Environmental Protection Technology Series
PHOSPHATE REMOVAL
IN AN ACTIVATED
SLUDGE FACILITY
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
EPA-670/2-74-061
August 1974
PHOSPHATE REMOVAL IN AN ACTIVATED SLUDGE FACILITY
By
Richard E. Finger
George J. Mason
Dale A. Carlson
Gary L. Minton
Municipality of Metropolitan Seattle
Seattle, Washington 98119
Project No. WPRD 247-01
Program Element No. 1BB043
Project Officer
John E. Osborn
U.S. Environmental Protection Agency
Region X
Seattle, Washington 98101
For
Advanced Waste Treatment Research Laboratory
Cincinnati, Ohio 45268
NATIONAL ENVIRONMENTAL RESEARCH CENTER
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
REVIEW NOTICE
The National Environmental Research Center—
Cincinnati has reviewed this report and approved its
publication. Approval does not signify that the
contents necessarily reflect the views and policies
of the U.S. Environmental Protection Agency, nor
does mention of trade names or commercial products
constitute endorsement or recommendation for use.
11
-------
FOREWORD
Man and his environment must be protected from the
adverse effects of pesticides, radiation, noise and other
forms of pollution, and the unwise management of solid
waste. Efforts to protect the environment require a
focus that recognizes the interplay between the components
of our physical environment—air, water, and land. The
National Environmental Research Centers provide this
multidisciplinary focus through programs engaged in
• studies on the effects of environmental
contaminant's on man and the biosphere, and
• a search for ways to prevent contamination
and to recycle valuable resources.
This report illustrates the combination of biological
and engineering controls that are necessary to accomplish
efficient phosphorus removal from a municipal wastewater
effluent.
A. W. Breidenbach, Ph.D.
Director
National Environmental
Research Center, Cincinnati
111
-------
ABSTRACT
Biological and chemical means of phosphorus removal were
studied at the Municipality of Metropolitan Seattle's
activated sludge facility in Renton, Washington. The
studies ranged from jar tests to full scale plant manipu-
lation. Only the secondary activated sludge system was
studied.
The results of these studies indicate that the observed
removal of phosphorus, in a soft water.area activated
sludge facility, is primarily biological in nature.
Some increase in removal can be encouraged by judicious
control of organic loading rates, air application rates
and excess sludge wasting rates. However, even under
optimum conditions, this biological mechanism fails to
consistantly reduce the phosphorus concentration to the
desired level of less than 1.0 mg/1.
Both ferric chloride and alum were found to be effec-
tive in removing phosphorus when added to the activated
sludge culture. Both can be controlled by automatic
means to give the desired concentration of phosphorus in
an activated sludge effluent. Addition of either chemical
produces more sludge mass. The density of the sludge
produced with alum will be greater, thus producing less
volume to handle. Alum is more effective than ferric
chloride in causing an increased density of the final
sludge.
Initial capital costs for either chemical are the same
except for the requirement of a larger volume storage
tank for the alum. Chemical costs will vary depending
on location of the supplier in relation to the plant.
Alum costs at the Renton plant are $16.80 per million
gallons and ferric chloride costs are $34.20 per million
gallons not including delivery. These figures represent
a cost of $2.88 and $5.50 per mg/1 phosphorus removed
per million gallons for alum and ferric chloride, re-
specively-
This report was submitted in fulfillment of Project
Number WPRD 247-01 and Demonstration Grant Number 17010 EDA
by the Municipality of Metropolitan Seattle under the partial
sponsorship of the Environmental Protection Agency.
Work was completed as of June 30, 1972.
iv
-------
CONTENTS
Section Page
Abstract iv
List of Figures vi
List of Tables vii
Acknowledgements viii
I Conclusions 1
II Recommendations 3
III Introduction 5
IV Plant Modifications for the Project 10
V General Plan 11
VI Materials and Methods 14
VII Biological Removal Studies 26
VIII Chemical Removal 43
IX Discussion 72
X References 81
-------
FIGURES
No. Page
1 Renton Treatment Plant Existing Facilities 9
2 Dye Study Plot - Detention Time 15
3 Total Phosphorus Manifold - Pump I 17
4 Total Phosphorus Manifold - Pump II 18
5 New Total Phosphorus Manifold - Pump I 19
6 New Total Phosphorus Manifold - Pump II 20
7 Chemical-Feed Control System 23
8 Chemical-Feed Control System 24
9 Chemical-Feed Control System 25
10 Full-Scale Biological Test Sampling Points 29
11 Time Relation Plot 33
12 Time Relation Plot 34
13 Linear Correlation Scatter Plot 35
14 Typical Potentiometric Titration Versus
0.10 N HC1 and End Point Determination 46
15 Diurnal Variation in Bicarbonate Assay 48
16 Diurnal Variation in Bicarbonate Assay -
24-hour Test 49
17 Sampling and Chemical Addition Points 52
18 Biological Floe Before Alum Feed 60
19 Biological Floe During Alum Feed 61
20 Biological Floe at the End of the Alum Feed 62
21 Biological Floe Before Ferric Chloride Feed 69
22 Biological Floe During Ferric Chloride Feed 70
23 Biological Floe at the End of the Ferric Chloride
Feed 71
VI
-------
TABLES
No.
1 Plant Design 8
2 Activity Schedule 12
3 Total Phosphate Analyzer Reagents 21
4 Variables Used in the Full-Scale Biological Tests 28
5 Metro Renton Treatment Plant Full-Scale Plant
Phosphorus Removal Weekly Report 31
6 Metro Renton Treatment Plant Full-Scale Plant
Phosphorus Removal Weekly Report 32
7 Average Results of Full-Scale Feed and Mixed
Liquor Solids Concentration Tests 37
8 Mean Percent Total Phosphorus Removed on
Wednesdays and Thursdays 39
9 Cyclic Variability of Selected Parameters and
Statistical Significance of this Variability 42
10 Ion Profiles 45
11 Analyses and Sampling Frequency—Full-Scale
Chemical Feed Tests 53
_ i
12 Average Value of G-rms Velocity Gradient(Sec )
and Mixing Times 54
13 Daily Averages and Totals—Full-Scale Alum
Feed Test 56
14 Alum Feed Test Solids Data 57
15 Daily Averages and Totals—Full-Scale Iron
Feed Test 65
16 Iron Feed Test - Solids Data 66
17 Summary of Full-Scale Chemical Feed Test Results 76
VI1
-------
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the support of Charles
V. Gibbs, Executive Director of the Municipality of Metro-
politan Seattle, Charles J. Henry, Jr., Director of
Operations and Gary W. Isaac, Renton Division Superinten-
dent.
Beyond-the-call-of-duty efforts of Barbara Strutynski and
other laboratory personnel in producing laboratory analyses
and the all-out efforts of Virginia Hunter, our typist,
are greatly appreciated.
The guidance and statistical analyses produced by Edward
Flagel of R. W. Beck and Associates made much of this
project possible.
The efforts and guidance of Dr. Peter G. Howe, Seattle
Community College, is greatly appreciated particularly
in reference to the use and handling of chemicals.
The cooperation of Allied Chemical Company and the
Van Waters and Rogers Company in supplying alum and
ferric chloride is sincerely appreciated.
To the many graduate students and other personnel of the
University of Washington who produced more than was
ever expected, we wish to express our sincere thanks.
Vlll
-------
SECTION I
CONCLUSIONS
1. Up to 50% phosphorus removal can be achieved by the bio-
logical process at the Renton treatment plant which does not
recycle anaerobic digester supernate to the plant influent.
2. Operating conditions that promote rapid biological
growth can increase the rate of phosphorus uptake signifi-
cantly.
3. While bench scale and jar tests were not adequate tools
in this study for predicting full scale behavior, particu-
larly relating to biological phosphorus removal, they can
be useful tools for predicting chemical effects on a full
scale facility.
4. Ferric chloride and aluminum sulfate react instantan-
eously with phosphorus in soft water areas. Both chemicals
reduce the pH of the mixed liquor which could be a problem
depending on the alkalinity and phosphorus content of the
water being treated.
5. Sodium aluminate causes an increase in pH in soft water
areas thus inhibiting the phosphorus complexing reaction.
6. Precipitation of phosphorus with aluminum is pH depen-
dent. The ratio of aluminum to phosphorus required to
achieve a fixed removal of phosphorus is at a minimum at
pH 6.0 and increases as the pH rises to 8.0. A similar
situation exists with iron at different pH's.
7. Alum effectively removes phosphorus when added to an
activated sludge system and causes a definite increase in
the density of the activated sludge. Alum is easy to handle
and presents a minimum of problems with respect to cor-
rosion and safety.
8. Ferric chloride is effective in phosphorus removal but
does not cause a significant reduction in the sludge den-
sity. Ferric chloride is difficult to handle due to its
extreme corrosiveness.
9- In the case of both ferric chloride and alum, the
estimated quantity of chemical required on the basis of
jar tests was significantly greater than was actually re-
quired. This resulted from a failure to account for the
reduction in mixed liquor phosphorus levels resulting from
-------
the recycle of return sludge having a very low soluble
phosphorus concentration.
10. There was no change in the required chemical to phos-
phorus ratios as the test proceeded. The recycle of return
-sludge having a significant ,level of metal salts, therefore,
did -not contribute to an enhanced phosphorus removal.
U. Alum costs were $16.80 per million gallons treated,
while ferric chloride costs were $34.20 per million gallons
treated at the Renton plant, not including shipping costs.
These represent costs of $2.88 and $5.50/mg/l phosphorus
removed/MG for alum and ferric chloride, respectively.
12. Capital costs for a complete storage and control
system for either chemical are approximately $26,000 for
a 24 MGD facility. The major portion of the costs is attri-
butable to analyzer and control equipment. This cost is
largely independent of the plant size.
13. Alum was the best chemical for removal of phosphorus
at the Renton plant on the basis of cost, ease of handling
and effect on settling. Costs are directly related to
removal efficiencies with alum costs .being approximately
half the ferric chloride costs to achieve equivalent phos-
phorus removals.
-------
SECTION II
RECOMMENDATIONS
1. Further research into the nature and control of biologi-
cal phosphorus uptake in full scale activated sludge fa-
cilities should be done. The answers obtained would add
to our basic fund of knowledge and aid in the develop-
ment arid selection of the most economical phosphorus re-
moval procedure.
2. Adequate mixing facilities for chemical addition should
be provided to insure efficient utilization of the added
chemical. Normal aeration basin mixing is not sufficient
to prevent excessive loss of the metal as the metal hyd-
roxide through hydrolysis reactions.
3. Activated sludge systems using chemical addition for phos-
phorus removal and having an alkalinity less than 150
ppm and a phosphorus concentration greater than 10 ppm
should have facilities for pll control. Otherwise, loss
of the biological culture can be expected.
4. A second phosphorus analyzer equipped with a high level
alarm should be provided at the outlet of the clarifiers
to provide backup control of the chemical addition.
5. A magnetic flow meter or variable speed proportioning
pump should be used to provide accurate measurement and
control of the chemical feed.
6. Total flow and phosphorus concentration at the chemical
addition point must be known to insure that the optimum
chemical to phosphorus ratio is maintained. Addition of
chemicals to the aeration basin will require that the
plant flow and the return sludge flow be summed to obtain
the correct flow for use on the chemical feed control
loop.
7. The chemical feed controller phosphate analyzer should be
equipped with high and low level alarms to protect against
autoanalyzer malfunction.
8. It is important to select a chemical feed point or points
which require a minimum number of controllers while pro-
-------
viding sufficient low energy mixing following the
chemical addition to allow for flocculation.
The use of polyelectrolytes in conjunction with alum or
ferric chloride feed should be investigated. Such a pro-
cedure could effectively reduce the effluent phosphorus
level by trapping colloidal floe which would otherwise be
discharged with effluent.
-------
SECTION III
INTRODUCTION
During the last ten years there has been a new awareness
of environmental problems and a growing concern about
man's effects on his surroundings. The effects of dis-
charging wastewater treatment plant effluents into our
streams and lakes has been one of the prime concerns. The
quantities of phosphorus in these effluents'has led to
major industrial research in an attempt to substitute other
materials for phosphorus in detergents. Many research
projects have been conducted at universities and munici-
palities to discover an economical method for removing
phosphorus in wastewater treatment plants. Most of the
studies have described methods of operation that are
applicable to only one particular plant with inadequate
information about the nature of the reaction to make the
proposed method transferable to other treatment plants.
The removal of phosphorus at the Rilling Plant in
Amarillo, Texas (1) and at the Hiperion Plant(2) can be
explained chemically. Both of these studies were con-
ducted in areas having water with a very high calcium and
magnesium content.
Since the Municipality of Metropolitan Seattle's (Metro)
Renton Treatment Plant is located in a soft water area,
it was felt that a demonstration of an efficient, econo-
mical method for removing phosphorus at this location
could provide useful information for helping other plants
in soft water areas meet the increasingtly stringent stan-
dards for phosphorus removal.
SCOPE OF THE PROJECT
Phosphorus removal at Renton was investigated in the
activated sludge process with no work performed in the
primary treatment area. Both biological removal and re-
moval through the addition of chemicals to the activated
sludge culture were studied.
The biological studies included laboratory experiments
prior to full scale application. Computerized statistical
techniques were applied to the data generated from both
the laboratory and full scale experiments. Both controlled
experimental techniques and long term studies were con-
ducted and analyzed during the full scale operation.
-------
Chemical studies were conducted in the laboratory using jar
scale and bench scale tests. Further pilot plant studies
were conducted to obtain information on the long term
effect of chemical addition. A control system for chemi-
cal dosing in proportion to the quantity of phosphorus
present was developed for the full scale test during which
ferric chloride and aluminum sulfate were used. This
report will be restricted to the discussion of the full
scale studies. Data on the jar, bench and pilot studies
is available on a loan basis from the Municipality of
Metropolitan Seattle.
PLANT DESCRIPTION
The Metro Renton Treatment Plant is an activated sludge
plant with aerated grit removal, primary settling and an
average summer daily flow of 22 MGD. The design flow dur-
ing this study was 24 MGD and has since been expanded to
36 MGD. The ultimate design capacity is 144 MGD. The
plant is located in a softwater area and receives pre-
dominantly domestic sewage. The only major industrial
complex discharging to the plant is a dairy product facil-
ity which provides approximately one third of the total
plant BOD loading.
The Renton Treatment Plant is designed to provide a high
degree of flexibility in the aeration system. Aerator
feed can be varied from plug through step to contact stabil-
ization while aeration can be varied from conventional to
tapered with a minimum of difficulty. Each aeration tank
has a total of twelve dissolved oxygen probes and is
divided into two, two-pass systems with respect to aera-
tion control. With this arrangement, any one of the first
six probes can be used for dissolved oxygen control in the
first two passes, and any one of the second six probes
can be used for control in the second two passes. The
aeration rate can also be controlled manually. The
dissolved oxygen probes provide a continuous dissolved
oxygen profile throughout the tank. The total volume of
each of the two four-pass aeration tanks is four million
gallons.
Mixed liquor is carried from the aeration tanks to the
final clarifiers through an aerated mixed liquor channel
having a volume of 0.58 million gallons. The return
activated sludge from the clarifiers can be varied from
6 MGD to 60 MGD and can be controlled either manually or
automatically as a percent of the influent flow.
-------
Effluent from the secondary clarifiers is collected and
chlorinated after which it is discharged to a chlorine con-
tact channel which provides 46 minutes detention time at
the design flow. The chlorine feed rate is controlled by a
continuous chlorine analyzer and the total chlorine residual
is normally held at 1.5 mg/1 at the control point. The eff-
luent from the chlorine contact channel is discharged to
the Green River.
Excess activated sludge can be wasted from the system as
either mixed liquor or return activated sludge. The waste
activated sludge is combined with the primary raw sludge
and scum and the combined waste sludges are pumped to the
Metro west side collection system. These combined wastes
then flow to the West Point plant where they are resettled
and anaerobically digested. This procedure eliminates the
necessity of maintaining digesters at the Renton plant and
thus eliminates the discharge of phosphorus rich digester
supernate to the plant influent.
The design data sheet for the plant at the time of the study
is presented in Table 1. A diagram of the Renton Treatment
Plant during the period of the study is presented in Figure
1.
-------
Table 1 - Design Data for the Renton Plant
CO
MtSIGW LG4IHMGS SCREENING EQUIPMENT (CONTJ PRIMARY SEDIMENTATION EQUIPMENT (CONTJ
TAWlurr STWia C/OD-CU.WDD «0- .D 0*tNl« IcrwUN MRS, INCWES-C*. Ml- L* Xt-IHIHO*) SOtMftCINUTCATtl NO Ptl r AW j
INDUSTRIAL ft.STts CMCJD-Ot.WWCTAJtt/a »»l[ft DEPTH PB.* *.» R£J-h J.W- LU lll.JJ - 3.«| StUT TROUCHTHS PtRTWt ' 1
NUv |6,OUS1BI*1 ARIAS SOO It 71 VUOCirv THRU SCREEN. ft/HC. -WStt. MN. 0.1 CM) DIAMETER ORIFICF.S W /TROUGH M
uoi inDusiBjAi ARUS an 11 n UNITS iNOfto*noH 7 u> uuemof MIS Pto TROUGH FEIHIL ^^ 3- a
GT±?(fc1l10YTVl*riW-S/Aa8'CU-'WI w
uurmSusin » R»fi-S«i«i5ifS «.o- Ul B: i.™ sussouK«to..ra«iisiMT.i««i«ciwY «.«• I i». ei
"wiil'SiWii^S.r"1'5'*™" 0 u,« '"VMT "••"•"•'•• ""'"• «D-U. «|.|lg «"TION EOUIPMENT - STANDAHD LOADING RATE
u», ,.,«,„« uus n:11;5 VSSV" „.».„:, S^M "K«, , ,u
«ioCK(Micw art cut CBXUO n(MPl *?.D- IM oi.i - ID.B I>IAO-IM.-OI> IMCMI INUTCATIS, NO. /TAW «
?aiWiVPc-cu'.*/iaBCii M.
KfLUENT SEWER
DlWHlR IMCXS-CN.
CAPt^irt' «CD- UCDCU M!£L*»
INLET GATES
Sia^FOT ».
INFLUENT CONTROL GAT£S
SCSEEN1NG EQUIPMENT
•">•«»
4*1 THC*NHi IMCMlt-CM.
CHUML wiDTh, FQT H.
TAWS
WJi-HR |
HIP* FUT-M. a.Ti- LI
3.3- 1. UNCTH FTE1-M |» . )|
S3 1 • O FJFlCTlVt OEPTW FUT M. li - i.t
s j.). L "U H't)S*uUCC*pn*w-MCC)-100BCU.»JD IB- 4H
C*P '8l.C*lB CU fTrVIlL CO MMIM Km E
»: m '«: ?S A,« 10 PRC^AI.ON.«HAVO-F.CU.M't«.'ir> t,i.[ »v o H F. MOUtS IS
. „ IAIK.U* «»'[ *v D * f ,
uJ. ,, * '.*..^S(,f- CU WDAS.-SOP. IOB- M
*' '"D I.^vir^cw''^*1'"8" " ii °H
D!SCh»BG[ PfiESSUW, PSI *CiSO,M
UJ »|0 SUPCUED. AV.D * F CU.fllC CU.M/CU.M.
»|B SJPPUESPE«FT.(» TA« CU.n.«IN-CU.***y(t
*|B SUPPLIED CU Fl.iLI-CU M/xC 0* 100 RCMOvfO
B-77T AERATION EQUIPMENT - HIGH LOADING
TAWS
NUMlE*
(4) U , 60-IN (Ml HCMI INtfTCAnS NO./TANK
;;ia\:sjffi «....
UWTH EACH PASS. FlfT-M.
™, *vEo*ct *ATIB DIPT*. FEIT-M.
"*' «Tt'»{* '<*•( AV 0 W F nOUW
VAI H'D9*UUC CAP iTAN< tfZ\> WOO CU.M'OAT
BOD LOIOINC IBS APPVlED lOD'IQOD CU FT1DA*-
„ OrtD- CU.M«C
M - «J»
1.0- J.S
».«- o.n
iu)- n mi- ».)i
RATE
<3e- t.n
mi nn
IHI.I- 170i
IU)
I1?.DDD-»4.I)
.i"«r i*'*
0? 1- l.li
Of 100 )*» • tli
SECONDARY SEDIMENTATION EQUIPMENT
'***.«« ' "«
DIAMfTIB FEET M
iV'BkCt (mCTlVEWini DCP CAP.. pji-KGfSQ.CM. UD- M
MASnSLUDUFOKI MAIM
NUMBER *
SIZE OlMMfTER. INCHCl-CM. O - JftJ
UNCtH. EiCH ffcCt MAIN. FlfT-M. ll.JBO- Si
MASIE RAS AUTOMATIC CONTKH VAL«
NUMICR j a
SIZE. INCXS-CM. a-mj
WAII *i«o UOWM C»T«H vuvc
NUMIER 1 *
sm. iNoas-c». Et-«.t
EFFLUENT WEIRS
NLWICR 1 •
*IDTM. FEn-M.IINIT|ALWtHCIPOlfni TYPC 7- 1.1 iff- fc»
DlSCH*RG£HFjtO FEfT-M.
AVERAGE OBI WATMEft FLOW 1. B 0.4 H.J- III
MINIMUM FLOW dtl- oil 0^3 • fcl
CHEMICAL TREATMENT EQUIPMENT
CH LOU I NATION
(VAPORAtOH CHtORlHATOn
KL*-B(fl FOR POSTCHlOfll NATION
CAPACITY RANGE PER UNIT, LIS-KC CL^O
RESIDUAL ccMTROiias
RtSIHJALMOhlTOti
NUMllI
CONTACT CHiNNfU
NU¥I(R
*lt>TH FEET-«.
II^CTH C*CPI CHANNEL FE£T M
«»T[B. MPTH, »V. O.W.f . FUT #.
M1O.1IO. TIME. AV. D.W.F.. MINUTES
DtCHLDRINATIOM
EVAf-OftAIOR- SUv/ONAtO«|
WLJMBCB
CHVOBINC PESIOUAl ANAiYjlftS
NUVUR
TOTAL TREATMENT
ASSUMED BOO REDUCTION. PERCENT
IDO IhtFfiulNI. IDODUSJDAY ICOKCJCAY
ASSUMLD SUS SOLIDS REDUCTION, P[R CtNT
SUS SOll OS IhlfTLUEM |QDO HS'fr 1000 «C/0
BOO 'OtttSl PI. SIP 1000 LBVD- 1930
-------
5-~~" OUTFALL
>6" EFFLUENT
iEWER
INFLUENT CONTROL
STRUCTURE
DIVISION
CHANNEL
AERATION
TANKS
r-PREAERATION
\TANKS
PUMP BLDG.
DISTRIBUTION
CHANNEL
PRIMARY
SEDIMENTATION
TANKS
CHLORINATOR
SULFONATOR
SECONDARY
TANKS
CHLORINE
CONTACT
CHANNEL
SEDIMENTATION
PRECHLORINATION-
MANHOLE
FIGURE I
RENTON TREATMENT PLANT EXISTING FACILITIES
-------
SECTION IV
PLANT MODIFICATIONS FOR PROJECT
The chlorine contact channel at the Renton plant has a
design detention time of 45.9 minutes. The addition of
chlorine at the head of the channel promotes flocculation
and settling of any solids that may pass through the
secondary clarification process. Low velocities in this
channel allows most of these solids to settle in the chan-
nel and, unless they are removed, anaerobic decomposition
will occur with release of previously entrapped phosphorus
to the effluent. A flushing gate was installed near the
discharge end of the channel to allow frequent flushing
of these solids to the influent sewer. This flushing
prevents phsophorus release to the effluent.
10
-------
SECTION V
GENERAL PLAN
The project plan called for the optimization of biological
phosphorus removal followed by the use of chemical addition
to the secondary treatment process to increase phosphorus
removal to the desired level. A final effluent phosphorus
concentration of 1.0 mg/1 or less was the desired goal.
Jar and bench scale tests were conducted for both the bio-
logical and chemical removal studies and the results of
these tests were applied to the full-scale tests. The
results of these jar and bench scale tests will not be
discussed except as they applied to the full-scale testing
activities. These data are available from Metro on a
loan basis. Table 2 shows the sequence of activity for
the full-scale plant tests.
BIOLOGICAL FULL SCALE STUDY
After obtaining background data on normal plant oper-
ating parameters, the plant was operated in such a fashion
that the results of the jar tests and anomalies in the
background data could be studied. In all cases an attempt
was made to integrate a controlled experimental approach
with real operating conditions. Particular attention
was given to operator-controlled variables. The control-
lable variables examined were return activated sludge rate,
mixed liquor solids concentration, waste activated sludge
rate, aeration rate and dissolved oxygen concentration.
The system was allowed to stabilize at a given set of
operating conditions for at least a week before intensive
sample collection and analysis was undertaken. The non-
controllable parameters which were monitored were effluent
biochemical oxygen demand BOD, primary effluent phosphorus
concentration, return activated sludge phosphorus concentra-
tion, effluent collection chamber phosphorus concentration,
final effluent BOD, final effluent phosphorus residual,
settling, primary effluent calcium concentration, primary
effluent magnesium concentration, and total plant flow.
In each case the data generated was computerized and s^a-
tistical techniques were used to determine any significant
relationships to phosphorus removal.
11
-------
Table 2. ACTIVITY SCHEDULE
ACTIVITY
AGENCY
October 1969 Plant detention time studies begun
November 1969 Plant switched from 2-pass to
4-pass aeration due to high flows
February 1970 Plant detention time study com-
pleted
April 1970 Background biological data collec-
tion
May 1970 Plant background data collection
completed
June 1970 Plant scale biological removal
tests begun
August 1970 Attempted full-scale duplication
of Hyperion work(2)
January 1971 Preparation for full scale chemi-
cal tests
April 1971 Construction of chemical feed
system
April 1971 Full scale biological tests
completed
June 1971 Full scale alum feed test
completed
August 1971 Full scale ferric chloride feed
test completed
August 1971 Statistical analyses of biolog-
ical removal completed
March 1972 Analysis of full scale chemical
data completed
April 1972 Final report written
Municipality of Metropolitan Seattle
Municipality of Metropolitan Seattle
Municipality of Metropolitan Seattle
Municipality of Metropolitan Seattle
Municipality of Metropolitan Seattle
Municipality of Metropolitan Seattle
Municipality of Metropolitan Seattle
Municipality of Metropolitan Seattle
Municipality of Metropolitan Seattle
Municipality of Metropolitan Seattle
Municipality of Metropolitan Seattle
Municipality of Metropolitan Seattle
R. W. Beck
Municipality of Metropolitan Seattle
Municipality of Metropolitan Seattle
University of Washington
12
-------
FULL SCALE CHEMICAL STUDIES
Aluminum sulfate and ferric chloride were selected for appli-
cation during the full scale plant study. The chemical
addition was controlled in direct proportion to the
mixed liquor supernate phosphorus concentration and the
wastewater flow. Intensive sampling and analysis was
conducted during these test periods to ascertain the
effectiveness of the chemical on phosphorus removal and
the effect of the chemical on the biota of the activated
sludge culture.
13
-------
SECTION VI
MATERIALS AND METHODS
Detention Time Studies
Dye studies, using Rhodamine WT dye because of its low
adsorbtivity on the activated sludge, were conducted to
accurately determine the detention times between various
points in the plant. The plant flow was held constant dur-
ing the tests and the fluorescence of the supernate from
centrifuged samples was measured with a Turner fluorometer.
A plot of fluroescence versus time from addition was pre-
pared such as is shown if Figure 2 and the time of peak
fluorescence was designated as the detention time for the
given flow.
The results of these studies were used to trace a plug of
sewage through the plant to. allow proper correlation of the
full-scale biological data. A multiple iteration calcula-
tion was perfomed with a computer to order the experimental
data with respect to flow through the plant.
Analytical Procedures
Sample analyses during the project involved the determina-
tion of a number of different parameters. Alkalinity,
biochemical oxygen demand and chemical oxygen demand
analyses were run according to the 12th edition of Standard
Methods (3). The chemical oxygen demand was run with a
fifteen minute reflux as this method gave a complete
reaction with the Renton sewage.
Total and volatile suspended solids were run by a modifica-
tion of the procedure found in the 12th edition of Standard
Methods (3). The modification involved the use of a glass
fiber filter pad in place of the asbestos mat. All other
conditions were as specified in Standard Methods (3).
Calcium and magnesium data for the routine daily analyses
were analyzed according to the 12th edition of Standard
Methods (3) by the eithylenediaminetetracetic acid titration
using EriochroiruS Black T and Murexide as indicator for total
hardness and calcium hardness, respectively. Potassium
cyanide was used as an inhibitor for the total hardness
test. Calcium and magnesium data for the full scale plant
tests were obtained by means of an atomic absorption spec-
trophotometer as were all iron data.
14
-------
X
o
o
-------
Sulfate analyses were run by the turbidimetric procedure
while chloride was analyzed by the mercuric nitrate titra-
tion procedure, both of which are presented in Standard
Methods (3). The aluminum analyses were performed by the
Eriochrome Cynaine R procedure (4).
Ortho-phosphate data was determined by the single reagent
method as presented in the 1970 edition of "Methods for
Chemical Analysis of Water and Wastes" published by the EPA.
Total (organic & inorganic) phosphorus data were analyzed
with a Technicon Autoanalyzer by a high temperature diges-
tion method. The flow diagrams and reagent compositions
for this analysis are contained in Figures 3 and 4. A new
autoanalyzer system was obtained during the later part of
the study and all total phosphorus data from the plant
scale tests were obtained using the new system. The flow
diagrams and reagent compositions for this system are
presented in Figures 5 and 6 and Table 3.
Chemical Feed Controller
As the first step in the preparation for the full-scale
chemical feed tests, it was necessary to design and con-
struct a chemical feed control system capable of delivering
the various test chemicals at a rate proportional to the
quantity of phosphorus present at the addition point. In
order to accomplish this,, it was necessary to measure the
flow and the phosphorus concentration at the feed point.
The product of these two parameters was used to generate
the control signal.
The ortho-phosphate concentrations in the mixed liquor
supernate was measured with a Technicon autoanalyzer system
equipped with a continuous filter for removal of the mixed
liquor suspended solids. Since the electrical signal gen-
erated by the autoanalyzer could not be easily altered
for use in the control system, the recorder was modified
by attaching a logarithmic cam to the drive shaft and a
pneumatic position indicator was used to generate a linear
concentration signal.
The chemical addition was to take place in the aeration
tank which required that the flow for use in the control
system be the sum of the plant flow plus the return sludge
flow. The return sludge flow was normally controlled as a
percentage of the plant flow and since a reliable totalizer
with which to combine these two flows was not available, it
was decided to use the total plant flow without the return
sludge flow. The error introduced by this action was
corrected for the most part by adjustment of the controls
16
-------
HELICAL
DIGESTOR
320"C
DILUTES AND NEUTRALIZES,
SAMPLE IN DIGESTOR
TO MIXING
CHAMBER
-X- NOTE'
THE PHYSICAL SET-UP OF THE
DIGESTER IS THE SAME AS FOR
THE KJELDAHL NITROGEN
ANALYSIS.
DIGESTION MIXTURE'
DISSOLVE l.5g V205( VANADIUM PENTOXIDE
IN 20Oml H20. ADD 100ml OF 70% HCI04
AFTER AT LEAST 10 MINUTES AND DILUTE
2 LITERS WITH CONCENTRATED H2S04.
PCI
0,110 ACIOFLEX
0.110 ACIDFLEX/ MIXTURE
>
DIGESTION
/
0.045
AIR
0.081
SAMPLE
0.110
NaOH
0.110
0.110
DISTILLED WATER
TO SAMPLER WASH RECEPTICAL
FIGURE 3
TOTAL PHOSPHATE MANIFOLD - PUMP I
-------
0.40
REDUCING REAGENT
^v
/
^ — nnim
F MI
\
0.40
X 0.110
X 0.45
*iwrc rnn c — ^>
SAMPLE FROM
1% AMMONIUM
AIR
MIXING CHAMBER
MOLYBDATE
oo
WITH CENTER INLETS
TO WASTE
0.81
FROM CALORIMETER
|
I
CALORIMETER
5 cm. FLOW CELL
660mm FILTER
REDUCING REAGENT:
0.2g SnCI2 PLUS 2.0g HYDRAZINE
PLUS 28ml CONCENTRATED HjSO,;
DILUTED TO 1 LITER WITH DISTILLE
WATER.
FIGURE 4
TOTAL PHOSPHATE MANIFOLD - PUMP 2
-------
DILUTES AND NEUTRALIZES
TO MIXING
CHAMBER
ACIDFLEX
ACIDFLEX
>
DIGESTION
MIXTURE
0.80
1.60
3.90
3.90
AIR
SAMPLE
No OH
2.00
2.00
DISTILLED WATER
DISTILLED WATER
TO SAMPLER WASH RECEPTICAL
FIGURE 5
NEW TOTAL PHOSPHATE MANIFOLD - PUMP I
-------
0.32 AIR
to
o
<
660 mu
FLOW CELL
660mu
WASTE
WASTE
1.00
'-00
SAMPLE
2.00 WATER
0.3-2 AIR
1.60 SAMPLE
3.90 MOLYBDATE
0.80 HYDRAZINE
FIGURE 6
NEW TOTAL PHOSPHATE MANIFOLD - PUMP 2
-------
Table 3. TOTAL PHOSPHATE ANALYZER REAGENTS
A. Digestion Mixture per liter
0.75 g V205
50 ml HC104
100 ml H20
Weigh out V2O5 into beaker (up to 4 liters)
Add water and allow to stand for 10 minutes.
Add HC104 than carefully add H2S04.
When the solution is cool (6 hours) bring to the de
sired volume with sulfuric acid. For analysis dilute
two parts digestion mixture with one part distilled
water.
Four liters are needed for each day's run.
B. Ammonium Molybdate (1%) (in 35 g) (3500 ml)
Dissolve 10 g of ammonium molybdate per liter of dis-
tilled water. Add 1 ml Levor IV wetting agent per liter,
C. Sodium Hydroxide
A stock solution of 40% NaOH is prepared by dissolving
400 g NaOH per liter. The working solution is prepared
by dilating 140 ml of the stock to 1 liter.
D. Reducing Reagent
.2000 g SnCl2
2.0000 g Hydrazine
28 ml H2SO4
Put solid reagent in 1 liter volumetric flask and fill
half way with distilled water. Add H2S04 and dilute to
volume.
E. 1000 ppm Phosphorus standard
(5.6234 g K2HP04 diluted to 1 liter)
21
-------
on the chemical feed control system.
The plant flow signal was combined with the phosphorus
signal with a Foxboro pneumatic multiplier. This product
could be adjusted using a Moore ratio relay to provide a
flexible chemical to phosphorus feed ratio; The ratioed
signal was fed to a Foxboro flow controller which controlled
the position of an air operated diaphragm valve. Chemical
flow rates were measured by a chemically resistant roto-
meter and this signal provided feed back to the flow con-
troller. The chemical was delivered to the aeration tank
by means of a centrifugal pump in the case of the ferric
chloride test and by air pressure in the case of the alum
test.
The entire control system with the exception of the roto-
meter and the diaphragm control valve was mounted inside
a weatherproof cabinet. The mixed liquor sample was pumped
to an overflow pot inside the cabinet for sampling by means
of a centrifugal pump which was mounted in the cabinet. A
schematic diagram and pictures of the chemical feed control
system are shown in Figures 7 to 9.
The unit was regularly calibrated by introducing a series
of standard solutions into the analytical stream after first
setting the base line using distilled water. During oper-
ation of the unit, baseline drift occurred as a result of a
variation in the quantity of colloidal suspended solids
which passed through the automatic filter. Adjustment of
the baseline to correct for this residual turbidity was
accomplished by introducing distilled water in place of the
ammonium molybdate reagent. There was a very slow change
in the residual turbidity with time and, as a consequence,
the baseline was adjusted twice daily. A baseline reading
was obtained by replacing the ammonium molybdate reagent
with distilled water. Following adjustment of the baseline,
the ammonium molybdate was reintroduced to the analytical
stream and automated control was resumed. This procedure
required approximately fifteen minutes.
22
-------
20 PSI AIR
1 ^
PHOSPHATE
RECORDER
WITH '
MOTION TRANS. 0*
SLIDE WIRE SHAF1
20 PSI AIR
SUPPLY
SIGNAL
BOOSTER
SUPPLY •*
3- 15 PS, PNEUMAT.C roMP!«?:fl™CE F,LOW
1
c A
SIGNAL
r ]
FOXBORO MULTIPLIE
DKirilMATir ANAI ClI
^ COMPUTER
1^3-15 PSI
&- u I-QI i
IR RECORDER
MOORE
FISHER
1 ' °
PUM
SEW
F.
10-50
MA
PEO
AGE
4:1
20 PSI AIR
SUPPLY
FOXBORO FLOW
CONTROLLER
RESET a PORP.
20PSI
AIR
SUPPLY]
CHEMICAL
STORAGE
TANK
TECHNICON
PUMP I
SUPERNATE
AMMONIUM
AMINONAPTHOLSULFONIC
TECHNICON
PHOTO-TUBE COLOR
METER WITH
50MM FLOWCELL
ELECTRONIC
SIGNAL
FIGURE 7
CHEMICAL FEED CONTROL
23
-------
Figure 8
Chemical Control Feed System
24
-------
Figure 9
Chemical Control Feed System
25
-------
SECTION VII
BIOLOGICAL REMOVAL STUDIES
Literature Review
A thorough literature search was conducted prior to the
start of the biological removal studies to provide a firm
theoretical base for these studies. This work provided
information which was useful for determining what para-
meters should be investigated and measured as well as help-
ing to eliminate needless duplication of effort. A complete
literature review was written and is available on a tem-
porary basis from the Municipality of Metropolitan Seattle*
There are two general schools of thought with respect to
biological or natural phosphorus removal in excess to
physiological requirements of the biological system (5-45).
Both recognize that biosynthesis during the treatment pro-
cess will result in the removal of some phosphorus which
is incorporated into the biomass. The quantity of phos-
phorus removed will be a function of the growth rate and
should represent approximately 1.5% of the new cell mass.
This will normally account for an uptake of phosphorus in
the range of 1-3 mg/1.
Additional phosphorus removal in excess 6<£ this natural
uptake has been reported at several treatment plants through-
out the country. (1, 2, 8-16) . This uptake has been attri->
buted variously to either biological uptake in excess of
physiological requirements, the so called "luxury uptake",
and to chemical precipitation of the phosphorus with one
or more naturally occurring cations in the sewage followed
by removal through adsorption and entrapment in the aera-
tion and secondary clarification units.
Experimental data have been collected which support both
of these hypotheses. In many cases, these data can be used
in support of either hypothesis with the subsequent result
that no clearcut conclusion has been arrived at. It seems
likely that either, both or neither mechanism may be oper-
able at any given plant. As a consequence, individual
evaluation of the conditions at each treatment facility is
required to ascess the most promising approach for maxi-
mizing the "natural" biological removal of phosphorus.
Biological Full Scale Tests
Full scale plant testing based on the results of bench and
jar scale tests, which are available from Metro upon request,
26
-------
were conducted in an attempt to maximize the naturally
occurring phosphorus removal in the activated sludge system.
Absolute control of most variables was difficult to achieve
in the operating facility although the degree of variabil-
ity for parameters such as dissolved oxygen, suspended
solids and detention time was minimized through operational
controls at all times that tests were being run. This
required special operating procedures regarding aeration
rates, return activated sludge rates and wasting rates.
Each of these variables were strictly controlled through
directives from the laboratory.
It was decided early in the project that a complete mathe-
matical analysis of all pertinent full-scale data would be
necessary if any valid conclusions were to be drawn con-
cerning biological removal of phosphorus. Statisticians
from R. W. Beck and Associates, and the University of
Washington helped develop the sampling and analyses tech-
niques to insure that the raw data would be consistent and
furnish valid information for statistical analysis. The
variables selected for sampling and analysis are listed on
Table 4.
A period of one week of operation under the selected condi-
tions was allowed before each sampling period. This was
done to develop as stable a sludge condition as possible be-
fore analysis was begun. The points at which samples were
taken are shown in Figure 10.
Grab samples were collected at 2-hour intervals during the
seven day studies, and brought to the laboratory for immed-
iate analysis, according to the techniques described in the
Materials and Methods section of this report.
Full scale plant operational reports using data obtained
from routine sampling at the treatment plant were produced
using a computer. These were two-page reports including the
basic data and several calculated parameters chosen to out-
line the biological treatment process. Linear correlation
and scatter plots were used as the first step in evaluat-
ing these data. Time relation diagrams, consisting of plots
of each variable against time, were also used to examine
the variability of each parameter. Cross correlation tables
containing simple linear correlation coefficients for every
possible combination of pairs of variables were developed
to provide a concise way of reviewing all the correlation
coefficients.
27
-------
Table 4. VARIABLES USED IN THE FULL-SCALE
BIOLOGICAL TESTS
1. Primary Effluent Biochemical Oxygen Demand(BOD) - mg/1
2. Primary Effluent Phosphorus - mg/1
3. Return Activated Sludge Phosphorus - mg/1
4. Effluent Collection Chamber Phosphorus - mg/1
5. Final Effluent BOD - mg/1
6. Final Effluent Phosphorus - mg/1
7. Settling - mg/1
8. Mixed Liquor Suspended Solids - mg/1
9. Primary Effluent Calcium - mg/1
10. Primary Effluent Magnesium - mg/1
11. Final Effluent Calcium - mg/1
12. Final Effluent Magnesium mg/1
13. Total Plant Flow - mg/1
14. Air Applied - 1000 CFM
15. Waste Activated Sludge Flow - MGD
16. Return Activated Sludge Flow - MGD
17. Primary Effluent Phosphorus - Ibs.
18. Effluent Collection Chamber Phosphorus - Ibs.
19. Waste Activated Sludge Phosphorus - Ibs.
20. Total Output Phosphorus - Ibs.
21. Phosphorus Accounted for - Ibs.
22. Percent Phosphorus Removed Secondary Process
23. Calcium Removed - Ibs.
24. Magnesium Removed - Ibs.
25. Lbs. BOD/lb. Solids Under Aeration
26. Cu. Ft. Air/gallon
27. Sludge Volume Index
28. Percent Total Phosphorus Removed
29. Total Phosphorus Removed - Ibs.
28
-------
0 PRIMARY EFFLUENT
(D RETURN ACTIVATED SLUDGE
© MIXED LIQUOR
0 EFFLUENT COLLECTION
CHAMBER
(D EFFLUENT
TION
PRIMARY —
CLARIFIERS
•>Wl_l_C.ls 1 \\J\\
AFTER CHLORINA-
T
^
_L
1
1
1
1
J
V V V \
(
t
' ~7~t
-s /
I/
t
t
>
SETTLED SEWAGE
/fv
f
AERATION
TANK
_
CO
CO
Q.
1
2S
<\J
CO
CO
o.
vj
ro
CO
CO
O.
-a-
^
CO
CO
Q.
I
czr
i . /
SECONDARY
CLARIFIERS
(TYPICAL)
CHLORINE
CONTACT
CHANNEL
FIGURE 10
FULL SCALE BIOLOGICAL TEST SAMPLING POINTS
29
-------
Stepwise multiple linear regression methods were intiated
for the examination of more complex relationships than
simple two variable situations. Results from simple linear
investigations and scatter plots were helpful in evaluating
the possible usefulness of multiple linear methods. Anal-
ysis of Variance was employed to determine whether varia-
tion in a parameter was of a level of significance beyond
that expected by chance. Analysis of Covariance provided
the same information with the added facility of removing
known sources of variability.
The results of all of these statistical analyses are avail-
able upon request from Metro. The programs for these
analyses were obtained from the "Biomedical Computer Pro-
grams" published by the Health Sciences Computing Facility
of the University of California, Los Angeles (46) . Samples
of operational reports, linear correlation scatter plots
and time relation diagrams are shown in Tables 5 and 6 and
Figures 11 through 13.
30
-------
5. METRO RENTON PLANT FULL SCALE
PLANT PHOSPHORUS REMOVAL WEEKLY
PROG NO. R01-1
INPUT WPC ROO
10/26/70
DATE RUN 02/12/71
PAGE 1
PRIMARY
EFFLUENT
BCD PHOS
HR MG/L
1 80
2 95
3 110
4 108
5 106
6 109
7 112
8 115
9 118
10 116
11 114
12 127
13 140
14 149
15 158
16 147
17 136
18 130
19 124
20 126
21 128
22 127
23 126
24 130
MINIMUMS
80
158
122
MG/L
9.4
9.6
9.8
9.1
8.5
9.5
10.5
11.0
11.5
10.5
9.5
9.6
9.8
10.3
10.9
11.2
11.5
11.5
11.5
11.2
11.0
11.0
11.0
10.0
ACTIVATED COLL
SLUDGE CHAMB
PHOS PHOS
MG/L MG/L
88
87
86
84
84
84
85
86
92
100
100
100
99
98
99
102
127
133
137
156
156
142
129
136
, MAXIMUMS,
8.5
11.5
10.4
84
156
108
10
10
10
9
9
9
8
8
8
8
8
6
5
4
4
5
6
6
7
7
7
7
7
7
AVERAGES
4
10
7
.4
.2
.0
.7
.3
.0
.6
.8
.9
.9
.4
.5
.3
.8
.7
.4
.2
.9
.0
.0
.0
.0
.3
.4
.7
.4
.7
FINAL
EFFLU .
BOD PHOS
HG/L
7.7
7.5
7.4
7.1
7.9
9.7
11.0
11.0
11.0
11.0
13.2
14.9
14.4
13.5
11.3
9.0
9.0
9.0
10.6
13.2
12.0
12.0
12.0
12.0
7.1
14.9
10.7
MG/L
11.7
11.3
10.8
10.2
9.9
9.7
9.4
9.1
9.2
9.5
8.9
8.3
7.0
6.0
5.7
5.4
6.0
6.4
6.8
7.0
7.0
7.0
7.0
7.0
5.4
11.7
3.2
SETTL-
ING
ML/L
462
431
406
334
356
320
330
339
333
334
345
374
419
413
391
437
552
585
570
693
504
481
474
456
320
693
433
MIX LIQUOR
SUSP PRIM.
SOLIDS CA MG
MG/L
1552
1485
1438
1428
1416
1400
1405
1417
1470
1521
1580
1601
1584
1570
1557
1537
1511
1515
1540
1540
1531
1564
1549
1504
1400
1601
1509
MG/L
15.0
15.0
15.0
16.5
18.0
17.5
17.0
18.0
19.0
17.5
16.0
16.5
17.0
17.5
18.0
17.5
17.0
17.0
17.0
17.0
17.0
17.0
17.0
18.0
15.0
19.0
17.0
MG/L
4.3
4.4
4.5
4.8
5.2
5.1
5.0
5.0
5.0
4.8
4.7
4.7
4.7
4.7
4.7
4.6
4.6
4.5
4.5
4.5
4.5
4.5
4.5
4.4
4.3
5.2
4.7
TOT. AIR WASTE
FINAL PLANT APPL- ACT
CA MG FLOW IED SLUDGE
MG/L
16.0
16.0
16.0
16.0
16.0
16.0
16.1
16.4
16.7
17.0
16.4
16.0
16.0
16.0
16.0
16.0
16.0
16.0
16.0
16.0
15.7
15.4
15.0
15.0
15.0
17.0
16.0
MG/L
4.5
4.5
4.5
4.5
4.6
4.6
4.6
4.6
4.7
4.7
4.6
4.5
4.5
4.5
4.5
4.5
4.5
4.5
4.3
4.2
4.1
4.1
4.0
4.0
4.0
4.7
4.4
' 1000
MGD CFM
13.0
15.0
12.0
21.0
26.0
2-8.5
29.0
28.0
26.0
25.0
25.0
24.0
22.5
22.0
23.5
24.0
24.0
23.5
22.5
19.0
17.0
16.0
14.5
11.0
11.0
29.0
21.3
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
23.0
MGD
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.10
0.04
RETURN
ACT
SLUDGE
FLOW
MGD
5.2
6.0
4.8
8.4
10.4
11.4
11.6
11.2
10.4
11.2
11.2
10.0
10.1
9.9
10.6
10.3
10.3
10.6
10.1
8.5
7.6
7.2
6.5
4.9
4.8
11.6
9.2
31
-------
PROG NO. R01-1
INPUT WPC ROO
10/26/70
io/26/7o PLANT PHOSPHORUS REMOVAL WEEKLY
EFFL WASTE % PHOS LB DOD/ CU FT
PRD1 COLL ACTIVAT. TOTAL PIIOS REI1OVED LB SOLIDS AIR/
EFFL C1IAMB. SLUDGE OUTPUT ACCT.'d SECOND, REMOVED UNDER GAL./
PIIOS PHOS. PHOS PIIOS FOR PROCESS CA MS AERATION DAY
HR
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
LB
1019
1201
981
1594
1843
2258
2540
2569
2494
2139
1981
1922
1839
1890
2136
2242
2302
2254
2153
1775
1560
1468
1330
917
MINIflUMS,
917
2569
1852
LB
1128
1276
1001
1699
2017
2139
2126
2055
1930
1356
1751
1301
995
331
921
1081
1241
1352
1314
1109
992
934
833
679
HAXMU!
679
2139
1361
LB
73
73
72
70
70
70
71
72
77
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
LB
1269
1414
1081
1786
2147
2306
2273
2125
1995
1981
1856
1661
1314
1101
1117
1031
1201
1254
1276
1109
992
934
847
642
LB
1201
1349
1073
1769
2037
2209
2199
2127
2007
1356
1751
1301
995
881
921
1031
1241
1352
1314
110°
992
934
833
C79
7
6
7
4
3
3
2
2
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.20
.04
.31
.40
.80
.10
.79
.79
.08
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
LB
-108.
-129.
-100.
S7.
433.
356.
217.
373.
498.
104.
-03.
100.
187.
275.
392.
300.
200.
196.
197.
158.
184.
213.
241.
275.
4
1
1
C
7
5
7
6
7
2
4
1
6
2
0
2
2
0
6
5
3
5
9
2
LB
-21.
-12.
0.
52.
130.
118.
96.
93.
65.
20.
20.
40.
37.
36.
39.
20.
20.
0.
37.
47.
56.
53.
60.
36.
7 0.
5 0.
0 0.
5 0.
1 0.
8 0.
7 0.
4 0.
1 0.
3 0.
9 0.
0 0.
5 0.
7 0.
2 0.
0 0.
0 0.
0 0.
5 0.
5 0.
7 0.
4 0.
5 0.
7 0.
06679
09563
09149
15830
19400
22117
23042
22650
20802
19004
17979
18976
19821
20811
23769
22879
21531
20099
13053
15495
14166
12950
11756
09477
1.020
1.577
1.971
1.127
0.910
0.830
0.816
0.845
0.910
0.915
0.915
0.952
1,016
1.038
0.971
0.952
0.952
0.971
1.016
1.204
1.346
1.423
1.577
2.083
1J-'J-' DATE RUN O:
PAGE 2
TOTAL
SLUDGE %
VOL. PHOS.
INDEX RE?1OVED
298
290
262
269
251
229
235
239
227
220
218
234
265
263
251
284
765
386
370
450
329
308
306
303
-24.
-17.
-10.
-12.
-16.
-2.
10.
17.
20.
9.
6.
13.
28.
41.
47.
51.
47.
44.
40.
37.
36.
36.
36.
30.
47
71
20
09
47
11
48
27
00
52
32
54
57
75
71
79
83
35
37
50
36
36
36
00
2/12/71
TOTAL
PHOS
RE-
MOVED
LB
-249
-213
-100
-193
-304
-43
266
444
499
208
125
260
525
789
1019
1161
1101
1000
832
666
567
534
484
275
IS, AVERAGES
0
77
27
612
2306
1448
67D
2209
1383
0
7
1
.00
.31
.69
-125.
498.
190.
1
7
3
-21.
130.
43.
7 0.
1 0.
3 0.
06679
23769
17333
0.316
2.033
.1.173
218
450
286
-24.
51.
19.
47
79
73
-304
1161
404
32
-------
10/ 25/ 70
OflTE RUN 03/ 02/ 71
U>
METRO RENTQN PLfiNT
OPERflTIONflL DfiTfi
TIME RELflTION DIP.GRflM - WEEKLY
FIGURE II
TIME RELATION PLOT
-------
10/ 25/ 70
DflTE RL'N C3/ C2/ 71
OJ
METRO RENTQN P
OPERflTICNP.L OPTP.
TIME RELP.7ICN DlflGRfiM - WEEKLY
FIGURE 12
TIME RELATION PLOT
-------
8/ 4/ 70 - 8/
70
o
o
Oo
UJ]
D
to
O.O
-Jo
CEo
o
o
ite cowKXfliiiw ucrnciCNT is
0.3121
nc REcaestioN LINE is
Y v -73.9055 .3.BO77 I
46 80 120- 1GOZOO240
PRItlflRY EFFLUENT BOD MG/L
CO
o
3-
o_
h-co
LU
o
co-
ThC CWREUU10N COEFrlCID.7 IS
o.ms
TtC RECRE6S1CW L1IC IS
Y B 4.14CO «0.16SS X
40 CO 120 160 200 240
PRIttflRY EFFLUENT BOD HG/L
FIGURE 13
LINEAR COORELATION SCATTER PLOTS
35
-------
The first series of tests was designed to study the effect
of feed method and solids concentration on phosphonus remov-
al. The combinations that were tried were plug feed-low
mixed liquor suspended solids concentration, plug feed-high
mixed liquor solids concentration, step feed-low mixed
liquor solids concentration and step feed-high mixed liquor
solids concentration. Each of these tests involved a mini-
mum of one week of operation at the desired conditions
before sampling was begun. Samples were collected and anal-
yses run every two hours for a period of seven days. The
average results of these tests are shown on Table 7.
Complete results can be obtained from Metro on a loan basis.
Examination of Table 7 suggests that the removal of phos-
phorus is directly dependent on the organic loading and
thus in theory on the biological growth. The pounds of BOD
per pound of mixed liquor solids during the first test was
twice that of the second test while the percentage and
pounds of phosphorus removed were also doubled. The step
aeration period shows that removal was better in the low
suspended solids test when the loading was slightly higher
and the wasting rate was significantly higher than during
the high solids test. These facts all agree with the
assumption that a large biological growth is necessary to
promote biological inclusion of phosphorus. This suggests
that both the loading and waste rates must be controlled
to obtain high biological phosphorus removal. An attempt
was made to determine the pounds of volatile solids pro-
duced per day to correlate this value with the phosphorus
removal. The results which are shown in Table 7 are not
completely consistent since test #2 shows a high solids
production combined with a low phosphorus removal. This
discrepancy was most likely related to errors in the mea-
surement of waste sludge flow which resulted in inaccurate
estimates of the pounds of'sludge produced. Variations in
the percent of phosphorus contained in the sludge could
also account for part of the apparent inconsistency.
As a consequence of these results, a full saale plant ex-
periment was designed so that a hypothesis concerning con-
trollable plant parameters could be tested. The hypoth-
esis was that waste activated sludge flow and aeration rate
had an effect upon the biological uptake of phosphorus.
These two parameters are often used to control the secondary
process so their importance in an operational model, if one
could be constructed, is evident.
These two variables are the only two of the twenty-nine
variables examined in this project which can be completely
controlled by plant personnel. They had not, however, been
36
-------
Table 7. AVERAGE RESULTS OF FULL SCALE FEED AND
MIXED LIQUOR SOLIDS CONCENTRATION TESTS
TEST
NO.
1
2
3
4
OPERA- PRIM. EFF.
TION
MODE
Plug
Flow
LOW
MLSS
Plug
Flow
High
MLSS
Step
Aera-
tion
Low
MLSS
Step
Aera-
tion
High
MLSS
BOD
rag/1
119
80
113
116
PHOS
mg/1
7.12
9.55
12.7
10.0
FINAL EFF.
BOD
mg/1
9.0
3.2
7.1
9.0
PHOS
mg/1
7.52
8.30
8.24
7.80
MLSS
mg/1
1497
2376
1789
2016
Plant
Flow
MGD
20.36
2175
2475
19.94
AIR
1000
cfm
23.0
35.2
17.5
22.8
Waste
Rate
MGD
0.044
0.05
0.09
0.04
Return
Rate
MGD
8.7
10.0
8.33
9.00
Ib.BOD
Lb MLSS
0.161
0.08
0.123
0.119
SVI
312
131
1C8
167
%
PHOS
Rem.
25.76
13.18
38.88
23.05
Ib PHOS
temoved/
Day
468
212
929
376
Ib. VSS
Produced/
Day
4,863
6,038
1.2,476
3,693
37
-------
controlled in any systematic way before this experiment was
defined, but had always been changed manually or automati-
cally in response to conditions in the plant. This elimin-
ated any possibility of determining if a change in another
variable was due to the fact that one of these controllable
variables had been changed.
The goal of the experiment was to observe the effect upon
the system when these two variables were changed in a
systematic manner. In light of this, the experiment was
designed so that sampling could occur during periods when
these two variables were held constant. The variables
were assigned values corresponding to three levels; low,
medium, and high since the absolute values would not be
relavent to any other plant while the relative values
could be applied to other plants in conjunction with opera-
tional data.
Each combination was held constant for seven days beginning
on Friday. Samples were collected and analyzed for all
variables on the following Wednesday and Thursday. The
sampling procedures for this experiment were almost the
same as for the weekly sampling schedule except that
samples were collected hourly instead of every other hour.
The results of this experiment were reported in plant re-
ports with exactly the same format as the weekly reports.
The raw data can be obtained on loan from Metro. Table 8
shows the nine experimental conditions and the average per-
cent total phosphorus removed for each sample period. It
was discovered after setting up for the combination of a
low air and low waste rate that it was not possible to
operate the treatment plant under these conditions. The
high solids concentration resulting from the low waste
rate in combination with the low aeration rate resulted in
inadequate oxygen concentrations in the aeration tank.
Therefore, operation under this set of conditions was dis-
continued before the equilibration period was completed.
The test involving medium air and low waste was not com-
pleted for the same reason.
Standard statistical methods were not applied to the data
since the experimental series could not be completed.
There is a great deal of variability in the data which makes
it difficult to come to any substantial conclusions without
the use of these methods. The experiment did, however,
demonstrate several things. As noted in Table 8, a con-
sistent increase in removal was recorded from Wednesday to
Thursday, Also there was better phosphorus removal at
higher aeration and wasting levels with an increase in the
aeration rate appearing to make a greater difference. It
38
-------
Table 8. Mean Percent Total Phosphorus Removed
On Wednesday and Thursday
\. Air
NV CU. ft./
\. gal.
Waste N.
Activated \.
Sludge \^
Ibs/day \
Low
2 PI K.
r oxo
Medium
7,400
High
14,000
Low
0.72 cu.ft.
per gallon
16.8
36.0
24.8
39.4
Medium
0.86 cu.ft.
per gallon
40.4
42.5
28.4
35.6
High Day
1.0 cu.ft.
per gallon
^ A *7 1*7*^/3
j*± . / we a .
41.7 Thurs .
40.1 Wed.
50.7 Thurs.
48.6 Wed .
44.4 Thurs.
39
-------
was not possible to make any definite conclusions because
of the amount of variability in the values and as a result
of our inability to treat the data statistically- If the
experiment could have been replicated so that several
values were available for each set of conditions, statisti-
cal methods would have been applied to the data to see if
the observed trends are meaningful.
In addition to the special sampling tests, normal sampling
and analysis of plant composite samples was conducted on a
routine basis during the period of the grant. Primary
effluent and final effluent samples were composited on a
flow paced basis by automatic samplers. The mixed liquor,
return activated sludge and effluent collection chamber
samples were hand composited on a time basis by the plant
operators. All samples were stored at 4 C after collection.
The variables shown on Table 5 page 32 were analyzed daily
over a period of eleven months. The daily averages for the
period December 1970 through April 1971, which can be ob-
tained from Metro, were selected for statistical evalua-
tion. It was felt that these data were the most valid as
a result of the valuable experience which had been gained
in collecting data prior to these dates.
Of the analyzed variables, only five were controllable by
plant operators. These were:
1. Return Activated Sludge Flow
2. Mixed Liquor Suspended Solids Concentration
3. Air Applied
4. Waste Activated Sludge by flow
5. Lbs. BOD/lb. Solids Under Aeration
To establish the importance of these variables to the bio-
logical uptake of phosphorus, a regression analysis was
employed. This method successively adds each variable to
a regression equation according to its mathematically deter-
minable importance in explaining variability in the depend-
ent variable, which was in this case the effluent collec-
tion chamber phosphorus. This variable was determined to
be the best indicator of the biological removal of phos-
phorus on the basis of the analyses which had previously
been performed during the research project. Regression
analysis showed that the controllable variables had the
following order of importance:
1. Return Activated Sludge Flow
2. Waste Activated Sludge Flow
3. Air Applied
4. Mixed Liquor Suspended Solids Concentration
5. Lbs. BOD/lb. Solids Under Aeration
40
-------
However, three-fourths of the variability in the effluent
collection chamber phosphorus was still unexplained after
these variables were introduced into the regression equation.
The failure to fully explain the variability in the effluent
collection chamber phosphorus implies that a large part of
this variability was due to factors which were not included
in the regression analysis. Identification of these factors
will be of significant importance if more complete control
of the biological phosphorus removal is to be achieved.
An examination of the data generated during this phase of
the grant showed that there was a weekly cycle in the phos-
phorus removal as well as in the primary BOD and the calcium
and magnesium removals. The phosphorus and calcium removal
were greater and the calcium removal was less during the
period Wednesday through Saturday than for the remainder
of the week. This cycle was statistically analyzed using
the Analysis of Variance. The average values during the
two periods as well as the results of the statistical
analysis are presented in Table 9.
As the table shows, phosphorus removal, magnesium removal,
and the primary BOD were greater, while primary phosphorus,
effluent phosphorus, and calcium removal were less during
the Wednesday through Saturday period than during the Sunday
through Tuesday period. All of these differences were sta-
tistically significant except for the calcium removal. The
difference in phosphorus removal averaged 0.63 mg/1 which
corresponds to an increase in the secondary phosphorus re-
moval from 20.6% on the Sunday through Tuesday period to
29.8% on the Wednesday through Saturday period. This
represents a net increase in the quantity of phosphorus
removed of 42% during the Wednesday through Saturday period.
The failure of the calcium removal cycle to coincide with
the phosphorus removal cycle makes it extremely unlikely
that the removal is chemical in nature. The correlation
between magnesium removal and phosphorus removal is in
agreement with the work reported by the City of Baltimore
(10) and Olson, et al. (13). The fact that the increases
in magnesium and phosphorus removal occur in conjunction
with an increase in the BOD entering the aeration system
strongly suggests that the phosphorus removal is biological
in nature. The inability to account for more than 25% of
the variability in the phosphorus removal indicates either
that factors other than those which were considered exert
a considerable effect on the biological removal or that the
statistical methods used were inadequate for examining
biological phenomenon.
41
-------
Table 9. CYCLIC VARIABILITY OF SELECTED PARAMETERS
AND THE STATISTICAL SIGNIFICANCE OF THIS VARIABILITY
VARIABLE
Primary Phosphorus mg/1
Primary BOD5 mg/1
Primary Calcium mg/1
Primary Magnesium mg/1
Effluent Phosphorus mg/1
Final Calcium mg/1
Final Magnesium mg/1
Calcium Removed mg/1
Magnesium Removed mg/1
Calcium Removed percent
Magnesium Removed percent
SUN-TUES
Average
mg/1
7.29
76.8
15.49
6.91
5.79
15.07
6.72
0.42
0.21
1.3
3.3
WED-SAT
Average
mg/1
6.79
93.2
15.89
6.88
4.66
15.72
6.35
0.17
0.55
1.07
8.0
F-FACTOR
3.9 sig.
95%
6.396
15.226
5.145
.024
16.753
12.301
5.647
2.318
4.170
2.000
4.915
42
-------
SECTION VIII
CHEMICAL REMOVAL
Chemical Removal Literature Review
As was the case for the biological removal studies, a thor-
ough literature search was conducted on chemical phosphorus
removal prior to the inception of the experimental work.
A complete literature review was written and is available
from Metro on a temporary basis.
All chemical removal schemes depend upon the formation of
an insoluble cation - phosphate compound which can sub-
sequently be removed by a physical separation process. The
selection of the appropriate cation for use in this pro-
cedure must be based on a number of criteria including avail-
ability, ease of use, lack of toxicity and, ultimately,
economics. The four cations which most closely meet these
criteria are aluminum in the +3 oxidation state, iron in
either the ferrous or ferric state and calcium.
In addition to the choice of chemical, there is also a
choice of chemical addition point. There are three basic
schemes for phosphorus removal which have been investi-
gated (47-130). These are: (1) addition of the chemical
to the raw sewage followed by the combined removal of
raw and chemical sludge in the primary clarifiers, (2) add-
ition of the chemical to the biological treatment system
followed by removal of the precipitated phosphorus in the
waste activated sludge, and (3) tertiary removal in which
the chemical is added to the secondary effluent and the
precipitated phosphorus is removed by a subsequent physi-
cal separation process.
Each of these separation schemes has several variations and
each offers distinct advantages as well as having distinct
disadvantages. The major advantage of chemical addition
to either the raw sewage or activated sludge is that it
makes use of existing clarification units and thus does
not require the construction of additional treatment
facilities. Addition of the chemical to the activated
sludge system places the further restraint on the chemical
that it must not cause the system pH to vary outside the
limits of the biological tolerances. Teriary removal
methods do not depend upon the removal efficiencies of the
secondary clarifiers and have the additional advantage
that they are dealing with a lower phosphorus concentration
43
-------
due to the prior biological removal. The chemical re-
quirements in that case may be lower than for the primary
and activated sludge removal schemes.
One of the major problems associated with all of the chem-
ical removal procedures is the additional sludge production,
Of the suggested chemicals, only lime can be effectively
recovered and reused. While this recovery and reuse does
not alter the chemical costs, it does minimize the quantity
of extra sludge being produced. Since effective phosphorus
removal does not occur with lime until the pH exceeds
10, this alternative is not available for chemical addition
to the activated sludge.
The major emphasis of this research work was directed
toward the use of direct chemical addition to the activated
sludge for phosphorus removal. On this basis, iron and
aluminum salts offered the greatest opportunity for in-
vestigation. It was hoped that this approach would result
in savings due both to the combined use of the secondary
clarifiers for activated sludge and chemical sludge removal
and to a reduction in the required chemical dosage as a
result of natural biological phosphorus removal prior to
the chemical addition.
Ionic Composition of Renton Wastewater ,
A series of analyses were performed to provide data regard-
ing the ionic composition of the Renton wastewater since
this data was necessary for the proper assessment of pos-
sible chemical reactions which might occur. The results
of these analyses for the primary and secondary effluents
are shown in Table 10. The anion deficiencies were assumed
to correspond to the bicarbonate concentration which is
in good agreement with the potentiometric titration data.
It was desired to measure the variation in the bicarbonate
content of the effluent with time in order to determine if
any correlation between bicarbonate and phosphorus removal
existed. The bicarbonate concentrations were determined
by potentiometric titrations. The validity of these titra-
tions to accurately measure bicarbonate levels was proved
by carrying out comparable titrations on a synthetic
effluent which was prepared to replicate the ionic com-
position shown in Table 10. An example of the potentio-
metric titration plot is presented in Figure 14.
A series of samples were collected over a one week period
including two intensive 24-hour samplings during which
samples were collected every two hours. A plot of the data
44
-------
Table 10. ION PROFILES
CATION
PROFILE
Na
K
Mg
Ca
Fe
NH4(as N)
PRIMARY
ppm
27.4
16.7
6.4
15.8
3.8
21.8
EFFLUENT
meq/1
1.19
0.43
0.53
0.79
0.14
1.56
Total 4.64
ANION
PROFILE
P04
Si°3
B0_
3
S04
Cl
NO3 (as N)
PRIMARY
PPHi
33.2
8.7
0.7
29.0
31.0
EFFLUENT
meq/1
0.52
0.3.1
0.10
0.58
0.86
FINAL EFFLUENT
BEFORE CHLORINATION
PPM
29.0
12.3
6.7
13.9
0.5
11.0
Total
meg/1
1.26
0.42
0.55
0.70
0.03
0.79
3.75
FINAL EFFLUENT
BEFORE CHLORINATION
ppm
24.4
5.7
0.7
28.0
31.0
2.4
meq/1
0.38
0.20
0.10
0.58
0.86
0.17
Total
2.37
Total
2.29
The Primary Cation/Anion balance shows a deficiency of
2.27 meq/1.
The Final Effluent Before Chlorination Cation/Anion balance
shows a deficiency of 1.46 meq/1.
45
-------
8r-
1
I
o.
FINAL EFFLUENT BEFORE
CHLORINATION 9-1-70
/ 2
TITER (milliequivolent per liter)
FIGURE 14
TYPICAL POTENTIOMETRIC TITRATION VERSUS
O.ION HCI AND END-POINT DETERMINATION
-------
from the one week sampling period is presented in Figure 15
Figure 16 shows the bicarbonate variation during the in-
tensive two day sampling. The results of these tests in-
dicate that there is a diurnal variation in the bicarbonate
content of the effluent ranging from 1.35 to 2.0 milli-
equivalents per liter with a minimum occurring at approxi-
mately 1300 hours. This variation appears to correlate
with the loading. There is no evidence to indicate that
this variation is correlated with phosphorus removal.
47
-------
00
II
I
o.
cr
UJ
t-
— 2.0
W
0>
V
Q.
cr
4)
OO.OO
10-25-70
•10-31 -70
O6.OO I2.OO
SAMPLING TIME (hours)
-H
I8.OO
FIGURE 15
DIURNAL VARIATION IN BICARBONATE ASSAY
24.OO
-------
2.O -
(D
UJ
OO.OO
06.0O I2.OO
SAMPLING TIME (hours)
18.00
24.OO
FIGURE 16
DIURNAL VARIATION IN BICARBONATE ASSAY- 24 HOUR TEST
-------
Full Scale Studies
The first full-scale test attempted to duplicate the re-
sults obtained at the Hyperion Treatment Plant (2). It was
reported that consistent and repeatable high phosphorus
removals could be obtained at the Hyperion Treatment Plant
whenever conditions in the aeration system could be adjusted
to achieve a pH rise in the aeration tank effluent and
partial nitrification in the secondary. Establishment of
these conditions at Hyperion required operation at plug
flow, high aeration rate and high mixed liquor suspended
solids levels in combination with an extended period of
aeration (approximately 8 hours), and the discontinuation
of digester supernate recycling.
The mechanism for this phosphorus removal was theorized
to be a calcium phosphate precipitation made possible by
the pH rise. The rise in pH was, in turn, proposed to be
caused by a C02 stripping reaction which could occur only
after the majority of the carbon metabolism had been com-
pleted (2). It was further proposed that the beginning of
nitrification indicated that the carbonaceous demand had
been satisfied and, as such, was a necessary condition for
the pH rise.
The flow regime, mixed liquor suspended solids, and aera-
tion rates were adjusted at the Renton treatment plant so
as to approach as closely as possible conditions at the
Hyperion plant. However, there were some limitations which
prevented attainment of an exact duplication of these con-
ditions .
There were two major differences which could easily have
affected the outcome of this test. The first of these was
the inability to provide a long aeration time. While the
aeration time at the Hyperion plant was close to 8 hours,
the maximum achievable aeration time at the Renton plant
was 4 hours, and this only during the daily low flow periods.
Even though a significant degree of nitrification was
achieved (50% conversion of NH3 to NO-j) , a pH rise through
the aeration system was not observed. This inability to
obtain a pH increase at Renton was attributed to the short
aeration time.
The second major difference was in the nature of the in-
fluent sewage which averaged 14 mg calcium per liter com-
pared with a value of 56 mg calcium per liter for the
Hyperion plant(2). Since the phosphorus removal mechanism
proposed for the Hyperion plant involved a precipitation of
50
-------
phosphorus as calcium phosphate, this difference was also
significant. Jar scale tests, were conducted in an attempt
to overcome the deficiencies in the full-scale tests.
However, the results obtained at the Hyperion Treatment
Plant could not be duplicated in either full or jar-scale
tests. Subsequent efforts were therefore directed to full-
scale testing of the effectiveness of chemical addition
for phosphorus removal.
The addition of alum and ferric chloride at the end of the
aeration tank was proposed to be the most effective pro-
cedure to follow during the full scale chemical feed studies
on the basis of the jar-scale chemical tests and full-
scale biological tests was to be controlled proportionally
to the quantity of phosphorus present at the addition
point. The major factor in the choice of the chemical
addition point was the highly variable and unpredictable
biological phosphorus removal in the aeration system.
Addition of the chemical to the primary could not take
this removal into account, which would have resulted in an
inefficient use of the chemical. Since addition at the
aeration tank outlet would occur following the biological
phosphorus removal, this feed point would automatically
compensate for any biological uptake and would therefore
result in the most economical chemical usage.
The same general conditions were maintained throughout both
tests. A description and schematic of the chemical feed
control system which was used for both tests has been pre-
sented in the "Materials and Methods" section. Figure 17
shows a diagram of the aeration system with the relative
position of the sampler-controller and the chemical addition
point. The position of the sampler-controller was located
upstream of the feed point to provide a detention time in
the aeration tank between the sampling point and the feed
point approximately equal to the time required to complete
the automated phosphorus analysis.
The frequencies of analyses were changed with experience
as it was determined that the initial analytical schedule
could not be maintained due to the quantity of analyses re-
quired. Samples of primary effluent, mixed liquor, return
activated sludge and final effluent were collected hourly
and brought to the laboratory for analysis. Samples col-
lected between 1800 and 0700 were stored in a 5 C refriger-
ator for analysis on the following day. Mixed liquor
settling was run hourly by the operations crew on a portion
of the samples which were delivered to the laboratory-
Additional samples were collected at points A through E on
Figure 17 every third hour (0900, 1200, etc.) for analysis
of pH and alkalinity. Alkalinity was measured on the super-
51
-------
(T) PRIMARY EFFLUENT
g) MIXED LIQUID
(3) RETURN ACTIVATED SLUDGE
PRIMARY —
CLARIFIERS
.UCIN 1
T
,
^^^
1
1
1
1
V
V V V \
(
/
r^ — t
/
Q
t
t
>
SETTLED SEWAGE
'TV
T^
AERATION
TANK
CO
Q_
^
1
2S
C\J
CO
CO
Q.
ro
CO
vQ
n
A
^^ '
^0
^s
2
^t
i
IT
^ 1 1
s®
tCHEM
'FEED
CONTROL-
LER
*FEED
<
csr • i ^
SECONDARY
CLARIFIERS
(TYPICAL)
CHLORINE
CONTACT
CHANNEL
FIGURE 17
SAMPLING AND CHEMICAL ADDITION POINTS
52
-------
Table 11. ANALYSES AND SAMPLE FREQUENCY-FULL SCALE
CHEMICAL FEuD TESTS
Chemical
Alum
Analyses
SS
VSS
BOD5
*A1
**Total P
Ortho
P04-P
pH
Settling
1
Primary Effluent
P
every 3rd hour
ii ii ii
4th
5th
hourly
—
—
"
2
Mixed Liquor
M . L .
every 3rd hr
n ii ii
—
—
hourly
—
hourly
n
3
Return Act. Sludge
RAS
every 3rd hour
n n n
—
—
hourly
--
—
"
4
Final Effluent
F
every 3rd hour
n ii 11
4th
5th
hourly
from 0800-2000
daily
_ ^
Ul
* For Fed., feed Fe test was run at the same frequency.
** Samples were stored for later analysis.
-------
nate following settling and decanting. The alkalinity and
pH of the samples collected between 2100 and 0600 were
measured by the operations personnel. Table 11 provides a
list of analyses run as well as the frequency for each
analysis at each sampling location. The frequencies in
this table represent the final minimum number of analyses
run with the initial schedule being more intensive.
Chemical addition was through a 40 inch long PVC tee which
had 1/4" holes at 5 inch intervals along the top edge.
This "diffuser" was situated directly above an aeration
header approximately forty feet from the end of the aeration
tank and four feet below the water surface. Mixing was
provided by the aeration in the aeration tank while low
energy flocculation was provided in the aerated mixed
liquor distribution channel leading to the final clarifiers.
Table 12 lists the average values of G (the rms Velocity
Gradients) and the time of contact for both mixing and
flocculation during both tests.
Average Value of G -
TABLE 12
— 1
rms Velocity Gradient (Sec ) and
Mixing Times
Alum Feed
Iron Feed
Chemical
Mixing
Chemical
Flocculation
G
Contact
Time
G
Contact
Time
137.8 sec l
8 min.
60.5 sec
30 min.
143.1
64.0
30
sec
8 min.
sec
min.
Alum, which was fed during the first full-scale test, was
stored in a rented tank truck which was pressurized to 25
psi to provide pressure delivery to the aeration tank. The
chemical feed rate was controlled by the chemical feed
control system, as discussed earlier. A minimum testing
period of fourteen days, which was at least two cell resi-
dence times, was selected in order to insure that equilib-
rium was reached. The alum feed test was conducted for a
total of seventeen days from June 14th through July 1st.
54
-------
As might be expected when placing a new system into oper-
ation, some initial problems were encountered. The major-
ity of these problems were related to the chemical feed
control system including:
1. Initial reagent imbalance in the phosphate analyzer
resulting in inaccurate phosphorus residual signals.
2. Mechanical problems in the computer section of the
controller which resulted in a considerable zero
point drift.
3. A sticking chemical feed control valve.
4. Limitations in the feed system which required that
the storage tank be depressurized and the alum
feed stopped whenever the tank was being filled.
All of these problems were corrected or eliminated either
during or at the completion of the alum feed test.
An aluminum to phosphorus ratio of 1.8 to 1 was initially
selected on the basis of the jar scale test results. This
ratio was increased to 2.0 to 1 beginning at 0800 on June
21st as a result of increasing effluent phosphorus levels.
It was held at that level throughout the remainder of the
test.
The raw data gathered during this test are available on a
loan basis from Metro. The daily totals for pounds of
primary phosphorus, mixed liquor supernate phosphorus, and
final phosphorus, as well as average daily aluminum to
phosphorus ratios, percent secondary phosphorus removals,
plant flows and effluent phosphorus concentrations are
presented in Table 13. The overall sums and averages of
these values over the entire test period are also pre-
sented in this table. Daily and average solids data for
the thirteen day period preceding the alum feed test as well
as for the period of the test are presented in Table 14.
As the data in Table 13 show, a significant removal of
phosphorus occurred between the primary effluent and the
mixed liquor. This removal continued to evidence the same
weekly removal cycle which was discussed previously. Over
the period of the test, the aluminum to phosphorus weight
ratio averaged 2:1. Secondary phosphorus removals
averaged 73% with the effluent phosphorus residual averag-
ing 2.1 mg/1. The data on this table is based on averages
generated from the hourly samples. Mixed liquor phosphorus
poundage was calculated using the sum of the effluent and
return sludge flows while all other poundages were calcu-
lated using the effluent flow.
55
-------
Table 13. DAILY AVERAGES AND TOTALS
FULL-SCALE ALUM FEED TEST
Effluent
Date
6-14
6-15
6-16
6-17
6-18
6-19
6-20
6-21
6-22
6-23,
6-24
6-25
6-26
6-27
6-28
6-29
6-30
Flow
MG
24
25
24
23
25
24
22
25
24
24
27
25
24
22
26
25
25
Ibs.
Prim
Total
P
1418
1732
1507
1193
1713
1774
1621
1876
1600
1636
1585
1593
1437
1570
1891
1710
1516
Ibs.
Soluble
M.L.
Ortho-P
843
847
553
375
892
1037
1352
1106
987
687
871
810
831
387
1203
989
525
Ibs.
Final
Total
P
649
242
251
189
474
472
530
588
352
299
486
403
461
343
598
499
466
Ibs.
Al
3493
1820
1647
648
1758
1541
1623
2039
2186
1403
1588
1629
1362
1134
1931
1641
1207
Average
Al-.P
ratio
(Wt.)
4.14
2.15
2.98
1.73
1.97
1.49
1.20
1.84
2.22
2.05
1.87
2.01
1.64
2.93
1.61
1.66
2.30
Average
Secondary
Total P
Kemoval
54.23
86.04
83.38
84.14
72.33
73.39
67.31
68.65
77.99
81.71
69.37
74.68
76.94
78.19
68.37
70.80
69.29
Average
Final
Total P
mg/1
3.2
1.2
1.2
1.0
2.3
2.4
2.9
2.8
1.8
1.5
2.2
1.9
2.3
1.9
2.8
2.4
2.2
Total 414 27,375 14,245
7,302
28,649
2.00
73.33
2.1
56
-------
Table 14. ALUM FEED TESTS-SOLIDS DATA
DATE
1071
June
1
2
3
4
5
6
7
3
9
10
11
12
13
Total
Average
14
15
1G
17
10
19
20
21
22
23
24
25
26
27
28
29
30
July 1
RAS
mg/1
5430
G18CT
5900
6320
70CO
6760
6400
6360
6920
6700
0900
6540
5300
35400
C569
5920
5340
6960
6,380
5520
5320
5740
6520
5020
7220
7600
7500
7400
6100
6160
6380
6620
6320
Total 116320
Average
6462
-AS VS
%
79
76
73
74
75
76
75
75
74
77
70
79
S3
76
74
60
69
67
64
64
64
6G
63
64
66
64
64
63
64
61
65
62
65
RAS WASTE
Ib
6,393
5,154
5,905
in,5/*-2
12,365
3,457
10,1/11
13,261
14,. '20
15,646
22,267
15,010
.14,020
154 ,490
11,804
12,037
12,025
17,994
13,623
17 ,034
12,135
11,960
12,507
11,376
17,462
15,479
23,143
22,035
1C-, 230
15,412
11, 7 "K
14,354
16,067
201,037
15,613
SS WASTE OVE£
V-EIP, Ib
3,621
4,431
5,668
4,62n
3,924
2, .671
3,075
1,115
4,704
3,300
1,540
2,092
2,002
42,766
3,290
2,640
3,504
3,225
2,704
3,000
3,240
2,30^
4,014
2,823
3,655
5,244
4,660
3,672
4,305
2,964
3,220
.3,330
2,231
60,904
3,303
TOTAL
ib WASTE
10,019
9,505
11,653-
15,162
16, 20^
11,131
13,216
14,076
19,132
10, 9/16
23,^07
17,010
16,030
197,756
15,173
15,477
15,609
21,219
21,327
20,114
15,375
14,356
16,521
14,204
21,117
21,723
27,«r>3
26,507
20,505
13,376
14,926
17,604
19,090
342,021
19,001
SVI
00
05
06
51
91
06
05
103
oo
104
02
113
115
92
114
85
03
79
81
77
75
66
73
63
60
67
62
63
61
61
56
66
72
SLUDGE
nuns .
1.14
1.10
1.16
1.96
1.10
1.16
1.05
0.97
1.09
0.96
1.22
n.oo
0.07
14.74
1.13
0.00
1.10
'1 . 2 0
1.26
1.23
1.30
1.33
1.51
1.37
1.50
1.47
1.49
1.61
1.59
1.6*
1.64
1.70
1..51
25.50
1.42
Ib.SS
T.1I7D. A"T, .
93,740
103,402
4,663
105,076
112,249
7,173
57
-------
Examination of the solids data leads to several interest-
ing conclusions. As would be expected, there was a definite
increase in the production of solids as compared to a
normal period when alum was not being added. This increase
was equal to 3830 Ibs. per day or approximately 25% greater
than for the reference period. The volatile content of the
sludge decreased at the same time from an average of 76% to
65%. It had stabilized in the range of 62%-64% by the end
of the test. This is in agreement with the fact that in-
organic solids were being added to the system.
The sludge volume index was also affected significantly by
the alum addition. The average SVI for the reference
period was 92 while the average during the addition period
was 72. The final SVI leveled off in the range of 60-65.
This decrease in the SVI is of significance, especially
when combined with the increase in solids production, since
it represents an increase in the settleability and thus the
density of the sludge after a 30 minute settling period.
Thus, even though there was a 25% increase in the pounds of
sludge produced, there was at least a 30% increase in the
sludge density compared to the reference period. Such an
increase would result in a decrease in the sludge volume
despite the increase in the pounds of sludge produced.
Other investigators have reported that alum addition
resulted in excessive turbidity and solids loss in the
effluent (93, 94). However, there was no appreciable in-
crease in the effluent suspended solids while there was an
improvement in the turbidity and transparency during the
alum feed test. This improvement was greatest during the
first few days of the test when turbidities of less than
1.0 JTU and transparencies of greater than 70 inches were
recorded. A malfunction in the feed controller which
resulted in a very aluminum to phosphorus ratio during the
first day of the test might have been partially responsible
for this very rapid improvement. This condition was
corrected and there was a slow decrease in the transparency
and increase in the turbidity through the first week of the
test, after which they stabilized at values which were
still better than normal.
Despite the better than normal transparencies and turbidi-
ties, there was no significant decrease in the pounds of
solids escaping with the effluent. This was attributed to
the large decrease in the SVI which, though indicating a
significant degree of flocculation and improvement in the
settleability of the majority of the floe, allowed straggler
floe, which was not captured by the rapidly settling sludge,
to escape in the effluent. As was stated previously, the
58
-------
degree of this problem was not sufficient to create an in-
crease in the pounds of solids in the effluent.
The solids removal during both periods averaged better than
94% and there were no indications of biological upset result-
ing from the alum addition at any time during the test.
Although the volatile content of the activated sludge de-
creased, it equilibrated at an acceptable level. BOD-
removals remained excellent with the effluent BOD- averag-
ing in the 3-6 mg/1 range on the basis of 5 daily flow
paced composite samples which are recorded in the plant
monthly report. Microscopic examination of the sludge
also indicated that no problems were occurring as can be
seen in Figures 18, 19, and 20 which are pictures which
were taken prior to, at the middle, and at the end of the
alum feed test, respectively. The sludge was well floc-
culated and the protozoa remained numerous and active
throughout the test.
During the period of the alum feed test a total of 414
million gallons of sewage were treated while a total of
50,700 gallons of alum (8.3% Al202) which corresponds to
138 tons of 17% A1203 were fed. The cost of alum during
this period was $50.55 per ton of 17% A120, f.o.b. factory.
The total chemical cost for the test was $6,920.00 or
$16.80 per million gallons treated, while the effluent
phosphorus concentration averaged 2.1 mg/1 and the secon-
dary phosphorus removal averaged 73.3%. On the basis of the
quantity of alum used and the quantity of phosphorus removed,
the treatment cost was $2.88 per mg/1 phosphorus removed
per million gallons, which corresponds to an average secon-
dary phosphorus removal of 5.8 mg/1. The total cost per
million gallons can be expected to increase as the level of
primary phosphorus increases, and as the desired effluent
phosphorus concentration decreases. The magnitude of this
cost increase in terms of dollars per mg/1 of P removed per
million gallons will be a function of the initial phosphorus
concentration and the desired final phosphorus residual.
In general, the reaction of aluminum in water containing
phosphorus can be viewed as a combination of two competi-
tive reactions. These two reactions are hydrolysis in
which the aluminum reacts with the water to form insol-
uble aluminum hydroxide and free hydrogen ions and phos-
phorus precipitation in which the aluminum reacts with the
phosphate to form insoluble aluminum phosphate. Both of
these reactions occur very rapidly with the rate for the
59
-------
Figure 18
Biological Floe Before Alum Feed
CO
-------
*
Figure 19
Biological Floe During Alum Feed
61
-------
Figure 20
Biological Floe at the End of the Alum Feed
62
-------
hydrolysis reaction being a function of the pH and the
aluminum ion concentration while the phosphate precipitation
reaction rate is a function of the aluminum ion concentra-
tion and the phosphate ion concentration.. Since a change
in the phosphorus concentration would not be expected to
affect the pH, the hydrolysis rate would be expected to be
constant while the rate of phosphorus precipitation would
vary in direct proportion to the phosphorus concentration.
For this reason, the cost to remove one mg/1 of phosphorus
from a given volume would be less at high phosphorus
residuals than at low phosphorus residuals. In addition
this cost would increase prohibitively as the desired final
phosphorus residual is decreased due to the almost total
loss of aluminum as the hydroxide.
The second full-scale test, during which ferric chloride
was fed, was conducted between August 9th and August 23rd,
1971. Prior to beginning this test, the mechanical and
chemical problems which had been encountered in the chemical
feed control system during the alum feed tests were correc-
ted.
The ferric chloride was stored in two 15,000 gallon vinyl,
above-ground swimming pools. The solution was protected
from contamination by a polyethylene sheet which was floated
on the liquid surface.
A centrifugal chemical pump was used to deliver the iron
solution to the aeration system. Since the capacity of the
pump was significantly greater than the required ferric
chloride flow rate, a bypass line, which went from the dis-
charge side of the pump to the storage tank, was installed
to prevent excessive pressure buildup when flow to the aera-
tion tank was limited by the diaphragm control valve. This
allowed the majority of the liquid pumped to be returned to
the storage tank while just enough chemical to satisfy the
controller demand was fed to the aeration system. This
system had the additional advantage of insuring homogeneity
of the ferric chloride solution due to the mixing caused by
the returned solution.
The corrections and modifications made in the chemical
feed control system were successful and the control problems
encountered during the alum feed test were eliminated. As
a result, a much closer control over the feed rate was main-
tained during the iron feed test. The major problems en-
countered during this test were related to the extremely
corrosive nature of ferric chloride coupled with a lack of
knowledge regarding compatible equipment on the part of
both the suppliers and the investigators. The problems
63
-------
which developed included:
1. The failure of all nylon fittings in the supply
system which required a postponement of the iron
feed test while they were replaced with PVC. Those
fittings had functioned perfectly throughout the
entire alum feed test, but began to fail within
three hours after exposure to the ferric chloride.
2. Decomposition of the feed pump seals which re-
quired that a new pump be installed before the
test could be continued. This pump was listed as
compatible with ferric chloride by the supplier.
3. Corrosion of the rotometer seals toward the end
of the test period. The rotometer had been used
during the alum feed test with no apparent decom-
position. It had been purchased with the under-
standing that it was compatible with both alum and
ferric chloride.
In addition to its corrosiveness, the ferric chloride
stained concrete and galvanized railing around the storage
tanks, killed the lawn where it was spilled and stained the
clothing and skin of the personnel working around the area.
While these characteristics are unrelated to the ability of
ferric chloride to remove phosphorus, they make it a
difficult material to work with or around.
At the beginning of the test, an iron to phosphorus ratio
of 3.6 to 1 was utilized on the basis of the jar scale test
results. The ratio was increased to 4.0 to 1 at 1700 on the
12th because the effluent phosphorus level was still above
2 ppm. This ratio was maintained to the end of the test.
The raw data collected during the iron test is available
from Metro on request. Daily averages of the iron to phos-
phorus ratio, plant flow, percent secondary phosphorus
removal, final effluent phosphorus concentration, total
pounds of primary phosphorus, mixed liquor supernate phos-
phorus, final phosphorus and pounds of iron fed are presented
in Table 15. This table also includes the overall sums and
averages of the above data for the test period. Average and
daily solids data for the test period and a thirteen day
reference period just preceding the test are presented in
Table 16.
As an examination of Table 15 shows, there was again a
significant phosphorus removal between the primary effluent
and the mixed liquor. The iron to phosphorus weight ratio
averaged 4.80 which significantly exceeded the ratio set
on the feed control system. This descrepancy was the
64
-------
Table 15. DAILY AVERAGES AND TOTALS
FULL-SCALE IRON FEED TEST
Effluent
Date
8-10
8-11
8-12
8-13
8-14
8-15
8-16
8-17
8-18
8-19
8-20
8-21
8-22
Flow
MG
19
22
22
22
20
20
22
22
22
21
22
21
20
Ibs.
Prim.
Total
P
1351
1481
1384
1298
1353
1342
1583
1344
1551
1136
1318
1492
1140
Ibs.
Soluble
M.L.
Ortho-P
1119
956
927
622
625
889
1114
753
937
500
804
1054
531
Ibs.
Final
Total
P
451
329
379
245
240
228
253
198
247
258
189
234
127
Ibs.
Iron
5297
4102
4516
1620
3355
4573
5262
3780
4343
3160
4171
4768
3010
Average
Fe:P
ratio
(Wt.)
4.74
4.29
4.87
2.61
5.37
5.14
4.72
5.02
4.64
6.31
5.19
4.53
5.67
Average
Secondary
Total P
Removal
%
66
77
72
81
82
82
84
85
84
77
85
84
88
.63
.77
.62
.13
.29
.98
.04
.29
.09
.32
.69
.32
.87
Average
Final
Total P
mg/1
2
1
2
1
1
1
1
1
1
1
1
1
0
.8
.8
.1
.3
.4
.4
.4
.1
.3
.5
.0
.3
.8
Totals 275 17,773 10,832
3376
51,958
4.80
81.01
1.5
65
-------
Table 16. IRON FEED TEST-SOLIDS DATA
DATE
1971
July
27
28
29
30
31
Aug.
1
2
3
4
5
6
7
8
Total
Average
Aug .
9
in
11
12
13
14
15
16
17
18
19
20
21
22
Total
Average
HAS
ing/1
5160
5600
4530
5340
5000
4340
4840
5480
5560
4760
4600
5000
5630
65970
5075
4970
5140
5360
6020
5960
5340
5060
5460
6010
6420
6360
6220
6300
6230
31900
5350
RAS VS
0,
74
77
78
76
73
75
75
75
79
80
78
77
76
76
76
75
75
73
67
72
64
66
64
63
63
55
63
58
67
RAS WASTE
Ib
11,139
11,676
12,037
15,537
15,429
9,049
0
10,512
12,934
-13,497
11,319
3 ,340
9 , 8 6 n
142,429
10,956
9,119
11,574
16,123
17,572
19,332
19,432
15,614
13,661
14,536
14,992
16,443
16,543
19,966
19,002
225,414
16,101
SS WASTE OVER
WEIR Ib
3,465
2,361
3,094
2,132
2,912
2,587
3,433
2,167
7,245
1,935
3,435
3,696
3,359
42,476
3,267
3,535
2,283
1,260
2,393
2,100
1,949
1,140
1,050
1,260
2,460
2,084
1,484
792
2,364
26,714
1,903
TOTAL
Ib WASTE
14 ,654
14 ,037
16,031
17,769
13,341
11,636
3,^88
12,679
2n,229
15,432
14,804
12,036
13,719
134,905
14,223
12,704
13,362
17,333
20,470
21,982
21,431
16,754
14,711
15,796
17,452
13,527
13,027
20,753
22,266
252,128
13 ,009
SVI
94
91
93
102
95
70
80
31
79
76
79
85
77
1102
85
63
78
84
38
87
86
94
86
84
86
93
115
115
38
89
SLUDGE
DEMS .
%
1.06
1.10
1.07
0.98
1.05
1.43
1.25
1.23
1.26
1.31
1.26
1.18
1.30
15.48
1.19
1.59
1.28
1.19
1.14
1.15
1.16
1.06
1.16
1.19
1.16
1.02
0.87
0.37
1..14
15.93
1.14
Ib.SS
UNO. AER.
92,285
95,906
3,621
98,492
111,597
13,105
66
-------
result of an incorrect allignment of this system. Since
the ratio was being set empirically based on the phosphorus
removal. Secondary phosphorus removals averaged 81% with
the effluent phosphorus concentration averaging 1.5 mg/1.
These data were calculated as was discussed previously for
the alum data.
The iron addition, as was the case with the alum addition,
resulted in an increase in the daily production of solids
and a corresponding decrease in the volatile solids content
of the sludge as compared to the reference period as is
shown in Table 16. The increased solids production was
4,440 pounds per day or approximately 25%. The volatile
solids content dropped as low as 55% after less than 13
days of iron feed and it appeared to be dropping further
at the time the test was discontinued. Such a drop in
the volatile content of the sludge could indicate a con-
version from a biological to a chemical floe which would de-
stroy the effectiveness of the biological system.
The sludge volume index was unchanged during the iron
feed test and the averages before and during the test were
in the range of 85-90. Since there was no change in the
SVI and thus none in the settled sludge density, the
sludge volume would be expected to increase by a factor
of 25-30% in response to the increased sludge production.
Such an increase in the sludge volume could be significant
in a plant which was already overloaded with solids.
The turbidity and clarity of the effluent improved during
the iron feed test as compared with normal conditions.
Turbidities stayed in the range of 1.5-2.0 JTU while the
effluent BOD- averaged in the range of 2-5 mg/1. This
improvement was accompanied by a decrease in the effluent
solids concentration. The decrease in the effluent solids
was attributed to the excellent flocculent characteristics
of ferric chloride which were not counteracted by a corres-
ponding decrease in the SVI. Thus the iron caused the
sludge to flocculate and helped to trap and remove the
straggler floe which would have escaped to the effluent
if the SVI had been decreased as well. Solids removal
during this period averaged better than 95%. Microscopic
examination of the floe showed no decrease in the biological
activity of the protozoal population.
-------
A total of 275 million gallons of sewage were treated during
the iron feed test while a total of 25,600 gallons of 43%
ferric chloride which is the equivalent of 70.8 tons of
anhydrous ferric chloride were fed. The cost of the ferric
chloride as delivered was $209 per ton, however, delivery
accounted for a significant part of the cost. The f.o.b.
the factory cost was quoted as $143 per ton of anhydrous
ferric chloride. Based on the cost at the factory, the
chemical cost for the iron addition would have been $34.20
per million gallons. The effluent phosphorus concentration
averaged 1.5 mg/1 and the secondary phosphorus removal aver-
aged 81.0%. When the chemical cost is converted to terms
of dollars per mg/1 of phosphorus removed per million gal-
lons, the result is $4.40. This corresponds to an average
removal of 6.2 mg/1 of phosphorus in the secondary. As was
discussed previously, this figure would increase as the final
phosphorus level was decreased as a result of kinetic con-
siderations.
68
-------
Figure 21
Biological Floe Before Ferric Chloride Feed
69
-------
Figure 22
Biological Floe During Ferric Chloride Feed
70
-------
'
'rf>
Figure 23
Biological Floe at the End of the Ferric Chloride Feed
71
-------
SECTION IX
DISCUSSION
This study involved the use of methods of research ranging
from straight literature review to full scale testing of
both biological and chemical methods of phosphorus removal.
The results obtained using each of these methods are dis-
cussed herein.
Biological Tests
Both bench scale and jar scale tests were in general agree-
ment. The uptake of phosphorus was not affected by dis-
solved oxygen levels within normal operating ranges (0.5-
4.0 mg/1) nor was there a correlation between organic load-
ing and phosphorus uptake. Both methods did indicate that
there is a direct correlation between the concentration of
the mixed liquor suspended solids and the removal of phos-
phorus. The jar scale test also showed some correlation
between the phosphorus uptake and the product of the solids
concentration and the hardness.
The results of full scale analyses showed that the removal
of phosphorus was directly related to the organic loading
and the magnesium uptake. That is, as the organic loading
increased, the removal of both magnesium and phosphorus in-
creased. These results are in agreement with previously
published reports of a correlation between biological phos-
phorus uptake and magnesium uptake. This strongly suggests
that the normal mechanism of phosphorus removal at the Ren-
ton plant is exclusively biological in nature. This con-
clusion is supported by the extremely low concentrations
of magnesium and calcium in the influent stream which are
below the equilibrium concentrations required for chemical
precipitation.
Despite the tremendous volume and variety of data collected
and analyzed, only 25% of the observed variability in phos-
phorus uptake could be statistically attrubuted to these
variables. This means that other more subtle factors were
responsible for the majority of the variation in the phos-
phorus removal. Further research needs to be done to deter-
mine the nature of those factors in order to provide the key
to complete control of biological phosphorus removal.
Since the normal phosphorus removal mechanism at the Renton
plant was of a biological nature, it would be expected that
operation under conditions which would promote rapid
72
-------
biological growth would provide maximum phosphorus uptake.
This was confirmed during one phase of the full-scale study
when operation under conditions of high waste rate and high
aeration rate was found to provide maximum phosphorus re-
moval. The magnitude of this removal will be directly
dependent on the ratio of BOD to phosphorus entering the
plant under a consistent mode of operation.
A comparison of the results obtained from the full-scale
tests and those from the jar and bench scale tests showed
that there was a significant difference between the two
types of tests. The jar and bench scale tests were subject
to numerous variables, including conditions related to the
scale difference, which would affect the results and their
relation to full-scale conditions. The advantages obtained
from having a small system with completely controllable in-
put parameters were outweighed by the disadvantages resulting
from the significant difference between the model and the
real systems. Jar and pilot scale studies were inadequate
for developing realistic data on the biloogical operation
of the full-scale facility.
Chemical Tests
The jar and bench scale chemical tests were in general
agreement and showed that alum, ferric chloride and ferrous
sulfate could be used equally well to remove phosphorus
from a biological culture. They also showed that alum was
significantly more effective than sodium aluminate under
the conditions encountered at the Renton plant.
Under the normal conditions of pH encountered upon the
addition of alum or ferric chloride to,the activated sludge
(i.e. pH 6.0 - 6.5) an aluminum to phosphorus weight ratio
of approximately 3.6:1 was required to reduce the final
phosphorus was found to occur instantaneously for both
salts. The iron to phosphorus ratio was found to be the
same for ferric chloride and ferrous sulfate, but there was
a significant reactio.n time required for ferrous sulfate
before complete precipitation occurred.
The aluminum to phosphorus ratio required for sodium alumin-
ate was found to be approximately twice that for alum. This
was attributed to the increase in pH to 8-9 which occurred
when sodium aluminate was added to the activated sludge.
The pH of the system gradually decreased with time to the
range of 7 with an accompanying increase in the phosphorus
removal, but the removal never reached the magnitude assoc-
iated with the addition of a comparable amount of alum.
Immediate adjustment of the pH to the 6.0-6.5 range resulted
in an aluminum to phosphorus ratio equal to the 1.8:1 value
73
-------
obtained with alum.
The importance of pH with respect to the ratio of aluminum
to phosphorus removal was further investigated in a series
of tests, the results of which can be obtained from Metro.
The data from these tests showed that the required aluminum
to phosphorus ratio increased from 1.6 to greater than 2.0
as the pH increased from 6 to 8. Such an increase can be
explained by considering the acid dissociation of the phos-
phoric acid species and the deprotonation equilibrium of
the hydrated aluminum species. Through these reactions,
the phosphate_species becomes more negatively charged (i.e.
H PO. —* HPO ~+H ) as the pH increases while the positive
charge on the hydrated aluminum species becomes less (i.e.
A1(H20)5OH -r>Al (H20) 4 (OH) 2 +H ) as the pH increases. The
increase in the aluminum to phosphorus ratio thus becomes
dependent on the ratio of charges on the two species and on
the loss of aluminum as an insoluble hydroxide with increas-
ing pH. A similar situation exists with iron, but the pH
range is lower due to the different equilibrium constants.
One factor which jar and bench scale tests were unable to
handle was the question of the effect of extended addition
of iron or aluminum to the biological system. This was
important because of the possible deleterious effects of
such additions on the biological system and because of the
need to determine whether there was a recycle effect on the
phosphorus removal which would result in a decrease in the
metal requirement for phosphorus removal with time. A
series of continuous pilot scale studies were conducted to
provide information on these subjects.
The addition of either iron as ferric chloride or aluminum
as alum had no detrimental effects on the biological system
although there was some change in the ratio of organisms
present. The alum addition caused a decrease in the sludge
volume index of the system while there was no change in the
SVI during the iron addition. There was no recycle or tur-
bidity effect observed with either chemical.
Unlike the biological jar and bench tests, the chemical jar,
bench and pilot scale tests were excellent methods for de-
veloping operational data on the operation of chemical phos-
phorus removal systems. The full scale tests were necessary
to obtain cost and chemical feed data for the full-scale
system, but the initial small scale tests allowed for the
selection of the most effective chemicals and addition
points at a minimum of cost and with a minimum of time.
74
-------
Both ferric chloride and alum functioned effectively during
the full-scale tests. The selection of the best and most
economical chemical on the basis of these tests is compli-
cated by a number of factors. Alum was a far more attrac-
tive chemical than ferric chloride from the standpoint of
ease of handling and lack of corrosiveness. Both chemicals
could be controlled equally well using the type of control
system discussed in the experimental section. A summary of
the relevant data for each test is presented in Table 17.
As would be expected, the addition of either chemical re-
sulted in an increase in the pounds of solids produced due
to the additional chemical precipitate. Although the in-
crease in sludge production during the iron test was greater
than during the alum test, the difference was not as great
as would be expected on the basis of the relative molecular
weights of the two elements and the actual quantities of
chemical added. This inconsistency was related at least in
part to errors in the measurement of the waste RAS flow.
The accuracy of this measurement was such that a significant
error might result. Another likely source of error was the
method used to calculate the increased sludge production,
which depended on a comparison of sludge produced during the
period of chemical feed to sludge produced during an arbi-
trary period preceding each test. The parenthasized figures
in Table 17, which are data for the control periods, show
that there were changes in loading and detention times be-
tween the control and test periods which could easily result
in errors in the calculated increase in sludge production
caused by the chemical additions. Under any circumstances,
the use of iron would be expected to produce approximately
one third more chemical sludge at the weight ratios used in
these tests.
The increased solids production was counteracted in the case
of alum by an accompanying increase in the sludge density
while no such density increase was observed with the iron
addition. Both tests were operated with beginning SVI's of
approximately 85 which might partially explain the lack of
change in the iron test. Some increase in the sludge den-
sity might have been obtained during the iron test if the
starting SVI had been higher. Since alum reduced the SVI
below 85 while iron was unable to do so, this increase would
be expected to be less than for the alum. Thus, the use of
alum would be expected to result in a significantly smaller
volume of solids than would the use of ferric chloride.
The average phosphorus residual in the effluent was some-
what less and the percent secondary phosphorus removal was
greater during the iron test than during the alum test. This
75
-------
Table 17.
SUMMARY OF FULL-SCALE CHEMICAL FEED TEST RESULTS
Parameter
Duration of test (days
Volume Treated (million gallons)
Volume of Chemical Used (gallons)
Weight of " " (tons)
Cost of Chemical Excl. shipping
Ferric
Chloride
14
275
25,571
(43%FeCl3)
70.8
$10,124
Alum
(Aluminum
Sulfate)
17
414
50,700
(8.3%A12O3
138
(17% A1203
$6,920
)
Cost of Chemical Excl. shipping
Average Chemical to Phosphorus
Weight Ratio
Sludge Volume Index-before
addition
Sludge Volume Index-after addition
$10,124
4.80:1
84
88
Final Effluent Total Phosphorus (ppm) 1.5
Average Secondary Removal (%)
Total Pounds Phosphorus Removed
Total Pounds Phosphorus Discharged
Treatment Cost ($/mg/l removed/MG)
Cell Residence Time (days)
Average Aeration Time (hours)
Average BOD Loading (Ibs BOD/
100 Ibs. MLVSS)
Increased Sludge Production
(Ibs. /day)
81
14,396
3,376
$5.86
5.8(6.7)
5.58(5.30)
24.6(25.6)
4443
$6,920
2.0:1
88
62
2.1
73
20,073
7,302
$2.88
5.7(6.7)
5.30(5.20)
27.4(29.3)
3829
difference can be explained as resulting from controller pro-
blems during the alum test and the fact that the ratio of
iron to phosphorus was greater than the ratio of aluminum to
phosphorus on a stochiometric basis. Such a difference could
be corrected by increasing the aluminum to phosphorus ratio.
This would, of course, result in an increase in the cost of
the treatment.
Although the full-scale tests generally agreed quite well
76
-------
with the jar and bench scale tests, the removals obtained at
the same metal to phosphorus ratio during the jar and bench
tests were significantly better than for the full-scale tests.
One of the major causes for this was a lack of adequate
mixing during the full-scale tests as demonstrated by the
G values at the mixing point. These values, which were in
the range of 150~ were significantly less than the value of
1000 sec which has been recommended (EPA Technology Trans-
fer Seminar-Seattle, Washington, 1972) . As a result, it
seems likely that significant quantities of metal were lost
through hydrolysis and precipitation in forms other than in-
soluble phosphate. If this were the case, a more effective
method of mixing would be required to provide the maximum
efficiency for phosphorus removal. The method used would
have to be adapted to the hydraulic configuration of the
system to which chemicals are being added.
The chemical feed and control system utilized during the full-
scale tests performed well considering the experimental
nature of the system, but certain improvements would be re-
quired to maximize the efficiency of the system. To provide
protection for the analytical equipment as well as the per-
sonnel servicing the system, the system should be installed
in a permanent structure. The portable cabinet served
adequately during the test period, but would not be satis-
factory for continuous use. The flow signal used during the
test period was the effluent sewage flow rather than the
sum of the effluent and return sludge flows because of the
difficulty associated with producing such a figure for a
short term test. Since the return rate was controlled as a
percentage of the effluent flow, it was possible to manually
compensate for this lack during the testing period while
adjusting the ratio set point. While this adjustment was
adequate for the testing period,a combined flow signal should
be provided for a permanent installation.
Another flow measurement critical to the proper control of
the chemical feed is the chemical flow rate. A transmitting
rotometer was used during the full-scale tests, however,
corrosion and sticking of the rotometer were fairly serious
problems. An alternative to a rotometer such as a magnetic
flow meter or a metering pump would be preferable.
The automatic analyzer used in the control system worked
very well. Such a system should be provided with a direct
signal from the colorimeter to the computer rather than
utilizing a logarithmic cam as was done in the experimental
system. The colorimeter signal should be converted to a
logarithmic function prior to use as a control signal if this
is not automatically done by the colorimeter. The concen-
77
-------
tration signal should be connected to both high and low
level alarms since these are necessary to give warning of
an analyzer malfunction as well as to protest against either
over or under feeding of the chemical. Although it is not
critical to the control of the chemical addition, it is
advisable to install a phosphorus analyzer in the final
effluent. Such an installation would provide back-up cover-
age and, if equipped with a high level alarm, would call
attention to any malfunction in the system.
The addition of both alum and ferric chloride caused a re-
duction in the mixed liquor pH from its normal level of
7.0-7.2 to 6.5 at the point where it entered the final clari-
fiers. The pH reduction was even greater immediately after
the chemical addition, but the biological activity resulted
in an increase to the 6.5 value during passage through the
mixed liquor distribution channel to the clarifiers. The
pH depression did not affect the biological activity with
respect to BOD5 and suspended solids removal, but it may have
been responsible for an inhibition of nitrification which
was observed during both full-scale tests. However, the
cell residence times during both tests were lower and the
organic loadings were higher than during the control periods,
which would also result in a decreased level of nitrifica-
tion during the test so it could not be determined whether
it could be achieved if it were desired.
While the pH reductions experienced during full-scale
tests were not excessive, this could be a problem in systems
having high phosphorus levels and/or low alkalinities.
Systems having alkalinities significantly less than 150
mg/1 or influent phosphorus levels greater than 10 mg/1
might find it necessary to control pH while adding chemicals.
One possibility would be the use of sodium aluminate in
combination with alum or ferric chloride to maintain the pH
at an acceptable level.
The addition of the two chemicals caused a decrease in the
volatile content of the activated sludge as a result of the
accumulation of chemical sludge. This decrease had stabil-
ized at an acceptable level during the alum feed test, but
the volatile content was still decreasing at the end of
the iron>feed test. This loss of volatile content of the
activated sludge during the iron feed could have correspon-
ded to a conversion from a biological to a chemical system
and, as such, would be a serious problem for continuous
chemical addition. The magnitude of this problem was in-
creased by the low BOD loading at the Renton plant. It
seems likely that no such problem would exist at a facility
where the loading and thus the biological growth rate is
higher.
78
-------
The design of a chemical feed system to be added to an
existing activated sludge plant would be similar whether
alum or ferric chloride are used. A larger storage capa-
city would be required for alum than for ferric chloride,
but the best procedure is to design the system so that
either chemical can be used depending on the results of ex-
perimentation. The piping for the system should be entirely
PVC and should allow for a wide flexibility in the selection
of dosing points. The chemical pump should be a heavy duty
model designed to handle corrosive materials such as ferric
chloride.
Fiberglass storage tanks provide the best answer to chemical
storage in terms of cost and maintenance. They can be used
to store either chemical and are much more reliable than
rubber lined steel tanks. If a minimum of seven days stor-
age is desired, a minimum storage volume of approximately
1000 gallons per million gallons of daily flow will be need-
ed assuming an influent phosphorus level in the 8 to 10 mg/1
range, and an aluminum to phosphorus ratio of 2.0:1. A
smaller storage volume will be required if ferric chloride
is selected, but sizing the tanks for alum provides greater
flexibility. The approximate cost of the chemical feed
control system is in the range of $10,000 to $12,000 inclu-
ding everything except the storage tanks. The automatic
analyzer system accounts for approximately $7,000 of this
figure. The number of chemical feed points in the plant
will determine the number of chemical feed control systems
required. Thus, if the feed point is the end of the aera-
tion tank, one system per aeration tank will be required.
However, if the aeration tanks discharge to a central dis-
tribution channel, provisions could be made to supply rapid
mix in this channel and thus minimize the number of control
systems required. It is important to insure that adequate
detention time remains after the chemical addition to in-
sure proper flocculation.
An analyzer for the final effluent would cost an additional
$7,000 but it would provide additional security with respect
to system malfunctions. The costs for the autoanalyzer sys-
tems are based on the system used during the full-scale
test. Less expensive systems may be available, but they
must have a prefilter capable of removing the quantity of
biological solids normally encountered in activated sludge
mixed liquor.
The major cost associated with chemical feed for phosphorus
control is the cost of chemicals. These costs in turn will
be strongly dependent upon the location of the chemical
producer with respect to the treatment plant. Since trans-
79
-------
portation costs are variable depending on location, no
attempt was made to provide data on shipping costs of the
chemicals. The costs of the chemicals are based on the cost
F.O.B. the factory in the case of both the alum and the
ferric chloride.
On the basis of the full-scale tests, alum was fed at an
average rate of 120 gallons (8.3% A12O3) per million gallons
while ferric chloride was fed at a rate of 98 gallons per
million gallons. This corresponds to a cost of $16.80 per
million gallons for alum versus a cost of $34.20 per million
gallons for ferric chloride under the conditions found at
the Renton treatment plant. The chemical costs based on
removal of phosphorus were $2.88 per mg/1 of phosphorus
removed per million gallons for alum and $5.86 per mg/1 of
phosphorus removed per million gallons for ferric chloride.
Either alum or ferric chloride can be effectively used to
remove phosphorus by addition to the activated sludge. Even
though the final phosphorus residual obtained during the
alum test was greater than the final phosphorus concentra-
tion during the ferric chloride test, alum is the chemical
of choice for use at the Renton plant. This choice is
based on the greater ease of handling, the enhancement of
the sludge settleability and the greater economy associated
with the alum. The additional cost required to bring the
residual to the same level as achieved with the ferric
chloride would not be great enough to make the ferric chlo-
ride less expensive and the cost factor is of necessity the
major factor. Implementation of the recommendation for
more efficient mixing would very likely reduce the cost of
both chemicals significantly, but it would not'change the
relative costs of the two.
The cost advantage of alum over ferric chloride is greatly
increased at the Renton plant since alum is produced within
20 miles of the plant while the nearest producer of ferric
chloride is over 1000 miles away in central California. It
will be necessary for each plant considering the use of
these chemicals for phosphorus removal to price the chemi-
cals based on their local situation. It is possible that
the transportation costs will change the economic order of
the two chemicals at which time the greater ease of hand-
ling and the better sludge settleability obtained with alum
become significant factors.
80
-------
SECTION X
REFERENCES
1. Vacker, D., Connell, C. H., and Wells, W. N. "Phosphate
Removal Through Municipal Wastewater Treatment at San
Antonio, Texas", Jour. Water Pollution Control Feder-
ation, 3J9, 750 (1967) .
2. Bargman, R. D., Betz, J. M., and Garber, W. F. "Nitro-
gen-Phosphate Relationships and Removals Obtained by
Treatment Processes at Hyperion Treatment Plant", pre-
sented to the 5th Int. Water Pollution Research Conf.
July 1970
3. "Standard Methods for the Examination of Water and
Waste Water", 12th Edition, American Public Health Assoc-
iation, New York, 1965.
4. Hill, Uno T., "Direct Photometric Determination of
Aluminum and Iron Ores", Analytical Chemistry, 28
No. 9 pp 1419-1423 (1956).
5. Lea, W. L. and Nichols, M. S. "Influence of Substrate
on Biochimical Oxygen Demand", Sewage Works Journal, 8_
435 (1936) .
6. Sawyer, C. N. "Biological Engineering in Sewage Treat-
ment" Sewage Works Journal, 16, 925 (1944).
7. Bogan, R. H. "The Use of Algae in Removing Nutrients
from Domestic Sewage", In "Algae and Metropolitan
Wastes", U. S. Public Health Service, Cincinnati, Ohio
(1961) .
8. Scalf, M. P., Pfeffer, F. M., Lively, L. D., Witherow,
J. L., and Preising, C. P. "Amenability Studies at
Baltimore, Maryland—Phosphate Removal by Activated
Sludge", Federal Water Quality Administration, Ada,
Oklahoma, 1968.
9. "Phosphate Removal at Baltimore, Maryland",
Jour. San. Eng. Div., Proc. Amer. Soc. Civil Engr.,
95, 817 (1969).
81
-------
10. City of Baltimore "Phosphate Study at the Baltimore
Back River Water Treatment Plant", Report No. 17010
DF V09/70, Federal Water Quality Administration, 1970
11. Preising, C. P., Witherow, J. L., Lively L. D., Scalf,
D. D.,DePrater, B. L., and Myers, C. H. "Phosphate
-Removal by Activated Sludge", Federal Water Quality
Administration, Ada, Oklahoma, 1966.
12. Wells, W. N. "Differences in Phosphate Uptake Rates
Exhibited by Activated Sludges", Jour. Water Pollution
Control Federation, 41, 765 (1969) .
13. Olsen, D. O., and Connell, C. H. "Co-uptake and Co-
release of Phosphorus and Magnesium by Activated
Sludge", presented at the 69th Annual Meeting of Chem.
Engr., (May 17, 1971, Cincinnati, Ohio.
14. Bargman, R. C., Betz, J. M., and Garber, W. F. "Con-
tinuing Studies in the Removal of Phosphorus by the
Activated Sludge Process", Chem. Eng. Prog. Sym., No.
107, 117, 1970.
15. Horn, T. A., DePrater, B. C., and Witherow, T. L.
"Phosphate Removal at Fort Worth, Texas", Water Wastes
Eng., 6_, 7, 40 (1969) .
16. Connell, C. H. "Phosphorus Removal and Disposal from
Municipal Wastewaters, Texas City, Texas", Project No.
17010DyB, Federal Water Quality Administration,
Cincinnati, Ohio.
17. Levin, G. V. and Shapiro, J. "Metabolic Uptake of Phos-
phorus by Wastewater Organisms", Jour. Water Pollution
Control Federation,. 37, 807 (1975).
18. Shapiro, J. "Induced Rapid Release and Uptake of Phos-
phate by Micro-organisms", Science, 155, 1269 (1967).
19. Moore, H. G., Higgens, R. G. and Fruh, E. G. "Surplus
Phosphorus Uptake by Microorganisms - Batch Tests with
Dilute Activated Sludge Cultures", CRWR-1, U of Texas,
1969
82
-------
2.0. Fruh, E. G., Morgan, W. E., Moore, H. G., Jr., Higgens,
R. G. and Shahan, M. L. "An Enzymatic Technique to
Detect Surplus Phosphorus Uptake by Activated Sludge",
presented to the 5th Int. Water Poll. Research Conf.,
July 1970.
2J" Yall, J., Boughton, W. H., Knudsen, R. C. and Sinclair,
N. A. "Biological Uptake of Phosphorus by Activated
Sludge", Applied Microbiology, 20, 145 (1970).
22. Randall, C. W., Marshall, D. W., and King, P. H. "Phos-
phate Release in Activated Sludge Process", Jour. San.
Eng. Div., Proc. Amer. Soc. Civil Engr., 96, 395 (1970).
23. Boughton, W. H. , Gottfreid, R. J., Sinclair, N. A., and
Yall, I. "Metabolic Factors Affecting Enhanced Phos-
phorus Uptake by Activated Sludge", Applied Microbiology,
22_, 571 (1971) .
24. Menar, A. B., and Jenkins, D. "The fate of Phosphorus
in Sewage Treatment Processes—-Part II—Mechanism of
Enhanced Phosphate Removal by Activated Sludge", U of
California, SERL Report No. 68-8, August 1968.
2§. Ferguson, J. F., and McCarty, P. L. "The Precipitation of
Phosphates from Fresh Waters and Waste Waters", Tech.
Report No. 120, Stanford University, 1969.
26. Ferguson, J. F., Jenkins, D., and Stumm, W. "Calcium Phos-
phate Precipitation in Wastewater Treatment", Chem.
Eng. Prog. Sym., No. 107 (1970)
27. Ferguson, J. F., Jenkins, D., and Eastman, J. "Calcium
Phosphate Precipitation at Slightly Alkaline pH Values",
presented at the 44th Annual Conference of the Water
Pollution Control Federation, October 4-9, 1971, San
Francisco, California.
28• Malhotra, S. K. "Metabolic Storage of Phosphorus in the
Activated Sludge Floe", No. OWRR-A-015-WISC, U of Wis-
consin, July 1970.
29. Smith, I. W., Wilderson, J. F. and Duguid, J. P. "Volu-
tin Production in Bacterium Aerogenes Due to Nutrient
Imbalance", Jour, of Bacteriology, 68, 450 (1954).
83
-------
30. Duguid, J. P-, Smith, I. W. and Wilkinson, J. F.
"Volutin Production in Bacterium Aerogenes Due to
Development of an Acid Reaction", Jour. Pathol. and
Bacteriology 6_7_, 289 (1954) .
31. Wiame, J. M. "The Metachromatic Reaction of Hexameta-
phosphate", Jour. Am. Chem. Soc., 69, 3146 (1947).
32. Roinstad, F. A. and Yall, I., "Volutin Granules in
Zoogloea Ramigera", Applied Microbiology, 19, 973 (1970)
33. Manthe, R. M. "Uptake and Release of Soluble Ortho-
phosphate in an Activated Sludge Plant", unpublished
master's thesis, Marquette University, 1970).
34. Pfeffer, F. M., Scalf, M. R., De^rater, B. L., Lively,
L. D., Witherow, J. L. and Preising, C. P. "Phosphate
Removal by Activated Sludge - Amenability Studies at
Pontiac, Michigan", Federal Water Quality Administra-
tion, Ada Oklahoma, 1968.
35. Scalf, M. R., DePrater, B. L. Pfeffer, F. M. , Lively,
L. D., Witherow, J. L. and Preising, C. P. "Phosphate
Removal by Activated Sludge - Amenability Studies at
Mansfield, Ohio", Federal Water Quality Administration,
Ada, Oklahoma, 1968.
36. Myers, L. H., Horn, J. A., Lively, L. D., Witherow, J.
L. , and Preising, C. P. "Phosphate Removal by Activated
Sludge - Amenability studies at Indianapolis, Indiana",
Federal Water Quality Administration, Ada, Oklahoma,
(1968)
37. Sillen, L. G., and Marten, A. E., "Stability Constants
of Metal-Ion Complexes" Pub. No. 17, The Chemical
Society, London, 1964.
38. Freidman, B. A., and Dugan, P. R. "Concentration and
Accumulation of Metallic Ions by the Bacterium Zooglea",
Develop. Industrial Microbiology, 9^, 381 (1968) .
39. Peter, G., and Wuhrmann, K. "Contribution to the Pro-
blem of Bioflocculation in the Activated Sludge Pro-
cess", presented at the 5th Int. Water Poll. Research
Conf., July 1970.
40. Stumm, W. and Morgan, J. J.,Aquatic Chemistry^, Wiley
Interscience, 1970 583 pages.
84
-------
41. Harold, F. M. "Inorganic Polyphosphates in Biology:
Structure, Metabolism and Function", Bacteriological
Reviews, 3Q_, 772 (1966).
42. White, L. A., and MacLeod, R. A. "Factors Affecting
Phosphate Uptake by Aerobacter Aerogenes in a System
Relating Cell Numbers to P Uptake", Applied Micro-
biology, 2JU 520 (1971) .
43. Oulton, R. E. "The Mechanism of Phosphorus Removal by
Activated Sludge", unpublished master's thesis, Uni-
versity of Maine, 1969.
44. Schnitzer, M. "Organo-Metallic Interaction in Soils",
Soil Science, 103, 183 (1967).
45. Ritel, Wilhelm The Physical Chemistry of Silicates
U of Chicago Press,1954.
46. Biomedical Computer Programs. W. M. Dixon, Editor
Health Science Computing Facility, Department of
Preventative Medicine and Public Health. School of
Medicine, University of California, Los Angeles.
Revised September 1, 1965.
47. Sawyer, C. N. "Some New Aspects of Phosphates in Re-
lation to Lake Fertilization", Sewage and Industrial
Wastes, 24, 768 (1952).
48. Owen, R. "Removal of Phosphorus from Sewage Plant
Effluents with Lime", Sewage and Industrial Wastes,
25, 548 (1953) . .
49. Lea, W. L. Rohlich, C. A., and Katz, W. J., "Removal of
Phosphate from Treated Sewage", Sewage and Industrial
Wastes, 2_6, 261 (1954) .
50. Henrikson, A., ^Laboratory Studies on the Removal of
Phosphates from Sewage by the Coagulation Process",
Schweizerische Zeutschrift fur Hydrologie, 24, 253
(1969) and 2_5_, 380 (1962) .
51.. Gulp, R. L. "Wastewater reclamation by Tertiary Treat-
ment", Jour. Water Poll. Con trol Fed., 35, 799 (1963).
52. Malhotra, S. K., Ute, G. F., and Rohlich, G. A. "Nut-
rient Removal from Secondary Effluent by Alum Floccu-
lation and Lime Precipitation", Intl. Journal Air and
Water Pollu. 8_, 487 (1964) .
85
-------
53. Rolich, G. A. "Methods for the Removal of Phosphorus
and Nitrogen from Sewage Plant Effluents", Advances^
in Water Pollution Research, Pergamon PressT207fT964)
54. Gulp, R. Lf/ and Roderick, K. E. "The Lake Tahoe Re-
clamation Plant", Jour. Water Poll. Control Fed., 38
147 (1960).
55. Siechta, A. F. and Gulp, G. L. "Water Reclamation by
Studies at the South Tahoe Public Utility District",
Jour. Water Poll. Control Fed., 39, 787 (1967).
56. Van Vuuren, L. R. J., Stander, G. J., Henzen, M. R. ,
Meiring, P. G. J., and Van Bierk, S. H. V. "Advanced
Purification of Sewage Works Effluent Using a Combined
System of Lime Softening and Flotation", Water Re-
search, 1, 463 (1967).
57. Emery, M. M. and Malina, J. F., Mr. "Phosphate Removal
Characteristics of Lime Softening Sludge", Center for
Research in Water Resources, University of Texas,
GRWR 32 (1968).
58. Gulp, R. L. "Wastewater Reclamation at South Tahoe
Public Utilities District", Jour. Amer. Water Works
Assn., 60, 89 (1968).
59. Dryden, F. D., and Stern, G. "Renovated Wastewater
Creates Recreational Lake", Environ. Sci. and Technol.
2_, 269 (1968) .
60. Ames, L. L., Jr. "Evaluation of Operating Parameters
of Alumina Columns for the Selective Removal of Phos-
phorus as Calcium Phosphate", Department of the Inter-
ior, Federal Water Quality Administration, Rept. TWRC-8
(1968).
61. "Nutrient Removal and Advanced Waste Treatment",
Symposium held April 29-30, 1969 at Cincinnati, Ohio;
sponsored by the Federal Water Quality Administration
62. Lyman, B., Ettelt, G. and McAloon, T. "Tertiary Treat-
ment at METRO Chicago by Means of Rapid Sand Filtration
and Micro-Stramers"m Jour. Water Poll. Control Fed.
41, 247 (1969).
63. Nesbitt, J. B. "Phosphorus Removal—The State of the
Art", Jour. Water Poll. Control Fed., 41, 701 (1969).
86
-------
64. Mulbarger, M. C., Grossman, E. , III, Dean, R. B., and
Grant, 0. L., "Lime Clarification, Recovery, Reuse, and
Sludge Dewatering Characteristics", Jour. Water Poll.
Control Fed., 4^, 2070 (1969).
65. Eliassen, R., and Tchobanoglous, G. "Removal of Nitro-
gen and Phosphorus from Wastewater", Environ, Sci. and
Technol., 3_, 536 (1969).
66. Seiden, L. and Patel, K. "Mathematical Model of Ter-
tiary Treatment by Lime Addition", Department of the
Interior, Federal Water Quality Administration, Rept.
TWRC-14 (1969).
67. Smith, R. and McMichael, "Cost and Performance Estimates
for Tertiary Wastewater Treatment Processes", Department
of the Interior, Federal Water Quality Administration,
Rept. TWRC-9 (1969).
68. Farrell, J. B., Salotto, B. V., Dean, R. B., and Tolli-
ver W. E. "Removal Recovery", Chem. Engr. Prog. Symp.,
97, 232 (1969) .
69. O1Farrell, T. P., Bishop, D. F., and Bennett, S. M.
"Advanced Waste Treatment at Washington, D. C.", Chem.
Engr. Prog. Symp., 65, 97, 251 (1969).
70. Stamberg, J. B., Bishop, D. F., Warner, H. D., and
griggs S. H. "Lime Precipitation in Municipal Waste-
waters", Chem. Engr. Prog. Symp., 67, 107, 310 (1970).
71. Gulp, G. L. and Ames, L. L., Jr. "Phosphorus Removal
from Wastewaters by Activated Alumina Columns", Chem.
Engr. Prog. Symp., 67, 107, 304 (1970).
72. Rose, J. L. "Advanced Waste Treatment in Nassau County,
N. Y.", Water and Wastes Eng., 7_, 2, 38 (1970).
73. Berg, E. L., Brunner, C. A., and Williams, W. T. "
"Single Stage Lime Clarification", Water and Wastes
Eng., 7, 3, 42 (1970).
74. Davis, S. "Alternatives for Phosphate Removal", Water
and Sew. Works, 117, 336 (1970).
75. Covery, J. J. "Treatment Techniques for Removing Phos-
phorus from Municipal Wastewaters", Program No. 17010,
Department of the Interior, Federal Water Quality
Administration (1970) .
87
-------
76. Ames, L. L., Jr., and Dean, R. B., "Phosphorus Removal
from Effluents in Alumina Columns:, Jour. Water Poll.
Control Fed., 42, R161 (1970).
77. Tossey, D. F., Fleming, P. J. and Scott, R. F. "Terti-
ary Treatment by Flocculation and Filtration", Jour.
San. Eng. Div., Proc. Amer. Soc. Civil Engr., 96,
SAl, 75 (1970).
78. Brunner, D. R., and Spoul, 0. J., "Virus Inactivation
During Phosphate Precipitation", Jour. San. Eng. Div.,
Proc. Amer. Soc. Civil Engr., 96, SA2, 365 (1970).
79. Humenick, M. J., and Kaufman, W. J. "An Integrated
Biological Chemical Process for Municipal Wastewater
Treatment", presented to the 5th International Water
Pollution Research Conference (July 1970) .
80. Ulmgren, L., "Erfarenheter Fran Svenska Reningsveric
Med Fosforredoktion", Avdeelnings Director Statens
Naturvardsvenk, Solna, Sverige (Swedish) (1970) .
81. Bell, G. R., Libby, D. V. and Lord, D. T. "Phosphorus
Removal Using Chemical Coagulation and a Continuous
Countercurrent Filtration Process", Program No. 17010
EDO, Department of the Interior, Federal Water Quality
Administration, (1970).
82. Stukenberg, J. R., "Biological-Chemical Wastewater
Treatment", Jour. Water Poll. Control Fed., 43, 1791
(1971) .
83. Jenkins, D., Ferguson, J. F., and Menar, A. B. "Chem-
ical Processes for Phosphate Removal", Water Research,
5_, 369 (1971) .
84. Shindala, A., and Stewart, J., "Chemical Coagulation of
Effluents from Municipal Waste Stabilization Ponds",
Water and Sew. Works, 118, 100 (1971).
85. Process Design Manual for Phosphorus Removal, Environ-
mental Protection Agency, Office of Water Programs
October 1971).
86. Stumm, W., and Morgan, J. J., "Chemical Aspects of
Coagulation", Jour. Amer. Water Works Assn., 54, 971
(1962). —
-------
87. Stumm, W., Discussion in Advances in Water Pollution
Research, Pergamon Press, 2T6(1964).
88. Recht, H. L., and Ghassemi, M., "Kinetics and Mechanism
of Precipitation and Nature of Precipitate Obtained in
Phosphate Removal from Wastewater Using Aluminum and
Iron Salts," Report No. 17010EK10, Department of the
Interior, Federal Water Quality Administration (1970).
89. Tenney, M. W. and Stumm, W., "Chemical Flocculation
of Microorganisms in Biological Waste Treatment",
Jour. Water Poll. Control Fed., 3_7_, 1370 (1965).
90. Thomas, E. A. "Phosphate Precipitation in the Sewage
Treatment Plant of Uster and Removal of the Iron-
Phosphate Sludge (1970 and 1966)", Vierteljahrsschr.
Natf. Ges Zurich (German) 111, 309 (1966).
91 Popel, J. "Die Elimination von Phosphaten", Stuttgarter.
Perichte zur Siedlungswasserwirtschaft (German) No. 16,
Munchen (1966)
92. Earth, E. F., and Ettinger, M. B. "Mineral Controlled
Phosphorus in the Activated Sludge Process", Jour.
Water Poll. Control Fed., 39, 1362 (1967).
93. Eberhart, W. D., and Nesbitt, J. B., "Chemical Precipi-
tation of Phosphorus in a High Rate Activated Sludge
System", Jour. Water Poll. Control Fed., 40, 1239
(1968) .
94. Bargh, E. F., Brenner, R. C., and Lewis, R. F., "Chem-
cal-Biological Control of Nitrogen and Phosphorus in
Wastewater Effluent", Jour. Water Poll. Control Fed.,
4_G, 2040 (1968) .
95. Wuhrman, K. "Objectives, Technology, and Results of
Nitrogen and Phosphorus Removal Process", Advances in
Water Quality Improvement, U. of Texas Press, 21 (1968).
96. Earth, E. F., Jackson, B. N., Lewis, R. F., and Brenner,
R. C., "Phosphorus Filter", Jour. Water Poll. Control
Fed., 41, 1932 (1969),
97. Zenz, D. R. and Pivnicka, J. R., "Effective Phosphorus
Removal by the Addition of Alum to the Activated Sludge
Process", Proc. 24th Inde Waste Conf., Purdue Univ.,
Ext. Ser., 135, 273 (1969).
89
-------
98. Popel, J., "Phosphorus Removal with Ferric Iron and
Aluminum", Advances in Water Pollution Research, Perg-
amon Press, 643 (1969) .
99. Echelberger, W. F., Jr., and Tenney, M. W., "Wastewater
Treatment for Complete Nutrient Removal", Water and
Sewage Works, 116, 396 (1969).
100. Leary, R. D. and Ernest, L. A., "Municipal Utilization
of an Industrial Waste for Phosphorus Removal",
presented at the 32nd Porcelain Enamel Institute
Technical Forum, University of Illinois (1970) .
101. Connell, C. H., "Phosphorus Removal and Disposal
from Municipal Wastewaters, Texas City, Texas", Pro-
ject No. 1701oDYB, Federal Water Quality Administra-
tion, Cincinnati, Ohio.
102. Grigoropoulos, S. G. , Vedder, R. C., and Max, D. W.,
"Fate of Aluminum—Precipitated Phosphorus in Activa-
ted Sludge and Anaerobic Digestion", presented to the
43rd Annual Conference, Water Poll. Control Fed. (1970)
103. Mulbarger, M. C. and Shifflett, D. G., "Combined
Biological and Chemical Treatment for Phosphorus Re-
moval", Chem. Engr. Prog. Symp., 107, 107 (1970).
104. Directo, L. S., Project Engineer, Pomona Research
and Development Facility, Private Comminication (1971)
105. Earth, E. F., Private Communication (1971).
106. Unz, R. F., and Davis, T. A., "The Microbiology of an
Activated Sludge Wastewater Treatment Plant Chemically
Treated for Phosphorus Removal", presented to the 26th
Ind. Waste Conf., Purdue Univ., (May 1971).
107. McAchran, G. E. and Hogue, R. D., "Phosphate Removal
from Municipal Sewage", Water and Sew. Works, 118, 36
(1971).
108. Long, D. A., Nesbitt, J. B., and Kountz, R. R.
"Soluble Phosphate Removal in the Activated Sludge
Process—A two Year Plant Scale Study", presented to
the 26th Ind. Waste Conf., Purdue Univ., (May 1971).
109. Mulbarger, M. C., "The Three-Sludge System for Nitro-
gen and Phosphorus Removal", presented to the 44th
Annual Conference, Water Poll. Control Fed., (October
1971).
90
-------
110. Derrington, R. E., Stevens, D., and Laughlin, J. E.
"Enhancing Trickling Filter Plant Performance by
Chemical Precipitation" presented to the 44th Annual
Conference, Water Poll. Control Fed. (October 1971) .
111. Green, 0., Eyer, F., and Pierce, D., "Studies on Re-
movinal of Phosphates and Related Removal of Suspend-
ed Matter and Biochemical Oxygen Demand at Grayling,
Michigan", Division of Engineering, Michigan Depart-
ment of Health, and the Dow Chemical Company (1967).
112. Wukash, R. F., "New Phosphate Removal Process", Water
and Wastes, Engineering" 5_. 58 (1968).
113. Johnson, E. L., Beeghley, J. H., and Wukasch, R. F.
"Phosphorus Removal with Iron and Polyelectrolytes",
Public Works, 100, 11, 66 (1969).
114. Cherry, A. L., and Schuessler, R. G. , "Private Company
Improves Municipal Waste Facility", Water and Wastes
Eng. , 8_, 3, 32 (1971) .
115. Malhotra, S. K., Parillo, T. P., and Hartenstein,
A. G. "Anaerobic Digestion of Sludges Containing Iron
Phosphates", Jour. San. Eng. Div., Proc. Amer. Soc.
Civil Engr., 9_7_, SA5, 629 (1971).
116. Rudolfs, W., "Phosphates in Sewage and Sludge Treat-
ment—Effects on Coagulation, Clarification, and
Sludge Volume", Sewage Works Journal, 19, 179 {1947).
117. Buzzell, J. C., and Sawyer, C. N., "Removal of Algae
Nutrients from Raw Wastewater with Lime", Jour. Water
Poll. Control Fed., 3J), RI6 (1967).
118. Albertson, 0., and Sherwood, R., "Phosphate Extraction
Process", Jour. Water Poll. Control Fed., 41, 1467
(1969) .
119. Schmid, L. A. "Optimization of Phosphorus Removal
with Lime Treatment", thesis presented to the Univer-
sity of Kansas at Lawrence, Kansas, 1969, in partial
fulfillment of requirements for the degree of Doctor
of Philosophy.
120. Schmid, L. A., and McKinney, R. E., "Phosphate Re-
moval by a Lime-Biological Scheme", Jour. Water Poll.
Control Fed., 41, 1259 (1969).
91
-------
121. Spohr, H., and Talts, A., "Phosphate Removal by pH
Controlled Lime Dosage", Public Works, 101, 63
(1970) .
122. Tenny, N. W., Schelberger, W. F., Jr., Coffey, J. J. ,
and McAloon, T. J., "Chemical Conditioning of Bio-
logical Slucges for Vacuum Filtration", Jour. Water
Poll. Control Fed., 4_2_, 121 (1970).
123. Schmid, L. A., "Progress Report-Phosphate Removal
Demonstration Project", Kansas State University,
(November 1970).
124. Weber, W. J., Hopkings, C. B., and Bloom, R., "Physi-
cochemical Treatment of Wastewater", Jour. Water Poll.
Control Fed., 4_2_, 83 (1970).
125. Zuckerman, N. M., and Molof, A. H., "High Quality
Reuse Water by Chemical-Physical Wastewater Treatment"
Jour. Water Poll. Control Fed., 42, 437 (1970).
126. Molof, A. H., and Zuckerman, N. M., "High Quality
Reuse Water from a Newly Developed Chemical Physical
Treatment Process", presented to the 5th Int. Water
Pollution Research Conference (Jul-1970).
127. Bishop, D. F., O'Farrell, T. P., and Stamberg, J. B.,
"Physical-Chemical .Treatment of Municipal Wastewater",
Department of the Interior, Federal Water Quality
Administration (1970).
128. Tofflemire, T. J., and Hetling, L. J., "Chemical-
Physical Waste Treatment Phase I: The Low Lime Pro-
cess", presented to the 44th Annual Conference,
Water Poll. Control Fed., (October 1971).
129. Lin, S. S., "Removal of Phosphorus by Dosing Aluminum
to Activated Sludge", unpublished PhD. dissertation,
University of Washington (1972)
130. Kjeldaas, R., "Phosphorus Removal by Iron Salts",
unpublished master'sthesis, University of Washington,
(1972)
131. Kerhum, G. D., "Regeneration of Nutrient by Zooplank-
ton, Proc. Verb. Cons. Int. Explor. Mer., 153, 142
(1967) .
132. Yall I., and Sinclair, N. A., "Phosphorus Utilization
by the Microorganism of Activated Sludge", Chem. Eng.
Prog. Symp., 107, 95 (1970).
92
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-670/2-74-061
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
PHOSPHATE REMOVAL IN AN ACTIVATED SLUDGE FACILITY
5. REPORT DATE
August 1974; Issuing Date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Richard E. Finger, George J. Mason, Dale A. Carlson,
and Gary L. Minton
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORG \NIZATION NAME AND ADDRESS
Municipality of Metropolitan Seattle
410 W. Harrison Street
Seattle, Washington 98119
10. PROGRAM ELEMENT NO. 1BB043
ROAP 21-ASO / TASK 09
11. CONTRACT/GRANT NO.
17010 EDA
12. SPONSORING AGENCY NAME AND ADDRESS
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Research 1971 - 1973
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Biological and chemical means of phosphorus removal were studied at METRO
Seattle's activated sludge facility in Renton, WA. The studies ranged from jar tests
to full-scale plant manipulation. Only the secondary activated sludge system was
studied. The results of these studies indicate that the observed removal of phos-
phorus, in a soft water area activated sludge facility, is primarily biological in
nature. Some increase in removal can be encouraged by judicial control of organic
loading rates, air application rates and excess sludge wasting rates. However, even
under optimum conditions, this biological mechanism fails to consistently reduce the
phosphorus cone, to the desired level of >1.0 mg/1. Both ferric chloride and alum
are effective in removing phosphorus. Both can be controlled by automatic means.
Addition of either chemical produces more sludge. Alum is more effective than ferric
chloride in causing increased density of the final sludge. Initial capital costs for
either chemical are the same except for larger storage tank for alum. Chemical costs
vary with location of supplier. Alum costs at Renton plant are $16.80/MG and ferric
chloride costs are $34.20/MG. These figures represent a cost of $2.88 and $5.50 per
mg/1 phosphorus removed per million gallons for alum and ferric chloride, respectively,
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
*Activated sludge process, *Precipitation
(chemistry),*Phosphorus, Process control,
Iron, Aluminum sulfate, Cost comparison
Renton plant (Seattle),
*Automatic dosing system
13B
B. DISTRIBUTION STATEMENT
19. SECURITY CLASS (ThisReport)'
UNCLASSIFIED
21. NO. OF PAGES
101
RELEASE TO PUBLIC
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
.S. GOVERNMENT PRINTING OFFICE: 197l4-657-581|/5302 Region No. 5-1 I
------- |