THE FATE OF PHOSPHORUS INSEWAGE TREATMENT PROCESSES

Part III

Post-Aeration, Anaerobic Digestion, andPreliminary Clean SolutionPrecipitation Experiments

by

Arnold B. Menarand

David Jenkins

December 1970

Sanitary Engineering Research LaboratoryCollege of Enginee~ring

andSchool of Public HealthUniversity of California

Berkeley

SERL Report No. 70-12

TABLE OF CONTENTS

Page

LIST OF TABLES . V

LIST OF FIGURES . vi

Chapter

I. INTRODUCTION . 1

Scope of Report . . . 1

Acknowledgments . . . . . . 2

II. POST-AERATION EXPERIMENTS . 3

Objectives . 3

Experimental Facilities . 3

experimental Design . . 4

Experimental Results . 5

Discussion . 8

III. ANAEROBIC DIGESTION EXPERIMENTS . . . 17

Objectives . 17

Experimental Rationale . 17

Materials and Methods . 21

Results . . . . . . 23

Supernatant Phosphate Concentration . 23

Treatment of Diqester Supernatant . 29

High CaCO3 Feeding Study . 29

Conclusions . 36

iii

TABLE OF CONTENTS (continued)

Chapter Page

IV. PURE SYSTEM EXPERIMENTS 39

Objectives 39

Experimental Procedure 39

Results 40

Discussion 44

V. SUMMARY AND CONCLUSIONS 51

APPENDIX

Analytical Techniques 52

REFERENCES 54

iv

LIST OF TABLES

Table Title Page

I. Mean Operating Data for Post-Aeration System . . . 6

II. Experimental Activity Products for Post-AerationData 11

III. Diqester Supernatant Phosphate Concentrations . . 19

IV. Characteristics of Sludges from Conventional andCombined Chemical-Biological Phosphate RemovalSchemes 20

V. Analysis of Feed to Laboratory Diciester 24

VI. Average Diqester Performance . 25

VII. Mass Balance on Total and Dissolved Ca and P forHiqh CaCO3 Sludge Feeding Experiment . . . . . . 35

VIII. Experimental Activity Products for DigestionExperiment Data . 37

IX. Batch Distilled Water Precipitation Experiments —

Initial Conditions 42

X. Batch Distilled Water Precipitation Data . 43

LIST OF FIGURES

Figure Title Page

1. Schematic Diagram of Post-Aeration ActivatedSludge Plant 4

2. Effect of pH on Post-Aeration Dissolved Phosphate 9

3. Empirical Correlation of [CaIPT and pH forExperimental Data 12

4. Effect of Mg on Phosphate Residual at pH 9.5 —

Laboratory, Batch, Clean Solution Experiments.(After Ferguson, Jenkins and Stumm [41) 14

5. Distribution of Phosphorus in Secondary TreatmentFlows 18

6. Schematic Diagram of Laboratory Anaerobic Digester • 22

7. Supernatant Phosphate and Calcium Concentrations . 27

8. Phosphorus Flowsheet for Post-Aeration and AnaerobicDigestion with Supernatant Recycle 28

9. Effect of Batch Aeration on Composition of DigesterSupernatants 30

10. Effect of Aeration on Supernatant pH, Phosphateand ~Calcium Concentrations 31

11. Phosphorus Flowsheet for Post-Aeration and AnaerobicDigestion with Recycle of Aerated and SettledSupernatant 32

12. Effect of Ca, pH and Alkalinity on DissolvedPhosphate Content of Anaerobic Digester 34

13. Residual Dissolved Phosphate in theCa - ~T - CT - H~ - H20 - (Mg) Systems vs.pH (98-hr values) 41

14. The Effect of pH and Time on Phosphate Removal 45

vi

Li/\

8Hd

~ [~AOWe~J ~~qdsoq~ UO wn!.Seub~J j~o ~.Daj43 ~u. a~ Hd ~ L~AOW~ e~qdSOq~

uo UO~4~41A~DUO3 e~uoq~~ ~LO ~~JJ-3 ~1~J 91

917 uo~~3~dpe.Ad a~qdsoq~WflLDL~3 ~LO ~ UO 5ULpee~ 1~s,c~D 40 p8343 •9J~~

e~An6id

(penuL~uoD) S3NflE)Id dO ISI]

I. INTRODUCTION

SCOPE OF REPORT

This report will cover three aspects of the general topicof phosphorus removal in wastewater treatment processes. The reportwill summarize experimental work conducted between August 1968 andDecember 1969 under a Soap and Detergent Association Donation insupport of research entitled “The Fate of Phosphorus in SewageTreatment Processes.” This work involved investigations into threemajor areas:

1. Phosphate precipitation from activated sludge effluent bya post-aeration scheme. This research was undertaken todetermine whether it would be possible to take advantage ofchemical precipitation phenomena observed in the enhancedremoval of phosphate by activated sludge treatment in thedesiqn of a tertiary waste treatment scheme for phosphateremoval.

2. The fate of phosphate in diqesters with special referenceto the fate of phosphate introduced into the digesters aswaste activated sludge from processes achieving enhancedphosphorus removal (coprecipitation of calcium ohosphate inactivated sludge) or employing a post-aeration process forphosphate removal. This research effort resulted from theobservation that significant releases of phosphate typicallytake place from the solid to the liquid phase in anaerobicdiqesters. Consequently, recycle of a siqnificant fractionof the phosphate that enters the diqesters takes place whenthe dinester supernatant is returned to the head end of thetreatment plant. The effectiveness of enhanced phosphateremoval in an activated sludge aeration basin would bereduced if the diqester supernatant returns a large part ofthe “already removed” phosphate to the plant in the digestersupernatant stream.

3. Preliminary investigations on the influence of inorganicwastewater constituents on the rate and extent of calciumphosphate precipitation in model systems representative oftypical wastewaters. This work was of a preliminary natureand since further work by one of the authors of this reportand by others appears to have siqnificantiy advanced theknowledge of this area of research, these data (which inthemselves do not present an adequate description of thephenomena encountered) will be discussed in the light ofthese more recent experimental data.

1

~DUP~SLSS~ ~ SLL~ ~OJ.oo~ uiu~.~ ~j o~ uo~ LD~dd~ ~ sse~dx~ s.Aoq~.n~ eqj

o~bL~.s~AuI ‘~Lfl~~d p11~ AO3e~ALO~J.D9COAd SPM suujuec P~A~Q .AOSSe.J.O.Ad ‘~ABbPU~ ~.DecO~Ad S~M ~U9N

g p~ou~j ~ uo~~~ossv ~.u~b~Ae;ea pu~ d~o~ eiç~. wo~j uo~.~uop ~u~ pe~oddns S~M pu~ ~ j~o ~LS~AeALufl ‘~o~~~oqq 1p.Ai2ese~j

61u.A9eu16u3 I~~LU~ eq~ ~ ~~Dfl~UO~ S~M UO~~bi4S~AU~ e~j

S.LN3NE~G31MON)f3~

II, POST-AERATION EXPERIMENTS

OBJECTIVES

The objective of the post-aeration experiments was to determinewhether it would be possible to take advantage of a calcium phosphateprecipitation mechanism that had been suggested in studies of theenhanced removal of phosphorus by activated sludge (Menar and Jenkins[lj)in the design of a chemical precipitation process for phosphate removal.The basis for the design rested in two observations made during thetesting of the calcium phosphate precipitation mechanism for enhancedphosphorus removal by activated sludge. These were:

1. That increased activated sludge aeration rates caused increasesin pH because of reduction of mixed liquor CO2 concentrationsby CO2 stripping in the higher air flows, and

2. that removal of the precipitated calcium phosphate was madepossible by its association with the flocculerit activatedsludge matrix.

It was postulated that a post-aeration process, in whichsecondary effluent was aerated, in which chemicals could be dosed ifnecessary to precipitate calcium phosphate, and in which a small amountof return activated sludge was fed to aid in the flocculation of theprecipitate, would provide a more suitable environment for calciumphosphate precipitation and precipitate removal than would a typicalactivated sludge aeration basin. This is because in such a basinthe CO2 concentration would he lower than in the aeration basin andbecause in the sludge settled from such a basin the microorganismconcentration would be so low that CO2 production and therefore pHreduction in the sludge blanket at the bottom of a clarification basinwould not reach levels that would cause a significant release ofphosphate from the underflow into the treated effluent.

EXPERIMENTAL FACILITIES

To meet these objectives a pilot-scale activated sludgeplant with post-aeration followed by a sedimentation basin was constructed and operated on settled sewage from the Sanitary EngineeringResearch Laboratory primary sedimentation basin. The aeration, postaeration, secondary sedimentation, and final sedimentation basinswere constructed from plastic bins (see Figure 1). The aeration basinconsisted of three 30-gal bins in series (with an effective aerationvolume of 225~)and the secondary sedimentation basin consisted of one45-gal bin in which a conical concrete bottom had been poured.

3

4

2, 3. Aeration basins.Secondary sedimentation basin.Post-aeration basin.

7. Final sedimentation basins.Return activated sludge.Influent (S~RL primary effluent).Waste activated sludge.Final effluent.Activated sludge effluent.Activated sludge to post-aeration for precipitate collection.

The effective volume of the secondary sedimentation basin was 902.The post-aeration basin consisted of one 30-gal bin with an effectivevolume of 752; and the final sedimentation basin was constructed fromtwo plastic bins in series — the first was a 222. bin (effectivevolume, 402). The term “final sedimentation” in this report willrefer to the sedimentation basin that follows post-aeration.

Influent sewage, return, and waste activated sludge werepumped by peristaltic pumps; flows between the aeration basin compartments, the secondary sedimentation basin, and the final sedimentationbasin were accomplished by gravity. The secondary effluent was pumpedto the post—aeration basin using a submersible pump. Samples ofinfluent, secondary effluent, and final effluent were taken every15 mm by a sampling pump operated on a cycle timer which pumped thesamples directly into refrigerated containers. Air was supplied bya 1/3 hp compressor capable of delivering 5.9 cu ft/mm at 0 psig.When a greater air rate was required it was possible to supplementthis source with a house compressed air supply.

EXPERIMENTAL DESIGN

During each of the experiments the performance of theactivated sludge process was characterized by organic loading(lb COD applied/lb MLVSS-day), hydraulic residence time (hr),

13

Figure 1. Schematic Diagram of Post-Aeration Activated Sludge Plant

1,4.5.6,8.9.

10.11.12.13.

5

mean cell residence time (days), substrate removal efficiency (%),and phosphorus removal efficiency (%). The performance of the post-aeration basin was characterized b~ phosphorus removal efficiencyresulting from Ca(II) and 0H or H ion addition as liquid calciumchloride, sodium hydroxide, or hydrochloric acid, resDectively.

Analyses of dissolved and total calcium, dissolved and totalorthophosphate, pH, alkalinity, and dissolved oxygen were performedon samples of influent, activated sludge effluent, and post-aerationeffluent. Dissolved magnesium was determined only on the activatedsludqe effluent and the post-aeration basin effluent. The activatedsludge mixed liquor was characterized by analyses of MLSS and MLVSS.Analytical methods where not according to Standard Methods [2] aredescribed in the Appendix.

The combined activated sludge, post-aeration basin units wereoperated at a given condition for at least 24 hr before any sampleswere taken, This was sufficient time to establish steady conditionssince the activated sludge section was operated without disturbanceand the only changes in operation that were made were in the post-aeration section of the treatment plant train, These changes usuallytook the form of alterations in chemical doses, changes in concentrationof dosing solutions, and flow rates. Prior to the start of each setof operating conditions the post-aeration basin and final sedimentationbasins were drained,

EXPERIMENTAL RESULTS

The mean operating data for the seven experimental runs aresummarized in Table I. During each experimental run, dosing chemicalswere added constantly at the indicated doses,

The activated sludge section of the plant behaved satisfactorilyat all times with substrate removal rates varying from 0.24 to 1.15 lb~COD/VSS-day and effluent dissolved COD concentrations ranging from32 to 51 mg/z. The effluent dissolved phosphate from the activatedsludqe section of the plant ranged from 0.2 to 0.37 mM, Since highremoval of phosphate p~~se was not the major experimental objective,it was desired to determine what phosphate removals could be achievedat given chemical doses.

The nominal hydraulic residence time of the post-aerationbasin was 2 hr (range 1.6 to 2.6 hr). Air was supplied to the basinat the rate of 4 cu ft/gal liquid treated and the temperature wasmaintained constant at 70°F. Various combinations of chemicaladditions were made with the objective of determining independentlythe effects of pH and the calcium concentration on phosphorus residuals.

TABLE I

MEAN OPERATING DATA FOR POST-AERATION SYSTEM

_______ Experiment_Number

ACTIVATED SLUDGE SYSTEM 1 2 3 4 5 6 7

Substrate Removal Rate lb COD/lb I1LVSS-dayMean Cell Residence Time daysHydraulic Residence Time hrMLVSS mg/2.MLSS Volatile Fraction

Influent Total COD mg/J~.Effluent Dissolved COD mg/9.Influent Total P mMEffluent Dissolved P mMInfluent Total Ca mMEffluent Dissolved Ca mMInfluent Total Mg mMEffluent Dissolved Mg mMInfluent Total Alkalinity mMEffluent Dissolved Alkalinity mMInfluent pHEffluent pH

0.564.55.91210

84

20841

0. 3~0.340.680.65

2.31 .46.86.5

0.3810.75.71910

84

21740

0.380.330.630.6

1~2.31.26.86.7

0.563.84.91620

86

24951

0.420.380.610.65

0.742.21.47.06.5

1.153.77.7

33687

18447

0.340.300.660.69

0.82.21.87.17.0

0.267.16.81170

83

12645

0.230.200.810.82

1 .242.71 .67.77.4

0.227.21.82670

79

17738

O .41O .37O .80O .80

1.133.21.37.77.6

0.2619.05.82650

81

1 9332

0.380.340.750.71

1.112.91.37.67.6

TABLE I (continued)

______ Experiment Number

POST AERATION SYSTEM 2 3 4 5 6 7 —

Hydraulic Residence Time hr 2.0 1.9 1.6 26b 2.3 1.8 1.9Aeration Rate cu ft/gal 4 4 4 4 4 4 4Temperature °F 70 70 70 70 70 70 70

Ca added mM 1.65 1.56 1.79 2.49 2.55 2.59 2.42Ca dose (Ca added and mM 2.30 2.16 2.44 3.18 3.37 3.39 3.13

Ca present)OH or H+*dose mM 3.24 6.27 2.63 0.67 0.79 12.8 9.82Effluent Dissolved Mg mM -- -- 0.78 0.84 1.29 0.10 0.48Effluent Dissolved P mM 0.042 0.036 0.17 0.19 0.21 0.008 0.026Effluent Dissolved Ca mM 1.3 0.63 2.24 3.04 3.60 0.73 0.57Effluent Dissolved Alkalinity mM 1.7 1.7 1.5 1.8 0.47 1.1 2.2Effluent pH -- 9.0 9.75 8.0 7.6 6.8 11.4 10.5

Sludge Solids mg/2. 175 160 123 102 39 190 126Sludge P % as P 7.5 12.1 6.8 6.0 2.1 5.3 7.3Sludge Ca as Ca 20.6 18.2 11.8 11.9 13.8 24.7 26.0Sludge Ca/P mole ratio -- 2.2 1.2 1.3 1.5 5.1 3,6 2.8

Influent Ca/[lg mole ratio -- 0.6 0.6 0.9 0.9 0.7 0.7 0.6Post Aeration Reactor Ca/fig mole ratio 1.3 0.6 2.9 3.6 2.8 7.3 1.2

a Assume

b Evidence of some low pH precipitation at hiqhest Ca/Mg ratio

8

DISCUSSION

A discussion of the post-aeration experimental data will bemade from both the practical and theoretical points of view. Withreference to the former aspect, it is evident that additions of lime(or chemical additions to simulate an addition of lime) to achievea pH in the region of 11 (as is common practice in phosphate precipitation in tertiary treatment processes) is not necessary (in thisparticular sewage) to produce phosphorus residuals of <2 mg/R. (thefigure often quoted as being the residual dictated by a treatmentobjective of “80% P removal~”). In this instance it was possible toachieve P levels of <2 mg/~:at pH 9 and at a total calcium concentration of 230 mg/k as CaCO3 — a calcium concentration which included65 mg/z Ca as CaCO3 already present in the sewage and thereforerepresented an additional dose requirement of only 165 mg CaCO3/2,.If this calcium dose is compared to the calcium (or lime) dose requiredto reach pH 11 (which would be the dose used if a typical tertiaryphosphorus precipitation process were to be employed), it can be deducedfrom the data of Buzzel and Sawyer [3] that a lime dose of approximately 270 mg CaCO3/~ would be required to reach pH of about 11 atthis alkalinity (approximately 1.4 mM/i). Thus for meeting theobjective of <2 mg P/~ (whi;ch can be achieved at a pH of about 9instead of at pH 11) a saving in lime dose of approximately 100 mgCaCO3/~ can be realized. Moreover, the effluent pH is in a rangein which it could be possibly discharged without adjustment downward — a step that would almost certainly be required if the effluentpH were 11.

The post-aeration experiments on the SERL sewage did illustratethat P residuals of much lower than 2 mg/i could be achieved byincreasing pH values to above 9. Indeed it appeared that phosphorusresiduals (at the relatively constant calcium, alkalinity, and magnesiumconcentrations used in these experiments) were largely controlled bypH (Figure 2).

The use of a small amount of waste activated sludge (approximately 1% of the return activated sludge flow) or approximately70 mg/i on the basis of volatile suspended solids (VSS) concentrationin the post-aeration basin ~as a precipitate collector for the precipitatefrom the post-aeration basin, was successful. A sludge with excellentsettling characteristics was obtained at all pH values. Throughoutthe post-aeration experiments the final effluent suspended solidsconcentration averaged 55 mg/i. There was no detectable release ofsoluble phosphorus from thepost-aeration basin sludge into the finaleffluent following clarification of the post-aeration basin effluent.This indicates that pH decreases (if any) caused by CO2 productionfrom the activated sludge (precipitate collector) in the postaeration basin sludge were not significant enough to produce anydissolution of calcium phosphate.

9

0.15C)

C)

ci~

P

C)

CC-,)C-F)

1=

(t:~D

C,)ci)

~ 0.05

0.20

0.10

0.01

7 8 9 10 11

Figure 2. Effect of pH on Post Aeration Disso~v~d Pnosp hat:.

10

In experiments 1, 2, 6, and 7 the computed experimentalactivity products indicate that CaCO3 (calcite) precipitation is likely(Table II). The experimental data indicate that this indeed tookplace. Thus, in these experimental runs the sludge contained thehighest percent calcium; these experiments showed the highest post-aeration basin solids content and effluent dissolved calcium concentrations from these four experiments were the four lowest of the sevenexperimental series.

It is not possible to indicate the nature of the precipitatedphosphate phase. However, from activity product calculations (TableII) it can be seen that it is very unlikely that dicalcium phosphate(DCP) was the solid phase (at pH values >8) because in most of theexperiments the liquid was undersaturated with respect to DCP. However,with respect to crystalline octacalcium phosphate (OCP) and hydroxylapatite (HAP) significant degrees of oversaturation exist. Lesserdegrees of oversaturation exist with respect to tricalciuni phosphate(TCP) in most of the experiments. It should be realized that whilethese figures indicate the possibility of precipitation of variouscalcium phosphate phases, they are based upon activity productsderived from crystalline materials and it has been shown by numerousworkers such as Ferguson et al. [4] and Leckie and Stunin [5] thatcalcium phosphate precipitatTin under conditions typified by theseexperiments results in amorphous or poorly crystalline solids phasesand that the formation of crystalline solids is kin~tically controlled.

Notwithstanding these observations, the data obtained bythese experiments when plotted on the empirical correlation between[Ca]PT* and pH (Figure 3) represent a perfect fit to the line derivedfrom previous wastewater calcium phosphate precipitation experimentsby Plenar and Jenkins [1]. For data from the post-aeration experimentsbelow pH 8, this empirical correlation appears to follow the linethat would be predicted if DCP were the solid calcium phosphate phasecontrolling solution dissolved phosphate concentration. Interestingly,the activity product data also indicate oversaturation with respectto DCP below pH values of 8. It is suggested therefore that in thispH range DCP is the controlling calcium phosphate solid phase.

The chemical characteristics of the SERI. sewage treatedby post-aeration resemble those of the wastewater studied by Schmidtand McKinney [6] who used lime precipitation to achieve pH 9 to 10in primary sedimentation basins wi thout recycle of precipitatedsludge. These authors reported that with low lime doses (approximately150 mg Ca(OH)2/L) and in th& reported pH range, polyphosphatespresent in the raw wastewater inhibited CacO3 precipitation and redirected

~ IE[H?04] + [H2P01] + [HPO~] + [P0]

C,)

C~

rC

D—

Ic’

~~

u)~

’c-)

(-)c

-)-s

CD

~-~

-UtJ

c))w

0 CD c-f

-S.

c-f -o

-~

01

t’J

-~

JO

1O

~)

0

ocx

~C

D~

~ C)

c-t

V)

C)

-~

I—4

—I

-U CD

CD

—~

m ><

-U m ‘-4 m -I

I C)

—4

~~~4

—I

-< -V 3:,

CD

C)

CD

-4 (I)

~1

-V CD

C,)

—I

m --4

I—’

CD

—I

~-~i

O~

01

Exp

erim

ent

No.

CD

-~

~J

CO

coco

01

Co

a~C

DC

DpH

CD

CD

CD

CD

CD

01

CD

Co

-~

o~o-~

o-~—

.‘—

.iD

ical

cium

r\)C

o.-

ico

01

Pho

spha

te(D

CP)

~T

rica

lciu

m4

~C

~)0,

4~

(~)

~

(-a.

)C

O‘•—

1C

Dci

Pho

spha

te(T

CP)

~0

1~

-~

-~

-~

Oct

acal

cium

01

CD

CO

Ww

4~

C~

CD

CD

CD

Pho

spha

te(O

CP)

4~

01

4~.

~4~

4~

3~

-~

o~co

-~

Hyd

roxy

lap

atit

eCC

)—

i03

—1C

D01

(HAP

)

--4

I m I-f

I-i

WC

D~

(51

(51

O~

••

Calc

ite(C

aCO

3)Co

-~

I.

12

C

7

• Post Aeration Dataa Digester Experiment DataaData from SERL Report 68-6

V

40

n4a

.

I1•

I

V

ban r

10_s

—910 L

.

••17•.

a

.

6 8 9 10 11pHofFigure a. Empirical Correlation [Ca]P7 and pH for Experifliehtal Data.

13

the calcium to calcium phosphate precipitation. Their hypothesis wassupported by their experimental data which showed that the amount ofCaCO3 precipitation was far smaller than that predicted by equilibriumdata and by previous wastewater treatment experiences.

The observations of Ferguson etaL [4] on the effect ofmagnesium on calcium phosphate precipitation are also pertinent to theresults of the~ These authors showed inlaboratory batch clean solution experiments that waters with lowCa/Mg ratios (<2) and high Mg contents an increase of Mg concentrationat pH 9.5 tended to decreasethe phosphorus residual (Figure 4),It ~as postulated that Mg interferes with the precipitation of calciteand redirects the calcium to preferentially form a calcium phosphatephase. It is suggested that in a batch system at this pH, initiallyprecipitated calcium phosphate will redissolve and CaCO3 will precipitateover the first 1 to 2 hours, Combining this observation with the factthat Mg is known to increase calcite solubility (Chave eta], [7]) andto inhibit CaCO3 phase changes (Bischoff and Fyfe [8]) provides arationale for these phosphate residual results.

Examination of the post-aeration experimental data fromexperiments 6 and 7 indicate that these were the only two in whichMg removal was experienced —~ this is in accordance with previousobservations in which Mg(OH)2 precipitation is not detected untilthe pH is greater than 10.3. Unfortunately at the time that thesepost-aeration experiments were being conducted, the importance ofMg was not fully realized and the measurement of Mg concentrationwas not made in the first two experimental runs. However, it may beassumed from the behavior of Mg in similar experimental systems thatno precipitation of Mg(OH)2 occurred in these two runs at pH 9 and9,75,

Post-aeration with chemical addition has been demonstratedto be a process by which phosphorus may be removed from the effluentproduced by an already existing activated sludge plants Using limeas a dosing chemical at pH values of 10.5 it should be possible toattain soluble effluent phosphorus residuals of 0.5 mg PR using thisprocess.

Higher residuals (lower removals) can be obtained by operatingat lower pH values; for example theeffluent criterion of <2 mg PRcould be met with this sewage by dosing with lime to achieve a pHof 9 to 9.5. It must be emphasized, however, that effluent chemicalcharacteristics (e.g., hardness, Ca/Mg ratio, alkalinity) will havea determining influence on both the phosphorus residuals attainableat a particular pH value and upon the best strategy for phosphorusremoval (e.g., selection of dosing location, •dosing chemicals, pH,etc.). In this vein it should be recognized that the post-aerationexperiments conducted here were made on a sewage with a very lowCa/Mg ratio (0.5 to 1) and a moderate alkalinity (1 to 2 mM).

14

2

U

C-)

0

0.025

0.0?0

0.015

0.010

0.005

0

Figure 4. Ef~ect of Mg on Phosphate Residual at pH 9.5 —

Laboratory, BatchCiean Solution Experiments.After Ferguson, Jenkins and Stumm [4].

0 10 20 30 40 50Time, hr

0 10 20 30 4Q. 50Time, hr

15

When considering phosphorus ~removal~ strategy one must workwiflin three major constra4rits: l):’the pi’esent treatment capability,2) the future effluent requirements, and 3) the water chemistry.On~ can conceive :of two possible- situations where post-aeration withchemical addition might be considered as a process alternative forphpsphorus- removal .

1. Existing primary-secondary treatment plant with anaerobicdigestion on which an additional effluent requirement of2mgP/tis:imposed.

While post-aeration -might be considered for this treatmentsituation, it would -probhbly not cpmpete well with a scheme

• in which phosphorus precipitants were added directly to thetreatment units because P residuals of this magnitude canbernet in this fashion without construction of additionalbasins as -would be necessary for a post aeration scheme.If this is the case, then- lime and/or other chemicalsdesigned to form a-calcium phosphate precipitate would notbe the chemicals of choice because of the release of phosphatefrom calcium phosphates ~that takes place in the secondaryclarifier sludge blanket -and during anaerobic digestion.Alum or ferrous or ferric salts added to the aeration basinwould yield the: best results.

2. Existing primary—secondary treatment plant with anaerobicdigestion on which an additional effluent requirement of0.5 mg.P/t is imposed. .

• Here -the use of chemical addition to primary or secondarytreatment units cannot be relied upon to produce• the desiredeffluent phosphorus residual. Moreover, the attainment ofsuch a low phosphate residual by chemical precipitation inthe secondary treatment unit would raise the danger that.phosphate starvation conditions might result in the biological treatment reactor. for this case a tertiary precipitation/flocculation/sedimentation process must be consideredand a post-aeration basin might offer a convenient reactorin-which to conduct the precipitation flocculation steps.While the choice of precipitant and operating pH-will belargely dictated by the chemical composition of the waste-water and by local chemical prices, the post-aeration basinoffers the capability for using the phosphate precipitantsalready present in the wastewater, i.e.,the calcium hardness.For example, if the wastewater were of the same compositionas that at-San Antonio, Texas where Ferguson [9] has suggestedthat the low pH-precipitation (pH 7.7 te 8.5) of phosphateto dissolved risiduals of cl mg P/s takes place in theaeration basinly the formation of a carbonate apatite, itshould be possible to devise a post-aeration scheme that willremove phosphate to these low levels without the danger

16

of release from sludgein the secpndary clarifier. In sucha scheme the activated sludge systemwould be operated toproduce complete ox~datlon of ammonia. The production ofN0~ (a strong acid) from NH3 (a weak base) would result inthe titration of some alkalinity and thek’efore an alkalinityreduction during secondary treatment. Precipitation of calciumphosphate cduld•be avoided in the aeration basin by keepingaeration rates low enough to maintain pH levels below 7.5In the post-aeration basin the pH would be raised to 7.8 to8.0 by aerationand:the precipitated calcium phosphate wouldbe removedby enmeshment in a small amount of waste activatedsludge fed into the post-aeration basin. The substitutionof carbonate into the precipitated carbonate apatite wouldbe reduced since a significant reduction in alkalinity willhave been achieved durinq secondary treatment. In this case,phosphate removal would be achieved without chemical additionbut with only a judicious adjusttnent of pH because sufficientcalcium was present in the wastewater. However, in soft, lowMg, low alkalinity wastewater the addition of Ca may benecessary to reach low phosphate residuals. Post-aerationwith Ca addition (as CaC~9, not as lime) and pH coptrol toapproximately pH 8 would be an ideal process of phosphateremoval in this type of situation. The use of post-aerationin conjunction with chemical additives other than thosedesiqned to produce calcium phosphate precipitation shouldbe considered. For example, it is conceivable that postaeration would offer an ideal physical arrangement for thehomogenous generation of Fe(III) from Fe(II) as proposed forphosphate precipitation i~y the Dow Process (Wukasch [10])and demonstrated by Singer and Stumm [111 to provide a moreefficient scavenger system for phosphate than does the directaddition of Fe(III).

III. ANAEROBIC DIGESTION EXPERIMENTS

OBJECTIVES

The general objective of these experiments was to determinethe fate of phosphorus during ahaerobic digestion. Special attentionwas paid to the behavior of precipitated calcium phosphate introducedinto the digester in the sludge from a postaaeration precipitationbasin. Experiments were also conducted to determine whether recycleof phosphate in the supernataht from a digester treating a mixtureof raw sludge, activated sludge, and post-aeration sludge could bereduced by supernatant aeratioR followed by sedimentation for solidsseparation prior to its return to the headworks of the plant. Preliminarydata wev’p obtained on the effect of dissolved calcium concentrations onthe dissolved phosphate content of anaerobic digester liquors.Confirmation of the empirical, reletionship between [CaIPT and pHfound previously by Menar and Jenkin~ [1] for anaerobic, high CO2envirormients was sought.

EXPERIMENTAL RATIONALE

In a traditional primary-secondary sewage treatment flow scheme,the anaerobic digester is the receptacle for sludge streams from theprimary and secondary treatment sections of the plant. In an activatedsludge plant, these streams comprise the underflow of the primaryclarifier (raw sludge) and the waste activated sludge stream. Twoliquid streams emanate from the anaerobic digester — the digestedsludge and the supernatant. The latter stream is generally returnedto the headworks of the plant and recycled through the treatmentprocess. Recycling of the supernatant stream leads to recycling ofphosphate that has already been “removedt’ by incorporation into theprimary and activated sludges. The magnitude of this recycling hasbeen estimated by Jenkins et al. L121 to be SO lb/day for a typicalprimary-secondary (activatid’ iTudqe) anaerobic digestion combinationwhich receives 1000 lb P/day (Figure 5): In comparison with wastewater,anaerobic diqestor supernatant is a concentrated waste stream — severalworkers have reported that its phosphorus content appears to be inthe neighborhood of 50 to 200 mg P/i (Table III).

If phosphate removal schemes are superimposed upon, or appendedto, typical primary-secondary sewage treatment processes and the anaerobicdigester remains as a receptacle for sludge streams, then even morephosphorus will be channelled to the anaerobic digester. Table IVillustrates the estimates of Jenkins et al. [l2j on the magnitude ofthis increase in phosphorus load to till anaerobic digester when calciumøhosphate precipitation occurs in the activated sludge process. These

17

DigesterSupernatant

Diqested Sludge

n d a ryEffluent

Removal 20%

co

1000RawSewage

Overall Phosphate

Figure 5. Distribution of Phosphorus in Secondary Treatment Flows. (Figuresin lb/day)

19

TABLE III

DIGESTER SUP ERNATANT PHOSPHATE CONCENTRATIONS

Total P Ortho PReference mg/i mg/9.

~ Johnson et, aL

Wukasch, [10] 100-200 —-

Barth et al, [14] 170 —-

Barth and Ettiñger~ 5 10 -—

[15]

Anon. (1969) FMC lS-~l2OCorporation, [16] (mean, 40)

TABLE IV

CHARACTERISTICS~ OF SLUDGES~ FROM CONVENTIONAL AND COMBINEDCHEMICAL - BIOLOGICAL PHOSPHATE REMOVAL SCHEMES

Phosphorus Process Solid~ Total Organic P Volume

Process Influent Effluent #/MG Precipitate Volatile #/ Conc. 1000mg/9. mg/2. #1MG #/MG MG % gal/MG

Conventional 13 12 1250 -- 875 8 5 3

Primary + Act. Sludge

a) Conventional 12 10. 700 -- 500 17 1.5 5.6

b) High Rate 12 9 850 -- 650 25 1.5 6.8

c) Enhanced*Phosphate 12 1.2 1200 500 500 90 1.5 17.6Removal

*Removal mechanism assumed tobe calciumphosphate precipitation.

21

authors indicate an increase from about 20 lb P/FIG waste treated toabout 90 lb P/MG waste treated when phosphorus removal is increasedfrom that expected from a typical primary-secondary waste treatmentprocess train to that ashteved by calcium phosphate precipitation inthe activated sludge process to produce a phosphorus residual of about1.2 mg/s. (such as that encountered at San Antonio, Texas).

It was the intent of this research to determine whether it wouldbe possible to conduct phosphorus removal by chemical precipitation ina scheme such as that described in the post-aeration experiments earlierin this report, and still retain the anaerobic digester as a sludgetreatment unit. To accomplishthis, the anaerobic digester was operatedon mixtures of activated sludge, primary sludge, and post—aerationsludge that would typically originate from a primary-secondary (activatedsludge)-post-aeration waste treatment scheme. The levels of phosphorusand calcium in the sludge and supernatant were followed over a 7-monthperiod. When it became evident that significant phosphorus was beingcarried into the supernatant (and would hence be subject to recyclethrough the treatment processes) effort was devoted toward devisingtechniques for supernatant treatment to decrease this phosphorusrecycling.

MATERIALS AND METhODS

The anaerobic digestion experiments were conducted in a10-liter laboratory fermenter which had an effective volume of 8liters (Fi8ure 6). The digester was located in a constant temperatureroom at 37 C. The digester was mixed by a centrifugal pump thatoperated for 2 1/2 mm every 10 mm. The entire digester and itsappurtanences were mounted on a balance so that rapid determinationof the sludge mass could be made. The digester was fed through anopening in its top; sampling and digested sludge withdrawal wasconducted through a valve at the bottom of the unit. The gas producedduring digestion was collected above 25% H250., in a gas holder locatedoutside the constant temperature room and kept under a slight positivepressure.

The digester was fed at an organic loading rate of 0.02 lbvolatile matter (Vm)/cu ft-day with a mixture of raw sludge, activatedsludge and post-aeration sludge that would be typical of the mixturesencountered in a primary-secondary treatment plant followed by post•aeration for phosphate removal. The digester was fed three timesper week. Before each feeding the digester was mixed well. Feedingand digested sludge withdrawal were accomplished simultaneously,and after feeding and sampling the valves were closed, and thedigester was mixed for a further 2 to 3 mm. The performance ofthe anaerobic digester was characterized by organic loading(lb Vm/cu ft-day), hydraulic residence time (days), Vm destructionrate (lb Vm removed/cu ft-day), gas production rate (cu ft gasproduced/lb Vm destroyed), pH, alkalinity, and materials balances

~as Line

Reci rculLine

Reci rculati onp unip

Sampling Valve

Fr

V

—I—-

GasHolder

Feedinq Funnelvi

Figure 6. Schematic Diagram. of Laboratory Anaerobic Digester.

23

on calcium and phosphate. Gas analyses were conducted once per month.Analyses of total volatile solids, total calcium9 and total phosphatewere conducted on all feedings and withdrawals from the digester.Alkalinity and pH measurements were made on supernatant samplesobtained by se~tling the sludge withdrawn from the digester and settledfor 48 hr at 4 C. When it was desired to conduct analyses of dissolvedorthophosphate and dissolved calcium, digester supernatant sampleswere filtered through 0.45~i membrane filters (Millipore HA) prior toanalysis.

RES ULTS

The anaerobic digester was operated for a period of sevenmonths — between July 1968 arid Februar~’ 1969 at an organic loading of0.02 lb Vm/Cu ft-day and a hydraulic retention time of 67 days.During this period, two series of experimeftts were conducted: betweenJuly 1968 and January 1969 the digester wäs.operated at steady stateon a feed that consisted of primary, secondary,. and. post-aerationsludges. in the mixture described in Table V; from January 1969 toFebruary 1969 the digester was fed withpost-aeration basin sludgerich in calcium carbonate so that the effect of changes in pH andalkalinity on the distribution of phosph~te in the digester supernatantcould be studied. Throughout the entire experimental period the digesterperformed satisfactorily. Table VI illustrates •the performance parameters and shows that satisfactory materials balances for phosphateand.calcium were obtained. The ëffiOiency of volatile matter destrUctionaveraged 70.7%. Gas production was 10.9 cu ft/lb Vm destroyed andgas analysis indicated a ~qas composition of 60% CH~ and 30% CO2.Alkalinity averaged 3080 mg as CaC03/~. for the first part of the experimental period when the digester~was being fed a mixture of primary,

• secondary, and post-aeration sludges. During the latter part of theexperiment the use of solely post-aeration sludge as a feed produced

•a s~ignificant increase, in the alkalinity of the digester contents.These latter data are excluded from the computation of this meanalkalinity value. During the first part of the experiment the pHof the digester contents varied between 6.9 and 7.4 and averaged 7.1.

SUPERNATANT PHOSPHATE CONCENTRATION

As indicated previously, the digester contents were wellmixed so that separate supernatant and digested sludge streams couldnot be withdrawn from the digester. Consequently an arbitrary definitionof supernatant was made. The mixed digested sludge stream withdrawnfrom the digester was allowed to settle for 48 hr at 4°C and theliquid withdrawn from the surface of the settled mixture was definedas supernatant. Because of the extreme opacity of the mixture it wasnot possible to obtain an accurate estimate of the volume of supernatant and digested sludge fractions. Therefore in later computationsit is assumed that the digested sludge represents 30% of the digested

24

TABLE V

ANALYSIS OF FEED TO LABORATORY DIGESTER

Volumetric Total Total TotalConsti tuent Fraction,% SOlids, % P, mg/i Ca, mg/i

Raw Sludge 56 6.0 180 2.0

Activated Sludge 8 1.0 225 320

Post-Aeration 8 32.4 1040 12900Sludge

Distilled Water 28 0 0 0

Mixture 100 6.0 200 1150

/

25

Organic loading lbVm/cu ft/day

Vm destroyed, per cent

Hydraulic residence time, days

Gas production, cu ft/lb Vm destroyed

Gas production, lb gas/lb Vm destroyed

Gas analyses CH4, %

a’r~ 0/k..V2, /0

Air +H2, %

Alkalinity,mgCaCO3/~

pH

0.02

70.70

67

10.93

0.77

60.00

30.00

10.00

7.1

TABLE Vi

A~IERAGE DIGESTER PERFORMANCE

Phosphate Balance

(initial + added) g P 5.28

(removed + remaining) g P 4.59

balance, % of completionb 86.93

Calcium Balance

(initial + added) g CaCO3 74.88

(removed + remaining) g CaCO3 71.30

balance, % of completionb 95.23

aExcluding period of CaCO3 - rich sludge addition.

b~ balance 0 P removed +~ ainin x 100- P initial + P added

Ca balance, % = P~~x 100Ca initial + Ca added

26

sludge stream and supernatant represents 70% of this volume. Thisassumption is based on data from well-operated anaerobic digestionunits where volume reductions from incoming to digested sludge ofapproximately 70% are commonly experienced.

Using the experimental, definition of supernatant, it wasfound that feeding post-aeration sludge to the anaerobic digesterproduced significant increases in the phosphorus content of the supernatant, The diqested sludge supernatant phosphorus increased fromapproximately 20 mg P/si to an average value of 68 mg P19. for thatperiod of the experiment when the digester had reached steady stateconditions with respect to phosphorus concentration. This representsalmost a 3.5-fold increase in supernatant phosphorus concentrationover digested sludge not containing the phosphorus-rich post-aerationsludge. Reference to Figure 7 clearly shows the increase of supernatant phosphorus during the first residence time during which thedigester was fed post-aeration phosphorus-rich sludge. It appears thatthe calcium content of the supernatant remains fairly constant throughoutthe experiment — averaging approximately 450 mg CaCO3/9, (4.5 mM)throughout the experimental period.

These data demonstrate that feeding post-aeration sludge richin phosphorus to a digester will result in the dissolution of some ofthe calcium phosphate — a conclusion that is somewhat expected sincea reduction in pH values from those experienced in the post-aerationbasin (approximately 9 and greater) to the pH of approximately 7encountered in the anaerobic digestion process can be expected toincrease the solubiiity of calcium phosphate.

Based on the experimental data and the assumed volumes ofsupernatant and digester sludge, ‘it is possible to construct a materialsbalance for phosphorus circulation through a typical plant in whichprimary-secondary (activated sludge) post-aeration for phosphorusremoval and digestion with recycle of the supernatant to the primaryportion of the plant are included (Figure 8). From this flowsheetit is evident that the digester supernatant would almost double thephosphorus input to the treatment plant for the case considered (thatof a post-aeration process in conjunction with an activated sludgeplant that together effect an overall 80% phosphorus removal). Ifit is assumed that primary sedimentation will remove approximately thesame fraction of phosphorus from supernatant that it does from rawsewage (i.e., 10%) then Figure 8 shows that the flow of phosphorusthrough the secondary treatment plant and to the post-aeration unitis almost doubled. This imposes a great load on the post-aerationunit which without supernatant recycle (for this particular example)would be called on to remove ~600 lb P/day while with supernatantrecycle the removal of >1400 lb P/day is required.

CD (J,

C)

:3 :3 c-t

-‘3 o

3~

OCD

CD~

~)

(i~

:3 0~

C-,

-4 0 C.

3 C~) C :3 C-)

CD :3 -~ 0 :3 (1)

I— 7-’.)

CD

C’)

CD

CD

CD

CDCt

’0

):3

Z

C~

t30

)C

D4:

-0

)71).

C-)

0) c_)

01

CD

(0(4

)

3ii

(a

C),

CD 3 (a

Sup

erna

tant

Pho

spha

te,

mgP/

9.~

-‘

7’.)

C.)

3~..

0,

~-~

CD

C,

CD

C,

CD

CD

CD

7-’)

CD

~-.

A~

(0cz

i-’

CD

CD

CD

C’)

C Cr7

CD 03 CD

‘-0

7—.

CD

CD

7--

.

7--.

(71

CD

7--

.03 CD

01 CD CD

Sup

erna

tant

Cal

cium

,mq

CaCO

3/9.

Figure 8. Phosphorus Flowsheet for Post-Aeration and Anaerobic Digestion withSupernatant Recycle. (Figures in lb/day)

na 1Effluent

Sludge Overall Removal = 80%

29

TREATMENT OF DIGESTER SUPERNATANT

Since a considerable release of phosphorus to the supernatantstream was observed during digestion, the possibility of treating thisstream for phosphorus removal was investigated. Aeration was selectedas the method of treatment, It was reasoned that aeration of thesupernatant would be a simple technique for raising the pH of thisstream since it would provide a method of removing C02, which in atypical digester atmosphere is approximately 1000 times greater thanits concentration in air, Also, preliminary experiments with batchaeration of two supernatantsfrom local wastewater treatment plantshad shown that aeration caused rapid (within 1 to 2 hr) decreases insoluble orthophosphate, soluble calcium, and increases in pH (Figure 9).Thus, within 2 hr of aeration, decreases of 77% and 55% in solubleorthophosphate concentrations were obtained for the Stege and Richmond,Calif. sewage treatment plant digester supernatants, respectively.

In the laboratory, digested sludge samples were withdrawnfrom the digester and split into two aliquots. The first aliquotwas aerated with diffused air for 2 hr and then allowed to settle for48 hr at 4°C. The second aliquot was settled for 48 hr at 4°C withoutprior aeration. After settling, the supernatants from both aliquotswere analyzed for total phosphate and total calcium. The data,presented in Figure 10, show that pH values of the supernatant wereincreased from approximately pH 7 to approximately pH 9. For theperiod when steady operation had been achieved (after 67 days) calciumconcentrations decreased from an average of 405 to 265 mg as CaC03/~and phosphorus concentrations decreased on the average from 68 to37 mg P/k. These decreases represent a reduction in supernatantphosphate of approximately 46% — a lower decrease than that experiencedin the batch aeration experiments on Stege and Richmond supernatantssince these data represent reductions experienced in total phosphorusresidual while the Stege and Richmond aeration experiments refer onlyto dissolved calcium and phosphate concentrations. These data indicatethat significant recycling of phosphorus would still result even ifdigester supernatant aeration followed by sludge separation priorto recycle were practiced (Figure 11). Thus, the post-aerationtreatment system would be required to remove approximately 1100 lb P/daycompared with the 600 lb P/day removal necessary without digestersupernatant recycle. Thus even with aeration and sedimentation ofdigester supernatant prior to recycle the phosphate removal load onthe post-aeration system would be almost doubled,

HIGH CaCO3 FEEDING STUDY

During the latter part of the anaerobic digestion experiments(for a period of 42 days) an attempt was made to determine the effectsof variations in calcium and pH on the dissolved phosphorus concentrationin the digester. The purpose of these experiments was to determinewhether previous observations on the relationship between {Ca]PT and

E

C)

~2.0

C)

Steq e San i~Dj~rictSj~ernatant

9J

C.-)3.

2.

1.

District SuDernatant

00

8.6

1 2 3 45

8.2

I I

~

.4 I01 2 3 4 5

7

Richmond Sanitary

7

0.3

0.2

0.1E

ci

DissolvedOrthophosphate

30

0 1

1.0

9.

8.

8.ci

7.

7..2

I-)

0 2Aeration Time, hr

0 1 2

Figure 9. Effect of Batch Aeration on Composition of Digester Supernatants.

100

90

80

t7~ 70

~60

•:~ 50

~ 40L~.

30

20

600

500

~ 400cj

‘U

300‘U

200

I I ~ I I I I I I I

9.0~ -~

0 Before Aeration8.0 ~- G After Aeration

1 a i a I I I I I I

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180Time, days

Figure 10. Effect of Aeratio~n on Supernatarit pH,Phosphate, and Calcium Concentrations.

RawSewage

Figure 11.

~Overall Phosphate Removal 80%

Phosphorus flow Sheet for Post—Aeration and Anaerobic Digestionwith Recycle of Aerated and Settled Supernatant. (Figures in lb/day)

C.)r’.)

Supernatant Treatment

WasteActivatedSludge

.33

pH could be substantiated and to determine the possible effect ondissolved phosphate that might be caused by practices such as liminga digester.

To avoid disturbing the digester by wide arid sudden pH changesthat could be caused by the addition of strong acid or strong base, thepH, alkalinity, and calcium concentrations In the digester were adjustedsimultaneously by the addition of CaCO3, added in the form of post-aeration basin sludge rich in CaCO3. This sludge was identical incomposition to that.described in the previous section. Since theiinterest in these experiments was in dissolved phosphorus and calciumconcentrations, it was necessary to separate the solids from the sampleswithdrawn from the digester, prior to analysis. To accomplish thisthe samples of digested sludge were inmiediately centrifuged at 15,000rpn for 5 mi n and then filtered through 0. 46~ membrane fi 1 ters(:Millipore HA). During this operation an upward pH change occurredfdue to the equilibration of the liquid from the digester CO2 concentrationtoward the atmospheric CO2 con4ntration) especially during filtrationwhen a partial vacuum was applied to the sample. Because of the pHchange, two measurements of pH were made — one inmiediately on samplingand a second after centrifugation and filtration. The former pHvalue was used in assessing the effects of pH on dissolved phosphorusconcentrations in the sludge digester because previous experiments onthe batch aeration of supernatant from Stege and Richmond sewagetreatment plant digesters (Figure 9) had shown that rapid pH changestook place prior to any changes in soluble phosphorus and calciumconcentrations. The centri fugation and filtration operation took amaximum of only 10 mm and reference to Figure 9 shows that duringthis time period no significant changes in soluble phosphorus andcalcium took place although the pH changes were rapid.

The digester performed excellently during the six-weekexperimental period, produci ng 11 cu ft gas/lb Vm destroyed at anorganic loading of 0.03 lb V~cu ft-day.

During the 42-day experimental period the addition of CaCO3rich post-aeration sludge affected the composition of the digestercontents producing a steady rise in alkalinity (from about 1 nIl/i, toapproximately 14 mM/z) and in pH from 7 to 7.3. Over the first20 days of the experiment there was a three-fold increase in dissolvedcalcium (from 3 mM/z to 9 nIl/t) and this increase was accompaniedIw• a decrease in dissolved phosphorus concentration from approximately1.7 mM/i to 0.3 nt1/~ (Figure 12). The data in Figure 12 conthined withmaterials balance data for total and dissolved Ca and P (Table VII)make it plain that during the first 20 days of the experiment calciumphosphate precipitation was taking place. In the latter part of theexperiment it appears that considerable .CaCO3 precipitation began toczccur in conjunction with the calcium phosphate precipitation. Moleratios of Ca/P in the solid increased from 1.5 to 1.8 in the first20 days and from LB to 5.5 during the final 20 days of the experiment.

=7.1 a

2

cii4-icr5-caU,0-ca--ocii

0U,U,

C

34 7.4

7.3

7.2

7.0

6.9

2.0

1.0

Q

30

Ca, pH and Alkalinity onPhosphate Content ofDigester.

Figure 12. Effect ofDissolvedAnaerobic

TABLE VII

MASS BALANCE ON TOTAL AND DISSOLVED Ca AND PFOR HIGH CaCO3 SLUDGE FEEDING EXPERIMENT

DATE DIGESTER P CONTENT DIGESTER Ca CONTENT Ca/P,mole

Total Dissolved Solid Total Dissolved Solid Ratio1969 mg mg mg mM mg mg mg mM in Solid

22 Jan 2301 522 1779 58 11878 3228 8650 86 1.524 2305 469 1836 59 11955 3043 8910 89 1.527 2280 476 1804 58 11835 3185 8650 86 1.529 2309 363 1846 60 14669 4484 10200 102 1.431 2302 250 2152 69 16536 6019 10500 105 1.5

5 Feb 2341 162 2179 70 19010 6913 12100 121 1.77 2391 326 2065 66 20813 8490 12300 123 1.8

10 2445 114 2331 75 22502 8954 13500 135 1.814 2415 119 2296 74 23529 8515 15000 150 2.017 2419 143 2276 74 28779 5103 23700 237 3.219 2431 120 2311 75 33428 6981 26400 264 3.521 2444 138 2306 74 37125 3900 33200 332 4.524 2443 125 2318 75 40397 6900 33500 335 4.526 2446 98 2348 76 43452 5143 38300 383 5.028 2482 138 2344 76 46452 4759 41700 417 5.5

3 Mar 2581 112 2402 78 46655 4817 41800 418 5.4

CA)01

36

Activity product data for each of the 16 experimental points arepresented in Table VIII. On the basis of these data it appears thatdigester liquid is highly supersaturated with respect to calcjte andone would therefore expect that this solid would form. Of the calciumphosphate phases enumerated, only supersaturation with respect toapatite appears to exist.

Unfortunately, maqnesium data were not taken during theseexperiments because at that time the importance of this cation inphosphate precipitation was not fully comprehended. The absence ofabsolute values for magnesium and the conduct of the experiment overa somewhat limited pH range makes comparisons with other workers difficult. If one makes the reasonable assuniption that the Ca/Mg moleratio was 5, then this digester appears to be representative of themodel of Ferguson [91 in which the probabl:e solids forming would bepoorly crystalline hydroxylapatite (pA 52) and calcite (pA 8.3).

The digester data fall perfectly on the correlation curveof [CaJPT vs. pH (Fiqure 3) that was derived from previous work byMenar and Jenkins [1) and are in excellent agreement with previousdata from anaerobic and high CO2 environments.

CONCLUS IONS

The general conclusion from these experiments is thatcalcium phosphate precipitates formed in a post-aeration basin (andindeed in any primary, secondary, or tertiary precipitation scheme)tend to redissolve at the lower pH values encountered in anaerobicdigestion and are released into the supernatant stream. This conclusionis an extension of the observation of Menar and Jenkins [1] thateven in the slightly lower pH (higher CO2) environment of a secondaryclarifier activated sludge blanket, a redissolution of calcium phosphatestakes place. Also, release of phosphate has generally been observedfrom all activated sludges claimed to be showing “luxury uptake,”when the sludges were exposed to anaerobic conditions and a loweringof pH. This conclusion makes it plai:n that the anaerobic digesteris not a good unit for the treatment of high phosphorus activatedsludge or post-aeration sludge if digester supernatant recycle iscontemplated and if high phosphorus removals are desired from thewastewater stream. Aeration of the digester supernatant caused precipitation of phosphate by stripping CO~ and raising its pH, but significant amounts of phosphate would still be recycled even if perfectseparation of precipitated phosphate from the aerated supernatantwere achieved.

These experiments have also shown that an increase in calciumconcentration in an anaerobic digester leads to a decrease in dissolvedphosphate concentration. The exploration of this technique of phosphorusremoval from digester supernatant was not pursued to a degree that wouldallow the conclusion that it offered a satisfactory phosphorus removal

37

TABLE VIII

EXPERIMENTAL ACTIVITY PRODUCTS FOR DIGESTION EXPERIMENT DATA

Experimental Activity Producta

pHDCP TCP OCP HAP Calcite

7.05 7.0 28.4 47.7 51.3 3.77.00 7.1 28.9 48.3 52.3 3.87.03 7.0 28.5 47.9 51.6 3.77.14 6.9 27.8 47.0 50.4 3.57.00 7.2 28.9 48.4 52.2 3.27.00 7.4 29.2 49.0 52.7 3.47.16 7.3 28.5 49.1 51.4 3.07.20 7.7 29.5 49.5 53.0 3.17.40 7.3 28.2 37.8 50.7 2.97.35 7.4 28.4 39.0 42.0 3.07.27 7.4 28.7 48.4 51.7 2.97.22 7.5 28.9 48.7 51.9 3.27.20 7.9 30.1 50.3 54.1 2.97.25 7.9 30.2 50.5 54.3 2.8

a For data points in Figure 12.

Saturation Activity Products

DCP 6.9TCP 26.0OCP 57.8HAP 46.9Calcite 8.4

[~jbu~ddi~s~~cqL~AOW~~uoww~e~q~ue0~9’~~O~Hdeq~~o~.pu~(e~d~~~,cL,Aood~

,c~q~qo~.~d)~~qdsoqdwnpi~~~j.dpe~Ado~.~(Ho)~~/b~ew~xo~dd~40U0L~.Lpp~aq~.S9A~0AU~s~u~u~Aedns~e~seb~.pwo~j(~uoww~pus)

e~t~dsoqd3°L~A0W~Aeifl..A03pa~SebbnSaweq~s~Aaq~ou’~juo~.~jWflflD~A~cqpe~~des~qPIflOMSp~OSeq~.(1~M0~AbL~SA’~AD.Ao3ew~

uo~ue~.ap~ez~w~u~wo~.eLDI~De~Asp~.~osq~M)uo~~~dpa~Ad~(~o3440eA!.~APO~U0~J9~icci)8°~Hdeq~.~u~snppPpu~(ep~ojqD

wnp1~6u~.sn)9~0~0~4~A~OWd/~Dai~6u~4snçp~S9A~0AU~~-~dns~e~sa6j.pwo~.je~qdsoqd40uo~~.~dpe~d~J04[~}uosnb~ed

~cqpe~se66nsewaipsqDnseuoaA!.suadxeu~pu~~.uep~je~i.uawwe~~~weipe~w,c~ww~e~Se~.S~Ms~q;u~~i~dsoqd40SU0~~~—ueDuoDqb~.q~~a[q~.ssods~~.j.~s~~seq‘peepulsseDo~.~d

efq~.S~a4~S~Wfl~~Dq~M~.u~~u~edns~.~e;se6~p40~uew~e~eq~.cqL~A0W~e~qdsoqd~uo~u~.doeq~.40s~‘s~e~se15ipu~sn~oqdsoqd40

e~3aq~u0)fAOM9Ai.SUe~.Xee.AOwU~‘~AeAeM0q‘[iiuosn&tedenb~uqDe~.

8E

IV. PURE SYSTEM EXPERIMENTS

OBJECTIVES

The objective of these preliminary experiments on calciumphosphate precipitation in distilled water solutions was to providesome information on the effect of such common wastewater constituentsas H~, CT*, Mg, and CO2 on the extent of calcium phosphate precipitationat concentrations and Ca/P mole ratios typical of wastewater treatmentsituations.

EXPERIMENTAL PROCEDURE

Batch precipitation experiments were conducted at 30°C usinga 25O—m~ initial liquid volume contained in 300-m2. BOO bottles asreactors. The contents of the bottles were mixed by water-drivenmagnetic stirrers. In experiments where the liquid was aeration withvarious gases, these gases were introduced through glass diffusersimmersed in the liquid so that they allowed a small clearance abovethe magnetic stirring bar.

The reagents for an experiment were pre-mixed in a 400-m2,beaker which had been flushed out with nitrogen gas. Deionized waterwas used for all dilutions and reagent preparation. One molar stocksolution of CaC12, MgCl2, Na2HPO~, and NaHCO3 were prepared usinganalytical grade reagents. Working solutions of the followingconcentrations were prepared by diluting the stock solutions withdeionized water just prior to their use in an experiment to: CaC12,0.1 M/~; MgCl2, 0.1 M/~,; Na2HPO~, O~O96 M/2.; NaHCO3, 0.05 M/2.

The order of reagent addition was identical in all experiments:CaCl2 solution was always added first to vigorously stirred deionizedwater; when magnesium was used, the MqCl2 solution was added next,then the NaHCO3 solution, and finally the Na2HPO4 solution. The mixturewas stirred for 1 mm and then the initial pH’ was measured andrecorded. After pH measurement the mixture was transferred into theBOO-bottle reactor, which was then placed in a constant—temperaturebath where mixing was continued. If aeration was being employed, thediffuser was immersed at this time and aeration started. Aeration wasby cylinder gases of either air, air + O~5% CO2, or air + 5% CO2. Thereactor contents were sampled by withdrawing liquid with a 50-m9.

*CT = [H2CO3J + fHCO3] + [C0}

39

40

pipette. The sample was filtered immediately through O.4S~~ membranefilters (Mi]lipore HA) and pH, orthophosphate, and calcium weredetermined immediately on the filtrate. Samples for analyses weretaken after L5 hr, after 48 hr, and after 96 hr.

RESULTS

The results of these experiments were of a preliminarynature and were desianed to screen out significant effects on calciumphosphate precipitation of some of the common wastewater constituentssingly and in combination. These results will be discussed in thecontext of later work in this area by one of the authors and others(Ferguson etal. [4]; Jenkins etal. [12]).

The current experimental results are somewhat confoundedby the inability to control pH closely during a precipitation experiment. The buffering components of the mixture were often the reactantsin the precipitation and therefore as these buffers were removed themixture became subject to pH chahges. However, if one surveysTable IX and Figure 13 it is possible to make a significant observation from the 96-hr dissolved phosphate residual values. In theCa - ~T - CT - H~ - H20 - (Mg) system the only samples with extremelylow phosphorus residuals (<0.05 mM/k) are those in which Mg is absent.In only one instance does the presence of Mg appear to decrease thephosphorus residual. This was in experiment 13 which when comparedto experiment 20 only differed in the presence of 0.8 mM/2. Mg in theformer. In both of these experiments the initial composition(4 mM/a, Ca, 0.5 mM/9. ~ 5 mM/~ CT), the aerating gas (air + 5% C0~),and the pH profile (pH 8.0 initially and pH 7.1 to 7.2 finally) wereall the same. However, the P residual in experiment 13 after 96 hrwas 0.31 mM/i and in the Mg-containing experiment 20 the P residualwas 0.08 mM/i.

Extremely high carbonate concentrations (30 mM/2,) appearto suppress calcium phosphate precipitation (independently of affectingthe pH during precipitation). For example, comparison of experiment23 with experiment 21 shows that the phosphorus residuals in experiment23 (with an initial composition of 4 mM/i Ca, 0,5 mM/i ~ 30 mM/z CT,4 mM/i Mg, and an initial pH of 8.4) was 0.38 mM/k while in experiment21, in which the only difference in initial conditions was the pH of8 and CT of 5 mM/2~ there was a P residual of 0.18 mM/i. The pA valuefor CaCO3 of 5,9 and the residual. calcium concentration of 1.4 mM/iindicate the occurrence of CaCO3 precipitation in experiment 23 butnpt in experiment 21 (residual Ca = 3.6 mM/i).

A similar pair of experiments (experiment 24 and experiment22) in which initial Mg was 8 mM and initial Ca/Mg mole ratio was 1/2does not confirm this behavior. In this instance it is possible thatthe very low Ca/Mg mole ratio was sufficient to almost completelyinhibit calcium phosphate precipitation but to allow CaCO3 precipitationto take p~ace in the case of experiment 24 with the high initial CTconcentration

41

0.50 ng~~______________ Initial P __________________

0.40L

0.30

0.20

~0.10. _______I~ v•. ~‘~~V\ \ % ‘~%~ \\ •~ ‘ S

~ 0.091’ indicates

- magneslum-iree

0.07 system.

0.06

0.05

j~ Experiment number0.04 ~ is indicated at

12 ~ phosphorus residual~ value.

0. 03 \\\\\\ \\\ \\ \ \

00.02

6.5 7 pH

Figure 13. Residual Dissolved Phosphate in theCa - ~T - CT - - H20 - (Mg) Systemsvs. pH (98—hr values).

42

TABLE IX

BATCH DISTILLED WATER PRECIPITATION EXPERIMENIS — INITIAL CONDITIONS

initial COfld~t1OflSa

Expt.No. Ca P CT Mg pH [ Gas

1 4.0 0,5 0 0 7.4 NONE2 4,0 0,5 0 0 7.5 NONE3 4.0 0.5 0 0 8.0 NONE4 4.0 0.5 0 0 8.1 NONE5 4.0 0.5 0 0 8.1 NONE6 4.0 0.5 0 0 7.5 AIR + 0.5% CO27 4.0 0.5 0 0 7.4 AIR + 5% CO28 4.0 0.5 1.0 0 8.2 NONE9 4.0 0.5 5.0 0 8.0 NONE

10 4.0 0.5 1.0 0 8.2 AIR + 5% CO211 4.0 0.5 5.0 0 8.8 AIR12 4.0 0.5 5.0 0 7,9 AIR + 0.5% CO213 4.0 0.5 5.0 0 8.0 AIR + 5% CO214A 4.0 0.5 5.0 4.0 8.0 AIR + 0.5% CO214 4.0 0.5 0 4.0 7.6 NONEl.A 4.0 0.5 1.0 4.0 7,8 NONE15 4.0 0.5 0 4.0 7.6 AIR + 0.5% CO~16A 4.0 0.5 1.0 4.0 7.8 AIR + 0.5% CO216 4.0 0.5 1.0 0.8 7.6 AIR + 5% CO217A 4.0 0.5 1.0 4.0 7.7 AIR + 5% CO217 4.0 0.5 1.0 2.0 7.7 AIR.+ 5% CO218 4.0 0.5 1.0 2,0 7.7 AIR + 5% CO219 4.0 0.5 1.0 2.0 8.0 AIR.+ 5% CO220 4.0 0.5 5.0 0.8 8.0 AIR + 5% CO221A 4.0 0.5 5.0 4.0 8.0 NONE21 4.0 0.5 5.0 4.0 8.0 AIR + 5% CO222 4.0 0.5 5.0 8.0 8.1 AIR + 5% CO223 4.0 0.5 30.0 4.0 8.4 AIR + 5% CO224 4.0 0.5 30.0 8,0 8.3 AiR + 5% CO2

a All concentrations are mM,

CDFFC)

LU

><Q

_

LU_JLU

cc~F-

~CDLU-j

-JFC,)

CDC)

Fcc

2(t~U)

04~)

‘I)C.)

0.

C.)

(t3

a)oc~

cccJaD

cL

ccc

~—

~—

cccccc

r~r-~

r~

ccc’~i

ccc~

C)

ccC

)ccc~

i~

cO

~-o

Dc~

)~

,~

dd

dd

dd

dd

dd

oddcc~

cC

Dccd

cd

d

~~

cccc~

~cc

C)

~~~

~a)

~-

r~~

-cc

cocc

C)

~r~

cc~

c-C

)cc

c-,

(\j

cccc

~-

t~cc

ccc~i

CD—

C)

COCO

CD0

)C\J

CDcc

cca

)0~

~~

~C

DC

DC

DC

DC

DC

DC

DC

DC

DC

DC

DcD

CD

CD

CD

CD

CD

CD

CD

CD

cDC

DC

DC

DC

DC

DC

DC

DC

D4~)

~~

ccLO

ccC

’C\J

OD

O)

cc~

a)c

cccr~

CD

co~

-C)c

Dr—

cccoC

Dccc)c

occ

C)cccccccc~

cccccc~

~

~cc

COC

OO

)r-~.

cc0

)cc

r-

~-

N.

0~cc

CD

COCO

N.

COcc

~~

—

s_c.

II

.I

•I

II

I~

cccc

N.

N.

N.

cccc

N.

C)

t-~CO

r-~N

.CO

N.

N.

ccN

.cc

ccaj

~—

_cc~

cc~

oD

ccccC

D~

tcoo)o

)ccccc)c

cccC

)0)

cccc

—~

-~

fCO

~-

c’~~C

)~

-~

-‘~f

cc~

cc~

COc..

•I

I•

•I

•I

II

I~

CD

CD

CD

CD

C)

CD

CDC

DC

DCD

CD

CD

CD

CD

CD

CD

CD

CD

CD

CD

CD

~cc

~~

N.c

ccccccccc

oD

~—

o)c

co-)c

oC

Da)

r.cc

•~:j

(__)

.1

II

.•

I•

II

II

•~cc

cccc

cccc

cccc

cccc

ccc~j

cccc

cccc

cc~

j-cc

cc~

-cc

0V

~4.~=i:

•a:c.

cc~

ccccN

.coo)C

Dcc~

cccccccc~

cco)

CD

r-~-c~

jcc~zI

>(r-

—i—

z—r

rr

s—r

r1

~C

dC

dC’.]

Cs.]C

s]C

s]LU

44

DISCUSSION

The results obtained in these preliminary experiments aresuggestive of data obtained recently by Ferguson etal. [4] in astudy of calcium phosphate precipitation in batch systems with chemicalcompositions similar to those studied in the preliminary experimentsand encountered in many wastewaters. In thqsç experiments it was foundthat the effect of pH in the Ca -, ~T — CT - W - H2O system wascomplex (Figure 14). While the general effect of increasing pHwas to decrease residual soluble phosphate, induction periods werelong at low pH values (pH 7.6 and pH 8). These induction periodscould be eliminated if previously precipitated calcium phosphate solidswere present (Figure 15) and were therefore most probably due tothe nucleation of poorly soluble phases followed by subsequent rapidcrystal growth. At pH 10 and 11 there was a very rapid precipitationof most of the phosphate, then the soluble phosphate concentrationincreased to a value that after about 1 hr remained virtually constant.Electron micrograph data indicate that this kinetic pattern was causedby an instantaneous precipitation of a calcium phosphate followed bya slower conversIon to calcite which is accompanied by the re-solutionof some of the initially precipitdted calcium phosphate — presumablya poorly crystalline carbonace—apatite

The effect of carbonate concentration on calcium phosphateprecipitation appeared to be exerted both kinetically and on the valueof the long term residual soluble phosphate (Figure 16). As carbonateconcentration is increased through the range coasnonly found in waste-waters, slower phosphate removal occurs and significantly higherphQsphate residuals are present at high carbonate levels even afterseveral hundred hours.

While the effect of carbonate on phosphate solubility isnot fully understood, it is believed that the occurrence of calciumcarbonate precipitation above pH 9 and the substitution of carbonatein apatite (increasing its solubility with respect to phosphate) areinvolved.

This work indicated that the errect of maqnesium on theprecipitation of calcium phosphate was exerted through its influence onthe nature of the calcium-containing solids but not through theformation of magnesium ccrbunate or phosphate precipitates. At pH8, Mg retarded the rate of phosphate precipitation and resulted inmuch higher phosphate residuals in samples that contained high Mgconcentration (Figure 1,’) This oehavior is indicated in experimentsconducted in the prel imi nary study reported herein and may be due tothe inhibition of nucleation and crystal qrowth — an interpretationthat has been suggested for simildr inhibitory effects of Mg in CaCO3

•phase changes. Magnesium may also ~ause a chdnge in the nature of thecalcium phosphate solid For exampie, Fergdson [9] has proposedthat in the presence of Mg, p tricalcium phospha.e can be precipitatedin systems that otherwise would precipitate hydroxylapatite. The

.24

.22

.20

.18

.16

E .14

.08

.06

.04

.02

—

_~~: TT

1 hr -

.1 -~

~ ~J____L.___J_.._‘~ 7 8 9 10 11

pH

Initial Conditions:

CT 3.8 mMCa 2mM

0.25 mM

Temp. 26 ± 1°C

\

125 hr

9

0 1 8 27 64 125 216

Time, hr

Figu~e 14. The Effect of pH and Time on Phosphate Removal.

343(31

0.25

0.20

0.15

0..1O

0.05

0.00

Time, hr.

Initial Conditions: Ca, 2 mM; CT~ 5 mM;

pH, 7.6; Temp. 25 ± 1°C

Indicates Addition of Crystals to Solution ofApproximately the Original Composition.

E

0

4-,

a.’C.-)

0C.-)

a)•1-~

CL

0

0 50 100 150 200 250 300 350 400

Figure 15. Effect of Crystal Seeding on Rate of Calcium Phosphate Precipitation.

0.24

0.22

0.20

0.18

~ 0.16

E

~; 0.14(e

C

cL

0.10

0.08

0.06

0.04

0.020

Tine, hr

Figure 16. The Effect of Carbonate Concentration on Phosphate Removal at pH 8

20 40 60 80 100 120 140

co0.24

0.22

0.20

0.18

E

~ 0.160)

Clv~ 0.140

0.12

0.10

0.08

0.0620 40 60 80 100 120

Time, hr

Figure 17. The Effect of Magnesium on Phosphate Removal at pH 8.0.

49

former solid is slightly more soluble with respect to phosphate andcalcium than the latter and would therefore result in higher phosphateresiduals in the presence of magnesium.

At pH values where calcium carbonate precipitation takesplace in the Ca - ~T - CT - H+ - H20 - Mg system, magnesium exerts arather different effect than that observed at pH 8 (Figure 4). Inthe experiments of Ferguson etal., [4] at pH 9.5, phosphate residualsincreased rapidly from a very low initial value to a concentrationthat appeared to be dependent on the Mg concentration. These resultswere interpreted to indicate that calcium phosphate and calciumcarbonate both precipitated initially. Then the initially precipitatedcalcium phosphate redissolved and calcium carbonate precipitated overthe first 1 to 2 hr. Since it is known that Mg increases the solubilityof calcite and inhibits calcium carbonate phase changes, the presenceof Mg in a system in which calcium phosphate and calcium carbonateboth precipitate would tend to produce lower P residuals than a systemin which Mg was absent (and in which phases change between calciumphosphate and calcium carbonate proceeded uninhibited).

It is possible to discuss the existing and proposed phosphateprecipitation practices in the context of the research conducted byFerguson etal. [4]. Phosphate precipitation by the addition of limeis practiced almost exclusively at pH values in excess of 10.5.Phosphate residuals of less than 0.5 mg P/z are common, but chemicaldoses are directly related to wastewater alkalinity and generallyexceed 400 mg/a. of lime. The work of Ferguson eta]. indicates thatwhere P residuals of less than 0.3 mg P12. are desired this high pHprecipitation process would probably operate most efficiently at shortcontact times (<0.5 hr) and with rapid solids separation devices suchas centrifugation followed by filtration. Since calcium carbonateprecipitation will take place at these high pH values a large fractionof the sludge from such processes will be calcium carbonate and theeconomic feasibility of recalcining for lime recovery should be considered on a case by case basis. The high pH lime addition processwould be a serious candidate to obtain low P residuals in wastewaterswith a high carbonate content (300mg as CaCO3/2.) and a high Mgconcentration (>24 mg/9..) since, at~pH values in excess of 11, thedeleterious effects of these components on calcium phosphate precipitationappear to be relatively small.

Schmidt and McKinney [6] have recently suggested thatphosphate precipitation can be conducted in primary sedimentationbasins by the addition of lime to achieve a pH value of 9 to 10.The results of the work of Ferguson etal. [4] sugqest that precipitation at this pH would be a preferred process where Ca/Mg ratiosare high (Ca/Mg <2) i.e., typically in an area with hard water andhigh Mg. This process would also be applicable where recycle ofprecipitated solids was impractical. At these pH values calciumcarbonate precipitation will take place and where high wastewater

050

jcarbonate concentrations exist excessive calcium carbonate precipitation may interfere with calcium phosphate precip-itation. Solids’separation in such a process should be rapid at these pH values becausephosphate release from precipitated solids occurs during the firstfew minutes after chemical addition. It would appear that solidsrecycle which seems to be an integtal part of the PEP process(Albertson and Sherwood (17]) may be desirable’ for flocculation butis probably not optimum for phosphate precipitation at pH values. of•• -

9tolo. ‘ UThe research of Ferguson et al. further helps to define the

conditions that were necessary for ca’TEium phosphate precipitation:at p14 values consistent with the activated sludge process. Thesuggestion that enhanced phosphate removal by activated sludge atSan Antonio, Texas is caused by calcium phosphate precipitation(Menar and Jenkins (1]) has been reinforced. The long nucleationtimes observed for calcium phosphate precipitation at low p11 valuesmay account for the absence of phosphate removal at” many locations,.while the recycle of precipitated 4oitds. with return activated sludge(eliminating the nucleation lag, time) may explain the effectivenessof the process where it does occur. The hindering effects of Ca andMg on calcium phosphate precipitatibn at low pH values may accountfor some of the differences in phosphate’ removalS observed at differentlocations; for example, the hi gh cqncentrations of these two componentshave been suggested by Ferguson (9)’ to be .the reason Why it ‘was’impossible for Menar and Jenkins (1] to obtain very low solublephosphate residuals in their activated sludge experiments at San Ramon,Calif. (1]. This is very possibly the same’reason that i.twas impossibleto obtain extremely low phosphate residuals in the SERL post~aerationexperiments discussed earlier in this report” Ferguson’~~,al:’ (4].suggest that j*osphate precipitation at pH’ values of 7.5 tciT.5 ‘ismost readily applicable in soft water areas, and in such areas it canbe used in co’mbination with activated sludge”treatment or ‘in a post—aeration system. In a typical low~hardriess”wastewater (e;g., Ca, 0.8 flt1/L;Mg, 0.4 rrtl/g.; alkalinity, 150 mg as CaCO3/t;’ bicarbonate ‘3 wills.; andphosphate, 0.3 mM/s.) the addition of 1 to 2 mM/s. Ca’ and 0.3 to 1 mM/s.strong base would raise the pHto slightly above pH8 and the’calcii.miconcentration to about 2 mM/s.. Under. these conditions, nucleationshould take place rapidly (within hours) and recycle of solids shouldresult in rapid crystal growth and effective phosphate removal tolow residuals (approximately 0.3 nIl/s.). Sludge production should besmall since calcium phosphate .prectpitation should take place in theabsence ,of calcium carbonate formation. .‘ ‘ - .

[

LC

V. SUMMARY AND CONCLUSIONS

The following conclusions can be reached as a result of, theresearch conducted here and later work discussed in conjunction withthe currently reported work:

1. It has been demonstrated that post-aeration with the useof a small amount of waste activated sludge as a precipitatecollector in the post-aeration basin is a viable processfor phosphate removal following the activated sludge process.

2. It has been dei~onstrated that the anaerobic digester wouldcause a significant recycle of phosphorus when used inconjunction with activated sludqe and a post-aeration processfor phosphorus removal.

3. It has been demonstrated that anaerobic digester supernatantaeration causes rapid precipitation of phosphate but thatseparation of the precipitated material is difficult andsignificant phosphorus recycle will still result even ifsupernatant aeration and separation of precipitated solidsis conducted.

4. It has been demonstrated that dosing of the digester supernatant with lime will cause decreases in the recycling ofphosphorus but a significant amount of recycling will stilloccur. In view of these findings it is recommended that theanaerobic digester not be considered as a sludge treatmentdevice if it is desirable to produce low (<2 mg PR) phosphorusresiduals and if it is impossible to eliminate digestersupernatant recycle.

5. Preliminary experiments on clean solutions have indicatedthat pH, carbonate, and magnesium have complex effects onthe rate and extent of cajcium phosphate precipitation atwastewater treatment concentrations. These data have beendiscussed in the light of later work by Ferguson etal.[4) and lead to the suggestion of new techniques for theremoval of phosphate from wastewater by calcium phosphateprecipitation as well as contributing information that helpsto explain previously observed results from such phosphateremoval processes.

51

APPENDIX

ANALYTICAL TECHNIQUES

Soluble Calcium, Soluble calcium was determined by EDTAtitration using hydroxynaphthol blue as indicator [18).

Total Calcium (sludge samples). Total calcium was determinedon a sample after ashing. A suitable sample volume (usually 20 m~)was evaporated to dryness on a steam table. The dried sample wasashed at 800°C for 30 minutes. The residue was heated on a steamtable with 10 m~ of 2N HC1 for 10 minutes and then transferred to a25O-m~ beaker using deionized water. The sample was neutralized toa pH of 7.0 with 0.lN NaOH and titrated with EDTA using hydroxynaphtholblue as indicator.

Chemical Oxygen Demand, COD was determined by the method givenin Standard Methods 12th Edition [19] using a silver sulfate catalyst.

Magnesium. Magnesium was det~rmined by the difference betweenthe total hardness and the calcium concentrations.

pj~. The pH measurements were made usinci a Beckman ExpandomaticpH meter.

Sludge Volume Index (SVI). SVI of the mixed liquor was determinedusing the method given. in Standard Methods l~th Edition [19].

Suspended Solids (sewage samples). Suspended solids weredetermined by the membrane (Millipdre) filter technique of Winnebergeret al. [20).

Suspended Solids in Activated Sludge. Suspended solids inactivated sludge were determined by filtration through a Whatman GF/Cglass filter followed by drying at l05°C:and weighing of the residue(Jenkins [21)).

Total Solids in Raw Sludge. Total solids were determinedgravimetrically on raw sludge following blending of the sludge maWaring blender,

Total Phosphate. Total phosphate was determined on a sampleafter alkaline ashing, A suitable sample volume (usually 10 m~)was treated with 1 m~ of 5% MgCl2• 6H2O solution and evaporated todryness on a steam table. The dried sample was ashed at 800°C for1 hr. The residue was heated on a steam table for 30 mm with 3 m~

52

53

strong acid molybdate solution and after cooling was transferred toa l00—m~ volumetric flask. Aminonaphthol sulfonic acid (4 ms~) wasadded and the volume made up to 100 m~. The absorbence of themolybdenum blue was read at 630 m~i after 6 mm using a 1-in cell ina Bausch and Lomb Spectronic 20. The concentrations of all reagentswere as given in Standard Methods 10th Edition [22].

Soluble Phosphate. Soluble phosphate was determined by filteringthe sample through washed O.4.5-p membrane filters and then proceedingas for the total phosphate.

Orthophosphate. Orthophosphate was determined using the aminonaphthol sulfonic acid (ANS) method outlined in Standard Methods10th Edition [22] with the following modifications:

1. Absorbence was read at 630 mu instead of 690 mu,

2. Three ms~ of strong-acid molybdate was used instead of 4 m~.

3. Six minute reaction time was used instead of 10 mm. Thetwo latter modificationswere made to reduce the possibilityof polyphosphate hydrolysis during the orthophosphate analysis.

Ortho + Hydrolyzable Phosphate. The sample was placed in a250-me beaker, 8 m~ of 8N H2S04 were added and the mixture diluted to40 m~. with distilled water. Two Hengar selenized granules were added~the beaker covered with a watch glass, and the mixture boiled gentlyfor 40 mm. After cooling, the sample was neutralized with 8 N NaOHto the phenolphthalemn end point. A few drops of 4.5N H2S04 wereadded until the pink color disappeared. The sample was then transferredto a l00-m~ volumetric flask and 3xn~. strong acid molybdate solutionwas added and mixed by swirling. ANS reducing agent (4 mz) was addedand mixed and then the mixture diluted to 100 m~ with distilled water.The absorbence was read after 6 mm on a Bausch and Lomb Spectronic 20at 630 mu using a 1-in cell. Reagent concentrations were as describedin Standard Methods, 10th Edition [22].

Total Hardness. Total hardness was determined by EDTA titrationusing eriochrome black T as indicatbr [23].

Volatile Suspended Solids in Activated Sludge. Volatilesuspended solids inactivated sludc~e were determined by combustionof the solids retained by a Whatman GF/C glass filter at 500°C for30 mm [21].

Diqester Gas Analysis. The amounts of CH4 and CO2 in the digestergas were determined by gas chromatography [24].

REFERENCES

1. Menar, A. B. and D. Jenkins. The Fate of Phosphorus in SewageTreatment Processes. Part II — Mechanism of EnhancedPhosphate Removal by Activated Sludge. SERL Report 68-6.Berkeley: Sanit. Eng. Research Lab., University of California.

2. Standard Methods for the Examination of Water and Wastewater,13th Edition. New York: American Public Health Association,1965.

3. Buzzel, J. C. and C. N. Sawyer. “Removal of algae nutrient fromraw wastewater with lime,” J. WPCF, 39, Rl6 (1967).

4. Ferguson, J. F., D. Jenkins, and W. Stumm. “Calcium phosphateprecipitation in wastewater treatment,” In Press. Water - 1970.American Institute Chemical Engineers.

5. Leckie, J. 0. and W. Stumm. Phosphate Precipitation. ProceedingsConference on Advanced Waste Treatment, University of Texas,Austin, Texas (1969).

6. Schmidt, L. A. and R. E. McKinney. “Phosphate removal by a lime-biological treatment scheme,” J. WPCF, 41:1259 (1969).

7. Chave, K. E., K. S. Deffeyes, P. K. Weyl, R. M. Garrels andM. E. Thompson. Science, 137:33 (1962).

8. Bischoff, J. L. and W. F. Fyfe. Amer. J. Sci, 266:65 (1968).

9. Ferguson, J. F. The Precipitation of Calcium Phosphates fromFresh Waters and Wastewaters. Ph.D. Thesis, StanfordUniversity. December 1969.

10. Wukasch, R. F. Phosphate Removal at Grayling, Mich. and LakeOdessa, Mich. Presented to Michigan Water Pollution ControlAssoc., Annual Conference, June 16—19, 1968.

11. Singer, P. C. and Stumm, W. Unpublished data.

12. Jenkins, D., J. F. Ferguson and A. B. Menar. “Chemical processesfor phosphate removal,” .In Press. Water Research. (1971).

13. Johnson, E. L., J. H. Beeghiy, ~and R. F. Wukasch. “Phosphorusremoval with iron and polyelectrolytes,” Public Works, 66 (1969).

55

56

14. Barth, E. F., B. N. Jackson, R. F. Lewis and R. C. Brenner.“Phosphorus removal from wastewater: direct dosing ofaluminate to a trickling filter,” J.WPCF, 41:1932 (1969).

15. Barth, E. F. and M, B, Ettinger. “Mineral controlled phosphorusremoval in the activated sludge process,” J. WPCF, 39:1362(1967). —

16. Progress Report Phase I. Development of process to removecarbonaceous, nitrogenous and phosphorus materials fromanaerobic digester supernatant and related process streams.Environmental Engineering Department, FMC Corporation.(1969).

17. Albertson, 0, E. and R. J. Sherwood. “Phosphate extraction process,”J, WPCF, 41:1467 (1969).

18. Jenkins, D. Analytical Methods. Berkeley: Sanit. Eng. ResearchLab., Univ. of Calif., March 1966.

19. American Public Health Association. .Standard Methods for theExamination of Water and Wastewater, 12th Edition, 1965.

20. Winneberger, J. H,, J. H. Austin, and C. A. Klett, “Membranefilter weight determinations,” J. WPCF, 35:807 (1963).

21. Jenkins, D. “An improved volatile suspended solids method,”Water and Waste Treatment J., London, England, Aug-Sept.1962.

22. American Public Health Association. Standard Methods for theExamination of Water and Wastewater, 10th Edition, 1955.

23. American Public Health Association. Standard Methods for theExamination of Water and Wastewater, 11th Edition, 1960.

24. Chan, D. B. The Hydrolysis Rate of Cellulose in AnaerobicFermentatTàn. SERL Report 70-3. Berkeley: Sanit. Eng.Research Lab,, Univ. of Calif.