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.

-------