphorus fluctuated widely and was a major factor in
control of chemical addition.)
amount drawn, some 2 percent to 3 percent of the total daily flow, is
probably greater than handled in most plants; consequently, the liquor
is rather dilute but if returned to the head of the plant untreated it
has a pronounced effect on the overall system. The effects of batch
alum coagulation, aeration, and settling on the quality of the super-
natant are shown in Figure 11.
Actually, these results were taken before supernatant treatment was fully
optimized. Efficiencies improved during the course of the study as con-
tinuous supernatant treatment was implemented, and treatment costs were
reduced to approximately 0.1* per thousand gallons of raw wastewater flow.
23
-------
2600
I8OO
1600
74
TOTAL SOLIDS SUSP SOLIDS
BOD,
LEGEND
UNTREATED
TREATED
I8OO
460
IOO
65
7.5
TR
COD
PHOSPHORUS
ALUMINUM
SULFIDES
FIGURE 11. RESULTS OF ALUM TREATMENT OF DIGESTER SUPERNATANT.
(Before and after values of pollutant concentrations (mg/1)
show what can be accomplished for a cost of less than 0.2£
per 1000 gallons of plant flow.)
In order to interpret plant performance during conventional operation,
actual loadings on several of the major treatment units of this standard-
rate trickling filter plant were calculated over a two year period of
time. The results are shown in Table 4.
Notice that hydraulic loadings on the clarifiers were considerably reduced
because the plant was deliberately operated at minimum recirculation.
From the very outset, it was considered highly important to be able to
operate clarifiers at average surface loading rates of 400 to 500 gpd/sq ft.
In a similar vein, the rate of wastewater flow through the plant was nor-
mally kept less than or as near 1000 gpm as possible. In order to do this,
operators took full advantage of storage capacity in the main wet well
and incoming sewers to reduce peak sewage flows.
24
-------
TABLE 4
LOADS ON BIOLOGICAL UNITS AND CLARIFIERS
(All calculated loadings exclude recirculation)
Trickling Filters
Ib BOD5/1000 cu ft/day
mil gal/acre/day
Clarifiers
Primary: gpd/sq ft
Final: gpd/sq ft
Typical
Design
10-20
2-4
900
800
Plant
Design
10.4
*4
*415
*410
Actual
Observed
14
5
450
440
*Designed to include 100% recirculation.
Finally, relatively stable flows allowed estimation of clarifier deten-
tion times at different flows, an important factor in predicting lag
time through the plant. Hyperbolic equations describing assumed plug
flow are plotted in Figure 12.
Although the plot is only an approximation of actual conditions (and
this is further compounded when assuming plant detention time equals
clarifier detention time, i.e., ignoring transit time in pipes, splitter
and junction boxes, and through the filter media), it did prove most
helpful in predicting period of passage through the system.
25
-------
FLOW(MGD)
PRIMARY CLARIF
2345
DETENTION (HOURS)
FIGURE 12. CLARIFIER DETENTION PERIODS AT DIFFERENT FLOWS.
(At a given rate of flow, detention periods in clarifiers
could be used to estimate travel time through the plant.)
26
-------
SECTION VI
ALUM TRIAL
The initial trial efforts using liquid aluminum sulfate (alum) took
place in the fall of 1970. The schedule called for a three-month period
during which alum would be added just ahead of the final clarifier, or
ahead of the primary clarifiers, or split-fed to both locations simul-
taneously. Figure 13 shows generally how time was allocated during this
test period.
STARTUP
PROBLEMS
FINAL
15
SEPT
PRIM-
ARY
BASELINE
8 OCT
SPLIT FEED
BASE
19
NOV
FINAL
DEC 9
FIGURE 13. ALUM TRIAL ACTIVITIES, FALL 1970.
(Baseline periods were used to dampen effects
of one arrangement before trying another.)
Alum was fed during some two-thirds of the period, the remainder of the
time being taken up with startup problems and baseline runs. Supernatant
was not treated during this phase and trickling filter effluent was not
chlorinated.
A program of jar testing was carried out before plant scale additions
began. Figure 14 shows a plot of phosphorus remaining versus alum fed
after the samples had been stirred and settled. To reduce residual
phosphorus to 0.5 mg/1 as P, a metal dose of two moles aluminum per mole
phosphorus was predicted from the jar testing, and this was reasonably
close to*what was found in plant-scale operations.
Typical startup problems delayed progress during the early part of this
test period. Usable data were not generated during the first two weeks.
Data taken from that point on is considered reasonably valid and repre-
sentative of plant-scale performance; however, plant operations were
27
-------
in
u)
DC 2
O
tr
O
£
(0
O
X
a.
o—
2 3
MOLE RATIO' AKUD/P
FIGURE 14. JAR TESTS OF PHOSPHORUS REMOVAL WITH ALUM.
(Raw plant inflow was treated and settled, then ana-
lyzed to yield results shown here.)
evolving towards fixed operating techniques during this period so there
is some trending in the records.
Physical startup problems involved repair of leaks in pipelines, recali-
bration of chemical delivery systems, modification of original treatment
equipment to improve its performance, and other physical improvements.
Careful observation during this initial chemical feed period resulted
in further modifications to maximize utilization of existing treatment
facilities as illustrated in the following example. When alum was fed
to wastewater, the resulting floe served as a highly visible tracer of
hydraulic patterns within settling basins. During dosage to the primary
clarifiers, it became apparent that one of the three suffered very poor
inlet distribution. Inlet baffling was designed and installed within a
period of two or three days and the situation was brought under control.
However/ this occurred during the second month of the three-month alum
28
-------
trial because the hydraulic shortcomings in the primary were not apparent
during the first month when alum was being fed ahead of the final clari-
fier.
Adjustments in operating technique were also made during this period. A
typical change was made after observing plant operations illustrated in
Figure 15. When the alum feed rate was adjusted only twice during a
24-hour period, the resulting effluent phosphorus concentration ranged
out of control during the evening peak load. This situation was corrected
by introducing a feed schedule involving four rate changes per day, and
finally five changes per day, both of which gave considerably better
performance even though the total gallons of chemical fed per day were the
same. Eventually, the standard operating practice was to change feed pump
settings four times per day according to a fixed schedule.
EFFLUENT PHOSPHOROUS
ALUM IN FINAL AI/P =• 1.9/1
MONDAY. SEPT Zl, I9TO
PLANT INFLUENT
8.9 COMPOSITE
LIQUID ALUM FED AT Z0.5 GPH
0.8 COMPOSITE
IOA NOON
MID-
NIGHT
FIGURE 15. PLANT PERFORMANCE WITH TOO FEW ALUM FEED ADJUSTMENTS.
(Effluent phosphorus peaked sharply when underdosing developed.)
During the first week or two of chemical treatment, it was assumed that
a daily composite sample of plant effluent would provide sufficient
29
-------
information for adjusting chemical feed rates. It became readily evident
that this was not the case. Hourly grab samples were thus taken around
the clock from the final clarifier effluent trough, analyzed for total
phosphorus, and the hourly concentrations plotted for each 24-hour period.
Review of these daily charts was the only manner in which sensitive con-
trol of chemical treatment could be assured.
Since the total pounds of phosphorus coming into the plant each day could
not be predicted beforehand, incoming phosphorus load was estimated and
chemical dosage was set to meet a pre-selected mole ratio of metal to
phosphorus. A given feed rate was maintained for at least four or five
days and, although the mole ratio would vary, it stayed generally within
the range desired.
Chemical feed produced a heavy blanket of floe in the clarifier following
the point of addition. The appearance and importance of this blanket was
most impressive. A blanket was invariably present during high efficiency
performance. This observation led to a major change in recirculation of
treated effluent and settled sludge from the bottom of the final clarifier.
For years, something near one MGD of recirculation was brought back to the
head of the plant, mainly during late night hours. During chemical feed,
however, this heavy recirculation almost totally evacuated the floe blan-
ket by the following morning. As the incoming morning phosphorus load
began to peak, the chemical feed rate would be raised. Contents of the
clarifier would become turbid and diffuse until after mid-day. Effluent
quality would temporarily deteriorate to a level less than satisfactory
to meet project objectives. Despite (or because of) these circumstances,
a floe blanket would begin to develop and by approximately 6 P.M. both
the blanket and the efficiency of treatment would be highly developed.
It was also determined that a high recirculation rate was not essential
to operate the rest of the treatment plant. Even during late night hours,
there was sufficient incoming sewage to supply adequate flow to the entire
system on a once-through basis. It was at this point that the timeclock-
actuated intermittent recirculation system was installed and used from then
on.
The workload involved in phosphorus analysis and the demands involved in
sampling effluent every hour around the clock proved a powerful stimulus
in automating this test. A Technicon Auto-Analyzer was made available
on a loan basis from the U. S. Environmental Protection Agency. It was
specially adapted for determination of total phosphorus in a continuous
sample stream of final effluent water. The performance of the automatic
analyzer proved highly reliable and test results were both accurate and
repeatable. From the day the automatic system was installed it became
an important focal point in control of plant operations.
Another dramatic performance characteristic in the final clarifier was
never realized before advent of chemical addition. This involved the
effect of wind on hydraulics through the vessel. When wind velocities
30
-------
reached or exceeded 20 mph, floe in the final clarifier was blown to the
downwind side of the tank and over the weir at that point. Under these
circumstances, the floe was considered a near-perfect tracer. Therefore,
it was concluded that high winds literally pushed water along with them
and caused high weir loadings on the downwind side of the final clarifier.
This occurred to a much lesser extent in the primary clarifiers, probably
because of their smaller size.
CHEMICAL FEED PRECEDING FINAL CLARIFIER
Figure 16 shows one of the better early days when alum was being injected
ahead of the final clarifier. Plant flow was 1.75 million gallons during
this day and recirculation was held at 100,000 gallons.
3.0
V
o>
c
£
J52.0
-------
The floe that developed on this particular day was large and heavy. On
earlier days in the run, pin-point floe, appearing exactly the same as
found in water treatment plants, was produced. The undesirable pin-point
floe characteristic ceased when the chemical feed pattern was properly
established.
Figure 17 summarizes plant performance during this entire period.
10.7
8.6
2.0
PHOSPHORUS
130
66
6.2
162
86
13
LEGEND
UNTREATED
PRIMARY EFFLUENT
FINAL EFFLUENT
B 0
SUSP SOLIDS
0.5
1.2
1.3
7.2
7.2
6.9
177
165
48
ALUMINUM
PH
ALKALINITY
FIGURE 17. RESULTS OF ALUM FEED PRECEDING FINAL CLARIFIER.
(Values are mg/1, except pH.)
32
-------
Although effluent phosphorus levels were not consistently reduced to desired
low levels, effluent BOD5 and suspended solids stayed under excellent con-
trol throughout the entire period. Other parameters seemed reasonable.
CHEMICAL FEED PRECEDING PRIMARY CLARIFIERS
After 15 days of alum feed preceding the final clarifier, the point of
chemical application was changed to just ahead of the primary clarifiers.
Figure 18 summarizes phosphorus removal for a typical 24-hour period of
feeding alum to the primary clarifiers. Performance fell below ^ex-
perienced in feeding to the final. Also, in this case the alum feed rate
had not been trimmed quite enough to level out an evening hump.
8.0
6.0
4.0
2.0
0
1
S55GPH
PLANT INFLUENT
8.1 COMPOSITE
i 22 5 GPH J,
0
o
, o o
3AM NOON
— •*'
\
El
FE
SUPERNATANT
"" 6800 GAL
33.5 GPH . 215
S~
./* o
6PM
\
rFLUENT PHOSPHOROUS
ED TO THE PRIMARY AI/P=32/I
THURSDAY , OCTOBER 8, 1972
GPH .
1 B.5 GPH
,12 COMPOSITE
MID-
NIGHT
0
6AM
*
9AM
TIME
FIGURE 18. EFFLUENT PHOSPHORUS PROFILE WHEN
FEEDING ALUM AHEAD OF PRIMARY CLARIFIERS.
(Even ignoring untreated supernatant, the
results of this approach were not outstanding.)
33
-------
Figure 19 graphs the overall performance of the treatment plant when
liquid alum was injected into untreated wastewater. These figures sum-
marize the eight-day October trial period. Although there was not suf-
ficient data to yield extensive information, these bar graphs reflect
overall operation of the plant at this time.
n.4
3.5
2.6
136
45
27
LEGEND
UNTREATED
PRIMARY EFFLUENT
FINAL EFFLUENT
PHOSPHORUS
BOD,
SUSP SOLIDS
0.3
1.4
0.9
7.2
6.7
6.9
173
78
54
ALUMINUM
PH
ALKALINITY
FIGURE 19. RESULTS OF FEEDING ALUM AHEAD OF PRIMARY CLARIFIERS.
(Values in mg/1, except pH.)
34
-------
One pronounced change observed during this feeding technique was a
considerable increase in the volume of primary sludge produced. It
appeared that some 50 percent more primary sludge was produced than when
the plant was run in a conventional mode without any chemical addition.
All three primary clarifiers contained highly visible and massive blan-
kets of sludge. These zones of compacting sludge sometimes reached as
high as the effluent weirs, although generally the sludge could be held
down and moved from clarifiers into the digestion compartments.
Primary effluent was murky or turbid most of the time and this character-
istic persisted in the wastewater all the way through the final clarifier.
The difference in effluent suspended solids in this phase of the trial
and in the preceding phase was not very great, but there was definitely
a difference in the gross appearance of the water. Unfortunately, an
adequate turbidimeter was not available at this point in the project so
valid information on this characteristic could not be documented.
From time to time it was possible to see a wispy floe blanket in the
final clarifier. This blanket had the same general shape and size as
the dense blanket produced when alum was fed at that point; however, in
this case, the blanket was very diffuse and appeared essentially as a
ghost of the highly developed blanket seen earlier.
After a week of feeding alum to the primary clarifiers, problems deve-
loped in the sludge digesters. These digesters were not accessible so
that the operators could not make a careful study of the situation at
several depths. However, it appeared that stratification developed in
the digesters and involved two layers of solids; one heavy large layer
on the bottom, and one fresh light layer at the top. They were separated
by a layer of relatively clear supernatant. This assumption could be
demonstrated fairly well by observation and tests as liquid was with-
drawn from digester supernatant zones. In addition, as the upper layer
grew the supernatant became darker, its solids increased to a one per-
cent concentration, and its hydrogen sulfide concentration increased.
This distressing situation developed in the short period of just slightly
more than one week. After some discussion, it was concluded that suffi-
cient data had been collected to indicate general performance during this
phase. More to the point, it appeared there was real danger of loss of
control of the digestion operation. Therefore, this phase was terminated
and the entire plant was put on a restabilization baseline operation.
Contents of the digesters were mixed with recirculating pumps during the
next two weeks. Digestion returned to normal and there was never a signi-
ficant recurrence of this problem in all of the months that followed.
In this phase, records show that the mole ratio of alum to phosphorus was
frequently more than two, yet effluent phosphorus concentrations were
consistently greater than one mg/1. To match performance achieved when
feeding ahead of the final clarifier, some one-third more alum was re-
quired in the primaries. When mole ratios approached three, final
35
-------
effluent phosphorus was finally driven down to about one mg/1/ but of
course chemical costs were quite high at this point and the digester
problems already noted began to develop.
SPLIT FEEDING TO PRIMARY AND FINAL CLARIFIERS
Phase three of the alum trial involved feeding chemical simultaneously
to both primary and final clarifiers. The operation began by feeding
100 percent to final clarifier; then the feed was split and an increasing
fraction was fed to the primary clarifiers while keeping total gallons
fed per day the same. Two days of 100 percent feed to the final clari-
fier re-established chemical treatment following baseline operations in
the preceding phase. Then, for ten days, 80 percent of the chemical
was dosed to the final clarifier and 20 percent to the primary clarifier.
At mole ratios near 2, the results were good. Phosphorus on a typical
day shown in Figure 20 was reduced from 9.2 to 0.3 mg/1 with recircula-
tion held at less than 100,000 gallons per day. The results of this
trial were as good or maybe slightly better than those gotten when 100
percent final feed was employed.
4.0
o
c 3'°
ZL
0
in
=>
0 2.0
0
a.
en
O
a.
< ..0
o
0
1C
4.1
16.4
" •
PLANT INFLUENT
9.2 COMPOSITE
,20% OF AL
3.9 GPH
15.6 GPH
^8O% OF AL
o
o
1
EFFLUENT PHOSPHOROUS
SPLIT FEED AI/P « 2/1
WEDNESDAY, NOV 4, 1970
UM TO RAW
4.7 GPH
18.8 GPH
UM TO FINAL
„ 0
J>"o
e *
4.1 J
16.4 "
_. — -<
1 2.7 GPH
10.6 GPH
e
^-B— .
^0.3 COMPOSITE
)A NOON 6P MID-
NIGHT
,4.1
nie.4
1
6A 9A
FIGURE 20. EFFLUENT PHOSPHORUS PROFILE DURING SPLIT FEED OF ALUM.
(This approach was as effective as feed to final, but required
more effort to control.)
36
-------
Next, a 70 percent final, 30 percent primary split was tried. The
results were not as good, over a 5-day period, as when a greater frac-
tion of chemical was fed to the final clarifier. Phosphorus concentra-
tion in the effluent rose to as high as 1.2 mg/1 during part of this
effort, due partly to a 15 percent cutback in total gallons of alum fed
per day. This was done in an attempt to see if economy in chemicals
would compromise plant performance, which it did. On the eighteenth
day of split feeding, there were early but unmistakable signs that diges-
ter problems were developing in a way similar to those reported under 100
percent feed to the raw inflow, necessitating termination of the trial at
that point.
GENERAL OBSERVATIONS
As reported, there were indications that the sludge digestion operation
was experiencing distress during much of this phase. However, methane
concentration remained consistently above 80 percent and gas production
remained vigorous. Alkalinity in supernatant would be near 1,000 mg/1
at the beginning of a draw; after 15,000 gallons it would drop to 500
mg/1, indicating dilution by wastewater being drawn from the clarifier
to the withdrawal area.
The concentration of digested sludge dropped from 8 to 6 percent during
the alum addition trial and there was an increase in the volume of sludge
produced. However, digested sludge dried quite rapidly, cracking in about
three days instead of the usual six or seven. Total drying time was
cut from the previous average of about 22 days to something on the order
of 8 to 10 days. The sludge dried in larger pieces and had a light grey
film over the top surface when it was dried.
Coliform organisms were present in the final effluent within the range
of 1,000 to 10,000/ml when conventional treatment was employed and chlori-
nation was temporarily discontinued. When using alum, and still without
chlorination, plant effluent showed never more than 800 total coliforms/
ml, nor more than 30 fecal coliforms/ml.
To achieve the kind of results shown in Figure 17 when feeding alum in
the final clarifier, chemical cost approximated 4.6
-------
SECTION VII
IRON TRIAL
Liquid ferric chloride addition was evaluated during the first three
months of 1971. The iron compound was actually fed and recorded a total
of 80 days over that period of time. Figure 21 shows the metal salt was
dosed either to the final clarifier or to the primary clarifiers. Split
feed was not considered necessary to study the basic characteristics of
the iron salt as a chemical additive. Digester supernatant was not
treated and effluent chlorination was not practiced during the iron trial
runs.
FINAL
JAN
PRIMARY
4 FEB
F
1
N
A
L
MAR 25
FIGURE 21. IRON TRIAL SCHEDULE, WINTER 1971.
(Chemical was fed either preceding final
clarifier or primary clarifiers. Split
feed was not practiced in this run.)
Outside temperatures were frequently near freezing during this season
of the year. Under these conditions, incoming wastewater temperature
decreased from slightly less than 70° F to near 60° F after Pas^^
through the trickling filter and final clarifier. While these tempera-
tures are not extremely cold, they were felt important enough to receive
due consideration.
39
-------
Figure 22 summarizes the results of jar tests with ferric chloride on
raw wastewater. In this bench scale trial, the iron salt decreased
total phosphorus residuals in the decant to less than one mg/1 when the
metal to phosphorus mole ratio was 1.5 or greater.
MOLE RATIO' Fe{IH)/P
FIGURE 22. JAR TESTS OF PHOSPHORUS REMOVAL WITH FERRIC CHLORIDE.
(After treating and settling, results similar to alum treatment
were recorded.)
40
-------
Plant performance did not parallel jar test results as it did with
aluminum, however. Figure 23 shows that more than a two to one ratio
of metal to phosphorus was required to even approach one mg/1 total
phosphorus in the plant effulent. This shortcoming was one of the domi-
nant characteristics of this trial period and will be discussed in more
detail later.
o>
£ 6
LU
3 5
to
§ 3
cc
o
a.
ID
i 2
a
TREATMENT IN
FINAL CLARIFIER
TREATMENT OF
PLANT INFLOW
JL
MOLE RATIO'
I 2
IRON (HI) TO PHOSPHOROUS (P)
FIGURE 23. PHOSPHORUS REMOVAL TRENDS WHEN FEEDING FERRIC
IRON TO FINAL AND PRIMARY CLARIFIERS.
(Dosage requirements proved to be higher than predicted
in jar tests.)
41
-------
Figure 24 indicates, in generalized figures, that phosphorus, BOD^ and
suspended solids were reduced to fairly low concentrations for the type
of treatment plant involved when iron was dosed to the final clarifier.
However, these were far greater than the concentration levels desired,
and in the case of effluent BOD5 and suspended solids actually represent
a deterioration from conventional treatment without chemical addition.
Further, iron and chloride concentrations were high enough to be con-
sidered undesirable. Alkalinity levels were appropriate for the chemical
additive being administered.
12.3
9.5
2.4
205
118
28
LEGEND
UNTREATED
PRIMARY EFFLUENT
FINAL EFFLUENT
PHOSPHORUS
BOD,
SUSP SOLIDS
0.7
59
7.O
IRON
47
63
102
175
I6O
75
CHLORIDE
ALKALINITY
FIGURE 24. RESULTS WHEN FEEDING IRON PRIOR TO FINAL CLARIFIER.
(Values in mg/1, except pH.)
42
-------
When iron salt was added ahead of the primary clarifiers, significant
reduction in iron concentration resulted as shown in Figure 25. Final
effluent phosphorus, BOD5/ and suspended solids concentrations were
similar to those seen when iron was fed to the final clarifier. Alka-
linity concentrations decreased sequentially through the plant as expected.
\z.\
5.0
1.8
PHOSPHORUS
179
90
38
LEGEND
UNTREATED
PRIMARY EFFLUENT
FINAL EFFLUENT
B 0 Dc
SUSP SOLIDS
0.7
10
3.3
50
no
no
IRON
CHLORIDE
180
125
85
ALKALINITY
FIGURE 25. RESULTS WHEN FEEDING IRON PRIOR TO PRIMARY CLARIFIERS.
(Values in mg/1, except pH.)
43
-------
A more sensitive display of some of these features is provided by
hourly profiles of typical treatment days. Figure 26 shows such a pro-
file for a typical day when iron was added to the primary clarifiers.
The mole ratio of metal to phosphorus was 1.9, and the resulting cost was
5C/1000 gallons treated. Total flow to the plant and effluent recir-
culation on that day were 1.3 million gallons and 90,000 gallons, res-
pectively. Effluent iron concentration was near 3 mg/1 for much of the
day. Note that the profile of the iron concentration very closely par-
allels the profile of effluent phosphorus, which averaged 1.6 mg/1 in
this instance.
TOTAL PHOSPHOROUS (P)
1.6 COMPOSITE
(INFLOW WAS 11.0)
EFFECT OF IRON (IE)
FED TO PRIMARY F«/P« 1.9
THURS.4 MAR.I97I
NOON
6A
FIGURE 26. PHOSPHORUS AND IRON LEAKAGE WHEN
FEEDING AHEAD OF PRIMARY CLARIFIERS.
44
-------
Figure 27 summarizes a typical day when iron was being fed to the final
clarifier. In this case, the mole ratio was somewhat higher at 2.33.
At a total plant flow of 1.2 MGD, this high chemical feed rate boosted
the cost to 6.4<:/1000 gallons of flow. Even so, effluent phosphorus
was determined at 1.4 mg/1 in the composite sample taken that day.
Concurrently, the effluent iron escaping the final clarifier had a com-
posite value of more than 6 mg/1.
PHOSPHOROUS (P)
DSITE
WAS IL3)
EFFECT OF IRON (EM
FED TO FINAL Ft/P« 233
TUE, 2 FEB. 1971
NOON
MID-
NIGHT
FIGURE 27. PHOSPHORUS AND IRON LEAKAGE WHEN
FEEDING AHEAD OF FINAL CLARIFIER.
45
-------
This distressing matter of iron leakage is displayed in more detail in
Figure 28. Clearly, when ferric chloride was injected prior to the final
clarifier/ considerably more iron escaped in the effluent than when treat-
ing the raw wastewater ahead of the primaries.
10
u
3
O 4
K
I 2
IRON LEAKAGE IN PLANT EFFLUENT
WHEN TREATING IN
FINAL CLARIFIER
WHEN TREATING
PLANT INFLOW
QS 10 \2 1.4 IB IS
MOLE RATIO' Ft (IH)/P
2.0
2.2
2.4
FIGURE 28. SUMMARY OF IRON LEAKAGE DATA.
(Most escaping iron was in colloidal
form, and it caused a distinct red
color.)
46
-------
Data from the lower curve are replotted by themselves in Figure 29. In
addition, the results of four days efforts with a polyelectrolyte are
also superimposed on the graph. During this brief period/ iron dosing
to the primaries was continued while the polyelectrolyte was added to
the final clarifier. There is clear indication the polymer did reduce
iron leakage by improving entrainment and settling of colloidal iron.
5 -
4 -
z
UJ
p I
IRON LEAKAGE WHEN TREATING PLANT INFLOW
o
IRON LOST IM
EFFLUENT
W/0 POLYMER
O
IRON LOST IN EFFLUENT
W/POLYMER IN FINAL
CLARIFIER
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
MOLE RATIO- F« (m)/P
2.4
FIGURE 29. REDUCTION OF IRON LEAKAGE WITH POLYMER.
(Efforts to optimize polymer feed were not carried
further once it became clear that benefits were
possible.)
47
-------
Even without use of polyelectrolytes, iron levels were reduce sequenti-
ally throughout the plant when ferric chloride was fed ahead > f the
primary clarifiers (Figure 30). Reductions were not sufficient, however,
because effluent concentrations of 3.2 mg/1 of iron still produced a
distinct red hue.
PRIMARY
EFFLUENT
8.9 mg/l
35% REMOVAL
FILTER
EFFLUENT
5.8 mg/l
64% REMOVAL
FINAL
EFFLUENT
3.2 mg/l
FIGURE 30. REDUCTION OF IRON LEAKAGE THROUGH TREATMENT UNITS.
(Incoming wastewater had an iron level of 0.85 mg/l. Ferric
chloride was fed just ahead of primary clarified. Data are
from a 30-day period.)
GENERAL OBSERVATIONS
Floe resulting from ferric chloride addition was smaller, denser and
less gelatinous than floe produced from alum. As already indicated,
however, effluent concentrations of several pollutants were not reduced
to the desired levels. This occurred in spite of dry weather which
decreased plant flow by some 20 percent. Average flow during the iron
addition trial was 1.3 MGD, and hydraulic loading rates throughout the
treatment plant were reduced accordingly. It was felt that reduced
wastewater flow, at least in some measure, compensated for lower wastewater
temperatures encountered during this period.
48
-------
Sludge digestion, although not supported by heating systems, continued
vigorously during this period. Sludge solids concentrations stayed near
six percent. The aluminum concentration in the sludge stabilized and
began to taper off as older sludge was removed from the digestion systems.
Iron concentration began to build in the digester solids.
After the alum sludge was substantially purged from the digesters, the
digested sludge produced during iron addition dried noticeably better
than digested sludge during aluminum addition. Dry weather drying cycles
were on the order of 9 or 10 days for iron versus 10 or 12 for alum
sludge. Supernatant during iron treatment was similar to that produced
when adding aluminum.
When ferric chloride was added to the primary clarifiers, a red film was
quickly established on the stones in the trickling filters. Some addi-
tional zoogleal film thickness had been noticed with aluminum, but iron
film was thicker and more visible due to its distinctive color. The
build-up apparently reached proportions where hydraulic action caused the
slime to begin to break up and slough off. This was followed by a
general sloughing of filter flora, an unusual event for this time of the
year in this standard-rate trickling filter system. There was no clear
evidence of subsurface ponding in either trickling filter, even in the
smaller filter which contained smaller rocks.
In summary, chemical costs for iron addition were higher than when alumi-
num was fed, largely because higher metal to phosphorus mole ratios were
required with iron. Despite the higher metal dose during the iron trial,
effluent concentrations of pollutants were considerably higher than in
the alum runs. Combining polymer feed to the final with iron addition
to the primaries improved overall effluent quality, but not to a point
to put iron in a competitive range with aluminum. This coupled with the
persistent iron leakage led to the selection of aluminum as the chemical
of choice for an extended plant run.
49
-------
SECTION VIII
EXTENDED ALUM RUN
A pictorial graph of activities during this phase is shown in Figure 31.
The extended alum run covered a total period of 11-g- months. During
the entire period liquid alum was fed into the treatment system just
ahead of the final clarifier.
0 N
25 9
ARTIFICIAL PEAK
NO CHLORINE
HIGH ALUM
A* M
J
FLOWS — H H-
HIGH ALUM
J A
1971
S
I
LOW
0 N
ALUM
D
FINAL
PERIOD
^8 F M
1972
FIGURE 31. EXTENDED ALUM RUN ACTIVITIES, APRIL 1971-MARCH 1972.
(During the entire year, alum was fed just ahead of final
clarifier. Chlorination was deferred until July and super-
natant treatment until August.)
Effluent chlorination was not practiced during the first three months of
this phase. Supernatant was not treated during the first four months.
In order to evaluate the effect of high flow rates, peak flow pumping
was practiced about midway through the ll^-month extended run.
During the period designated High Alum, the aluminum/phosphorus mole
ratio approximated 1.7/1.0. The ratio was decreased to about 1.5/1.0
during the Low Alum period.
During the last three months, shown on the graph as Final Period, the
entire operation was optimized and much of the data taken then are given
detailed attention in this section of the report.
The bar graphs presented in Figure 32 give generalized results of the
treatment efforts during the entire ll^-month extended alum run. The
51
-------
I7O
11.4
8.6
155
115
110
68
72
0.49
PHOSPHORUS
4.5
B 0 Dr
7.0
SUSP SOLIDS
TURBIDITY
0.3
3.7
i.8
100
IIO
190
LEGEND
UNTREATED
PRIMARY EFFLUENT
FINAL EFFLUENT
ALUMINUM
SULFATE
ALKALINITY
FIGURE 32. OVERALL PERFORMANCE DURING EXTENDED ALUM RUN.
(Values (mg/1) represent operations during an ll^-month
period.)
average mole ratio of Al/P during the entire phase was 1.6, resulting in
a chemical cost of 5.1C/1000 gallons of flow treated. Detailed daily and
monthly average data for the extended alum run are given in Appendix C.
Flow during this period averaged 1.6 MGD. This resulted in overflow
rates of 445 gallons per day per square foot in the primary clarifiers,
and a corresponding figure of 420 for the final clarifier. Other perti-
nent data are shown in Table 5.
52
-------
TABLE 5
PLANT PERFORMANCE DURING EXTENDED ALUM RUN
Influent Effluent
PH
Temperature (°F)
Dissolved Oxygen (mg/1)
Total Solids (mg/1)
COD (mg/1)
Ratio BOD5/COD
Volatile Fraction,
Suspended Solids
7.2
70's
0.2
690
370
0.46
0.86
6.7
60's
7.2
475
42
0.11
0.90
Long periods of stable operation were logged to develop the results which
have just been presented. Figure 33 depicts a typical day when the entire
system was under good control. Flow through the plant on that day was
1.38 million gallons; recirculation was 70,000 gallons.
8
9
TOTAL PHOSPHOROUS (as P) in m
M * «
2C
SUPERNATA
TREATEC
0— O— 0—
NC
PLANT INFLUENT
12.0 COMPOSITE
5 GPH
NT I
1
^^-o-o-o-
EFFLUENT PHOSPHOROUS
ALUM IN FINAL AI/P- 1.63
WEDNESDAY, AUG.I8 . 1971
CHEM COST - 5.2 «/IOOO GAL
2Z5GPH
0.66 COMPOSITE 1
t o-«-»-^
2Q5GPH 13.5 GPH
- ^^O"^O^^
•ON 6P EJi'^T t"°>
JJ
. TYPICAL DAILY PERFORMANCE DURING OPTIMIZED CONTROL.
53
-------
EFFECT OF SUPERNATANT TREATMENT ON PERFORMANCE
r
To achieve the kind of performance illustrated in Figure 33, supernatant
treatment was required. Typical results when supernatant treatment was
not treated are shown in Figure 34. In this instance, 30,000 gallons of
untreated supernatant were returned to the head of the plant during the
daylight hours. The result was a noticeable increase in effluent phos-
phorus commencing about seven hours after the onset of supernatant return
and a relatively poor composite value of 0.8 mg/1 effluent phosphorus
for that day.
a
(o» P) in mg/1
01
3SPHOROUS
*
0.
z2
p
2C
30,000 GA
RETUR
__
PLANT INFLUENT
14.6 COMPOSITE
5 SPH
. RAW SUPERNATANT
— or
EFFLUENT PHOSPHOROUS
ALUM IN FINAL AI/P- L52
MC
CHE
22.5 6 PH
>— Ow
^Q— O Q Q |
NDAY. JUNE 7 . 1971
M COST • 6.0«/IOOO GAL
M3GPHI 115 6PH
l
as COMPOSITE 7
NOON 6P MID- 6A
NIGHT
FIGURE 34. EFFECT OF UNTREATED SUPERNATANT ON PLANT PERFORMANCE.
If alum treatment of the main flow of wastewater had not been underway,
the returned supernatant would have increased effluent phosphorus by
about 2 mg/1 above the conventional treatment effluent phosphorus level
of 8 mg/1 without supernatant return.
54
-------
A combination of untreated supernatant and rainfall infiltration caused
the situation illustrated in Figure 35. Wastewater flow through the
plant was increased to 1.87 million gallons on this day/ while recircu-
lation was reduced to 30,000 gallons. Analysis of the composite sample
for effluent phosphorus on this day showed a relatively high concentra-
tion of 1.0 mg/1.
a
i.
TOTAL PHOSPHOROUS (atP)inr
N * «
20
10.000 GAL
RAW SUP
INFILTRATIQ
PLANT INFLUENT
9.0 COMPOSITE
3GPH
N FROM 2* RAIN
^O^fcO***1"*
EFFLUENT PHOSPHOROUS
ALUM IN FINAL AI/P-1.9
MONDAY, JUNE 21, 1971
CHEM COST • 4.3 t/IOOO GAL
22.56 PH
ZOSSPHj 13.3 GPH
T
1.0 COMPOSITE 1
n""-° — o— o— <
"~*
NOON 6P MID- 6 A
NIGHT
FIGURE 35. PLANT PERFORMANCE PROBLEMS FROM
INFILTRATION AND UNTREATED SUPERNATANT.
EFFECT OF LOW WASTEWATER TEMPERATURES ON PERFORMANCE
Table 6 shows some typical low temperature performance data during the
extended alum run.
55
-------
TABLE 6
TYPICAL LOW TEMPERATURE PERFORMANCE
Temperature (°F) Flow Phosphorus (rng/1)
Date Raw Final (MGD) Raw Final
1-23-72 60 56 1.77 11.0 0.5
1-24-72 60 56 1.83 12.8 0.9
1-25-72 60 56 1.73 10.0 0.5
1-30-72 58 54 1.74 11.5 0.5
2-3-72 60 56 1.61 8.6 0.4
2-7-72 64 56 1.78 12.2 0.4
Prior to the dates listed in the table during December 15 - January 18,
there were five days when temperature in the final effluent was less
than 49° F. These occurred during the holiday season when all but the
skeleton operating crew were away from the plant. The chemical feeding
operation was shut down for approximately ten days, and it took several
days to bring it back into proper operating balance once chemical feed
was resumed. This was, unfortunately, precisely the period of time when
the effluent water temperatures were the coldest. In looking back over
the small amount of data available during this period, no particular
observations can be made one way or the other regarding the effect of
these coldest wastewater temperatures on plant efficiency.
EFFECT OF PEAK FLOW RATES ON PERFORMANCE
There was very little rainfall during the last three months of the study
and peak flows through the plant never exceeded 1.83 MGD. Minimum flows
were near 1.35 MGD. Figure 36 illustrates that even within this small
range, the rate of flow affected concentration of phosphorus in the
effluent.
This rate of flow can be re-expressed in terms of average hydraulic load
on the final clarifier as shown in Figure 37. As overflow rates approach
500 gpd/sq ft, effluent phosphorus begins to edge upwards.
56
-------
a.
O
I
o
lL
UJ
ilillllll
liiiiiin
niiiiiii
iiiiinii
iiiiinii
10
1.4 1.5 1.6
FLOW (M G D)
8.0
FIGURE 36. EFFECT OF PLANT FLOW ON EFFLUENT PHOSPHORUS.
These overall data were supported by some specific test runs made late
in October and in early November, 1971. In these cases, the daytime
flow was boosted to a rate of 2.5 MGD (approximately 160% of average
daily flow) for periods of four hours at a time. Unless chemical feed
rates were increased to compensate for the higher rates of flow, effluent
phosphorus composite values increased to 0.8 or 0.9 mg/1 for the entire
day involved. However, if compensating chemical feed rates were care-
fully administered, effluent phosphorus concentrations remained at
levels of 0.3 or 0.4 mg/1 as they had been when the plant was operated
at a reasonably constant flow of 1,000 gpm.
57
-------
(C
O
X
a.
VI
O
CL
SOO 320 340 360 380 400 420 440 460 480 500
CLARIFIES OVERFLOW (SAL/SO FT/DAY)
FIGURE 37. RELATION OF EFFLUENT PHOSPHORUS
TO FLOW THROUGH THE FINAL CLARIFIER.
EFFECT OF VARYING METAL TO PHOSPHORUS MOLE RATIO ON PERFORMANCE
During the final period of the extended alum run, the plant usually
delivered effluent phosphorus concentrations of 0.5 mg/1 or less when
mole ratios of aluminum to phosphorus of at least 1.5 were utilized (Figure 38)
The benefit of utilizing mole ratios much in excess of 1.5 was minimal.
58
-------
0.
•
o
-------
a.
M
o
EC
O
a.
o>
O
X
O.
UJ
3
4.0
4.6 4.7 5.2 5.5 5.8 6.1
COST OF ALUM (t/IOOO GAL FLOW)
6.4
6.7
7.0
FIGURE 39. COST OF CHEMICAL INJECTED FOR
VARIOUS LEVELS OF EFFLUENT PHOSPHORUS.
EFFLUENT SULFATE LEVELS
It was expected that final effluent sulfate concentrations resulting
from the aluminum sulfate (alum) injection would be inversely propor-
tional to effluent phosphorus levels. This proved to be the case as
shown in Figure 40. Of course, sulfate concentrations themselves also
varied according to rate of wastewater flow. On a strict weight
basis, about 5 pounds of sulfate were added to plant effluent for
every pound of aluminum added during treatment.
60
-------
[TTTT1TTTT
-------
K
O
fc
O
UJ
u.
u.
UJ
20
4O
60
80 100 120 140
ALKALINITY (mg/1)
160
ZOO
FIGURE 41. RELATIONSHIP OF EFFLUENT ALKALINITY TO EFFLUENT PHOSPHORUS,
(Reactions removing phosphorus also reduced alkalinity.)
Plots were also made to try and correlate effluent phosphorus with the
following parameters: turbidity, suspended solids, aluminum, and BOD5.
In all these cases, the residual values of the parameters named were
so low that the correlation either did not exist or it was not apparent,
Correlation with effluent phosphorus could not be shown for effluent pH
or temperature either.
SLUDGE PRODUCTION
Digester operation was normal during the extended alum run. Digested
sludge production was approximately 950,000 gallons per year, or 1700
gallons per million gallons of wastewater flow. Solids concentration
in the digested sludge was about 5 percent yielding an average sludge
production on a mass basis of 685 pounds of digested solids per million
62
-------
gallons of wastewater flow. This compares with typical unit volume and
mass sludge production figures of 850 gallons and 410 pounds, respective-
ly, when the plant is operated in a conventional manner without chemical
addition.
63
-------
SECTION IX
PILOT-SCALE FILTRATION AND CARBON ADSORPTION
Facilities for the pilot plant testing described here were shown pre-
viously in Figure 8 and discussed in Section IV of this report. The
system was capable of handling flow rates between 300 and 1500 gallons
per day. Flow control was accurate and easy to maintain at a given rate.
Sampling points were numerous and permitted a number of composite samples
to be compounded each day. Figure 42 indicates the length of the three
separate runs made with the pilot system.
RUN 1
JULY 1- SEPT 6
RUN 2
SEPT 8-DEC 21
1971
RUN 3
JAN 2 -MAR 30
1972
FIGURE 42. SCHEDULE OF PILOT TESTS.
(Best data were from high rate
trials in Runs 2 and 3.)
Run 1 utilized a throughput rate of 0.26 gallons per minute (375 gpd).
This feed rate was dictated by some physical constrictions in the tubing
and other facilities that made up the hydraulic pattern. The constricted
operation resulted in surface flow rates through the filter and carbon
on the order of 2 to 3 gallons per minute per square foot. The data
generated during Run 1 seemed reasonable and indicated some trends.
However, since the flow rates were unrealistically low, the data had
very little significance and is not reported or discussed here.
Runs 2 and 3 were conducted at a flow rate of 0.5 gallons per minute
(720 gpd). In both cases, the mode of operating the multi-media
granular filter was identical. The continuous rate of flow through
that unit was equivalent to 5 gallons per minute per square foot.
Direction of flow was vertically downwards through the coal, sand, and
garnet.
65
-------
The superficial velocity through the carbon columns was the same in
Runs 2 and 3, being 6.05 gallons per minute per square foot. However,
the direction of flow was upward in Run 2 and downward in Run 3. In
terms of treatment efficiency, there was no practical difference between
these two modes of operation, so the data for Runs 2 and 3 are combined
and presented together in this section. Generalized results are shown
in Figure 43.
42
21
0.5
0,15
0.03
COD
PHOSPHORUS
SUSP SOLIDS
1.8
'INFLUENT
FILTER EFFLUENT
0.8
r
CARBON EFFLUENT
TURBIDITY
ALUMINUM
FIGURE 43. PERFORMANCE OF MULTI-MEDIA FILTER AND CARBON COLUMNS.
(Influent to this system is effluent from the wastewater treat-
ment plant. All values are in mg/1. The COD figure of one mg/1
is a statistical aberation, as discussed in the text.)
66
-------
The figures shown here have been normalized to represent a central ten-
dency or "reasonable average" of all the data taken during the test
period of some 200 days duration. Generally, within each family of data,
there were a few extreme values which rendered arithmetic averages un-
realistic. On the other hand, most of the families of data were not
geometrically normal in the strictest sense either. Therefore/ the
values shown here have been rationalized and probably represent magni-
tudes somewhere between geometric and arithmetic means. Effluent COD
values reported for the carbon column are statistically valid; they
result from a group of "zero" or less-than-discernable values, combined
with another group of readings near 5 mg/1. The result is an "average"
of one mg/1 even though analytical limitations preclude accuracy at
that low level.
In very broad terms, filtration through the multi-media unit reduced
incoming pollutants to less than half their incoming concentration.
Following that, if the carbon system was used with a full 14-foot bed
depth, COD was reduced to about 5 percent of its concentration in the
filter effluent, and other pollutants were reduced to about 25 percent
of levels in the filter effluent.
When operated at a flow rate of 5 gallons per minute per square foot,
the filter seemed very stable and reliable in its performance. Back-
washing was done on a scheduled basis every 24 hours, and headlosses
were on the order of 9 feet by the time backwashing was begun. About
4 percent of the production was used to backwash in the early stages of
the pilot plant operation. Later this was reduced to 2 percent and that
backwash fraction was retained for the duration of the pilot studies.
Attempts to reduce backwash further were not undertaken because 2 percent
seemed a reasonable fraction and pilot scale indications of smaller
fractions might not be altogether reliable.
Operating four carbon columns in series, each having 3.5 feet of carbon
in them, was equivalent to having a single bed 14 feet deep with accurate
and accessible sampling at the quarter points. A surface flow rate^of
6.05 gallons per minute per square foot is equivalent to a superficial
velocity of 0.81 feet per minute. Based on plug flow, this represents
an empty bed carbon contact period of slightly over 17 minutes. The
carbon used in this study occupied about 50 percent of the volume of
columns, however, so actual contact time was on the order of 8 or 9
minutes.
When the direction of flow was upwards in Run 2, carbon in the first bed
was fluidized to the extent that there was something like a 10 percent
expansion. The three beds follox^ing were expanded 5 percent. This
upflow rate is probably somewhat greater than used in plant-scale prac
tice today, but there did not seem to be any effect on the operating _
efficiency in terms of removing pollutants. There were some difficulties
during backwashing in keeping carbon in place in the relatively small first
column. This appeared to be related to the relatively high solids storage
in that small diameter tube.
67
-------
Specifically, during backwash the carbon there seemed to bridge and hold
together. Even with very careful manual washing the carbon tended to
migrate upward in plugs until it got very near the top of the lead column.
Then the carbon would break up and, in a considerable shower of turbulence,
some of the carbon would pass out of the column with the wash water. The
small column was not outfitted with a breaker bar, or similar device,
which could relieve such problems at plant scale.
Figure 44 illustrates COD levels typically found in wastewater passing
through the 14-foot test column. These data indicate the rate of removal
was a straight line in the lower part of the curve, and they infer that
a zero COD could be gotten by extrapolating beyond the data.
25
20
15
COD
(mg/l)
10
10
15
BED DEPTH (FT)
FIGURE 44. COD LEVELS IN WASTEWATER PASSING THROUGH 14-FOOT CARBON COLUMN.
(This degree of COD removal was maintained until carbon became exhausted
and breakthrough occurred.)
68
-------
Note that over half the residual COD removal occurred in the first
column. Also, visual observations showed nearly all pinpoint alum floe
was removed in the first column. This probably contributed to the back-
washing problems mentioned previously.
Figure 45 is an analysis of the bed depth versus service time prior to
COD breakthrough. A COD level of 10 mg/1 was established as breakthrough
concentration in this case. That value was sharply apparent (as a func-
tional breakthrough concentration) in the test data.
SERVICE
TIME
( days)
IOO
75
50
25
7
X.O.I
BREAKTHRU
ORUN2
a RUN 3
5 IO 15
BED DEPTH(FT)
20
FIGURE 45. SERVICE TIME UNTIL COD BREAKTHROUGH
AT VARIOUS DEPTHS OF CARBON.
(Continuous throughput was 6 gpm per sq ft, and
exhaustion occurred near loadings of 0.434 Ib
COD per Ib carbon.)
Some of the reported "expansion" mentioned during upflow operation in
Run 2 was probably related to the presence of biological growth^and
alum film on the carbon.
This did not occur or was not apparent during
69
-------
Run 3. To whatever extent bioactivity developed, major dissolved oxy-
gen demands never occurred. Dissolved oxygen into and out of the system
remained near 6 mg/1 throughout the study.
After the end of Run 3 the pilot facilities were operated for several
more months* although data generated during this post-project interval
are not analyzed or presented here. It is probably proper to report
that the system continued to operate as it had during Runs 2 and 3.
70
-------
SECTION X
DISCUSSION
Ten subsections are discussed in this section, each addressed to a
particular aspect of the investigation reported. These subsections are
intended to be reasonably comprehensive if read apart from the main
body of the report. To promote this, some key data are either repeated
or re-expressed. Persons needing more detailed information should refer to
pertinent sections of the main body of the report and Appendix C as well as
these discussions.
OVERALL RESULTS
Best results in terms of effluent enhancement occurred when alum was
fed just ahead of the final clarifier. This approach proved to be a
stable and manageable operation in this plant. When temporary upsets
developed, they would generally be preceded by an hour or so of increas-
ing cloudiness in the water in the final clarifier. Thus, operators s
responding rapidly to a potential upset could adjust their treatment
strategy to correct problems before poor quality effluent escaped the
plant.
The second best approach, in terms of effluent quality per unit cost,
was to split feed alum at an approximate 4:1 ratio, i.e., 80 percent
ahead of the final clarifier and 20. percent ahead of the primary clari-
fiers. There did not seem to be any unique improvement in the effluent
using this approach nor were savings in chemical costs indicated. There
fore, split feed seemed a more complicated operation without redeeming
benefits in this particular installation.
Feeding metal salt ahead of the primary clarifier did not result in very
effective treatment and generally led to sludge digestion problems
within a couple of weeks. It is very likely that effluent quality could
have been improved if more time and effort had been applied (as it was
in feed to the final clarifier) ; such improvements would have no prac-
tical value, however, if they caused digester failure in the plant
being studied.
Experiences at this project indicate the best way to establish chemical
dosing patterns is to estimate hourly variations in h°s
71
-------
consistent throughout the entire period of this investigation.
An automatic analyzer proved ideal to monitor effluent phosphorus.
This apparatus greatly reduced labor required for sampling and analyses.
Furthermore, it provided a continuous and reliable record of phosphorus
concentrations in treated wastewater. Relatively short-term perturba-
tions were very evident, although they might not necessarily affect 3-hour,
6-hour, 12-hour, etc. composite samples compounded for analysis. These
minor perturbations are important if they persist several days, thus
indicating a pattern. Such a pattern would indicate feed schedule
changes to realize a real improvement in performance.
Changes in chemical dosing rates should generally remain within the
established overall daily dosage rate. This daily rate, based on pre-
selected mole ratios of metal to phosphorus, should usually provide suf-
ficient metal to accomplish the treatment goals set forth. The really
important factor is to take the pre-selected daily ration of metal and
feed it to incoming wastewater at rates which allow the most effective
utilization of the metal.
In this study, aluminum was preferred over iron, but this choice is
reported here without prejudice regarding the potential of the two metals
at otLer locations. As in any wastewater system, Richardson's sewage
comprises a highly complex aqueous system. Many of its prime characteris-
tics may never be clearly known, but they probably influence the choice
of metal salt. In this case, alum was the better choice. Iron fell
short in terms of efficiency and caused color in the effluent (even
though aluminum leakage occurred, it was colorless). Sorptive capacity
in an activated sludge plant might lead to an opposite course. Probably
the key feature is to have the capability of feeding either compound at
several points; then, under different circumstances, the better choice
can be made.
Leakage of colloidal iron while feeding ferric chloride is certainly not
surprising considering the type of plant involved in this study. There
was no granular media filtration to polish the final effluent. The
secondary treatment system in this plant has far less sorptive capacity
than in bio-flocculation inherent with activated sludge. Passing iron-
bearing water over the organic film in a trickling filter bed had very
little effect, nor was it expected to. The only defense against iron
leakage in this plant was gravity settling.
Of course, when iron was fed, residual chlorides increased as did sulfates
when liquid alum was being injected. Both of these followed generally
stoichemetric levels. Chemical selection will be affected if either
extraneous ion is of concern in the receiving water involved.
Aluminum leakage always occurred during both trial and extended alum
runs. It was on par with the escaping iron colloids when iron was dosed
ahead of the primary clarifiers, except it was not visible. If cases
12
-------
exist where increased aluminum residuals are to be avoided, then this
element would be a key consideration in selection of metals to be fed.
The ultimate solution to colloid leakage lies in recognizing that the metal
is probably bound with pollutants such as phosphorus and oxygen consuming
compounds. If any of the three require further reduction, then scone treat-
ment operation which can remove colloids would be required. Probably the
most natural initial consideration would be given to an effluent filtra-
tion process of some sort.
Polymers were not given a rigorous trial during this investigation pri-
marily because when alum was dosed at an optimized feed rate/ the use of
polymers did not improve efficiency or reduce the cost of treatment.
As in opting for aluminum, the choice not to use polymers in this study
is reported without prejudice to their potential in other situations.
Cold weather operations were never underway long enough to draw any
convincing long-term observations in this study. The data that are
available indicate low water temperatures neither compromised nor domi-
nated reaction efficiencies and kinetics to the degree that might be
expected. Certainly, cold weather effects were less than sometimes seen
in water treatment plants. An important difference in considering this
historical precedent is that, in the case of treating fresh water supplies,
the aqueous system is of a less complex nature than the waters involved
in this study. In wastewater treatment, new colloids and natural poly-
electrolytes are formed during treatment, and a variety of complex reac-
tions stem from addition of metal salts; these may not occur in water
treatment. In summary/ although sufficient data for clear definition are
lacking, there are strong evidences that wastewater temperatures in the
high 40's and low 50's have relatively small effect on precipitation
with metal salts.
A far greater effect on operating efficiency lies in control of hydraulics
in settling basins. There are clear indications that average daily
clarifier surface loadings exceeding 500 gallons per square foot will
reduce capture of colloidal phosphorus and other pollutants. Further-
more, instantaneous overflow rates of 1,000 gallons per day per square
foot or more caused temporary poor performance which was great enough
to adversely affect average daily performance. These features appeared
important enough to warrant discussion in the next two subsectxons.
EQUIPMENT AND FACILITIES
From a physical sense, the clarifiers available in this plant are not
particularly efficient. Per present day knowledge of design and opera-
tion they have shortcomings as follows: they are relatively shallow
(the final clarifier ranges from a 6-foot sidewater depth to 8-foot at
the center) and they have some deficiencies in control of inlet veloci-
ties (particularly in the case of the primaries). Furthermore, in all
instances these clarifiers are outfitted with standard rakes that may
73
-------
not be the most efficient to handle the increased volume and weight of
sludge produced by chemical addition. More efficient tanks would pro-
bably handle higher peak liquid and solids loading rates. Very possibly,
clarifiers outfitted with tube settlers could have performed markedly
better.
Flocculation facilities at this plant were far from sophisticated. The
only flocculation energy provided was whatever hydraulic turbulence was
available in the flow distribution systems themselves. This meant there
was no practical way to control the energy gradient within the floccu-
lation operation. Fortunately, descending hydraulic energy levels are
inherent in a system of this type. The basic aims of proper flocculation
were served whether or not actual control of the operation was available.
Standard-rate trickling filters comprise about 55 percent of the existing
trickling filter facilities in the United States. However, there is no
clear indication that different results or new problems would have re-
sulted if high-rate trickling filters had been used at the plant under
study.
Sludge digesters of the Imhoff tank type are the simplest used in current
practice today. The three parallel digestion tanks at Richardson are in
no way better or worse than others of their type around the country.
In plants having heated digesters with more effective mixing, it may
be possible to accelerate the digestion of combined alum-biological
sludges.
Sludge drying beds in this plant are quite marginal in their capacity.
Only one square foot per capita is available. Consequently, as discussed
in the subsection devoted to sludge handling, a constant struggle was
required to stay abreast of increased sludge production and disposal
requirements resulting from chemical addition.
CLARIFIER PERFORMANCE
Sludge recirculation was practiced in two important different ways during
this study. Both deserve individual attention.
When feeding alum just ahead of the final clarifier, a heavy sludge
blanket could be built in that clarifier by careful reduction and control
of sludge withdrawal. Since the tank involved here was circular with a
centerwell inlet, the solids contact blanket assumed the shape of a dough-
nut surrounding the centerwell. This blanket of floe extended about one-
third the radius toward the peripheral launder. Floe in the doughnut
was continuously recirculating, and its motion traced a pattern of tur-
bulence that seemed ideal for low energy flocculation.
The second type of sludge recirculation occurred in the primary clari-
fiers when alum was being fed just ahead of the final clarifier. The
74
-------
sludge withdrawn from the bottom of the final was returned to the primary
clarifiers where it had a profound and important influence in improving
primary settling. For this reason, the results of primary settling are
quite atypical, and overall plant performance seems to have been improved
by the return of this "spent" sludge which evidently still had consider-
able usefulness left when admixed with incoming wastewater.
To serve both purposes, it appeared important to keep recirculation rates
from the final clarifier low. This technique preserved the solids con-
tact doughnut in the final clarifier and enhanced performance in the
primaries by delivering a rich seed sludge there and keeping hydraulic
overflow rates to a minimum.
The heavy floe produced by metal addition proved to be excellent tracer
material. In fact, for practical studies of basin hydraulics, this floe
was the most effective tracer that any of the investigators in this project
had ever observed.
One other condition observed in the final clarifier merits discussion.
It was apparent that distribution of colloidal COD within that settling
tank was directly related to the extent and concentration of floe parti-
cles. A COD profile showed this pattern and indicated that if a large
skirt were added outside the existing centerwell and about halfway to the
outer wall, that considerable additional reliability would result in
retention and capture of COD and suspended solids. This would, in effect,
enclose the low energy flocculation process.
SLUDGE PRODUCTION, DIGESTION, AND DRYING
Slightly less than one percent of the total plant flow of wastewater was
withdrawn into the sludge handling system when chemicals were added as a
part of treatment. This amounted on a yearly basis to between 900,000
and 1,000,000 gallons of digested sludge (about 2,600 gallons per day)
and approximately 3,500,000 to 5,000,000 gallons of digester supernatant
(about 10,000 to 14,000 gallons per day). The digested sludge was with-
drawn to sand beds for drying and the supernatant recycled (either with
or without separate alum treatment) to the plant's influent wet well.
As described earlier, the three parallel digesters are not heated and do
not employ mixing in the usual sense of the word. Stirring is provided
in each digester by truss members attached to an extension of the vertical
shaft driving the clarifier mechanism above. These truss arms rotate at
three revolutions per hour and constitute all the mixing provided in the
digesters.
On a population basis, the volume in these digesters equates to approxi-
mately 3 to 3.5 cubic feet per capita. Considering the amount of sludge
digested and withdrawn (but excluding supernatant), displacement time
through the digesters for solids averaged between 5 and 6 months during
-------
the project. This long residence time probably explains why, when low
wintertime temperatures developed, the digesters were able to either con-
tinue effective digestion at a reduced pace or perhaps store organic
matter until warmer temperatures stimulated bacteria to accelerate their
activities.
Cold weather dropped temperatures in the digesters to as low as 65° F,
while in the summer they rose to nearly 85° F. In the same period, maxi-
mum and minimum values of pH ranged between 6.5 and 7.4. Digested
sludge consistency did not vary greatly, however, remaining on the order
of 5 to 6 percent throughout the year when alum was being administered.
The volatile fraction of digested sludge was consistently near 45 percent.
Prior to chemical addition, annual sludge production averaged about half
that experienced during chemical precipitation. Total production of
digested sludge was in the range of 450,000 to 500,000 gallons per year
(about 1,300 gallons per day). Nonchemical sludge had the tendency to be
slightly more concentrated, within the general range of 6 to 7 percent
solids by weight.
Generally speaking, digested sludge during iron addition was somewhat
heavier than during alum addition. There are no irrefutable records to
support this because iron was only added for the better part of three
months while at least twice that long was required to establish a clear
pattern of the effects of a particular chemical.
As anticipated, both aluminum and iron eventually concentrated in the sludge
during their respective dosing periods. Final concentrations of the metals
in the digested sludge were on the order of 3 to 4 percent by weight. Phos-
phorus also concentrated in the digested sludge and reached levels on the
order of one percent by weight. There was never any indication that
phosphorus resolublized in the supernatant, although it was present in the
range of 100 mg/1 as colloidal solids.
At one square foot of space per person served, drying bed capacity at this
facility was critically short. This drying capacity was considered stan-
dard design practice at the time the plant was built, but subsequent ex-
perience has proven there is very little, if any, reserve in such a drying
bed system. During the course of operating the treatment plant prior to
chemical addition, there were occasional periods when drying capacity was
marginal. With chemical precipitation, the volume of sludge generated
doubled, so certainly the logistics of getting sludge on and off the beds
became more critical and shortage of drying space became a greater problem.
Consideration was given during the course of the experiments to digging
an emergency lagoon; however, as it turned out the operating crew was
able to handle sludge just fast enough to avoid this emergency procedure.
On two brief occasions, weather conditions prevented full utilization of
available drying beds, causing sludge to accumulate in the digesters until
finally it backed up through the sludge hole in the floor of the primary
76
-------
clarifiers. These conditions resulted in discoloration of primary effluent
and severe deterioration of the quality of that water. Both of these in-
stances were short in duration and had little or no effect on overall
performance. They did indicate what could happen if conditions causing
them were not corrected.
Although the volume of digested sludge doubled when chemicals were added,
no net increase in drying bed capacity was required if the weather cooper-
ated. This stemmed from the fast drying nature of the combined chemical-
biological sludge. Conventional digested sludge was, over a long time
average/ dried and stripped from the beds in a period of 20 to 22 days,
weather permitting. Under the same circumstances, the combined chemical-
biological sludge could be dried and removed on 8 to 10 day cycles.
Clearly, the additional volume of sludge could be handled provided rainy
weather was not encountered and assuming additional manpower was always
available.
SUPERNATANT TREATMENT
By the time the study period was completed, the operating staff at this
treatment plant had accumulated over five years' experience in treatment
of supernatant returned from anaerobic digesters. As described earlier,
the original system was a relatively crude batch-type approach and chemi-
cal feed was not optimized in those early years. This subsection offers
comments on how the overall system progressed and what the best arrange-
ment seemed finally to be.
The volume of supernatant at this treatment plant could and does vary
considerably from week to week. It is generally near one-half to one
percent of total plant flow. This is probably more supernatant than is
drawn at most biological treatment plants, but drawing such volumes has
historically helped overall treatment efficiency at Richardson. Super-
natant is usually drawn during parts of either two or three days per week.
On these days, about 10,000 gallons of liquor is decanted from each of
the three digesters, totaling 30,000 gallons or approximately two percent
of plant flow that day. If weather or other operating conditions dictate..
supernatant is drawn only twice per week or perhaps just once. Also,
operators draw more than 10,000 gallons from a given digester if on-bita
sampling and observation indicate a larger draw is called for. For
these reasons, then, supernatant withdrawn per week can range from as
little as 30,000 gallons to as much as 100,000 gallons or more.
On a typical day when supernatant is being returned at the rate of 30,000
gallons, it will contain colloidal phosphorus of 100 jng/1. This would
cause an obvious increase in total phosphorus in the raw wastewater if
supernatant were not treated. The increase would be on the order of 2
mg/1, representing something like a 15% surcharge in phosphorus inflow
concentrations. As already shown in Figure 34, untreated supernatant
leads to higher phosphorus values in the plant effluent. Suspended
77
-------
solids and other pollutant concentrations increase too, and generally in
similar proportions.
On the other hand, when supernatant was treated (either continuously or
on a batch basis), there was never any problem in overall plant operation
because of its return to the raw inflow. Typical before and after pollu-
tant concentrations in treated supernatant were shown previously in Figure
11. To treat supernatant, aluminum was dosed at an average M/P mole ratio of
1.8/1.0. This comment infers operators made an estimate of phosphorus
concentration in supernatant each time which was not so; they added to
meet an assumed demand. That estimate would not be confirmed until the
next day. To be more specific, operators batch treating supernatant would
add 1.5 gallons of liquid alum per 500-gallon tank. In earlier trials,
they had been adding 3 or 3.5 gallons, and it took a while before the
operators recognized they could cut their alum dose and still obtain
essentially the same results.
At the higher dosage just mentioned, supernatant treatment cost 40* to
50£ per thousand gallons of liquor. Once the reduced dosage was con-
firmed, the unit cost dropped to about 23£. The same chemical feed rate
was maintained when the supernatant treatment system was converted to a
32 gpm continuous operation.
Ultimately, chemical costs were more related to the amount of supernatant
drawn within a given week than to any other factor. If 50,000 gallons were
drawn, the average cost of treating supernatant would be 0.1C per thousand
gallons of wastewater treated in the main plant; if 100,000 gallons of
liquor were drawn in a given week, the cost would double to 0.2£ per
thousand gallons of wastewater handled in the main plant. In either
case, the cost was considered modest and the improved plant performance
well worth the time and effort.
PILOT ADSORPTION AND FILTRATION
The primary goal of this project was to evaluate plant-scale chemical
addition as a means of upgrading overall plant performance. Pilot-scale
carbon adsorption and multi-media filtration studies were ancillary oper-
ations to the main effort. They were undertaken to see what improved
performance could be expected and what size units would be required if
these processes were added to the present system.
Without any question, it seemed clear that filtration was a reliable and
reasonable method to improve capture of colloidal and other suspended
solids in the treatment plant effluent. For example, this solids capture
system would have been an excellent backup during the periods of time
that oxidized iron colloids were escaping with the plant effluent and
causing a highly undesirable red color. (This conclusion does not infer
that the pilot units were being operated during the iron trial. They
were not. But filter performance on finely divided alum solids indicates
iron particles would have been caught.)
78
-------
Tertiary pilot filtration data show suspended solids were reduced during
main-plant alum treatment from a level of 7 mg/1 to 1 or 2 mg/1 as a
consistent matter of course. By allowing head losses to build to eight
or ten feet prior to backwashing, 24-hour filter runs were routinely
achieved. Filter rates of 5 gallons per minute per square foot were
used on a continual basis. The pilot filter consistently reduced COD
values by half and phosphorus levels to below 0.25 mg/1.
The study unearthed no new information on theoretical or applied fil-
tration technology. Rather, the test filter was run for extended periods
of time until performance data established design parameters to a high
degree of confidence.
Generally/ these same comments can be made about the carbon absorption
system. The four columns were run in series for three month periods.
Their performance definitely indicated that full-sized carbon adsorption
units following tertiary filters could be expected to reduce pollutants
to near trace levels.
The proposed loading rates to achieve this type of results in full-scale
operation would be 0.43 pounds COD per pound of activated carbon. In
units 14 feet deep, this would entail a flow rate of 6 gpm per sq ft
for a period of 60 to 90 days. It made very little difference in the
pilot studies whether carbon columns were operated downflow or upflow.
Either orientation would probably serve in plant-scale units.
In looking at the entire pilot system, there is clearly a range of
effluent polishing available. The multi-media filter alone can produce
water of a defined quality which will meet the water quality requirements
of many communities. An important side benefit with filtration is the
backup or safeguard function it provides to an existing treatment plant.
If still better effluent is required, carbon can be used as an adjunct
to filtration. Data show that, within the carbon adsorption process, a
range of effluent quality is possible. For instance, the bed depth
could be chosen to suit different needs, or some value besides 10 mg/1
of COD might be used as a breakthrough concentration.
In summary, the pilot studies generated valuable data and experience
which can be utilized in upgrading the Richardson Treatment Plant as
future requirements dictate.
DRAINAGE FROM SAND DRYING BEDS
In Richardson, for every ten gallons of digested sludge put on drying
beds, about one gallon reappears eventually as underflow and is returned
to the head of the plant. Of this ten percent underflow return, well over
three-fourths of that volume will come through in the first 48 hours.
This return factor may or may not be typical of experience at other treat-
79
-------
ment sites, but whatever the return fraction, some observations on the
nature and treatment of that underflow .may have broad application.
Over a period of years the local staff observed that if certain precau-
tions were taken, sand bed underflow would be relatively clear and pure.
If these precautions were ignored that liquor would be dirty, strong,
and fully as undesirable as its close kin, digester supernatant.
Probably the strongest single conviction which developed out of this long
period of operation and observation is that the sand layer (in this case
the upper six or seven inches of the drying bed itself) should never be
allowed to get too thin to do its job. Each time a load of sludge is
stripped from the beds, as much as 1/4 inch of sand can be taken off with
the sludge. This attrition is neither unusual nor controllable; it is a
fact of life in the operation of sand drying beds and ought to be recog-
nized as such. Therefore, if the beds are to continue to perform pro-
perly, the sand should be replaced at frequent intervals. The replace-
ment sand should be of medium to coarse grain, rather than a very fine
beach or field sand which tends to plug and impede percolation down into
the underdrainage system.
If the sand layer is properly maintained, it should provide good fil-
tration and surface biological action. It should yield a clear under-
flow water with qualities near those expected from a primary or possibly
a secondary treatment plant.
The preceding remarks dealt with the quality of underdrainage water. An
equally important factor is the rate of dewatering. Observations at
this plant indicate the sand layer needs attention between each loading.
This attention consists of fluffing the sand with a rake or a small power
driven implement. Then the scarified surface should be allowed to dry
for a day or so before receiving another load of digested sludge. Leav-
ing the bed open and fallow is probably against human nature; the natural
tendency is to draw a new load of sludge onto the bed quickly so that the
drying cycle can begin anew. However, at Richardson, if the sand is
fluffed and allowed to dry, it seems to have more adsorptive capacity, a
higher permeability, and a rejuvenated capacity for purifying the under-
flow water.
THINGS THAT DID NOT WORK
No long-term research and development effort can be completed without
making a variety of mistakes. The mistakes made on this project are
explained in sufficient detail in this subsection that others might
benefit.
The final clarifier sludge recirculation sampler shown in Figure 7 was
mechanically reliable. It was easy to adjust and simple to maintain. It
cost very little. As a mechanism, it was a success. Unfortunately, the
80
-------
flow of water it sampled was in no way representative, so resulting sample
portions wera not only nonrepresentative, they were unrealistic and mis-
leading. False samples were taken for a period of months before the
situation was recognized. The problem lay in the fact that, when using
20-minute pulse blowdowns from the bottom of the final clarifier, the
rate of flow and character of the fluid was highly variable over the 25-
second blowdown cycle. Different schemes were tried, without success,
to obtain representative samples of this stream. It was finally concluded
the only way to get a good representative recirculation sample would be
to combine all the 90,000 gallons per day in one tank, stir it, and then
take the sample from that total mixed volume. This was, of course, un-
realistic and a satisfactory recirculation sample was never gotten,
making it impossible to reliably calculate the impact of the recircula-
tion flow on raw wastewater characteristics.
Very early in the investigation an attempt was made to measure flow
through a trickling filter by gaging water head in the center column.
This also did not work. The center column was tapped and a transparent
calibrated water gage mounted on the outside. Water level could be
determined by an operator standing at the edge of the filter, so this
was not the problem. Rather, practical failure resulted from the fact
that the water level fluctuated continuously over an approximate three-
inch range, and that range seemed to fluctuate within a larger and less
frequent harmonic. Furthermore, the head-discharge curve for the trickling
filter distributor (which is, in effect, an orifice system) was highly
sensitive to the variations in head imposed. The whole approach repre-
sented no more than a gross evaluation of the amount of water going through
the distributor.
Standard one-inch xvater meters were installed to measure underdrainage
flow from the sand drying beds. This seemed like a very obvious and
practical idea as the flow throughout the many years it has been sampled
at Richardson had always appeared to be sufficiently clear and clean to run
through conventional water meters without problems. It was not recog-
nized that the small sump pumps used to drive underdrainage through the
meters would pick up loose fine sand in the bottom of the collection sump.
The pumps hurled these abrasive particles into the water meters causing
severe wear and other maintenance problems. (Physical limitations at this
site made it impractical to have deep sumps which would prevent sand
particles from being picked up by the pumps.) The water meters lasted
long enough to obtain good flow readings for typical sludge bed drainage
periods. When they failed, measurement of underdrainage flows was termi-
nated.
The final clarifier in this treatment plant also serves as the chlorine
contact chamber. Under normal circumstances, there were no problems in
collecting a sample of effluent partway through the clarifier and pumping
it to the automatic chlorine analyzer. However, when alum was being
added, the floe generated in the final clarifier necessitated filtering
the sample through a traveling paper filter or some similar mechanism
81
-------
prior to injection into the analyzer. Failure to filter this sample flow
caused the instrument to clog.
For the first few months of operation, a dissolved oxygen analyzer was
stationed in the plant effluent channel. A continuous recorder was
attached and an excellent strip chart record of effluent dissolved
oxygen was produced. Unfortunately, colloidal aluminum particles men-
tioned previously also affected the sensing section of this apparatus.
In fact, it became very difficult to keep the sensing probe in operable
condition for more than a few hours. Fortunately, it was also unneces-
sary to have a continuous record of dissolved oxygen concentrations.
Grab samples at all hours showed oxygen tensions were continuously at. the
level of 5 to 7 mg/1. Entering an "average" value for oxygen concentra-
tion for the day based on these grab samples was an easy and justifiable
procedure.
An unexpected and serious problem developed when alum was being fed to
untreated wastewater entering the plant. After floe developed in the main
wet well, wastewater was pumped through a pipeline to the primary clari-
fiers. In this pipeline, there was a venturi meter which included a
differential pressure cell utilizing mercury columns. Somehow, particles
of alum migrated to and congregated in the mercury chamber, clogging the
tubing to the transmitters. This problem which also plagued the final
clarifier recirculation meter throughout the project was never satis-
factorily resolved.
Two mistakes were made in the design and provision of polymer addition
facilities. The first was installation of a manual dispensing system.
At about the time this project began, and certainly in the two or three
subsequent years, there have been several automatic polymer dispensers
offered for this type of service. Several of these appear to be highly
reliable and reasonable in cost and would have been a considerable im-
provement. Another miscalculation was the decision to install the polymer
feed system without shelter. Designers and operators alike agreed that
discomforts of wet or windy weather during polymer mixing made the cost
of simple shelter very reasonable.
UNANSWERED QUESTIONS
From time to time, addition of alum ahead of the final clarifier caused
a luxuriant foam in the final effluent. The foam was present on the
water traveling around the launders of the clarifier and was more apparent
in the plant effluent ditch where an 8-foot head loss was expended in
intense turbulence. The foam was generally white and contained fairly
high concentrations of colloidal pollutants including phosphorus compounds.
The foam seemed to be intensified by chlorination. Curiously, this foam-
ing tendency was not persistent; it probably occurred during approximately
20 percent of the project. Some consideration was given to making use of
this foaming tendency in a foam separation process to be pilot tested on
site. Time and funds did not permit this prospect to be evaluated, however,
82
-------
Without being able to present clear correlation, it seemed that chlorina-
tion had a deleterious effect on phosphorus removal by alum addition.
Within broadest limits, the more chlorine added, the more alum was re-
quired to maintain effluent phosphorus concentrations at a consistent
level.
The techniques of sampling, or more probably the techniques of analyzing,
prevented a reasonable mass balance of nitrogen in the system during the
study. Typical ammonia nitrogen values seemed reasonable, some 25 mg/1
entering the plant with half that much in the final effluent.^ Nitrite
nitrogen was scarcely ever present in significant amounts. Nitrate
nitrogen concentrations in the final effluent were usually low and
appeared reasonable. When all these were added together, however, they
missed balancing with Kjeldahl nitrogen by a considerable amount. Since
nitrogen considerations were not part of this study, the matter was never
checked out in detail.
COSTS
The greatest single cost factor in this project was the cost of chemicals.
To this must be added some allowance for increased sludge handling and
disposal costs. Capital equipment costs were both minor and singular to
this particular situation; other plants should apply equipment costs
described here to their own situations with considerable caution.
The average cost for chemicals during the final extended alum run was on
the order of 5.1* per thousand gallons ($51 per million gallons). This
figure is based on buying alum at a rate which placed the cost of tnvalent
aluminum, or Al (III), at 25* per pound plus some 12* per pound to transport
the liquid a distance of 300 miles. These operating-oriented fibres were
derived from the following costs expressed in the commercial world of
liquid alum sales: the base cost of 17 percent aluminum oxide was $43 per
ton and freight in minimum truckload lots of 42,000 pounds was 51* per
hundredweight. These figures also translated into another very usable
form of 17.3* per gallon of liquid alum. Each gallon of alum weighed
some 11.1 pounds of which 0.485 pounds was trivalent aluminum.
There may be value in moving towards the more universal expression of
chemical cost per pound of phosphorus removed. In this case, the cost
appTies to reduction of phosphorus from 11.4 to 05 ag/1. With a removal
o? 10.9 mg/1, a plant flow of 1.55 MGD, and an Al/P mole ratio of 1.6,
figures out to 36 cents per pound of phosphorus removed.
Additional operating costs included relatively minor items such as $200
or $300 per year for electricity, about $250 per year for miscellaneous
replacement parts for pumps and other working parts, and some allowance
for paint, grease, and maintenance manpower.
83
-------
No additional operators were required to run the plant when chemical
addition was brought into the treatment scheme. The laboratory was
already staffed to support conventional treatment and did not need any
additional analysts to support chemical treatment. The additional lab
staff utilized were required only for the purpose of supporting the
project and would not be necessary on a strict operational basis.
Identification of capital cost for chemical addition equipment becomes
confusing in this case if the reader tries to separate the equipment
needed for chemical addition at a typical plant from the actual equip-
ment supplied to support this project. For example, a $3,000 zetameter
and automatic pH and dissolved oxygen monitors were purchased strictly
for research purposes; they would not be needed in a plant upgraded to
permit chemical addition. Some other improvements were made because it
was believed they would improve overall plant operation; an example of
this is the addition of a new $7,000 automatic chlorinator when the
original smaller manual unit was still capable of doing a fairly good
job of disinfection. The $4,000 spent on magnetic meters to monitor raw
and treated supernatant flow also would probably not be necessary in a
typical operating plant. For this project, the capitalized costs (10
years at an interest rate of 5.5 percent totalled something less than
2^/1,000 gallons treated.
In summary, total cost for chemical addition at Richardson amounted to
It to 8^/1,000 gallons treated, of which something just over 5^/1,000
gallons was consumed in chemicals.
84
-------
SECTION XI
ACKNOWLEDGMENTS
Mr. Derrington, Director of this project, was Water Superintendent at
the City of Richardson, Texas; he is presently Operations
Manager for Wastewater Treatment at the North Texas Muni-
cipal Water District, Wylie, Texas.
Mr. Stevens, Project Manager, is Assistant Director of Utilities at
the City of Richardson.
Mr. Laughlin, Associate Director, is a partner in the consulting engineering
firm of Shimek-Roming-Jacobs & Finklea, in Dallas, Texas.
A number of people from the United States Environmental Protection Agency
made notable contributions to this project:
Mr. Richard Brenner, National Environmental Research Center, Cincinnati,
Ohio, was Project Officer for the duration of the study;
he probably devoted more time to this project than any
other person besides those directly involved in its
prosecution.
Mr. Edwin Barth, National Environmental Research Center, Cincinnati,
Ohio, was active in the inception of the project and
encouraged organization and pursuit of work reported here.
Mr. Patrick Tobin, Office of Research and Monitoring, Washington, D.C.
gave administrative support to the project as it drew
towards a conclusion.
Mr. Larry Kamphake, National Environmental Research Center, Cincinnati,
Ohio, set up the automatic phosphorus analyzer and
provided considerable assistance in refining other
analytical techniques used by the control laboratory.
Dr. William Duffer, Robert S. Kerr Water Research Center, Ada, Oklahoma,
and other members of his staff helped instruct project
personnel in special analytical techniques and in methods
of laboratory quality control.
Mr. Robert Crowe, Technology Transfer, Washington, D. C., and members
of his staff helped disseminate information generated
and brought an influx of suggestions from related re-
searchers during the course of this study.
85
-------
Mr. George Putnicki and Mr. Robert Hiller, Region VI, Dallas, Texas
and other members of the Regional Staff provided local
support and encouragement.
Three individuals were engaged as expert consultants and, from time to
time, met with the project staff to review progress and offer suggestions;
Dr. C. H. Connell, Consultant, Bertron, Texas; formerly Professor of
Preventive Medicine, University of Texas Medical Branch,
Galveston, Texas.
Mr. I. W. Santry, I. W. Santry & Associates, Dallas, Texas.
Mr. C. L. Shimek, Shimek-Roming-Jacobs & Finklea, Dallas, Texas.
Pilot facilities for granular media filtration and carbon column adsorp-
tion were provided through general coordination of Neptune MicroFloc,
Inc., represented by Mr. Howard Shireman, Dallas, Texas. Carbon columns
and the granular carbon were supplied to the project by ICA America/ Inc.;
their Mr. Paul Stubbe and Mr. James Black provided advice and on-site
assistance in getting these units in service and interpreting the data
generated. Mr. Shireman provided the same type of service for the support
hardware facilities as well as for granular media filtration equipment.
Electronic data processing support was supplied primarily through Mr.
Donald Wilson of Compuroute, Garland, Texas. Mr. Wilson set up the
original statistical programs and other elements of hardware including
printout formats and interface coordination with microfiche equipment.
During final stages of the project, Mr. David G. MacKenzie, Teledyne
Geotech, Garland, Texas, provided active support. Mr. MacKenzie refined
some of the statistical software, advanced and refined microfiche trans-
lation operations, and generated a variety of computer plots such as
probability and scatter diagrams. He is the author of the flow chart
shown in the appendix dealing with data processing.
Finally, the Mayor and City Council of the City of Richardson gave their
continued strong support and encouragement during the project. A number
of departments within the City Staff were involved in the project, in-
cluding the office of the City Manager, general administrative personnel,
and a variety of people in the Water Department such as operators, main-
tenance specialists, laboratory personnel, and administrators. This
entire group provided uncounted hours of dedicated support.
86
-------
SECTION XII
REFERENCES
1. "Kinetics and Mechanism of Precipitation and Nature of the
Precipitate Obtained in Phosphate Removal from Wastewater
using Aluminum (III) and Iron (III) Salts," Recht, H. L. and
Ghassemi, M., Water Pollution Control Research Report 17010 EKI
04/70, FWQA, April, 1970.
2. "Chemistry of Nitrogen and Phsophorus in Water," AWWA Committee,
Jour. AWWA, pp. 127-140, February, 1970.
3. "Phosphate Removal: Summary of Papers," discussion by Theis,
T.L., et al, Jour. Sanitary Engineering Division, ASCE, pp. 1004-9,
August, 1970.
4. "State of the Art of Coagulation," AWWA Committee, Jour. AWWA,
pp. 99-108, February, 1971.
5. "Aluminum and Iron (III) Hydrolysis," Bilinski, H. and Tyree, S. Y. ,
Jr., Jour. AWWA, pp. 391-2, June, 1971.
6. "Colloids Complicate Treatment Processes," Dean, R. B., Environ-
mental Science and Technology, pp. 820-4, September, 1969.
7. "Chemical Interactions in the Aggregation of Bacteria Biofloccu-
lation in Wastewater," Busch, P. L. and Stumm, W., Environmental
Science and Technology, pp. 49-53, January, 1968.
8. "Standard Methods for the Examination of Water and Wastewater,"
APHA/AWWA/WPCF, 13th ed., 1971, New York, N. Y.
9. "Methods for Chemical Analysis of Water and Wastes," Environmental
Protection Agency, 1971, Cincinnati, Ohio.
10. "Laboratory Quality Control Manual," Environmental Protection Agency,
Kerr Water Laboratory, 1971, Ada, Oklahoma.
87
-------
SECTION XIII
APPENDICES
A. LABORATORY OPERATIONS AND FACILITIES
FIGURE 1 ANALYSES PERFORMED ON COMPOSITE SAMPLES
PAGE
91
92
B. DATA PROCESSING
FIGURE 1 LABORATORY BENCH SHEET
FIGURE 2 FLOW DIAGRAM OF EDITING PROGRAM
FIGURE 3 PROGRAMING SEQUENCE USED TO PRODUCE
DAILY REPORTS
FIGURE 4 TERMINAL PROGRAM DEVELOPED FOR
STATISTICAL STUDIES
95
96
97
99
100
DETAILED DATA SUM>iARY FOR
EXTENDED ALUM RUN
TABLE 1 DETAILED SUMMARY OF MAJOR PARAMETERS
DURING EXTENDED ALUM RUN
TABLE 2 MONTHLY AVERAGE VALUES OF ADDITIONAL
CHEMICAL PARAMETERS DURING EXTENDED
ALUM RUN
TABLE 3 MONTHLY AVERAGE VALUES OF PLANT OPER-
ATING PARAMETERS DURING EXTENDED ALUM RUN
103
104
116
117
89
-------
APPENDIX A
LABORATORY OPERATIONS AND FACILITIES
There were thirteen sampling points in the treatment plant, and up to
twenty-three analyses might be run on composite samples taken at any
one of them. Figure 1 summarizes the stations and analyses involved.
Approximately 140 items of data were reported each sample day.
A staff of three analysts worked five days per week, usually Monday
through Friday. The samples they analyzed were collected Sunday through
Thursday. Occasionally, they worked on weekends to analyze samples
collected Friday and Saturday. A fourth analyst, working about 30 hours
a week, helped handle the heavy laboratory load generated by the demon-
stration project. A staff of this size would not have been required to
support the treatment plant if the only aim were plant operation rather
than investigation; one analyst would be sufficient for a plant of this
size and type under normal operating conditions.
Some of the equipment used in this project might be of interest to others.
A Technicon Auto-Analyzer was used to determine total phosphorus in the
plant effluent. An optimized apparatus was under development by the
Technicon Company at the time the necessary individual components were
assembled and put into service in this study on a trial basis by EPA.
Since then, the company has made the equipment available on a commercial
basis. A Hach Model 2100 A Turbidimeter was obtained for the last nine
months of the study; this instrument was most helpful in providing accu-
rate measurement of final and intermediate effluent turbidity. Two Bausch
& Lomb Spectronic 20's of the solid state type were used in the laboratory;
one was set up for and dedicated to phosphorus analyses. The other
Spectronic 20 was used for a wide variety of colorimetric analyses. An
American Instrument microstill was used for determination of total Kjeidahl
nitrogen. All of this equipment served the project well.
Other equipment made available for the laboratory included a Zetameter
(not used to its fullest potential in this particular project as dis-
cussed in Section X) and a Taulman Turbitrol brand jar test apparatus
(whose performance was generally satisfactory).
Laboratory procedures were performed generally in conformance with the
classic analyses described in standard references (8) (9). At the time
the project began, some of the analyses in these recent references were
in a tentative form. Subsequently, however, they have been refined as
reported in the references.
Other treatment plants considering chemical addition to enhance plant
pe-formance need not undertake all of the analyses shown in Figure 1.
For operating control, those required would include all the analyses
91
-------
FLOW
TOT SOL
TOT VOL SOL
SUS SOL
SUS VOL SOL
SET SOL
BOD
00
COO
PHOS
TKN
NH3
NO 2
NO 3
A| tf
FE
AL
S
SO^
CL
PH
TEMP
TURB
COLI
FECAL
R
A
W
•
•
•
•
•
•
•
•
•
•
•
•
•
e
•
•
•
p
R
1
M
E
F
F
•
•
•
•
•
•
•
•
•
•
•
F
1
L
T
E
F
F
•
•
•
F
1
N
A
L
E
F
F
•
•
•
•
•
•
•
•
•
•
•
•
•
R
E
C
1
R
C
•
S
L
U
D
6
E
1
•
S
L
U
D
6
E
2
•
S
L
U
D
6
£
3
•
R
A
W
S
U
P
1
•
•
•
R
A
W
S
U
p
2
•
•
•
R
A
W
S
U
P
3
•
•
•
T
R
E
A
T
S
U
p
•
•
•
B
E
D
0
R
A
N
9
O
FIGURE 1 - APPENDIX A. ANALYSES PERFORMED ON COMPOSITE SAMPLES.
(Thirteen sample stations and up to 25 analyses per station
were involved. About 140 items of data were generated on a
typical day.)
92
-------
done previously (in a conventional operation) plus those related to
phosphorus and the metal salt being added. Analyses for alkalinity
would be required if terminal concentration, after treatment, was 50
mg/1 or less.
A concerted effort was made during this project to insure reliable data
reporting. Towards this end, a laboratory quality control program was
established and maintained throughout the project. Many laboratories
run duplicate samples to verify precision and standard solutions to check
the accuracy of analyses. In this case, the results of these procedures
were recorded and handled statistically to verify whether or not the
efforts of all the analysts and the reliability of the apparatus were
acceptable. This technique (10) indicated data reported here are valid.
Bacterial tests included both total and fecal coliform analyses. These
tests were the only bacteriological work undertaken during the course of
this study. A special room was dedicated to this work. A full time
bacteriologist was not available, however, and one of the regular anar
lysts performed coliform analyses one or two times a week. Nearly all
of these tests were performed on treated effluent in an attempt to differ-
entiate the degree of coliform removal afforded by baseline operation
and routine chlorination versus addition of chemicals with and without
chlorination.
The jar test apparatus was used heavily at the beginning of the trial
runs with iron and aluminum. The apparatus was also used when various
polymers were under preliminary consideration. However, proof testing of
polymers was done on a plant-scale basis and did not require extensive
jar testing. Overall, the jar test apparatus was used only during 5 or
10 percent of the project.
93
-------
APPENDIX B
DATA PROCESSING
It is no overstatement to say that generation of 140 items of data per
day can soon lead to problems in data processing. .Considerable time and
effort were devoted during early states of this project to devise a
reasonable and effective way to handle data; i.e. , to accumulate it in an
accessible form and retrieve and manipulate it in a meaningful way. Some
of the key elements of the data processing program are reported here for
the interest of others involved in similar operations.
A fundamental decision was made quite early: data generated in the
laboratory should be recorded only once. That record, with related
calculations, was to be entered on bench sheets used by the analysts
themselves. A typical bench sheet (for ammonia nitrogen in this case)
is shown in Figure 1. The analyst is required to fill in the date and
then the calculations for derivation of ammonia concentrations at any of
11 sampling stations. Stations where a particular test was not done
were omitted as evidenced by blanks for trickling filter effluent and
recirculation flow in the example being used. After the analyst had
prepared the samples according to the particular procedure involved, the
absorbance (for example, 425 millimicrons for ammonia) was measured and
recorded. Then all calculations for these samples could be carried
forward and recorded under "Results". The "Results" column includes an
obligatory decimal point. Values derived there are rounded off and re-
corded finally under the column entitled "Computer".
At the beginning of a given calendar day, all of the analyses were assem-
bled into a complete set of blank bench sheets. The packet would be split
up and sheets given to those who would perform particular tests. When
the analyses were completed, the packet was re-assembled and represented
the entire laboratory effort for that particular day.
No manual transcription of the original laboratory results into daily,
weekly, or monthly summaries was necessary during course of the project.
Rather, all laboratory data were transcribed from bench sheets directly
onto computer cards. The column numbers for these cards are shown in
Figure 1 and those familiar with electronic data processing will quickly
appreciate the intent and extent of this operation. Every month or so,
the packets of bench sheets would be card punched and carded data trans-
lated onto magnetic tape for processing with any of several programs.
95
-------
DATf
NH3
STAT/OM
FACTO*
0/i. (ST/0M
fOHPUTfR
ff
fti Tfff ffVt.
fffL
*/ VOtff /
Stuoezf t
SUP /
Tiff AT SUP
1 1
FIGURE 1 - APPENDIX B. LABORATORY BENCH SHEET.
(This is a reduced photograph of the Qh x 11
inch sheet that went directly from analyst to
computer cardpunch operator.)
The first program (Figure 2) consisted of an editing run involving such
features as whether or not given data entries fitted within pre-selected
maximum/minimum limits. Any values which exceeded these limits were
printed out for further review to determine whether they had been proper-
ly derived and transcribed. The editing program included the ability to
seek and identify other anomalies and recorded data typical of most pro-
grams in which data processing is undertaken.
96
-------
eEAD AND LIST
MIMIMUW
VALUES FOR
HEADEfc CARD
PROGRAM CHECK
READ AND LIST
MIMIMUNJ VALUES
FOB ALL PACAMETERS
AT ALL MEA&UKMBKJT
ALL \
VAI UF^ > wo
IKJ SEQUENCE/
S . '
WRITE
tee OR.
MESSAGE
I Y»6
>
EEAD AMD LIST
MAXIMUM VALUES
FOB ALL PAEAMETEB.
AT ALL MEASURMENT
* TAT IOWS
MO Q6.T(l FOR.
'HIS DAY
MORMAL DATA
IB THIt> A\
TYPE. O \-
DATA SET/
103X^5
IS RAIN,
&AL OF FLOW
GAL OF
STBeNSTH OF
IU CAUSE
' IS DETAIL
' DATA WITHIKJ
.MAX AKJD MISJ
\ LIMITS
[^
LEGEND
D6CI5IOW POIMT
OPEEATlOKi
COWMECTOE OE END POINT
STATEMEUT
FIGURE 2 - APPENDIX B. FLOW DIAGRAM OF EDITING PROGRAM.
{Unproofed daily data, cardpunched directly from ana-
lysts' bench sheets, was reviewed by this general pro-
gram before further processing.)
97
-------
Another data processing program was called the Daily Report Program
(Figure 3). This program took edited and reconciled data and organized
it into a condensed matrix along with extended calculations. The print-
out, for a given day, fit onto a conventional 11 x 14 inch computer print-
out sheet. This single sheet showed every item of data recorded for a
particular day in the condensed data matrix. This section of the daily
printout constituted a "card image" of the raw input from the daily
bench sheets. In addition to the card image section, daily reports also
had results of over 100 calculations based on the data processed on that
particular day.
Loadings in pounds per day at any of 12 sample stations were computed
for suspended solids, suspended volatile solids, BODr, COD, and phos-
phorus. Ratios were computed for BOD5/COD and volatile solids fractions
at tnese same .sample stations. Both primary and overall percentage re-
movals were computed for suspended solids, 300^, phosphorus, total Kjeldahl
nitrogen, and settleable solids. Hydraulic and solids loadings were cal-
culated for both primary and final clarifiers, with and without recir-
culation. Trickling filter hydraulic and organic loadings were calculated
in several different ways, as were loadings to the anaerobic digesters.
The amount of metal salt and polymers being added including mole ratios,
pounds per day, and milligrams per liter were calculated and recorded.
Cost of chemicals used were computed in cents per thousand gallons. All
of these numbers were included on the single computer printout sheet for
each day.
Selected series of Daily Reports could be fed into the final program,
the Statistical Report (Figure 4). Unlike Daily Reports, a Statistical
Report was relatively long. It covered 36 conventional computer output
pages. On a given page, for example, a parameter such as incoming BOD^
concentration would be analyzed and displayed. The Statistical Program
computed and reported the following: the number of occurrences being
analyzed, maximum and minimum values, arithmetic mean, standard deviation,
standard deviation divided by the mean, and the mean plus or minus one
and two standard deviations. In addition to the above tabulation, the
page showed tabulated class interval printouts for five selected sample
stations in the plant. The limitation to five stations was dictated
by space limitations on the printout paper. These class interval tabu-
lations were cumulative probability tables which showed the group limits
of the class intervals and the plotting position of each, the population
within each group, exceedence values within groups (95% Gaussian normality),
and cumulative frequency. These tables could be used by an experienced
observer as a histogram. Also, technicians could take data from these
tables and plot probability graphs very rapidly to examine normality of
distribution..
The data processing system as described to this point was pursued for a
period of many months. At this point, even the highly condensed formats
had generated an enormous amount of computer output paper. It became a
98
-------
DAILY REPORT
SET «-&£ R.CAO / e \
YES
IMITIAL EJEdT COMTEOL < TAPB /
[
/
IOO c
T^i. COUNTERS "t^S*"1" /^Fll^Y0^
-^ AklD AVEtASES "*» ^ ^ /
o»»yviB»
EUD FILE
TAPES AMD
WBITE
COMPLETION
WESG,
, ,
^~«AyT«»)_MS_ WJJT. _/ ^M^ \_!
YES Ifi^l ^ TTjo — '
p.,p ^,. | 1 N0
BUM
I6I\_
"©
-ES
•^\_/P»,izAMETeK \ YE6 1 OrTIOMfkL SUMMARY DISPLAY
\ '•»* _ /H0 [PeesencTgP »TA. HJTER.KIAL. _
LOPY
eo^t
• TAPE 1
OM TAPB 1
VI
TAPE IF ADOtTIOMS ARI TO
&E MAOe 1 Life OAT
PARAMETEK \
. frfi,
f too
s
LOAD
TEMP
PATE
STOEASE
FOR
/
YB&
n*i
TOO
'
CHLC
i^AL
/ PAEAUETEE
^
X«5
TWO
PAOiB
EJBC.TS
PO
X — I '
J
3CIME AMD
CULATIOW
\ KJO
WEtTE
ME&fi.
^
CALCULATE CH6UICAL FEED
FEBPIU (iAL, L»| 00.
UOLB / DAY
WRITE
HB**P|KJG»
F 1.0-L
>
RE&C?
/ COCIWCT \
•LO^i HO I EKEOB I ^
O* TO
MCA METER.
UUMMR
1 '
i
i t
CALCULATE LIU
THEOU&H TEEAT
WKITE
OF C4C.E
T LOADIWGS
UEWT SYSTEMS
/iUMMAI
I MC
IY\
V
)
OUTPUT
LjklE
ves
t MO
_/LAST\
\ LIUE /
WEITE
SUMMAEY
OATA OkJ
TO.PE
Ii^ CALCULATE COVltEuTeATlOU WEITE
(Lfc/PAYj IPJ 10 FLOW STBlAtJO * LIUES
(£)
'
LEGEND
D6CISIOW POIMT
OPECATIOM
COWUECTOE oe EMP POINT
TOCTCAW STATEMENT
FIGURE 3 - APPENDIX B. PROGRAMING SEQUENCE USED TO PRODUCE DAILY REPORTS
"n Edition to a card image of all pertinent data, printout xncluded
condensed results of over a hundred extended calculations.)
99
-------
STATISTICAL REPORT
EEiD PATA CABDS
THAT SCU6CT (7UTPUT
&er
400
/usrafN
>-jo
-------
real burden, for example, to compare a statistical program output for
several different operating runs. The next step was to convert all of
the paper output into microfiche.
A basic microfiche format was developed as follows: using a 14 x 16
space card, daily reports were arranged to place the seven days of one
week in the vertical fourteen-space column. This left blank spaces
between each day and these spaces were available for summary runs which
accumulated and reported back arithmetic means for any period of time
selected by the operator. With a week's data so arranged in a vertical
format, the balance of the microfiche card would accommodate fifteen
more weeks, giving essentially three months of data per microfiche card.
Once the observer became accustomed to this particular format, he could
compare a very large number of cards in a short time. This particular
style also let him scan Mondays or Saturdays, for instance, in a very
direct manner.
Conversion of Statistical Reports to microfiche formats required a
slightly different approach. As mentioned earlier, 36 computer sheets
were required to provide a Statistical Report for a given period of time.
Using 12 of the 14 available vertical positions on a fiche, a Statistical
Report could be completed in a vertical line three microfiche long. As
many as 16 Statistical Reports could be placed side by side. If the
selection of periods for Statistical Reports was properly made, a rapid
and very instructive survey of specific items could be made. For example,
if the observer wanted to compare effluent phosphorus concentrations
(mg/1) for various periods of time when different plant operations
were being performed, all he would have to do is go vertically to the
proper position and then read horizontally across all of the operating
periods involved.
Some of the data generated in a project of this type will not fit the
classic arithmetic patterns of distribution and normality. This ^s
particularly true where active zero end constraint is involved, such
as observations of dissolved oxygen in treated effluent. Project person-
nel quickly learned to judge whether or not microfiche data was geomet-
rically normal. However, these assumptions could not be presented with-
out more rigorous testing of the data. To do this, data which appeared
to be geometrically normal was plotted on log probability paper. If a
straight line was obtained by this manipulation, the geometric mean was
selected and reported as the proper "average" for this family of data.
If a family of data exhibited a Gaussian distribution, plotting the data
on arithmetic probability paper allowed verification and selection of an
"average". Sometimes data were irregular in terms of either of these
classic distributions, and in those cases all characteristics were studied
and finally the best approximation of an "average" was extracted and re-
ported.
Late in the project, computer probability plots and scatter diagrams
were generated to speed up the process of displaying analyzed data.
101
-------
Some of these plots are shown in the body of this report. Except for
the speed and convenience of computer plotting, these were not essential
to the project. On the other hand, the basic collection and manipulation
of data by electronic computers were absolutely mandatory to successful
analysis of this investigation.
102
-------
APPENDIX C
DETAILED DATA SUMMARY FOR
EXTENDED ALUM RUN
103
-------
TABLE 1 - APPENDIX C
DETAILED SUMMARY OF MAJOR PARAMETERS DURING EXTENDED ALUM RUN
APRIL 1971
•~rr
A
T
E
1
2
3
^
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
AVJ*
RW
FLOW*
(MGD)
1.46
1.54
1.45
1 46
1.22
1.39
1.36
1.56
1 17
1 66
1.22
1.41
1 40
1 26
-i rt r
. .
0
1 43
1 46
1 45
1 59
1.49
1 47
1 AQ
BODS (me/1)
RW
188
174
154
162
124
109
132
70
79
172
146
171
172
14?
PE
98
86
74
_
.
48
60
5g
4S
106
98
69
73
82
74
FE
t
16
19
12
_
10
8
10
10
8
11
12
3
4
.
11
RW - Raw Wastewater
PE - Primary Effluent
FE - Final Effluent
COD (mg/1)
RW
384
384
364
452
532
532
_
_
348
428
484
516
44ft
329
324
293
356
416
397
483
376
424
.
40
PE
232
152
152
265
_
_
238
274
224
281
27 2
219
234
194
282
194
722
286
234
229
.
_
23;
FE
100
_
_
70
60
130
?16
126
_
118
107
99
14?
ILL
92
51
43
48
37
_
75
24
40
36
.
' 86
SS (mg/1)
RW
Hfi
—
_
58
784
?44
182
268
_
186
_268
178
368_j
24 8_
_
186
160
118
156
133
218
152
194
182
_
_
196
PE
66
_
_
70
92
82
_
9S
_
_
6.2
142
118
74
98
_
56
94
88
88
36
_
88
76
188
82
_
_
89
FE
?5
_
_.
17
5
14
28
18
-
_
n
23
24
16
17
_
6
13
7
n
3
_
18
7
5
4
_
_
14
TP (rng/1
RW
7,1
.
_
13.8
14.6
5.1
13.9
14.fi
-
-
14,3
18.2
17. 4
n.s
14.8
_
_
10.2
12.3
9 7
12.1
n.?
_
_
10.1
12.8
13.6
11.8
-
-
13.
PE
9.3
_
_
12.8
12.8
13.3
-
14.0
-
-
14.1
17.2
8.6
1?.4
14. S
_
_
9.9
10.3
6.7
fi.8
6.0
-
_
7,5
10.7
8.
7,5
-
-
_
10.
P)
FE
6.0
_
-
10.9
8.2
9.9
ll.fi
11.8
-
-
17..5
7.fl
5.6
9.0
1(1.?
_
_
7.3
3.6
0.4
0.3
0.3
-
_
1.4
1 .
0.
0,
-
-
_
5.
MOLE
RATIO
(Al/P**)
0
.
-
0
0
o
0
0
0
0
0
0
0
0
0
0
0
0
2.98
4.72
2.76
1.97
-
-
,44
1.91
1.88
2.18
-
-
_
2.36
SS - Suspended Solids
TP - Total Phosphorus
Flov7*-Design = 1.5 MGI
V**- Tot. Phos. in RW
-------
TABLE 1 - APPENDIX C
DETAILED SUMMARY OF MAJOR PARAMETERS DURING EXTENDED ALUM RUN
MAY 1971
'!>'
A
T
E
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Avg.
RW
(MGD)
1.37
1.48
1.27
1.36
1.40
1.34
1.26
1.38
1.33
1.30
1.30
1 26
1.29
1.37
1.35
1.34
1.32
1.33
1.39
1.31
171
1.49
1 56
1.37
BODc; (mg/1)
RW
_
185
15?
?30
155
181
.
—
_
_
.
14?
240
175
1,88
120
183
177
PE
_
126
89
11 5
88
140
_
_
—
.
.
116
132
106
122
84
134
114
FE
_
12
7
13
9
16
_
_
_
_
_
8
14
10
8
10
34
13
COD (mg/1)
RW
329
458
376
528
364
376
388
388
520
426
31?
43?
516
446
349
372
466
447
482
269
.
375
410
PE
255
318
283
26,6
245
258
249
237
250
210
200
320
252
314
256
301
278
257
262
211
_
_
298
258
FE
51
20
39
?0
43
67
31
44
31
39
20
35
60
23
39
32
24
44
39
31
M
_
51
37
SS (mg/1)
RW
112
142
16.fi
??8
244
166
284
148
374
238
108
164
164
282
110
134
312
252
234
144
_
..
138
197
PE
i
92
144
100
1?,0
146
76
112
86
72
82
88
182
84
112
106
100
176
106
98
78
_
.
_
170
111
FE
7
4
ft
10
10
9
10
12
4
5
6
15
'4
3
12
11
11
9
6
5
_
_
.
18
8
TP (mg/1 P)
RW
11.5
15.4.
13.2
10.1
10.2
11.7
14,5
13.9
9.2
13.6
12.1
14,2
14.7
171,4
8.2
_
12,2
11.9
16.1
9.8
-
_
-
10.5
12.2
PE
10.5
17.9
10.6
10.4
7.9
10.0
10.2
11.2
7.6
6.2
7.2
„
9.6
9.5
8.8
_
10.5
9,2
9 9
8.3
-
_
-
12.4
9.6
FE
0.9
O.1?
0.7
O.1)
1 .3
1.5
0.8
1,0
0.9
0.3
0.7
0.3
0.4
0.5
_
0.5
0,7
0.8
0.8
-
_
-
0.9
0.7
MOLE
RATIO
(Al/P**)
1.97
1.44
1.85
2.26
2.17
?.03
1.43
1.56
2.40
1.62
2.35
1.59
1.56
1.86
2.86
_
_
_
1.83
1.99
1 .13
2.12
-
_
-
1.89
1.90
RW - Raw Wastewater
PE - Primary Effluent
FE - Final Effluent
TP - Total Phosphorus
Flow*-Design =1.5 MUD
p*ft_ Tot. Phos. in RW
105
-------
TABLE 1 - APPENDIX C
DETAILED SUMMARY OF MAJOR PARAMETERS DURING EXTENDED ALUM RUN
JUNE 1971
A
T
E
1
2
3
4
5
6
7
8
9
10
11
1 2
13
14
15
16
17
18
1 Q
20
21
22
23
24
25
26
27
28
29
30
31
Avg.
RW
(MCD)
1.52
1.50
1.44
1.38
1.33
1.48
1.35
1.32
1.31
1 ^n
1.29
1.37
1.31
1.33
1.30
1.34
1.33
1.84
1.32
1.37
1.34
1.32
_
1.30
1.47
L.36
1.36
1.38
BOD
RW
181
177
143
116
162
163
230
158
208
714
194
113
171
122
114
210
138
_
.
187
_
175
187
168
5 (me
PE
124
96
81
67
105
_
137
109
125
1%
140
74
88
67
87
126
97
_
—
131
_
117
107
107
/I)
FE
9
7
4
4
7
7
7
5
8
9
9
4
5
17
8
11
6
_
—
8
_
9
6
7
cor
RW
360
504
376
408
588
49?
428
345
310
404
TH
400
388
247
368
293
439
380
_
_
334
388
490
395
) (nig
PE
253
253
257
276
_
78?
247
282
212
273
76?
261
237
191
208
218
203
234
_
•n
247
294
751
76?
?.4fi
a)
FF.
32
43
_
36
39
fin
66
31
35
47
43
40
47
43
51
32
38
40
_
**
43
66
^
35
_
44
SS
RW
186
248
148
168
196
128
198
200
114
172
166
180
148
106
164
116
242
146
_
_
136
196
746
166
189
(mfi/1
PE
104
106
56
90
.
108
124
86
116
96
102
74
66
76
82
124
122
_
_
96
124
108
96
109
)
FE
9
9
.
11
16
?
4
9
11
13
9
12
12
3
19
16
10
18
_
_
6
19
17
6
11
TP (
RW
10.6
10.5
11.0
10.4
14.6
T2.5
11,9
9.2
9.5
13.1
11 .'i
11.6
8.7
10.9
9.0
15.6
13.2
10.9
„
_
10.2
12.9
10.3
17. -7
11.4
mg/1
PE
9.9
9-0
9.8
8.3
12.1
17..7
9.6
8.3
8.0
10.3
9.3
9.4
7.9
6.6
7.5
13,7
9.6
8,9
_
8.1
10.9
7.9
8.9
9.4
P)
FE
0.7
0.8
0.6
0.4
0.8
0.4
0.5
0.4
0.4
0.6
O.fi
0.6
0.4
0.3
1.0
1.0
0.5
0,5
_
0.5
1.4
0.4
0.3
0,6
MOLE
R ATTH
(Al/P**)
1.92
1 .97
1.96
2.64
1.52
1.84
1.97
2.57
2.98
1.73
?.Rfi
2.01
2.74
2.52
1.87
1.50
1.71
2.12
_
2.75
1.63
9.71
1.82
?.na
RW - Raw Wastcwatcr
PE - Primary Effluent
FE - Final Effluent
SS - Suspended Solids
TP - Total Phosphorus
Flow#-Design =1.5 MCD
P**- Tot. Phos. in K;,'
106
-------
TABLi: 1 - APPENDIX C
DETAILED S1/M.V.AKY OF MA.lOtt PARA>£TERS DURING EXTENDED ALUM RUN
JULY 1971
D
A
T
E
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Avg.
RW -
PE -
FE -
™ «)D
FLOW--- ;
(MGD) RW
1.40 235
1 .33
1.30
1.32
1.37
1.38
1.39
1.31
1.23
1.31
1.33
1.32
1.32
1.32
1.34
lrA3
1.34
1.35
1.35
1.54
1.40
1.41
1.54
1.80
1.74
1.63
1.40
176
216
218
_
167
245
258
225
245
211
259
265
228
158
215
157
135
91
206
? (mg/U
PE FE
COD
P,, 1
KW
1
119 5 : 340
107
100
—
—
_
_
212
208
165
185
186
i8i
161
109
91
101
78
143
5
3
5
7
6
8
13
R
5
6
6
5
6
3
6
6
7
7
.
_
6
380
404
425
344
368
432
381
499
360
379
46.4
356
439
404
368
384
408
235
214
_
379
_1';'S'
_ PE _
252
220
?00
232
178
313
249
257
245
227
778
268
279
196
166
235
238
149
134
_
_
225
fl)
FE
44
52
52
38
36
43
51
44
48
51
44
54
32
36
47
32
35
37
37
39
_
_
43
SS (mg/1)
RW
132
176
176
?06
148
148
154
130
166
98.
130
184
144
160
190
114
136
186
126
98
_
_
150
PE
—
96
86
78
6.4
64
84
84
96
66
72
98
88
82
64
32
84
100
56
40
_
_
77
FE
7
a
4
4
8
6
5
4
14
8
9
4
5
"?
6
7
8
9
9
11
_
_
7
TP (mg/1 P)
RW
12.0
13.4.
U-2
n.8
10.5
9. 1
11.7
10.0
11.8
10.3
10.7
10.5
in. 3
11.8
10.4
9.3
11.1
10.0
5.0
•5,?
_
_
10.4
PE
-S^JL
8.1
6.5
5.1
9,2
7.6
7.7
7.6
7.3
8.4
8.6
8.0
6.5.
7.1
8.7
8.6
4.1
4,9
-
-
7.4
FE
0.6
_
0.6
0.4
0.4
0.4
0.3
0.7
0.5
0.6
0.7
0.5
0.7
0.5
0.5
0.5
_
.
0.2
0.7
0.6
0.4
0,7
_
-
0.5
MOLE
RATIO
(Al/P**)
1.84
_
1.75
2.02
1.63
2.12
3.26
2.02
2.33
1.99
2.28
2.54
2.06
2.24
1.95
2.21
_
—
2.80
1.98
2.01
3.44
3.24
-
-
2.29
Raw Wastowater SS - Suspended Solids
Primary Effluent TP - Total Phosphorus
Final Effluent Flow*-Design = 1.5 MGE
P**- Tot. Phos. in RW
107
-------
TA151.K 1 - APPF.NHIX C
DETAILED SUMMARY OF MAJOR PARAMETERS DURING EXTENDED ALUM RUN
AUGUST 1971
D
A
T
E
1
2
3
L,
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Avj-,.
RW
FLOW"
(MGD)
IWDci Cmg/1)
RU'
1.46 133
1.56
153
1.55 173
1.50
1.34
1.35
1.27
1.52
1.33
1.38
1.43
1.38
1.46
1.92
1.74
1.62
1.70
1.58
1.58
1.63
1.56
1.50
1.51
1.51
1.64
1.44
1.32
1.52
1.43
1.52
1.48
1.50
150
125
137
310
160
217
75
113
168
150
190
142
183
177
315
130
I1)?
.
190
328
_24Q
179
_££^
76
cor
KW
2 < 283
92 i 3 1 313
_
126
90
106
117
92
121
97
73
113
110
96
90
119
141
151
97
]?1
—
120
185
156
113
5
8
3
3
13
5
7
4
4
11
8
9
11
3
6
1
2
6
.
.
3
6
7
6
332
376
282
348
445
470
400
390
263
411
316
536
328
512
3 §4
543
372
_3jO_9_
_
.
370
392
388
381
) Cn^/1)
PE
190
FE
ir
219 j 51
225 ! 40
235
220
244
334
255
258
333
202
250
234
260
230
250
?67
281
290
270
_
.
255
264
h280_
250
35
35
35
58
35
50
46
49
49
53
41
60
48
4^
76
55
47
.
_
39
58
45
47
SS (mg/1)
RW
,26
174
_
146
144
102
146
158
188
332
84
128
126
196
116
128
16?
184
134
-18_6_
.
—
182
132
?10
158
PE
74
90
_
94
98
104
128
112
100
108
66
70
82
98_
96
76
l?f>
98
110
102
-
—
90
114
104
143
FE
7
11
_
13
9
7
11
13
10
7
5
14
6
_i
15
10
?fl
14
13
15
_
11
10
8
11
'IT (mg/1 P)
RW
9.1
9.7
9.2
10.0
10.4
9.5
12.0
11.0
11.4
11.5
9.3
11.6
10.2
12.0
10.2
11.5
in.fi
8,5
9.4
11,9
_
10.8
12.4
1?.?
10.6
PE J VE
8,7
7.6
8.0
8.1
6.6
7.6
10.6
7.7
9.1
8.2
8.0
10,2
8.6
9.6
8.0
9.6
a.1)
7,9
8.3
_
^
^
9.8
10.0
7.9
8.2
0.0
0.6
0.8
0.8
0.6
_
0.6
1.1
0.7
0.6
0.6
0.7
1.1
1.0
0.7
1.1
0.6
(1.9
0,9
0.8
1.3
_
_
0.8
1.1
0.9
0.8
MOLE
RATIO
(Al/P**)
2.74
2.05
2.17
2.07
2.22
_
2.54
1.94
2.04
1.90
1.95
2.25
1.65
1.79
1.63
1.92
2.11
1.94
2,41
2.01
1.81
_
_
2.36
1.64
1.72
2.04
RW
PL
FE
Raw Wastcwater
Primary Lf.fluent
Final Effluent;
SS - Suspended Solids
TP - Total Phosphorus
Flow*-Design = 1.5 MGD
P**- Tot. Phos. in RW
108
-------
TABLE 1 - APPENDIX C
DETAILED SUMMARY OF MAJOR PARAMETERS DURING EXTENDED ALUM RUN
SEPTEMBER 1971
D
A
T
E
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Avg
RW
FLOW*
( MGD)
1.46
1.38
1 50
1.60
1 45
1.40
1.18
1.37
1.30
1 40
1 77
1.41
1.43
1.41
1.53
1.17
1 18
1 43
1 34
1.41
1 29
1.69
1 65
1 62
1 81
1.61
1.58
1 51
1 Al
1 76
1.44
BODS (mg/1)
RW
258
230
285
232
250
225
260
262
260
262
168
168
165
135
100
100
153
160
223
155
160
201
PE I FE
149
145
212
182
160
155
190
165
175
_14JL
88
98
84
81
70
67
104
104
94
120
110
128
7
6
17
13
7
9
9
6
6
4
4
4
1
4
3
3
3
4
5
5
3
6
COD (mg/1)
RW
369
342
352
163
555
336
384
428
432
450
352
400
427
436
395
273
290
431
30Q
378
384
386
PE
272
252
338
353
285
263
265
289
261
272
230
231
229
240
229
230
220
26
25
26
30
26
FE
47
54
75
63
63
65
48
40
51
43
40
55
4g
56
63
32
_
47
55
44
40
48
_5_1_
ss (
RW
138
168
130
138
170
174
114
116
140
154
114
140
164
140
198
94_
^
_
110
150
MJ.6
116
166
138
mg/1)
PE 1
1
114
102
1?4
150
126
122
8?
96
100
104
58
88
96
100
102
10Q
_
24
140
120
112
124
.
JJ24_
FE
14
15
17
14
15
11
8
7
13
7
2
1?
12
21
7
10
_
_
15
13
20
8
12
—
11
TP (rap, /I P)
RW
9.6
12.0
9.7
2.4
10.8
12. 2
4 3
16.0
12.1
13.4
11.3
14.0
11. ?
11.1
8.7
8,7
-
_
10.2
11.4
10.2
11.4
11 .5
_
11.5
PE
8.2
8.8
7.3
11.2
9.3
10.8
9 6
13.4
_
9.2
9.1
7.2
5.4
fi.fi
5.5
5.3
-
-
7.6
10.2
-
8.4
8.8
_
8.8
C.
FE
0.7
0.9
1.6
1.?
0.7
0.9
.
n,s
1.3
0.6
n.3
0.2
_
_
0.7
0.0
n.fi
0,9
0.1
-
-
0.4
0.
0.6
0,4
0.
_
0.
MOLK
RATIO
(Al/P**)
2. 17
1.87
2.28
2.12
2.09
1.95
_
1 Rl
1.35
1.82
1.51
2.34
_
_
1.94
1.96
2.16
2.11
7.99
-
-
1.89
1.72
2.01
1.93
2.14
-
1.97
l.x-1 ^r> 1 i f] Q
RW - Raw Wastc-waLer
PE - Primary Emine
FE - Final Effluonl
TP - Total Phosphorus
Flow*-lVsii;n = 1.5 MCD
p;;«_ Tot. Phos. in RW
109
-------
TAB1-E 1 - APPENDIX C
DETAILED St'MMARY OF MAJOR PARAMETERS DURING EXTENDED ALUM RUN
OCTOBER 1971
D
A
T
E
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
2 ft
27
28
29
30
31
AvK.
RW
( MC;D)
1.25
1 49
1.57
1.78
1.76
1.85
1.78
1.63
1.61
1.57
1.52
1.51
1.49
1.60
1 55
1 50
1.56
1.76
1.25
2.44
1.74
1.75
1.72
1.72
7 10
1.85
1.95
1.85
1.88
1.84
1.78
1.78
BOD
RW
145
75
146
80
155
79
184
147
194
148
182
170
80
75
62
93
96
89
10?
85
—
_
130
117
. (,.,„
PE
74
60
61
60
80
68
717
95
114
130
160
96
50
60
40
83
63
54
58
64
_
_
66
84
/I)
FE
4
4
4
4
1
0
5
7
7
4
3
4
5
1
4
5
4
3
3
_
_
4
3
coc
RW
204
173
221
332
1fi4
100
388
148
15?
11'
376
174
86
101
11?
476
223
548
770
272
_
—
248
290
(mg/
PE
146
135
167
206
709
741
292
770
333
108
268
248
175
.
104
151
173
158
174
155
_
_
200
203
'!)
FE
38
35
47
42
14
14
44
1?
19
5?
43
58
43
39
43
15
46
18
36
11
_
„
31
40
SS
RW
204
78
56
110
186
7?
100
86
?60
220
252
160
188
52
212
178
154
204
718
96
_
_
170
155
(mg/1
PE
84
50
76
110
1?0
fiO
118
700
16?
718
200
1?4
260
50
110
154
114
74
708
84
—
98
128
)
FE
10
9
13
12
8
9
4
0
.
—
20
28
2
6
?5
17
16
77
19
_
35
19
TP (
RW
3.7
5.5
6.7
7.6
9.1
8,9
11,9
n,o
10.2
9,7
12,0
10,0
2.4
3,2
3.5
7.7
9,7
12.4
8,4
6.1
_
9,6
8.1
mg/1
PE
5.4
5.7
6.5
6.5
7. a
8.4
9,5
10,2
10.0
9,5
9,P
3,8
4.0
3.9
7,fi
7,6
5.8
6,6
fi.?
_
9,6
7.2
P)
FE
0.1
0.4
0.5
0.5
0.7
0.5
1,0
o,S
0.7
n,7
0,7
1,1
0.5
0,5
0,8
_
1 .7
1 ,n
0.4
0.1
0.4
—
_
o,9
0.6
MOLE
RATTO
6.28
3.16
2.63
1.88
1.60
, fi1
1,46
1.35
1.74
i «n
_
.
1.95
1.50
8.81
3.39
4.34
—
.
9.75
1.37
1.29
2.06
9.14
•
_
2.13
2.69
RW
PE
FE
Raw Wastewatur
PriLiarv Kl'iiluenC
Final Effluent
SS - Suspended Solids
TP - Total Phosphorus
Flow*-Design = 1.5 MCD
P*»- Tot, Phos. in RW
110
-------
TABLE 1 - APPENDIX C
DETAILED SUMMARY OF MAJOR PARAMETERS DURING EXTENDED ALUM RUN
NOVEMBER 1971
"D"
A
T
E
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21 _,
22
23
24
25
26
27
28
29
30
31
Avg.
RW
FLOW*
(MGD)
1.77
1.60
1.68
1.61
1.46
1.62
1.50
1.60
1.55
1.45
1.50
1.53
1.26
1.34
1.37
1 .48
1.68
1.68
1.49
1.53
1.38
1.51
1.56
1.44
1.46
1.35
1.38
1.40
1.42
1.49
1.50
BODS Cmg/1)
RW
123
98
102
98
_
145
210
717
177
140
.
233
167
169
225
102
.
144
144
147
169
204
134
157
PE
80
81
61
56
_
87
100
fl9
108
103
.
1?4
109
80
100
7?
_
.
107
103
100
.
109
149
122
97
FE
3
3
2
1
_
2
4
?
4
3
.
7
2
2
4
3
.
.
1
2
4
3
9
18
4
COD Cmg/1)
RW
268
265
268
238
31A
420
3fi4
389
305
^
^
371
375
608
392
.
.
•m
384
348
345
409
445
360
PE
184
I
211
203
176
216
223
??n
214
210
_
274
242
223
192
__
_
222
223
209
_
_
226
339
336
.
255
FE
23
43
42
31
43
37
47
39
44
_
51
43
39
35
_
_
31
31
31
_
56
68
87
_
44
SS (mg/1)
RW
172
148
126
460
—
«.
150
14$
80
334
94
^
172
m
276.
106
—
—
100
260
86
•»
«
134
148
13fi
17?
PE
110
130
152
394
_
_
110
76
66
84
90
„.
98
94
148
84
—
—
118
124
68
_
M
_
86
168
Ififi
_
1?S
FE
12
34
32
43
_
—
10
9
15
15
67
_
19
?3
26
—
«
13
22
9
^
_
_
6
_
%
-
23
TP (mg/1 P)
RW
11.5
10.2
il,5
10.2
12.0
12.4
V? . ?
12,2
12,0
11.2
12.0
n 4
11.4
10,2
_
12.0
13.4
1?.?,
_
..
_
11.0
-
11.0
_
11.7
PE
8.8
8.4
3,6
7.6
8.1
6.9
in.n
7,5
8,1
8,0
10.4
q.fi
6.1
7,0
-.
8.9
8.3
7.8
-
_
-
—
8,9
-
13.0
_
8.5
FE
1.2
0.9
0.9
1.1
0T8
0T8
n.7
0,4
0.4
0,9
0.8
0,5
0.7
.
_
_
0,4
0.6
1.0
-
_
_
_
0,7
-
2.4
.
0.8
MOLE
RATIO
(Al/P**)
1.30
1.62
1.37
1.61
2,03
1.33
1.40
1.49
1,47
2,43
1.61
-1,",
1.38
1,54
.
.
2,20
1.31
1.19
-
_
.
^
2,3?
-
1.61
'
1.62
RW - Raw WasUewater
'PE - Primary Effluent
FE - Final Effluent
TP - Total Phosphorus
Flow*-Design =1.5 MGD
P»»- Tot. Phos. in RW
111
-------
TABLE 1 - APPENDIX C
DETAILED SUMMARY OF MAJOR PARAMETERS DURING EXTENDED ALUM RUN
DECEMBER 1971
MH>C
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Avg.
RW
FLOW*
(MGD)
1.28
2.72
2.64
2.60
3.51
3.20
2.99
2.79
3.17
3.38
3.75
3.57
3.57
3.27
3.14
?^S&
2.68
2.70
2.45
2.50
2.36
2.28
2.28
2.23
1.92
1.80
2.00
1.89
1.89
1.85
1.95
2.61
BODS (mg/1)
RW
106
49
.
54
4fi
46
32
27
„.
44
39
in
52
SS
_
57
194
63
_
_
.
.
.
_
60
PE
_
.
.
.
.
.
.
_
^
_
_
_
_.
_
.
_.
.
.
.
_
.
FE
11
5
.
?0
6
•?
2
8
„
_
_
n
18
S
_
5
4
3
_
.
.
.
_
_
8
COD (rog/1)
RW
353
201
?5T
H4
If?1?
224
_
_
91
92
129
103
86
„
115
2L6
178
.
_
.
_
167
PE
^
^
.
.
.
_
_
_
_
_
_
_
_
^
_
.
_
.
FE
197
53
29
46.
44
36
40
w
_
51
48
41
35
IS
_
20
36
.
_
.
—
51
SS (mg/1)
RW
180
80
110
T>4
42
84
_
_
•}R
78
4fi
70
4ft
_
28
.
.
^
.
_
80
PE
.
_
.
.
.
_
.
_
_
_
_
_
_
_
,
_
_
.
FE
32
39
32
35
28
•.
.
•^1
34
Ifi
20
22
_
6
_
_
_
.
_
27
TP (mg/1 P)
RW
11.8
5.1
4-1
4 ?
4.7
?,7
2,6
.
4,?
5.1
4.?
6.7
—
5.7
6.5
7.6
_
_
.
_
5.3
PE
.
_
.
.
w
.
—
^
—
^
.
^
.
^
.
^
FE
2.0
1.5
1-2
1.8
2.1
!,•>
1,6
.
2.4
1 .8
2.2
_
0.9
1-5
0.8
_
_
.
_
1.6
MOLE
RATIO
(Al/P**)
1.75
1.91
2,45
1.97
2.11
_3JL_
3.21
.
9 Tft
1.54
i q^i
1.37,
_
1.89
K63
1.47
_
„
.
^
?.08
RW - Raw Wastcwater SS - Suspended Solids
PE - Primary Effluent TP - Total Phosphorus
FE - Final Effluent Flow*-Design =1.5 MGD
P**- Tot. Phos. in RW
112
-------
TABLE 1 - APPENDIX C
DETAILED SUMMARY OF MAJOR PARAMETERS DURING EXTENDED ALUM RUN
JANUARY 1972
D
A
T
E
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Av>>.
RW
( MGD)
1.80
1.85
1.99
1.85
1.90
1.98
1.99
2.11
2.02
1.87
1.91
1.83
1.72
1.63
1.69
1.67
1.84
1.70
1.75
1.65
1.75
1.17
1.77
1.83
1.73
1.73
1.78
1 75
1.17
1.74
1.71
1.77
BODS (mg/1)
RW
_
113
127
85
145
98
.
132
86
75
107
109
.
.
91
129
189
125
114
164
174
159
200
190
152
161
133
PE
_
80
104
_
72
63
_
99
74
53
67
63
_
_
90
.
90
86
64
_
97
87
99
108
133
102
99
82
FE
_
2
_
2
3
_
_
5
5
2
4
4
_.
.
2
2
4
4
2
_
1
2
2
3
4
3
3
COD (mg/1)
RW
—
221
416
206
214
270
_
250
322
228
206
?34
_
_
288
352
328
276
250
_
326
376
402
434
321
302
387
300
PE
_
159
330
164
167
182
.
_
177
204
186
176
182
_
_
192
_,
218
206
166
^
211
219
236
179
188
22
22
20
FE
_
58
39
31
38
48
_
31
35
38
34
36
_
_
40
68
39
47
28
_
.
35
47
39
43
39
_
27
40
40
SS (mg/1)
RW
_
632
138
98
102
108
.
_
144
112
7fi
118
100
_
_
84
108
194
102
212
_
.
1?0
n?
146
460
1??
_
136
154
164
PE
_
602
92
66
68
44
.
_
110
116
«?
52
62
_
_
80
»
112
78
60
_
_
92
106
88
78
60
_
_
98
96
107
FE
_
468
22
13
13
14
.
-
-
14
13
6
8
_
.
12
32
9
11
_
-
_
in
n
7
8
1
_
_
_
2
36
TP (mg/1 P)
RW
_
12.0
9.0
6.4
6.4
6.8
.
-
8.9
9.7
fl.fl
7.7
7.6
_
8.8
10.8
8.8
9.4
10.1
-
_
11 .0
12,8
10.0
9.6
8.9
_
_
11.5
10.8
9,4
PE
_
11.0
9.6
5.8
6.5
5.3
_
-
7.3
7.8
7.8
6.3
6.1
_
_
6.7
8.4
7.9
7.2
5.1
-
_
7.9
8.7
7.0
•?.?
6.4
_
_
7.8
7.6
7.2
FE
_
_
2.4
1.4
1.2
1.2
.
-
1.1
1.5
1.4
i.l
0.8
-
_
0.7
1.2
1.0
1.0
0.6
-
_
fl,1)
0.9
0.5
0.5
0.6
_
-
0.5
0.6
1,0
MOLE
RATIO
(Al/P**)
_
1.64
1.48
2.23
2.17
1.96
_
-
2.03
1.46
1.57
,_ 1.88
2.02
-
-
2.48
1.33
2.07
1.88
1.96
-
-
1.87
1.32
1.79
1.87
1.96
-
-
1.82
1.68
1.84
RW
PE
FE
Raw WnstewaUer
Primary t£f:luent
Final Effluent
TP - Total Phosphorus
Flow*-Design =1.5 MGD
P*«- Tot. fhos. in RW
113
-------
TABLE 1 - APPENDIX C
DETAILED SUMMARY OF MAJOR PARAMETERS DURING EXTENDED ALUM RUN
FEBRUARY 1972
A
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Avi;
RW
FLOW*
( MGD)
2.01
1.82
1.61
1 51
1 71
1.72
1.78
1.69
1.73
1.62
1.77
1.74
1.73
1.75
1 73
1.71
1.74
1.63
1.64
1.61
1.60
1.72
1.75
1 74
1.70
1.71
1.59
1.70
1.45
1.70
BODq (rng/1)
RW
141
143
140
166
167
210
176
156
177
152
185
151
158
149
164
147
127
175
199
186
161
PE FE
68
88
55
103
85
98
112
78
117
104
92
130
121
119
108
143
62
97
153
143
154
106
6
2
3
5
6
5
1
2
.
2
1
2
3
1
4
6
5
4
10
-
6
5
.
4
COD (mg/1)
RW
334
370
769
302
414
372
440
332
„
31?
440
392
298
36?
546
345
389
303
S76
460
.
PE
194
233
178
190
240
212
255
238
_
31?
255
250
714
280
783
306
230
260
278
•nn
.
.
FE
59
44
48
?4
56
36
44
52
_
40
33
31
74
_
25
40
40
36
36
.
39
40
_
.
382 i 248 39
SS (mg/1)
RW
120
172
114
114
754
150
140
90
_
104
142
96
136
94
_
188
156
144
220
198
.
108
118
164
_
_
144
PE
96
14
68
90
100
94
102
108
_
114
122
104
106
126
_
150
118
128
138
T84
_
170
132
lift
_
_
116
FE
36
20
8
7
14
8
7
6
-
7
11
7
10
14
_
37
6
12
7
51
_
_
13
14
8
_
..
14
TP (
RW
9.6
10.8
8.6
12.7
12.2
10.5
10.5
10.7
.
-
n.4
12.8
11.6
12.8
-
_
_
11.5
12.0
12.2
12.8
10,8
_
_
12.1
12.0
12.7
-
-
11.4
mg/1
PE
6,1
7.7
6.3
7.6
9.4
7.3
6.7
7.0
_
-
7.9
8.4
7.3
8.2
7.8
-
_
10.0
9.8
10.2
8.8
9 0
_
_
10.3
10.1
7.9
-
-
P)
FE
2,0
0.8
0.4
.
0.2
0.4
0.3
0.4
0.4
_
-
0.4
0.5
0.4
0.4
0.4
-
_
0.2
0.4
0.5
0.5
O..1?
_
-
0.7
0.7
0.5
-
-
8.3J 0.5
MOLE
RATIO
(Al/P**)
1.61
1.58
2.24
_
_
1.67
1.43
1.75
1.71
1.91
_
-
2.51
1.38
1.54
1.42
-
-
_
1.97
1.61
1.48
1.38
1.65
-
-
1.90
1.70
1.68
-
-
1.71
RW - Rav Wastewater SS - Suspended Solids
PE - Prunarv Effluent I'l' - T°'-^ Phosphorus
FE - Final Effluent Flow*-Ues ipi -1.5 MCL
[>;,-«•_ Tot. I'hos. in l\\\
114
-------
TABLE 1 - APPENDIX C
DETAILED SUMMARY OF MAJOR PARAMETERS DURING EXTENDED ALUM RUN
MARCH 1972
""D '"
A
T
E
1
2
3
4
5
6
7
8
9
10
11
12'
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Avf, .
RW
FLOW"-
(MGD)
1.65
1.54
1.63
1.68
1.71
1.63
1.54
1.65
1.59
1.52
1.46
1.55
1.53
1.49
1.55
1.46
1.37
1.49
1.57
1.50
1.55
1.49
1.52
1.58
1.50
1.39
1.54
1.40
1.35
1.31
1.26
1 S9
UW - Raw Wa
PE - Primal^
T'E - Final
I»DS (HE/D
RW
250
147
.
164
248
'65
790
.
_
275
224
218
162
.
.
190
215
•180
190
293
143
183
158
201
180
_ZQS_
-.lowa
• !•££
ifllu
PE
181
114
.
138
744
.
_
175
171
132
.
.
160
140
110
90
265
60
175
119
138
115
_L.46_
Ler
lucnt
ent
FF.
5
6
_
7
6
8
.
_
.
_
.
.
3
4
3
4
3
8
8
5
2
5__
COD (mg/1)
RW
508
320
_
336
591
411
340
431
.
412
432
392
44a
338
_
.
382
370
384
396
340
.
286
310
286
368
294
380
PE
340
273
_
332
360
286
265
270
.
274
314
314
368
326
_
_.
361
378
245
317
313
.
208
.
248
257
227
299
FE
51
43
_
47
47
58
4?
49
.
_
41
46
49
^4
43
_
_
40
52
47
47
59
.
22
35
31
40
20
44
SS (ni<7,/l)
RW
190
102
_
216
?16
164
280
914
.
_
146
140
176
156
90
_
_
128
128
162
122
84
-
118
118
84
164
106
150
PE
162
106
_
166
176
222
104
.
_
98
130
116
148
138
_
_
164
182
158
132
118
_
206
184
10?
110
10?
144
FE
7
7
_
17
4
5
6_
1?
_
7
a
4
1?
7
_
„
13
16
18
6
4
_
15
4
5
10
17
9
TP (rnc/1 P)
RW
10.8
12.2
_
12.5
•n,4
13.4
13,4
10.0
.
_
12.5
11 ,8
12.6
11 6
11.3
_
_
13.0
13.8
13.8
13.4
17?
_
^
8.7
12. Oj
8.9
10.6
10.7
11 9
PE
9,2
S.I
_
10.6
11,0
9.0
10,8
7,0
_
_
9 4
10.1
8,9
8.1
_
_
11.0
12.8
13.8
10.0
10.0
_
6.8
6.7
5,3
6.1
5.7
9,1
FE
0.4
0,3
_
0.4
o.s
0.4
0.5
0,4
_
0.3
0 3
0.4
0 4
0.4
_
_
0.4
0.5
0.6
0,4
0,5.
_
_
0.1
0.2
0.2
0.2
n.?.
0 4
MOLE
RATIO
(AI/P*-:M
L74
1-68
_
1.71
1 42
1.50
1.40
1 96
_
1.88
1.73
1.65
1.72
1.88
_
_
1.78
1 51
1.45
1.50
1-68
-
..
2 90
1.65
. 2. -50.
1.76
1.62
_
1.76
SS - Suspended Solids
TP - Total Phosphorus
Flows-Design = 1.5 MCE
ptf:t_ Tot. plios. in RU
115
-------
TABLE 2 - APPENDIX C
MONTHLY AVERAGE VALUES OF ADDITIONAL CHEMICAL
PARAMETERS DURING EXTENDED ALUM RUN
MONTH
APRIL 1971
MAY 1971
JUNE 1971
JULY 1971
AUG. 1971
SEPT. 1971
OCT. 1971
NOV. 1971
DEC. 1971
JAN. 1972
FEB. 1972
MARCH 1972
+•*•+
EOT. Al
mg/1)
RW
1.0
c
/
C
r
.3
£
2
.3
.3
FE
1.1
1.8
1.4
1.6
1.4
2.1
2.6
1.7
1.4
2 2
2.1
JU8
so4
( mg/ 1 )
RW
59
81
85
93
1 109
121
119
102
135
122
118
118
FE
97
169
162
164
181
212
196
189
T54
182
209
221
TOT. ALK.
mg/1 as
CaCO-i )
RW •
195
i
17
25
12
??
27
27
FE
n
13
9
10
10
11
7
12
6
11
13
13
NH3-N*
mg/1)
RW
20
20
20
18
18
20
12
18
7
16
2.0
20
FE
13
12
10
11
10
10
6
1Q
5
10
14
13
N03-N*
(mg/1)
FE
2
3
3
2
3
3
4
3
2
2
1 -
1
RW - Raw Wastewater
FE - Final Effluent
N* - mg/1 as N
116
-------
TABLE 3 - APPENDIX C
MONTHLY AVERAGE VALUES OF PLANT OPERATING PARAMETERS
DURING EXTENDED ALUM RUN
MONTH
APRIL 1971
MAY 1971
JUNE 1971
JULY 1971
AUG. 1971
SEPT. 1971
OCT. 1971
NOV. 1971
DEC. 1971
JAN. 1972
FEB. 1972
MARCH 1972
RECYCLE
FLOW*
(MGD)
.46
2.1
.06
.06
.06
.71
.57
.85
.45
.86
.82
.73
PH
RW
7.1
7.3
7.3
7.2
7.4
7.2
7.2
7.4
7.3
7.3
7.3
7.2
FE
7.0
7.0
7.0
7.0
6.9
6.5
6.6
7.0
7.2
6.9
6.7
6.7
D.O.
(mg/1)
RW
1.2
.5
.2
. I
.1
.1
.3
.3
.9
.5
.3
.4
FE
6.6
6.7
6.2
6.6
6.3
6.5
6.9
7.4
7.7
7.6
7.7
6.7
WATER
TEMP.
(°F)
RW
73
76
81
84
84
83
78
74
63
61
65
67
FE
70
74
84
82
83
81
74
69
60
57
62
66
C12
FEED**
LB/DAY)
0
0
0
100
100
110
98
95
100
100
97
100
TOT. C12
RESIDUAL
(mg/1)
FE
0.0
0.0
0.0
1.1
.9
1.5
1.7
1.6
1.4
1.6
1.2
1.3
FECAL
OL1 FORMS
No. /ml)
RW
30t
99t
113t
116t
306t
108t
18t
93t
35t
80t
85t
62t
FE
0
770
925
0
0
0
0
0
0
0
0
0
RW - Raw Wastewater
FE - Final Effluent
* - Settled sludge recirculation flow from final
clarifier hopper to raw wastewater wet well.
** - Chlorine fed to trickling filter effluent
just prior to entering final clarifier.
117
*U.S. GOVERNMENT PRINTING OFFICE: 1973 546-310/75 1-3
-------
SELECTED WATER
RESOURCES ABSTRACTS
INPUT TRANSACTION FORM
w
ENHANCING TRICKLING FILTER PLANT
PERFORMANCE BY CHEMICAL PRECIPITATION
Derrington, R. E., Stevens, D. H. and Laughlin, J. E.
Richardson, Texas, City of
5. Report
8. Performing
o.
S800685
11010 EGL
13. Type cf Repot and
Period Covered
12. Sponsoring
'Environmental Protection Agency
Environmental Protection Agency, Report No. EEA-670/2-73-060,
August 1973.
Two years of plant scale studies indicated metal addition was an effective effluent
polishing technique at this conventional waste-water treatment plant. Effluent phos-
phorus (P), five-day BOD and suspended solids were reduced to 0. 5, 5, and 7 mg/1
respectively. Aluminum sulfate was more effective than ferric chloride. Alum
addition ahead of the final clarifier proved the best arrangement. An optimum mole
ratio (metal/phosphorus) of 1. 6 developed; this ratio shows moles of aluminum, fed
per mole of incoming total phosphorus. Chemical costs, of which one-third was for
transportation, were 5 cents per 1,000 gallons of flow treated, or 36 cents per pound
of phosphorus removed when in the 96 percent reduction range. Chemical addition
doubled the volume of digested sludge but dewate ring on sand beds took half as long
as previous conventional operations. During this demonstration the treatment system
received some 1. 6 mgd of typical domestic discharge, essentially its design loading.
Hydraulic loading on clarifiers was minimized by drastic reduction of recirculation
lows.
. Descriptors * Biological Treatment, * Chemical Precipitation, #Oxygen Demand,
* Phosphorus, * Suspended Solids, * Tertiary Treatment, * Trickling Filters,
1= Wastewater Treatment, Activated Carbon, Biochemical Oxygen Demand, Chemical
Dxygen Demand, Coagulation, Colloids, Data Processing, Dispersion, Diurnal
Distribution, Domestic Wastes, Feeding Rates, Filtration, Flocculation, Laboratory
Tests, Nutrient Removal, Sedimentation, Sewage Treatment, Sludge Disposal, Texas,
Tracers, Treatment Facilities.
17b. Identifiers
* Richardson (Texas)
05D
19. Sfearit? 'Class,
(Report)
20. Security Class.
------- |