EPA-R2-72-064
DECEMBER 1972
Environmental Protection Technology Series
Calcium Phosphate Precipitation
in  Wastewater Treatment
                                  Office of Research and Monitoring

                                  U.S. Environmental Protection Agency
                                  Washington. DC. 20460

-------
            RESEARCH REPORTING SERIES
Research reports of the  Office  of  Research  and
Monitoring,  Environmental Protection Agency, have
been grouped into five series.  These  five  broad
categories  were established to facilitate further
development  and  application   of   environmental
technology.   Elimination  of traditional grouping
was  consciously  planned  to  foster   technology
transfer   and  a  maximum  interface  in  related
fields.  The five series are:

   1.  Environmental Health Effects Research
   2.  Environmental Protection Technology
   3.  Ecological Research
   4.  Environmental Monitoring
   5.  Socioeconomic Environmental studies

This report has been assigned to the ENVIRONMENTAL
PROTECTION   TECHNOLOGY   series.    This   series
describes   research   performed  to  develop  and
demonstrate   instrumentation,    equipment    and
methodology  to  repair  or  prevent environmental
degradation from point and  non-point  sources  of
pollution.  This work provides the new or improved
technology  required for the control and treatment
of pollution sources to meet environmental quality
standards.

-------
                                                          EPA-R2-72-064
                                                          December  1972
                CALCIUM PHOSPHATE  PRECIPITATION
                    IN WASTEWATER  TREATMENT
                               by

                          A.  B. Menar
                          D.  Jenkins
                       Grant #17080 DAR
                        Project  Officer

                      Warren A.  Schwartz
           U.S.  Environmental Protection Agency
           National  Environmental  Research Center
                   Cincinnati, Ohio  45268
                           for the

            OFFICE  OF RESEARCH AND  MONITORING
          U.S. ENVIRONMENTAL PROTECTION AGENCY
                  WASHINGTON, D.C.   20460
For sale by the Superintendent ol Documents, U.S. Government Printing Office, Washington, D.C. 20402 - Price $1.60

-------
                       EPA Review Notice
This report has been reviewed by the Environmental Protection Agency,
and approved for publication.  Approval  does not signify that the
contents necessarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commercial  products
constitute endorsement or recommendation for use.

-------
                              ABSTRACT
This investigation examined the precipitation of calcium phosphate both
from chemically-defined solutions representative of wastewater composition
and from wastewater.  The steady state solid phase that controlled
dissolved phosphate residual was an amorphous tricalcium phosphate.   The
solubility of this phase, determined from chemically-defined systems,was
used with success to predict dissolved phosphate residuals  from both
chemically-defined systems and wastewaters.  Suspension recycle was found
to result in lower dissolved phosphate residuals, but poor  suspension
settling below pH 10 made this process difficult to maintain.   Suspension
settling was enhanced by Mg(OH)2 precipitation but not by CaC03 precipi-
tation.  In wastewater of moderate alkalinity and hardness, a phosphate
removal in excess of 80% was consistently achieved at pH 9.5 with  lime
doses of, at the most, 200 mg/ji as CaC03.  The overall phosphate removal
performance was dictated by the performance of the precipitation reactor
and its ensuing sedimentation basin.  Phosphate-containing  particles  that
escaped sedimentation could not be removed by filtration because they
dissolved rapidly during the recarbonation process that necessarily
precedes the filtration step.

This report was submitted in fulfillment of Project Number  17080 DAR
under the sponsorship of the environmental Protection Agency by the
Sanitary Engineering Research Laboratory, University of California,
Berkeley, California 94720.
                                   in

-------
                              CONTENTS

Secti on                                                        Page
    I    Conclusions 	    1
   II    Recommendations  	    5
  III    Introduction   	    7
   IV    Phosphate Removal from Wastewaters by Calcium
           Phosphate Precipitation	    9
    V    Mechanism of Calcium Phosphate Precipitation	    15
   VI    Approach Rationale   	    21
  VII    Experimental Equipment and Procedures 	    23
 VIII    Precipitation  of Calcium Phosphate in
           Chemically-Defined Systems   	    31
   IX    Precipitation  of Phosphate in Wastewater  	    57
    X    Acknowledgments	    71
   XI    References	    73
  XII    Glossary	    77
 XIII    Appendices	    79

-------
                                FIGURES

                                                                Page

 1    Lime Requirement to Reach  pH 11  as  a  Function  of
         Wastewater Alkalinity 	     10

 2    Schematic Flowsheet for Experiments on  Chemically-
         Defined Systems	     24

 3    Schematic of Calcium Phosphate  Precipitation with
         Suspension Recycle  	     25

 4    Treatment Systems  Used in  Studies at  SERL  Pilot Plant   .     27

 5    Dependence of Steady State PT Concentration on
         Solution Composition and pH  in Batch  Reactors  ....     33

 6    Dependence of Steady State Pj Concentration on
         Detention Time  in CSTR  Experiments	     34

 7    Effect of pH on pA  for Various  Calcium  Phosphate
         Solids	     36

 8    Effect of pH on pA  of Tricalcium Phosphate	     37

 9    Critical  Ionic Concentrations in Batch  and CSTR
         Experiments	     38

10    Effect of Mg on pA  Ca3(PQk)2	     40

11    Influence of Mg on  Steady  State Concentrations of
         Ca, Mg, CT, and  PT at pH 9.5	     42

12    Effect of CT on pA  Ca3(POl|)2	     44

13    Effect of CT on Steady State Concentrations of
         Ca, Mg, and Py  and on Amount of  Precipitated
         Carbonate at pH  9.5	     45

14    Effect of Cj on Steady State Concentrations of
         Ca, Mg, and Py  and on Amount of  Precipitated
         Carbonate at pH  11.0	     46
                                                             i   i
15    Effect of pH on Ca/Pj Mole Ratio of the Precipitate
         and on Carbonate Incorporation  Into  the Precipitate  .     47

16    Effect of Suspension Concentration  on Effluent
         Phosphate Concentration in Precipitate  Recycle
         Experiments	     53


                                   vi

-------
                                                                Page

17    Steady State Ca and Cj Concentrations  During
         Precipitate Recycle Experiments  ............   53

18    Effect of Suspension Concentration  on  Rate of Calcium
         Phosphate Precipitation in Batch Experiments   .....   54

19    Effect of Suspension Concentration  on  pA Ca3(POlt)2
         in Precipitate Recycle Experiments   ..........   55

20    Phosphorus Removal Performance of the  Entire Process
         Train During Pilot Plant Runs on Wastewater ......   60

21    Comparison of Lime Slurry Addition  to  Dissolved
         Lime Concentration in Wastewater Experiments   .....   62

22    Change in Dissolved Ca and Py Concentrations in
         Reactor Compartments - Wastewater Experiments  .....   63
23    Dependence of pA Ca3(POif)2 on pH f°r Wastewater
         Experiments ......................   64

24    Settling Properties of Calcium Phosphate-Carbonate
         Suspensions (Series I)  ................   67

25    Settling Properties of Calcium Phosphate-Carbonate
         Suspensions (Series II) ................   68

26    Critical Ionic Concentrations as Determined by Walton
         et jj]_. and by This Investigation  ...........   84

27    Effect of Unit Processes on pH ..............   90

28    Effect of Unit Processes on Total Phosphate
         Residuals (Runs 1, 2, and 3)  .............   91

29    Effect of Unit Processes on Total Phosphate
         Residuals (Runs 4 and 5)  ...............   92

30    Effect of Unit Processes on Dissolved Phosphate
         Residuals (Runs 1 and 2)  ...............   92

31    Effect of Unit Processes on Dissolved Phosphate
         Residuals (Runs 3, 4, and 5)  .............   93

32    Effect of Unit Processes on Dissolved Calcium
         Concentration (Runs 1, 2, and 3)  ...........   94

33    Effect of Unit Processes on Dissolved Calcium
         Concentration (Runs 4 and 5)  .............   95

34    Effect of Unit Processes on Alkalinity ..........   96

                                   vii

-------
                                TABLES

                                                                Page

 1   Component Concentrations  in Wastewater
        (From Ferguson [3])	    21

 2   Initial  Composition of  Batch Systems  Used  to
        Define Time to Reach  Steady State  	    32

 3   Initial  Concentrations  in CSTR Experiments  to
        Determine Effect of  Magnesium 	    41

 4   Initial  Conditions for  CSTR Experiments  to  Determine
        Effect of Carbonate  (CT)  	    43

 5   Initial  Conditions for  CSTR Experiments  to  Determine
        Effect of Carbonate  on Phosphate and  Other
        Component Residuals  	    43

 6   Comparison of Predicted  and Experimental Residual
        Phosphate Values in  Chemically-Defined Systems   ....    50

 7   Comparison of Predicted  and Experimental Residual
        Phosphate Values in  Chemically-Defined Systems
        Corrected for Calcium  Carbonate Precipitation  	    51

 8   Operating Conditions and  Results for  Wastewater
        Experiment 1   	    58

 9   Operating Conditions and  Results for  Wastewater
        Experiment 2	    58

10   Operating Conditions and  Results for  Wastewater
        Experiment 3	    59

11   Operating Conditions and  Results for  Wastewater
        Experiment 4	    59

12   Operating Conditions and Results for  Wastewater
        Experiment 5	    60

13   Comparison of Predicted and Average Experimental
        Dissolved Phosphate  Residuals for  Wastewater
        Experiments	    63

14   Initial  Component Concentration for Precipitate
        Separation Experiments, Series I   	    66

15   Initial  Component Concentration for Precipitate
        Separation Experiments, Series II  	    66

                                  vi ii

-------
                              SECTION  I

                             CONCLUSIONS
The general objective of this investigation was to develop  a method of
predicting the performance of phosphate removal processes that  involve
the precipitation of calcium phosphate.  The development of this  predic-
tive method was based on a steady state precipitation model.  The model
was developed by conducting experiments in chemically-defined systems and
then examining the utility of these results in predicting dissolved
phosphate residuals obtained by precipitation of calcium phosphate from
wastewaters.

It became evident as the work progressed that a major determining factor
in the efficiency of calcium phosphate precipitation processes  was the
ability to separate the insolubilized phosphate.  The objectives  of the
work were therefore expanded to include an examination of some  of the
chemical factors that influenced precipitate separation by  sedimentation,
and of some of the process considerations that determine the efficiency
of particulate phosphate removal.
 CHEMICALLY-DEFINED SYSTEMS
The effect of common wastewater components (calcium, carbonate,  magnesium)
and precipitation conditions (pH and reaction time) on dissolved phosphate
residual was investigated in chemically-defined systems in both  CSTR  and
batch reactors.  At chemical component concentrations  typical  of waste-
waters, steady state levels of dissolved phosphate were reached  in
continuous flow reactors after nominal residence times of 10 min and
were maintained for at least 200 min - an average residence time range
typical of wastewater precipitation processes.

The data from chemically-defined systems suggested that the nature  of
the steady state phase was one with the stoichiometric and solubility
characteristics of tricalcium phosphate Ca3(POit)2(pA = 23.56).   This
solid was amorphous and did not have a distinct X-ray diffraction
pattern.  However, later calcium phosphate precipitation experiments  in
chemically-defined solutions in which precipitate recycle was  conducted
showed the presence of a solid with the X-ray diffraction pattern of
tricalcium phosphate (Ca3(P04)2-4H20).  Using the solubility of  trical-
cium phosphate (pA = 23.56) to predict residual dissolved phosphate
concentration in chemically-defined systems gave agreement within 25%
of experimental values for systems in which the initial magnesium concen-
tration was less than 2 mM (200 mg/£ as CaC03)  and the initial carbonate
concentration was less than 4 mM (400 mg/£ as CaC03) - neither of which
is rarely exceeded in wastewaters.

-------
the  recycling of precipitate suspension in chemically-defined systems
was  found to enhance the removal of dissolved phosphate and therefore
lower the phosphate residual for a given pH and solution composition.
Difficulties were encountered, however, in these experiments because of
poorly settling suspensions which resulted in the inability to maintain
high suspension concentrations for long periods.  Studies in the  critical
area of the settling properties of calcium phosphate-carbonate suspensions
conducted in chemically-defined systems revealed that initial magnesium
concentration and the pH of precipitation were the most important factors
Influencing the separation of these suspensions by sedimentation.  Increases
1n the pH to between 9.0 and 11.0 improved the settling of the suspension.
The beneficial influence of magnesium concentration on suspension settling
was exerted most strongly at pH values of greater than 10.5 where
precipitation of gelatinous Mg(OH)2 was likely.  Suggestions in the
literature that calcium carbonate precipitation aids in precipitate
separation by producing a dense suspension were not supported by  this
investigation.

In light of the stated findings it appears that high (>90%) phosphate
removal by calcium phosphate precipitation followed by suspension
settling is only possible at pH values of greater than 10.  It would
also appear that to achieve the same degree of phosphate removal  at a
pH of 9 to 10 the precipitate must be coagulated either by cationic
polyelectrolytes, ferric chloride, or alum.


PHOSPHATE REMOVAL  FROM  WASTEWATERS
Investigations of phosphate removal from wastewaters  by calcium phosphate
precipitation were conducted at the Sanitary Engineering Research
Laboratory (SERL) wastewater treatment facility.   The phosphate removal
performance of two treatment schemes was examined.  The first involved
primary sedimentation, activated sludge, lime precipitation,  recarbonation,
and clinoptilolite sorption; the second, lime precipitation following
primary sedimentation.

Values of residual dissolved phosphate in wastewater precipitated  with
lime at pH values between 9.5 and 11.0 could be predicted to  within
25% of their experimental values if these predictions were based on the
initial concentrations of dissolved lime rather than  on the total  lime
added and if a correction for calcium complexation was made.   This was
because the efficiency of lime dissolution in the slaking operation was
poor and lime continued to dissolve throughout the several compartments
of the precipitation unit.                                         ,

An overall phosphate removal of greater than 80%, to achieve  a residual
of less than 2 mg P/a, could be consistently achieved from this waste-
water, which was one of average alkalinity (240 mg/j>  as CaC03) and a
typical Ca/Mg mole ratio of 3, by precipitating activated sludge effluent
or primary effluent with lime doses of 200 mg/t, as CaC03 to achieve a
pH of 9.5.

-------
The overall phosphate removal performance of the entire treatment train
was dictated by the combined performance of the precipitation reactor
and the ensuing precipitate separator (in this case a sedimentation
basin).  Any calcium phosphate particles that escaped sedimentation were
dissolved in the low average residence time (5 min) recarbonation basin
that necessarily preceded the clinoptilolite sorption columns.  This
observation has important ramifications in the removal of phosphate by
calcium phosphate precipitation.  It means that improvement of the overall
phosphate removal of a calcium phosphate precipitation process cannot be
achieved by post-filtration of the effluent from the solids separator
of such a process.  Filtration of such an effluent will require a prior
downward pH adjustment, but this pH adjustment will cause a rapid
dissolution of calcium phosphate particles making it impossible to remove
them by filtration.  These observations emphasize the importance of
solids separation in calcium phosphate removal and suggest this general
topic  as a fruitful area for investigation.

-------
                             SECTION  II

                           RECOMMENDATIONS
The work reported herein has demonstrated that short-term calcium  phos-
phate precipitation processes can be modelled successfully on  the  basis
of the formation of an amorphous tricalcium phosphate.   Indications  that
this steady state solid phase can be made to transform  and grow  into a
more insoluble calcium phosphate phase were obtained from precipitate
recycle experiments.  It is recommended that these experiments be  further
pursued to determine the factors that influence the formation  and  growth
of more insoluble calcium phosphate solids in wastewaters since  the
formation of these solids will allow the attainment of  lower dissolved
phosphate residuals.  Investigations of these phenomena should proceed
along several lines including:  1) the role of recycled precipitate  in
providing crystal growth opportunity, 2) the role of precipitate in  a
sludge blanket clarifier in providing sites for crystal growth,  and
3) the possible use of calcium phosphate clinker from lime regenerated
by recalcining to provide material on which calcium phosphate  crystal
growth may occur.

An important conclusion of this investigation is that the phosphate
residuals in effluents of calcium phosphate precipitation processes
cannot be reduced by post-filtration because prior downward pH adjustment
will dissolve the phosphate-containing particles.  It is important
therefore to devote considerable effort to improving precipitate removal
by sedimentation processes, especially since it is possible to produce
low dissolved phosphate residuals at pH values of 8.5 -9.5 (but at
these pH values precipitate separation is difficult).  The factors that
influence the flocculation and settling properties of calcium  phosphate-
carbonate suspensions produced at pH values below 10 should be investigated.
Such an investigation should consider the surface properties of  such
suspensions and their possible modification by coagulants to improve
their settling properties.

-------
                             SECTION  III

                            INTRODUCTION
 PHOSPHATES  IN WASTEWATERS
The conventional combination of primary sedimentation and secondary
biological treatment processes with anaerobic digestion for sludge treat-
ment can be relied upon to remove between 2 and 3 mg P/£ from domestic
wastewater of average strength.  With the current domestic wastewater
total phosphate content averaging about 10 mg P/a, such removals  typically
account for some 20% to 30% of the incoming phosphate in domestic waste-
water.  About half of the total phosphate in domestic sewage is derived
from synthetic detergent builders so that even if the phosphates  in  these
products were to be completely replaced by nonphosphate containing
compounds, a sewage with some 5 mg P/Jl would result.  Treatment of such
a wastewater by current primary and secondary treatment schemes would be
expected to leave a residual of some 2 to 3 mg P/i.  In instances where
phosphate has been identified as a nutrient limiting the growth of aquatic
photosynthetic organisms it is generally agreed that growth control  over
these organisms by controlling phosphate concentration may only be exerted
when phosphate levels on the order of 50 yg/£ or less are reached.   It
would therefore appear that in such instances treatment further than the
conventional primary and secondary biological methods would be needed to
produce such phosphate levels.  Since complete elimination of phosphate
from synthetic detergents is neither an impending nor a likely event and
would not reduce raw sewage phosphate to less than 5 mg P/a in any case,
the development of processes that achieve high phosphate removals is
justified.

Of the treatment processes suggested for increasing phosphate removal to
a level greater than that possible by primary and secondary treatment,
those using precipitation with metal ions are the only ones to have  found
wide application and to be economically feasible.  Of the chemical
precipitants commonly used (ferric and ferrous iron, aluminum and calcium
salts) lime has been the most common, possibly because of a combination
of its cheapness, the capability for its regeneration, and its familiarity
in the field.  Most calcium phosphate precipitation schemes involve
raising the pH of the waste stream to 11 or higher.  Under these  conditions
low phosphate residuals (<1 mg P/a) and readily settleable suspensions
(supposedly because of concomitant Mg(OH)2 precipitation) are obtained.
However, to achieve these high pH values on a variety of typical  waste-
waters Nesbitt [1] has reported that lime doses of between 280-720 mg/£
as Ca(OH)2 are required since the lime dose is largely determined by the
wastewater alkalinity.  Moreover, pH adjustment of the treated effluent
(commonly by recarbonation) is mandatory following these high pH
precipitation processes.

-------
Several authors (Menar and Jenkins [2], Ferguson [3], Ferguson et al.
[4], and Jenkins et al_. [5]) have suggested and demonstrated experimentally
that phosphate residuals of 1 mg/fc or less can be obtained in wastewaters
and synthetic systems representing wastewaters under some  conditions  at
pH values far below 11.  Most of these experiments have, however, been
conducted in batch systems with reaction times and concentration ranges
different from those that exist in a typical  lime precipitation unit  for
removing phosphate from wastewater.

It was the purpose of this investigation to determine the  behavior of
chemically-defined systems representative of wastewater when subjected
to phosphate precipitation by the addition of calcium salts (including
lime) at various conditions of pH and under the physical constraints  of
reaction time and reactor design typical of a wastewater phosphate
precipitation process.  The overall aim of such experiments was to develop
a predictive method that would allow the determination of  chemical dose
and phosphate residuals from chemically-defined systems and test it on
wastewater.
OBJECTIVES
The general objective of this research  was  to develop a method for
predicting the performance of phosphate precipitation processes involving
the addition of calcium as a precipitant.   Such  processes  include but are
not necessarily restricted to:   lime addition; the addition of calcium
salts plus strong base; and the use of existing wastewater calcium hard-
ness as a phosphate precipitant accompanied by aeration for upward pH
adjustment.  The general objective was  reached by the fulfillment of the
two following specific objectives:  1)  the  development of  a predictive
method based on calcium phosphate precipitation experiments in chemically-
defined systems containing calcium, magnesium, orthophosphate, carbonate,
hydrogen ion and water, and 2)  the testing  of the results  of these experi-
ments on domestic wastewaters under differing conditions of precipitation
to determine the utility of the predictive  method.

It became evident as the work progressed that a major determining factor
in the efficiency of calcium phosphate precipitation processes was the
ability to separate the insolubilized calcium phosphate.  The objectives
of the work were therefore expanded to include an examination of some of
the chemical factors that influenced precipitate separation by sedimentation,
                                    8

-------
                                SECTION IV

                   PHOSPHATE REMOVAL FROM WASTEWATERS BY
                      CALCIUM PHOSPHATE PRECIPITATION


The precipitation of calcium phosphate from wastewater has most commonly
been conducted by the addition of unslaked or slaked lime to provide
simultaneous increases in both calcium and hydroxyl ion.  Most calcium
phosphate precipitation schemes involve raising the pH of the waste stream
to at least 10.5 (and often above pH 11) because in this pH range experi-
ence has shown that low dissolved phosphate residuals (<1 mg P/£) are
obtained together with a settleable precipitate.


Perhaps the earliest observation of the use of lime for phosphate preci-
pitation in the waste treatment field was by Rudolfs [6].  He concluded
that phosphate removals exceeding 90% — to reach residuals of approximately
0.3 mg P/a — could be obtained together with excellent flocculation by
lime addition to reach a pH of higher than 10.1.


Because phosphate precipitation processes using calcium have been conducted
at high pH values, lime requirements have been largely dictated by the
alkalinity of the wastewaters and have borne little or no relationship to
its phosphate concentration.  Thus Wuhrmann [7] states that the lime
requirement for phosphate removal at pH 10.5 to 11 is equal to 1.5 times
the carbonate hardness.  Mulbarger et^ a]_. [8] and Buzzell and Sawyer [9]
present a summary figure (Figure 1) of lime dose required to attain pH
11.0 as a function of wastewater alkalinity.  The wide variation in lime
doses to produce low dissolved phosphate residuals (0.4-2.4 mg P/s,) is
indicated by Nesbitt's [1] summary of the literature in which it is
revealed that doses ranging from 280 to 720 mg/i Ca(OH)2 were employed
by various investigators.      »

From a survey of operating data in tertiary lime precipitation plants at
Blue Plains, Washington, D. C.; Pomona, California; Lebanon, Ohio;
Las Vegas, Nevada; and S. Lake Tahoe, California, Seiden and Patel [10]
have concluded that residual dissolved phosphate concentration will be
about 0.15 rag P/A at pH 11.  Notwithstanding the conclusion that high pH
operation produces readily settleable precipitates, it is Nesbitt's
opinion that a minimum phosphate residual can only be obtained if a
filtration step (necessarily requiring recarbonation) follows lime

-------
   500
   400
5  300
S
 CM
   200
   100
                               •   Boston, Mass.
                               o   Marlboro, Mass.
                               A   Clinton, Mass.
                               A   Leominster, Mass.
                               •   Hudson, Mass.
                               a   Schenectady, N.  Y.
                                      After Buzzell & Sawyer  [9]
                50
100
150         200        250
  ALKALINITY, mg CaC03/fc
300
350
400
   FIGURE 1.  LIME REQUIREMENT TO  REACH pH 11 AS  A FUNCTION OF WASTEWATER ALKALINITY

-------
precipitation.  It is interesting, however, to note that if such  a
filtration step is contemplated for the removal of calcium phosphate
particles from the effluent of a high pH precipitation unit it will be
necessary to lower the pH value of the effluent (possibly by recarbonation)
prior to filtration.  The effect that the fall in pH has on the dissolu-
tion of calcium phosphate particles has not been discussed by authors who
suggest filtration following lime precipitation to enhance phosphate
removal.

Wuhrmann [7] has indicated that the calcium carbonate and apatite preci-
pitates that form at high pH values do not settle readily, and he suggests
the addition of 1 to 2 mg/Jl Fe(III) to aid flocculation of the precipi-
tates.  In pilot plant post-precipitation or tertiary treatment of acti-
vated sludge effluent by this method (268 mg/s, Ca(OH)2, 1 to 2 mg/£
Fe(III), pH 11.1) phosphate reductions from 3.2 to 0.4 mg P/a were
obtained.  Results that indicate the carryover of phosphate in particles
were obtained by Owen [11] who precipitated phosphate from a high-rate
trickling filter effluent.  Using 720 mg/2, Ca(OH)2, a dissolved phosphate
residual of 0.13 mg P/a was obtained, while the total phosphate residual
was 1.7 mg P/a.  Reducing the lime dose to 360 mg/s, produced a soluble
phosphate residual of 1.5 mg P/£ and a total residual of 2.6 mg P/i.

The use of alum and lime at pH of 10 to 11 has been reported by Spiegel
and Forrest [12] at Amarillo, Texas, where the wastewater phosphate
concentration was reduced to at least 0.4 mg P/a.

Experience at the South Tahoe Public Utilities District Plant [13,14] has
shown that the addition of 400 mg/a CaO to activated sludge effluent to
raise the pH to 11.5 will produce a phosphate residual of 0.3 mg P/a by
plain settling of the precipitate.  However, the process flow sheet has
included multimedia filtration subsequent to the sedimentation of the
precipitate.

A Densatop pilot plant operated at the Los Angeles County Sanitation
District's Pomona facility for conditioning activated sludge effluent
prior to treatment by ion exchange showed the beneficial effect of
filtration in decreasing phosphate residuals [15].  When operated between
pH 10 and 11 the Densator produced an effluent containing an average  total
phosphate of 1.85 mg P/a and an orthophosphate of 0.2 mg P/a.  Filtration
changed these values to 0.11 and 0.09 mg P/4, respectively.  Operation  at
pH values in the range 9.1 to 9.8 produced mean unfiltered residuals of
5.5 mg P/a total phosphate and 0.86 mg P/a orthophosphate, which were
reduced to 0.65 and 0.48 mg P/a by filtration.

In summary, high pH (>10.5) lime precipitation processes, therefore,
appear to produce low residuals (0.5 mg P/a}, but separation of the
colloidal precipitate may be difficult and require the use of coagulants.

There are several processes and suggested processes for calcium phosphate
precipitation that involve the use of pH values lower than 10.5.  In one
                                   11

-------
of these, the phosphate extraction process (PEP) proposed by Albertson
and Sherwood [16], a key feature is the recycle of previously formed
chemical sludge.  The process involves lime addition to raw wastewater
in a flocculator-clarifier to achieve a pH of 9.5 to 10.5, and recycle
of clarifier underflow to maintain a solids concentration of 500 to
2000 mg/Si in the flocculator.  The clarifier effluent which contains
residual total phosphate of about 2 to 3 mg P/a is fed to a completely
mixed activated sludge aeration basin which further reduces the phosphate
content to 1-2 mg P/a and is claimed to provide an overall process BOD
and suspended solids removal of 90% to 95%.  Chemical sludge recovery by
thickening, dewatering, combustion (calcining), and scrubbing is proposed.
It is also proposed that pH adjustment following lime treatment to a
range that is suitable for activated sludge operation can be achieved by
the biologically produced C02 in the completely mixed aeration basin.
This proposition is supported by studies at the South Tahoe Public
Utilities District Plant [14] which indicate that when recalcined lime
was added to the primary clarifier to produce a primary effluent of
pH 11 the aeration basin contents some 5 ft into the aeration basin had
a pH of 8.5 to 8.6.

Buzzell and Sawyer £9] expressed doubts that an activated sludge unit
could be effectively operated on a primary effluent that has been treated
with lime to a pH of 10 to 11 since insufficient degradable organic carbon
would be present.  In addition, it is possible that the precipitation unit
might reduce phosphate concentrations to the point that the ensuing
biological process would be phosphorus limited; it also is possible that
carryover of solids from the flocculator process could significantly
decrease the phosphate removal of the process.  Humenick and Kaufman [17]
have used the foregoing rationale to suggest that a chemical-biological
process combination for achieving high BOD and phosphate removals should
use a high-rate biological treatment prior to a chemical precipitation
unit.

The recirculation of chemical sludge in the PEP process besides aiding
the flocculation of wastewater particulates, enables total phosphate
residuals of about 2 mg P/i to be achieved at approximately one pH unit
lower than without precipitate recycle and consequently with about one
half the lime dose of a conventional lime precipitation process.  The
presence of recycle undoubtedly increases the rate of calcium phosphate
precipitation by providing nuclei or growth sites for precipitate
formation and growth. -The provision of nuclei for the precipitation of
apatite has been shown by Stumm and Leckie [18] and Ferguson e_t a\_. [4]
to strongly influence the rate of phosphate removal under similar condi-
tions in chemically-defined systems.

The significance of sludge recycle has been demonstrated by work at. the
South Tahoe plant [14] in which introduction of lime mud recycle from the
chemical clarifier to the lime rapid mix reduced the mean effluent
phosphate concentration from 0.22 to 0.16 mg P/£.  Recycling of scrubber
water from the sludge and recalcining furnaces (which contained a total
phosphate concentration of 12.8 mg P/z, of which 12.3 mg P/a was parti-
culate) caused a reduction of the effluent orthophosphate from 0.16 to
                                    12

-------
0.09 mg P/X,.  When these recycle streams were removed from the  rapid  lime
mix basin and again introduced to the primary clarifier,  the  average
effluent phosphate concentration increased from 0.09 to 0.31  mg
Schmidt and McKinney [19] conducted laboratory tests  of a chemical-
biological scheme that is essentially similar to the  PEP process  except
that it includes no provision for precipitate recycle.   These  workers
used a lime dose of about 150 mg/x. Ca(OH)2 to reach a pH of 9.5  in  raw
sewage precipitation.  At this pH these authors claim that polyphosphates
present in the raw sewage prevent the precipitation of calcium carbonate
by adsorbing onto calcium carbonate nuclei.  Since the major consumption
of lime in a precipitation process is devoted to calcium carbonate  preci-
pitation, a reduction in lime dose will be obtained by this phenomenon.
No full-scale or pilot-plant application of this work has been reported.

There are several activated sludge treatment plants in the United States
at which phosphate removals are achieved that are far in excess  of  those
predictable by normal biological means.  There are two general schools  of
thought concerning the mechanism of these removals — that which  proposes
a biological mechanism ("luxury uptake") and that which proposes  a  chemical
precipitation mechanism.  Recent studies by Ferguson e_t ^1_. [20]  have
demonstrated that the high removals of phosphate experienced at  the
Rilling Road Treatment Plant of San Antonio, Texas can be explained on
the basis of a combination of calcium phosphate precipitation  and reactor
configuration (aeration basin design).  This demonstration was based on
the previous work of Menar and Jenkins [2] who explained the so-called
luxury uptake of phosphate in terms of calcium phosphate precipitation.
They proposed that precipitation occurred in an activated sludge  aeration
basin at a point where C02 stripping caused the pH to increase to the
point that calcium phosphate precipitated.  The behavior observed at
San Antonio [21] could not be reproduced by a pilot plant operated  by
Menar and Jenkins [2] at San Ramon, California.  Thus, at San  Antonio
soluble phosphate residuals of 1 to 2 mg P/2. were obtained at  pH  7.8
to 7.9 while at San Ramon, even at pH values above 8.5, soluble  phosphate
concentrations of greater than 10 mg P/l remained.  Ferguson and McCarty
[22] explained this difference in behavior based on the effects  of
carbonate and magnesium on phosphate precipitation proposing that in the
lower alkalinity and magnesium concentrations at San Antonio (compared
to San Ramon) more rapid calcium phosphate precipitation to form a  more
insoluble solid occurred.  Ferguson et_ al_. [20] proposed that  apatite
formation occurred significantly only during residence times typical of
activated sludge treatment in the plug flow aeration basin of  the Rilling
Road activated sludge plant at San Antonio and not in the more back-
mixed basins of the East and West plants.

Calcium phosphate precipitates formed in primary, secondary, or  tertiary
treatment units will tend to redissolve at the lower pH values encountered
in anaerobic digestion.  Indeed, redissolution of calcium phosphates
precipitated in activated sludge aeration basins is found even in the
slightly lower pH (higher C02) environment encountered in the  sludge
                                    13

-------
blanket of a secondary clarifier [2].   Release  of phosphorus  has been
generally observed from all  activated  sludges claimed  to be showing
"luxury uptake"  when the sludges were  exposed to  anaerobic conditions
and lowering of  the pH.
                                  14

-------
                              SECTION V

          MECHANISM  OF CALCIUM  PHOSPHATE  PRECIPITATION
 INTRODUCTION
Although the use of lime for precipitating  calcium phosphate from waste-
water has been suggested and tested widely,  the mechanism of calcium
phosphate precipitation from this  medium  is  not well defined.  For the
conditions commonly encountered in wastewater, hydroxyapatite (Ca5OH(POit)3)
is the thermodynamically stable solid  calcium phosphate phase.  The
concentrations of dissolved calcium and phosphate in lime-treated domestic
wastewaters, however, far exceed the equilibrium values of crystalline
hydroxyapatite.  It is therefore apparent that either precipitation
kinetics or the presence of more soluble  calcium phosphate phases control
phosphate residuals in wastewaters treated  with lime.
 PRECIPITATION OF  CALCIUM PHOSPHATE  FROM
 CALCIUM PHOSPHATE  SOLUTIONS
A considerable body of experimental  work  in  chemically-defined solutions
has shown that the initial  solid phase  that  appears in the rapid preci-
pitation of calcium orthophosphate from basic solutions is structurally
noncrystalline.  The noncrystalline  structure is metastable and, if
allowed to remain in contact with the preparative solution, it will
spontaneously convert into  a crystalline  product with a chemical
stoichiometry and X-ray diffraction  pattern  typical of the mineral
apatite.

Recent kinetic studies by Stumm and  Leckie [18] have defined three  phases
in calcium phosphate precipitation.  These were:  nucleation —the
formation of amorphous calcium phosphate  from solution; phase transfor-
mation — the slow transformation of  amorphous calcium phosphate into
crystalline apatite; and crystal growth of crystalline apatite.  Phosphate
is removed from solution during nucleation and crystal growth but not
during phase transformation.  Since  the dissolved phosphate residual in
equilibrium with crystalline apatite will be lower than that coexistent
with the amorphous nucleating phase, Stumm and Leckie [18] suggested that
low wastewater phosphate residuals might  be  obtained by recycle of
preformed crystalline apatite to a precipitation reactor.   Their sugges-
tion was based on the observation that  the presence of apatite crystals
largely eliminated the phase transformation  step between nucleation and
crystal growth.  Ferguson ejt al_. [4] recently have demonstrated that the
addition of such preformed  precipitate  to calcium phosphate carbonate
                                   15

-------
solutions supersaturated with respect to calcium phosphate, eliminates
lag periods prior to precipitation that were observed in batch reactors
not containing added precipitate.

There have been various reports of the nature of the calcium phosphate
phase that precedes crystalline apatite.  Strates jst ^1_. [23] suggest
that dicalcium phosphate dihydrate (CaHPOi+-2H20) first crystallizes and
rapidly converts into apatite.  McGregor and Brown [24] maintain that
octacalcium phosphate CattH(P0lt)3 is the metastable precursor of apatite.
Booth and Coates [25] suggest that dicalcium phosphate dihydrate is the
first crystalline product to appear and it transforms into octacalcium
phosphate, Ca^HKPOiJa, which itself then transforms into apatite.
Walton et_ al .  [26] found that over the pH range 6.3 to 9.04 calcium
phosphate "[presumably apatite) was preceded by a metastable precursor
of the stoichiometry but not the structure of tricalcium phosphate,
Ca3(POj2.

The stable crystalline apatite phase formed from pure calcium phosphate
solutions, while yielding X-ray diffraction patterns typical of apatite
(Ca5(P0lt)3OH) ,  may deviate significantly in stoichiometry from the Ca/P
mole ratio of 1.67 predicted from this formula.  Such calcium deficient
and other nonstoichiometric apatites have been reported over the entire
range of Ca/P  mole ratios from 1.33 to 2 [27] with crystal structures
indistinguishable by X-ray diffraction from well-crystallized apatite.

Several theories have been advanced to explain the existence of these
nonstoichiometric apatites - thus some workers [28] explain low Ca/P
ratios by postulating the presence of adsorbed phosphate on the solid
surface, possibly in the form of surface complexes.  Rootare et al .
[29] believe that these surface complexes control the phosphate concen-
tration in a solution in equilibrium with apatite and are formed by
hydrolysis of the apatite surface thus:


      Ca10(POit)6(OH)2 + 6H20 $ 2Ca2+ + 2HPOJ" + 4[Ca2HPOtf(OH)2]      (1)


                  Ca2HP0lt(OH)2 * 2Ca2+ + HPOJ" + 20H"                (2)
The evidence for the existence of such a complex is indirect — and
largely based on a change in Ca/P mole ratio in solution from 1  to 1.67
as the concentration of an apatite suspension is decreased.

McConnell [30] suggests the simultaneous precipitation of lower Ca/P
ratio solids together with apatite, e.g., brushite (CaHPO^,  Ca/P =: 1)
and octacalcium phosphate (CaitH(P0lf)3, Ca/P = 1.33).  However, this
suggestion would not be valid if the apatite formed had no detectable
phase impurities.  Other investigators suggest the formation of apatite
with a calcium deficient, defective lattice [31] as well as  the isomorphous
substitution of ions or groups for those normally present in the apatite
lattice.
                                   16

-------
 PRECIPITATION OF  CALCIUM PHOSPHATE FROM SOLUTIONS
 CONTAINING  MAGNESIUM  AND CARBONATE
Domestic wastewaters contain a wide variety of organic  and  inorganic
compounds, many of which might influence the precipitation  of  calcium
phosphate.  However, hydrogen ion (pH), magnesium,  carbonate,  and, to
a lesser extent, fluoride (because of its low concentration in waste-
waters), appear to be the most significant.  Their  effects  will be
reviewed in the context of the three steps  of precipitation (nucleation,
phase transformation, and crystal growth) defined by Stumm  and Leckie [18].

Since pH influences the relative abundance  of the various phosphate
species and components of the carbonate system, it  is important in
determining the range of stability of various calcium phosphate solids.
However, for pH values above 7 with solution Ca/P ratios of >0.5  the
precipitation of monocalcium phosphate (Ca(H2POit)2) and its monohydrate
(Ca(H2POi.)2'H20) or dicalcium phosphate (monetite,  CaHPOiJ  and its
hydrate (brushite, CaHPOi+'2H20) are not important.   In  fact for commonly
encountered wastewater Ca/P ratios, hydroxyapatite  is the stable  phase
at all commonly encountered process pH values.

pH influences the stability of formed precipitates  in the sense that at
high pH values rapid precipitation of unstable solids results, whereas
at lower pH values slower precipitation of more stable  solids  may occur.
Walton et_ aj_. [26] reports that hydroxyl ion is directly involved in the
formation of the critical nucleus in calcium phosphate  precipitation and
that the rate of nucleation is a function of [OH~]2.  Eanes ejt ail_. [32]
report that a pH increase causes a reduction in the rate of crystalliza-
tion from the amorphous phase.  They report that conversion to apatite
was completed in 2.75 hr at an initial pH of 8.0, while at  a pH of 9.8
complete conversion took place in 6.5 hr.  The apparent stabilizing
effect of the higher pH upon the noncrystalline phase is probably due to
the incorporation of impurities into the precipitate during the more
rapid high pH nucleation phase.

Magnesium has been reported to influence both the rate  of calcium phos-
phate precipitation and the nature of the precipitated  solid.  It is not
likely, however, that magnesium phosphate or carbonate  solids  will form
for the ranges of magneisum, carbonate, and phosphate concentrations
typical of wastewaters.  Magnesium inhibits the nucleation  of  calcium
phosphate, possibly due to the competition  between  magnesium and  calcium
ions and to the formation of the strong magnesium phosphate complex
MgHPOlf[33].  The presence of magnesium at calcium phosphate crystal
growth sites has been postulated to slow down the transformation  of
amorphous calcium phosphate phases into apatite [4].  Newesely [34] has
reported that anhydrous tricalcium phosphate and Mooney and Aia [35]
have reported that beta tricalcium phosphate (whitlockite)  form instead
of apatite in the presence of magnesium. Whitlockite is often found
with magnesium as a minor component -with  contents of  about 6 to 8
atom percent.  Other metals, e.g. manganese and ferrous iron,  also
                                   17

-------
stabilize whitlockite.  Ferguson [3] has proposed that the presence of
Mg/Ca mole ratios of greater than 5 stabilize beta tricalcium phosphate
in sludge digester liquors and proposed that in wastewaters of differing
Ca/Mg ratios either apatite or a stabilized beta tricalcium phosphate
may control the dissolved phosphate concentration.

In systems containing carbonate and phosphate the effect of Mg on calcium
phosphate and carbonate precipitation appears to be complex and pH
dependent.  Magnesium is reported to interfere with the precipitation of
the calcium carbonate solids, calcite and aragonite, by inhibition of
nucleation and lattice distortion.   Calcites containing magnesium as a
minor component exist.  These magnesian calcium carbonates are more
soluble than calcite and their solubility increases with their magnesium
content over the range 0-20%, from pKSD 8.4 to pK<-D 6.3 (Chave et al.
[36]).                                           SP

Recent work in laboratory batch systems at phosphate, carbonate, and
magnesium concentrations typical of wastewaters [4] showed that besides
forming calcium carbonate and phosphate solids, carbonate and phosphate
interacted in the formation of these respective solids.  Carbonate has
been reported by Ferguson e_t al_. [4] to inhibit both the nucleation and
phase transformation of calcium phosphates.  Newesely [34] reported that
carbonate at 0..8 mM disturbed the cyrstallization of calcium phosphate,
and Ferguson [3] reported that no crystalline calcium phosphate is formed
from solutions of typical wastewater constituent concentrations containing
10 mM carbonate.  However, the presence of carbonates in bone, dentine,
and enamel (all of which are apatite) and the existence of minerals such
as francolite and cellophane suggest the possibility of forming mixed
calcium carbonate—phosphate compounds [28].  Rapid precipitation, as
encountered in wastewater calcium phosphate precipitation processes,
results in mixed calcium phosphate—carbonate solids because the discrim-
.ination between these radicals becomes difficult and recrystallization
is a slow process.  Simpson [37] has reported that high partial pressures
of C02 destabilize apatite and has  observed the formation of octacalcium
phosphate at C02 partial pressures  of between 0.01 and 0.1 atm. and at
pH values below 8.1.
 SURFACE PROPERTIES OF  CALCIUM PHOSPHATE SUSPENSIONS
It is generally observed that calcium phosphate-carbonate suspensions
settle well whe'h produced at pH values greater than about 10 but that
the opposite is true for suspensions produced at lower pH values than
this.  'The stability of colloidal calcium phosphate—carbonate suspensions
can be attributed tjo electrpstatic repulsive forces created by' a surface
charge that originates from ionization of various groups on the solid
side of the solid-liquid interface by isomorphous substitution, or by
ion adsorption.  Somasundaran [38] has suggested surface hydrolysis as  a
possible mechanism of.charge development on apatite:
                                   18

-------
                      P0j~ + H20 J HPOJ" + OH"                       (3)


                      HPOif + H20 £ H2PO; + OH"                      (4)


                      H2PO; + H20 Z H3PO^ + OH"                      (5)


                         Ca2"1' + OH" Z CaOH+                          (6)


                        CaOH+ + OH" t Ca(OH)2                        (7)


At low pH values reactions (3), (4), and (5) will proceed in the forward
direction and (6) and (7) will proceed in the reverse direction, while
at high pH values the reverse is true.  It is also evident from reactions
(3) through (7) that the charge on the solid surface depends on pH of the
solution.  Somasundaran [38] found the pH at the zero point of charge
(PHzpc) for apatite to be 6.  Stumm and Morgan [33] argue that the pHZpc
of salt-type minerals depends in a complicated way on the pH and also on
the concentration of all potential-determining ions.  The presence of
precipitate impurities changes the pHzpc because the zpc of solids should
correspond to the pH of the charge balance (electroneutrality) of potential-
determining ions [39].

Calcium phosphate precipitation in wastewater treatment normally involves
raising the pH to 10.5 or higher.  At these pH values Mg(OH)2 also preci-
pitates and its formation has been proposed as a factor in improving the
settling properties of the calcium phosphate-carbonate suspension.  It
is possible that specific adsorption of Mg(OH)2 on the negative suspension
surface [33] takes place.  It is also possible that bridging occurs
between Mg(OH)2 and the surface phosphate groups leading to a three-
dimensional network [40].


SUMMARY
Precipitates are formed in solutions as a result of nucleation followed
by crystal growth.  In calcium phosphate precipitation, however, crystal
growth does not usually follow the nucleation process immediately, but
is preceded by a lengthy phase transformation step.  During the later
step, the nucleated amorphous calcium phosphate is transformed into a
crystalline phase with no net phosphate removal from solution.

Hydrogen ion concentration exerts a significant effect on each of the
three phases of calcium phosphate precipitation.  An increase in pH
speeds up nucleation, while at the same time appears to cause a reduction
                                   19

-------
in the rate of crystallization.  In addition, pH exerts  a significant
effect on the surface properties of the suspension.

Magnesium appears to inhibit the nucleation of calcium phosphate,  to
slow down the formation of apatite from amorphous calcium phosphate,  and
to stabilize the formation of tricalcium phosphate.   It  also tends to
interfere with the precipitation of calcium carbonate.  At pH values
above 10.5, Mg(OH)2 precipitates and appears to be involved in producing
a readily flocculable calcium carbonate-phosphate suspension.

Carbonate appears to inhibit both the nucleation and the phase transfor-
mation of calcium phosphate.  A high partial pressure of CO^ reportedly
destabilizes apatite and enhances the formation of octacalcium phosphate.
                                   20

-------
                             SECTION VI

                        APPROACH  RATIONALE
The approach to the problem of predicting  the dissolved phosphate resi-
duals obtainable by calcium phosphate precipitation from wastewater was
first to study a model  or chemically-defined system representative of
the relevant components of wastewater.  The results of this study were
then applied to the prediction of dissolved phosphate residuals obtained
experimentally from wastewater.

The complex and variable nature  of domestic wastewater makes the use of
simple models difficult though necessary.  This work used the model
system, Ca - Mg - Pj - Cj -  H+ -  H2Q^ where CT = H2C03 + HCOg + COi"
and PT = H3P04 + H2PO; + HPOif +  POij  with which previous authors [3,4,
22] have had some success in modelling waste treatment situations for
calcium phosphate precipitation equilibria and kinetics.  It must be
realized, however, that since domestic sewage contains species such as
polyphosphate, fluoride, and organic acids, as well as clay minerals  and
organic particulate solids,  the use of a simplified system excluding
these cannot provide an exact representation of the actual situation.

The components of the model  system vary in domestic sewage over the
approximate ranges indicated in Table 1.  The pH in domestic wastewater
is typically within the range 6.5 to 8, averaging about 7.5, but wastewaters
subject to calcium phosphate precipitation may have pH values of up to
11.5.  In the model chemically-defined system, the range of chemical
conditions studied were purposely made wider than those cited in Table 1.
                               TABLE 1

            COMPONENT  CONCENTRATIONS  IN WASTEWATER
                        (from Ferguson  [3])
Component

Calcium
Magnesium
Carbonate
Phosphate
Range
mM
0.
0.
2
0.
5-5
25 - 1
- 8
1 -0.5
Typical
Value
mM
1.5
0.5
4.0
0.3
                                  21

-------
The precipitation model described by Stumm and Leckie [18] was adopted
as the basis for study of calcium phosphate precipitation.  It was postu-
lated, on the basis of the literature on calcium phosphate precipitation
and a knowledge of the reaction times used in wastewater calcium phos-
phate precipitation processes, that a steady state amorphous calcium
phosphate phase controlled the residual dissolved phosphate concentration.
The validity of this steady state postulate was first tested and then the
nature of the steady state phase was investigated.

The activity product variation of various phases with pH and solution
composition was tested.  The stoichiometry of the steady state phase was
determined.  Attempts were made to obtain crystallographic information
from X-ray powder diffraction patterns of the aged precipitate.

The results obtained from the chemically-defined systems were applied to
the precipitation of calcium phosphate by lime from wastewaters.  Since
it became evident, as the work progressed, that precipitate separation
was a major factor in determining the total phosphate residual obtainable,
the study devoted some attention to the chemical factors that influence
the removal of precipitate by sedimentation.

Limited attention was also directed toward the effect of precipitate
recycle on dissolved phosphate residual.  This was prompted by the
precipitation model  and results of previous investigators who attributed
low dissolved phosphate residuals to the growth of crystalline calcium
phosphates.
                                   22

-------
                             SECTION VII

              EXPERIMENTAL  EQUIPMENT  AND PROCEDURES
 EXPERIMENTAL EQUIPMENT
The experimental phases of the study  involved runs both in chemically-
defined systems containing reagent grade  chemicals dissolved in distilled
water and in domestic wastewater from the SERL pilot plant.  For the
chemically-defined systems both continuous  flow and batch reactors  were
employed.  The experiments with sewage used a continuous flow system
onlv.
 Chemically-Defined Systems —Continuous  Flow
 Experi ments

Continuous flow experiments on chemically-defined systems were conducted
in a single acrylic plastic CSTR with  inside  dimensions of 10 in. x 10
in. x 10 in.  The reactor contents were mixed at 71 rpm by a stainless
steel stake and stator stirrer mounted vertically and driven by a 1/8 hp
DC motor (Bodine Co.) controlled by  a  Speed Controller (Minarik Co.).
Power input was measured by a torquemeter that could be attached to the
vertical stirrer shaft.  Power input was such that an average velocity
gradient of approximately 75 sec"1 was maintained in the experiments.
To prevent precipitation in the feed lines and feed storage vessels prior
to the reactor, it was necessary to  prepare the chemical solutions in
three separate batches and then mix  these in  a well-defined fashion just
preceding their entry into the reactor.  The  bulk of the feed, which was
either stored in 55-gal stainless steel drums or in a 1000-gal redwood
tank lined with a plastic sheet contained the calcium, magnesium, and
orthophosphate components.  This solution was prepared by adding stock
solutions of reagent grade CaCl2, MgCl2, and  H3P04 to distilled water
(Figure 2).                  '

After 30 min aeration to mix the solution and to produce a uniform pH
of 3.5, the solution was pumped to a constant head tank using a stainless
steel centrifugal pump.  The solution  then flowed by gravity through a
1/2-in. PVC pipe and a Brooks Full View Rotameter into the reactor.  A
concentrated NaOH solution (for pH adjustment) was pumped into the
influent stream through a "T-joint"  20 in. ahead of the reactor inlet.
At a point 2 in. before the reactor  inlet the carbonate component (Cj)
was added as a concentrated solution of sodium carbonate.

The reactor pH was monitored continuously using a Radiometer Model 22
pH meter and a combined calomel-glass  electrode.  The reactor effluent
                                  23

-------
ro
P.
1

CaCl2
MgCl2
Solutions


Constant
Head Tank

$r
\
k. r
— i
Pump
                                                             Pumps
                                               Stake and
                                               Stator
                                               Stirrer
                                                 CSTR
                                       Rotameter
                                                                                  Effluent
                                                          Sedimentation Basin
                                                                                  Sediment
                                                                                  Drai n
        FIGURE 2.   SCHEMATIC FLOWSHEET FOR EXPERIMENTS ON CHEMICALLY-DEFIED  SYSTEMS

-------
discharged through 1 1/2-in. PVC pipe.   Samples for analysis of dissolved
Pp C-j-, Ca, and Mg were taken from the  reactor effluent pipe after steady
state had been reached and filtered immediately through 0.45-y membrane
filters (Mi Hi pore HA).  When it was desired  to collect precipitate
samples for X-ray diffraction analysis, the reactor effluent was collected
in a 30-£ sedimentation basin.

In continuouous flow experiments employing a  recycle of suspension, a
circular 12-4 concave-bottomed sedimentation  basin followed the completely
mixed reactor in the flow-scheme illustrated  in Figure 3.  Suspension
was recycled to the reactor from the underflow of the sedimentation basin
using a peristaltic pump (Sigmamotor Model T65).  In these experiments the
mean hydraulic residence times were 42  min in the reactor and 34 min in
the sedimentation basin.  Sufficient feed solution for several days of
continuous operation was prepared by adding reagent grade CaCl2, MgCl2,
and H3P04 to 3000 a of well-mixed tap water held in a 1000-gal polyethy-
lene-lined redwood tank.  The performance of  the recycle reactor was
monitored on the basis of grab samples  taken  from the sedimentation basin
overflow.
      3,000-4
    Feed Water
       Tank
                                                             Effluent
                                                     12-4
                                                     Clarifier
                             Suspension Recycle

 FIGURE 3.   SCHEMATIC  OF CALCIUM PHOSPHATE PRECIPITATION
               WITH SUSPENSION RECYCLE
 Chemically-Defined  Systems  - Batch Experiments

Batch experiments  were  conducted on chemically-defined systems when it
was demonstrated that they could be expected to yield information that
was identical  to that obtained in continuous flow, steady state system.
In these batch experiments the reactor was  a magnetically stirred 4-4
beaker.  Predetermined
mixed rapidly  into 4 4
amounts  of Na2C03 and
of distilled water in
                                             PQk stock solutions were
                                             batch reactor and the  pH
                                   25

-------
of the mixture adjusted immediately to the  desired  value with NaOH
solution.  A mixture of the desired amounts  of CaCl2  and MgCl2 stock
solutions were then added rapidly to the  stirred  solution  and at that
time the measurement of reaction time was started.  A 15-min reaction
time was employed in these experiments on the  basis of results from
continuous flow studies.   During this reaction time the pH was monitored
and maintained constant by NaOH addition.  Samples  for the analysis of
dissolved Cj, Pj> Mg» and Ca were withdrawn  after 15  min and immediately
filtered through 0.45-y membrane filters.


Wastewater  Experiments

Continuous flow wastewater precipitation  experiments  were  conducted in
a 3-compartment epoxy-coated galvanized steel  reactor of which each
compartment was 15 in.  x 15 in.  x 15.8 in. water  depth.  The contents
of each compartment were stirred by straight blade  turbine paddles mounted
on a vertical shaft.  The precipitation unit was  an integral part of a
chemical-biological process train that, in  addition to lime addition,
precipitation reactors, and two sedimentation  basins, contained a
recarbonation basin and clinoptilolite columns for  ammonia removal
(Figure 4).

Sufficient lime slurry for 24-hr operation was prepared by adding
quicklime to 100 a of tap water in a 30-gal  plastic bin.   The lime
slurry was made up to a concentration of  either 300 mM or  600 mM, depending
on what pH of operation was desired.  Lime  slurry batches  were analyzed
for CaO immediately after preparation and immediately prior to the
preparation of a fresh batch.   Lime doses were computed on the basis of
these analyses combined with measurement  of the volume of  lime slurry
pumped.  The lime slurry was introduced into the  wastewater feed at a
point 2 in. ahead of the  first reactor and  at  a rate  that  was determined
by a pH controller in the reactor, preset to the  desired pH value.

After passing through the three reactor compartments, the  wastewater
entered a circular concave-bottom sedimentation basin.  The performance
of the precipitation unit was  monitored by  analysis of samples which
were collected continuously from the effluent  of  the  sedimentation basin.
Samples of both influent and effluent were  stored at  4°C between sampling
and analysis.  In addition to the analyses  conducted  on the chemically-
defined systems, measurements  of total orthophosphate and  dissolved
hydrolyzable phosphate plus orthophosphate  were made  on samples of
influent, lime precipitated effluent, the recarbonated effluent, and the
clinoptilolite column effluent.  Each experimental  run with the waste-
water treatment plant lasted for at least four days.
                                   26

-------
                                           Lime




1,

1
A
A
Primary Activated rl , ..
Sedimentation Sludae Flocculation
<
Sedi
menta
tion
                                                                  Recar
                                                                  bona-
                                                                  tion
                  Sedi-
                 menta-
                  tion
                                                             Ammonia
                                                             Sorption
                Primary
             Sedimentation
 Activated
  Sludge
ro
—i
               Lime
                   Flocculation    tion
Sedimenta^  """  Floccu-  Sedl"-
          Recar- lation  menta-
          bona-           tion
          tion
Ammonia
Sorption
                   FIGURE 4.  TREATMENT SYSTEMS  USED  IN STUDIES  AT SERL PILOT  PLANT

-------
 ANALYTICAL  PROCEDURES
 Sampling Procedures

 Sampling procedures  varied with  the system under investigation and  have
 been discussed previously with descriptions of the experimental  systems.


 Analytical Methods

 Chemical and physical  analyses were generally performed in accordance
with Standard Methods  [41], the FWPCA Methods for Chemical Analysis
 [42], or the SERL Analytical Methods [43].  The following brief descrip-
 tion of analytical methods is presented to indicate method selection
 and any novel method modification employed.

 Soluble  Calcium.    Soluble calcium was determined by EDTA titration
 using hydro>
-------
     .   Six min reaction  time was used instead of 10 min.  The two
        latter modifications were made to reduce the possibility of
        polyphosphate hydrolysis during the orthophosphate analysis.

Qrtho +  Hydrolyzable  Phosphate.  A 50-ms, sample was placed in a
125-m«, Erlenmeyer flask and 1 ml of strong acid solution (310 ml cone.
\\2SQk to 600 mil distilled water) was added.  The sample was boiled  gently
for 30 to 40 min.  Phenolphthalein (3 drops) was followed by 1 N NaOH
until a pink color developed.   This was discharged by the addition  of
strong acid solution. The  sample was cooled and diluted to 50 mi and
the analysis continued as for orthophosphate.

Inorganic Carbon.   Inorganic carbon was determined with a Beckman
Carbon Analyzer, Model 215A.

A1 ka 1i ni ty .   Alkalinity  was  determined by titration with standard acid
to a pH of 4.3.

Suspended Solids.   Suspended solids were determined by the membrane
filter technique of Winneberger ejb al_. [44] using 0.45-y membrane filters.

Suspension  Settling  Tests.  Settling tests were conducted on  reactor
effluent by collecting it in a  4-& beaker and allowing it to settle
quiescently for 30 min.   The settling properties of the suspension  were
evaluated in terms of the total phosphate removal achieved during this
period by taking samples  for total phosphate analysis before and after
settling, using a pipette to sample 1 in. below the surface of the
reactor effluent.

X-Ray Powder Diffraction  Analysis.  Suspensions for X-ray powder
diffraction analysis were removed from the bottom of the sedimentation
basin following the CSTR  and centrifuged at about 1000 rpm.  After
decanting the supernatant,  the  suspension was quickly frozen by immersion
in a dry ice-n-butyl alcohol mixture.  The frozen suspension was
lyophilized for a period  of 48  hr.  The X-ray powder diffraction pattern
was determined using a Norelco  diffractometer and scanning the spectrum
from 4 degrees to 60 degrees.   Because early experience demonstrated
the presence of noncrystalline  materials, lyophilization was later
replaced by air-drying.

Lime Purity.    A 0.5-gram finely pulverized sample was dissolved in
100 ma of a 10% sugar solution  after one hour on a shaking machine.
The solution was filtered and a 25-ms, aliquot was titrated with 0.1 N
sulfuric acid, using phenolphthalein indicator.

Total Ca(OH)2.   Total Ca(OH)2 was determined on a sample of lime
siurry after ju min refluxing with HC1.  A suitable sample volume
(usually 10 ml) was added to 25.00 mi of 1 N HC1 solution and the
mixture was refluxed for  30 min.  After cooling to room temperature the
excess HC1  was back-titrated with 1 N NaOH using phenolphthalein
indicator.
                                   29

-------
                            SECTION VIII

              PRECIPITATION OF  CALCIUM  PHOSPHATE  IN
                    CHEMICALLY-DEFINED  SYSTEMS
 INTRODUCTION
Batch and CSTR work in chemically-defined  systems  took the form of a
series of experiments in which one or another of the components of the
Ca - Cj - Pj   H+ - Mg - H20 system was  varied singly to determine its
effect on the residual phosphate or other  component concentrations.  The
range over which individual  component concentrations were varied was
selected to cover the range  commonly  encountered in wastewater treatment.
Nominal residence times of 15 min were generally used in CSTR experiments
and a reaction time of 15 min was used in  the batch experiments.  Some
kinetic studies employed variable detention  (CSTR) and reaction (batch)
times.
 TIME  TO ESTABLISH  STEADY STATE
Prior to conducting experiments  to determine  the effect of various solu-
tion components on the phosphate residuals  obtained by calcium phosphate
precipitation, it was necessary  to determine  the time required to reach
steady state values of phosphate residual  that were representative of
these precipitation processes.   From the  previous discussion of the phases
involved in the precipitation of calcium  phosphate from aqueous solution,
it might be surmized that steady state  dissolved phosphate residuals
might be reached rather rapidly  — at least  within a few minutes of mixing
reactants to produce a supersaturated solution.  The period over which
this steady state dissolved phosphate level is maintained is of importance
since it is necessary to know whether measurements made soon after the
initial attainment of steady state are  representative of the conditions
that might exist in phosphate, precipitation reactors of longer residence
times.

To establish the time required to reach steady state and the duration of
this steady state condition, a series of  batch and CSTR experiments were
conducted under conditions  typical  of wastewaters.  Three series of batch
experiments at 23°C were conducted for  systems containing Ca and Pj only,
Ca, Pj, and Cy, and Ca, Pj, Cj,  and Mg.   The  initial concentration of
each of these components is summarized  in Table 2.  Each of these systems
was examined at pH values of 8.0, 9.0,  10.0,  and 11.0 -values that were
maintained constant throughout the experiments by addition of NaOH
                                  31

-------
                               TABLE 2

         INITIAL  COMPOSITION OF  BATCH  SYSTEMS USED  TO
                DEFINE  TIME  TO REACH STEADY  STATE

Run
I
II
III
mM
Ca
2.0
2.0
2.0
PT
0.5
0.5
0.5
CT
-
2.0
2.0
Mg
-
-
2.0
Figure 5 shows  that  in each of the batch experiments steady state was
reached in less than 10 min.  Neither the introduction of carbonate
(Run II) nor carbonate and magnesium (Run III) into the system lengthened
the time required  to reach steady state, although changes in the  levels
of steady state phosphate concentration were observed.

CSTR experiments at  pH 8 were conducted with nominal reactor detention
times varying from 3 to 200 min and with the initial concentrations  of
reactants reported in Figure 6.  Because of their time-consuming  nature
and because of the results of the previous batch experiments, the CSTR
runs were only  conducted for a limited range of conditions.  Thus only
solutions containing Ca and Pj were tested in CSTR runs because batch
experiments had indicated that the presence of Mg and Cj did not  affect
the time taken  to  attain steady state.  Also since the batch experiments
indicated a slightly slower rate of attainment of steady state at pH 8,
the CSTR experiments were conducted at this pH value.

In all of the CSTR experiments, steady state dissolved phosphate  concen-
trations were reached in less than a 10-min mean residence time and  main-
tained for up to at  least 200 min — a residence time typical of a phosphate
precipitation process.  On the basis.of these results, reaction times of
15 min (batch or CSTR) were selected as being truly representative of the
steady state conditions existing in nonrecycle wastewater calcium phos-
phate precipitation  processes.
THE NATURE OF  THE  STEADY  STATE  SOLID  PHASE
It is important to determine the solubility characteristics and the nature
of the steady state solid  phase that forms under conditions representative
of wastewater calcium phosphate precipitation processes.  An insight into
the stoichiometry and solubility behavior of this solid was obtained by
                                  32

-------
  10
    -1
                         pH 8.0
Run No.
I 0
II Q
III A
Initial Conditions"
mM
Ca
2.0
2.0
2.0
PH
PT
0.5
0.5
0.5
CT
2.0
2.0
Mg
2.0
9.0
              S-:^—.-!).-
                       J-
                                Batch, 23°C
         ;-%_
  10
    -2
   _.Q	El-
               pH 10.0
                                  J§	1
       a
       oVnt
               PH n.o
          to	-a	-Q—
                                   E
  10
    -3
                    15
                       30
45
                          TIME, min
FIGURE 5.
DEPENDENCE OF STEADY STATE PT CONCENTRATION
 ON  SOLUTION COMPOSITION AND pH IN BATCH
 REACTORS
                         33

-------
   0.6
   0.5
0.
   0.4
   0.3
   0.2
   0.1
1 1
V D D
a
V ° o
I
Run
A
o
a
0
Initial Conditions, mM
Ca Pj Mg CT
3.7 to 4.2 1.0 to 1.1 0.0 0.0
3.7 to 4.0 0.5 to 0.6 0.0 0.0
1.8 to 2.1 1.0 to 1.1 0.0 0.0
2.1 to 2.2 0.5 to 0.6 0.0 0.0
0 " ° 20° to 24°C
£* CSTR
V V A &

i i
—
<& <
\
                        50
         100
DETENTION TIME (V/Q), min
150
200
     FIGURE 6.  DEPENDENCE OF STEADY STATE PT CONCENTRATION ON DETENTION TIME
                  IN CSTR EXPERIMENTS

-------
 analysis of both steady state solid and liquid composition of batch and
 CSTR experiments.

 CSTR experiments were conducted between 19° and 23°C with chemically-
 defined solutions  in which initial component concentrations were in the
 ranges:  Ca, 2.1 to 2.4 mM; Mg, 0 to 0.6 mM; PT, 0.3 to 0.5 mM; CT, 0 to
 14.8 mM; pH 8  to 11.  The initial mole ratios of the various components
 varied as follows:  Ca/P, 3.9 to 6.7; Ca/Mg, 2.1 to 4.8; Ca/Cj, 0.2 to
 2.4.

 Batch experiments  with 15 min reaction time were conducted at 19° to
 23 C using initial solution concentrations in the range:  Ca, 0.4 to
 5.5 mM; Mg, 0.1 to 0.5 mM; PT, 0.3 to 0.4 mM; CT, 0.8 to 2.7 mM; pH 8
 to 11.  The initial mole ratios of the various components varied as
 follows:  Ca/PT, 1.1 to 17.2; Ca/Mg, 6.9 to 12.9; Ca/Cj, 2.0 to 7.3.

 In all of these experiments nucleation proceeded without detectable lag
 period and steady  state dissolved phosphate levels were reached within a
 10-min period.  The nature of the steady state solid can be circumstan-
 tially deduced, and a useful tool to predict steady state dissolved
 phosphate residual can be obtained by computing the activity product of
 various postulated solids from the steady state dissolved component data.
 This procedure was initially performed for selected experimental data for
 the several solids that have been postulated as apatite precursors and
 therefore might be expected to be present under steady state conditions
 prior to transformation to an equilibrium solid.  Figure 7 is a plot of
 the negative logarithm of the activity product (pA) versus pH for the
 solids, dicalcium  phosphate (DCP) (CaHPOiJ, octacalcium phosphate (OCP)
 (Ca^FKPOiJo), hydroxyapatite (HAP) (Ca5(OH)(POlf)3) and tricalcium phos-
 phate (TCP) (Ca3(POlt)2).  It is evident from Figure 7 that only the
 activity product for TCP appears to be constant over the entire pH range
 (8 to 11) of the experiments.  Following this evidence further, Figure 8
 presents a plot of pA of TCP for all data points (both batch and CSTR)
 collected in chemically-defined systems.  The regression line fitted to
 these data points  has the form:
                         pA = -0.064 pH + 24.15
(8)
indicating the virtual constancy of the activity product with pH over the
range pH 8 to 11.

This value of the negative logarithm of the activity product of TCP which
averages 23.56 is somewhat lower than values reported in the literature
which range from approximately 25 to 29 [45-47].  By using the plotting
technique of Walton ejt al_. [26] (Appendix A), which determines the
stoichiometry of a nucleating solid phase, Figure 9 shows that the Ca/P
mole ratio in the precipitate is 1.44.  This value is very close to the
stoichiometry of TCP which would give a theoretical Ca/P mole ratio of
1.5
                                    35

-------
38

36

34

32

30

28

26

24&
22
         9
                        pH
                                             11
FIGURE 7.
          EFFECT OF pH ON pA FOR VARIOUS CALCIUM
           PHOSPHATE SOLIDS
                       36

-------
   25

1 1
Reactor
Batch o
CSTR A
Initial Conditions, mM
Ca
0.4-0.5
2.1-2.4
Mg
0.1-0.5
0.0-1.9
PT
0.3-0.4
0.3-0.4
CT
0.8-2.7
0.0-14.8
L
   243-pA = -0.064 pH + 24.15   ©
o
Q.

 CO
 (O
O


-------
co
CO
                n
                     ©O
12
                        19° to 23°C

                        15 min
                                     &A3&A
Reactor
Batch o
CSTR A
Initial Conditions, mM x 103
Ca
0.4-5.5
2.1-2.4
Mg
0.1-0.5
0.0-0.6
PT
0.3-0.4
0.3-0.4
CT
0.8-2.7
0.0-14.8
13
Ca/P = 1.44
                                               i
       A
                  A
                                                                   A
                                                          A
                                                             18
                      19
20     21
              FIGURE 9.   CRITICAL IONIC CONCENTRATIONS IN BATCH AND CSTR EXPERIMENTS

-------
X-ray powder diffraction analysis  of the  solids  produced in CSTR experi-
ments failed to reveal the presence  of any  crystalline calcium phosphate.
For experiments conducted at pH 10 and higher,  crystalline.calcite (CaC03)
was detected.  The X-ray analyses  indicate  that the steady state calcium
phosphate solid is amorphous in nature - a  fact that  is supported by the
lower pA values obtained here for  TCP in  comparison to those values
reported in the literature.


EFFECT  OF  SOLUTION COMPOSITION  ON  STEADY STATE
PHOSPHATE  RESIDUAL
Even though the evidence produced in the previous  section  tends to support
the presence of TCP as a steady state solid  phase, there is a considerable
range of variation in the values of the activity  product of TCP at each
of the pH values examined.  While such variation  (over  about 1 to 2 orders
of magnitude in pA) is'common in equilibrium data  derived  from hetero-
geneous systems, it is necessary to determine whether these variations
are in any way related to the concentrations of other important solution
components — especially to the concentration of Cj and  Mg.  In the data
analysis that follows, the effect of both of these components on the
activity product of TCP and composition of the steady state solid phase
is examined.
 Effect  of Magnesium

Previous reports have indicated that magnesium concentration  is  an impor-
tant factor in calcium phosphate precipitation under conditions  typical
of wastewaters.  Its effects have been classified as equilibrium effects
(in which the solid calcium phosphate phase is altered by  the presence of
magnesium) and kinetic (in which the rate of attainment of equilibrium
or steady state phosphate residuals is influenced by magnesium concen-
tration).  An additional "process effect" can be visualized,  i.e., in
which magnesium influences the separability of precipitated solids.

The influence of magnesium on steady state dissolved phosphate residual
was tested in five series of CSTR experiments for which the initial
component concentrations are listed in Table 3.

For the range of magnesium concentrations (more correctly  Ca/Mg  ratios)
commonly encountered in wastewaters, the pA of TCP does not appear to be
affected by magnesium concentration.  Thus Figure 10 shows that  only when
initial Ca/Mg ratios were less than 0.6 (commonly most wastewaters have
Ca/Mg ratios in the range of 1.5 to 6) did the activity product  of TCP
increase significantly.  These results were supported by an additional
series of CSTR experiments in which the steady state levels of dissolved
PT, CT, Ca, and Mg were measured as a function of varying  initial
magnesium concentration in the range 0.5 to 10.9 mM.  In this series of
experiments the pH was maintained constant at 9.5, and the initial
component concentrations were constant at Ca, 2.2 mM; Py,  0.38 mM, and
                                   39

-------
   25
   24
O
Q-

 co
 to
O
   23
   22
Run
A
E
O
e
Initial Conditions
mM x 103
Ca
2.2
2.1
2.2
2.2
Mg
0.6
2.3
3.6
10.8
PT
0.4
0.4
0.4
0.4
CT
2.7
4.1
2.6
2.4
Ca/Mg
3.7
0.9
0.6
0.2
            19° to 23°C

            15 nrin

            CSTR
                                     10
11
                             pH
         FIGURE 10.  EFFECT OF  Mg ON pA Ca3(P04)2
                            40

-------
                               TABLE  3

          INITIAL CONCENTRATIONS  IN CSTR EXPERIMENTS
                TO DETERMINE EFFECT OF  MAGNESIUM
Experiment
No.
1
2
3
4
mM
Ca
2.2
2.1
2.2
2.2
PT
0.4
0.4
0.4
0.4
CT
2.7
4.1
2.6
2.4
Mg
0.6
2.3
3.6
10.8
Mole Ratio
Ca/PT
6.0
5.8
6.1
6.1
Ca/Mg
3.7
0.9
0.6
0.2
Ca/CT
0.8
5.1
0.8
0.9
Cj, 2.5 mM.  The data in Figure  11  indicate that no significant increase
in residual dissolved phosphate  occurred until the initial magnesium
concentration was greater than 2 mM resulting in an initial Ca/Mg mole
ratio of less than 1.

As in previous experiments, X-ray powder diffraction analyses failed to
detect crystalline calcium phosphate in any of the solids derived from
these CSTR experiments.   Crystalline calcite  (CaC03) was detected in
suspensions produced from solutions containing initial calcium concen-
trations of 2.1 to 2.2 mM and initial magnesium concentrations of 0.5 mM.
Suspensions produced from solutions in which  the initial calcium concen-
tration was 2.1 mM and the initial  magnesium  concentration was 2.3 mM
failed to show the presence of crystalline calcite even after 1000 hr of
contact between the suspension and the solution from which it was produced.
After 4000 hr of solution-suspension contact  a crystalline calcium
carbonate, designated as Calcite III, was detected in the suspension.


 Effect  of  Carbonate

The effect of carbonate on the precipitation  of calcium phosphate from
wastewaters has been interpreted both in terms of a competition between
the calcium phosphate and calcium carbonate solids for calcium ion and
by the modification of calcium phosphate solids such as apatites by the
incorporation of carbonate.  Both of these interpretations predict that
the effect of increasing carbonate concentration will generally be to
increase dissolved phosphate residual at a given calcium dose or conversely
to require a higher calcium dose to reach the same phosphate residual.

Experiments on chemically-defined solutions were undertaken in CSTR
systems with 15 min nominal residence time to determine the significance
of the effect of carbonate on dissolved phosphate residuals at concen-
trations and under conditions that typified wastewater calcium phosphate
                                   41

-------
 12

 10

 8
 6

 4

 2
 0
Initial Conditions, mM
Ca
2.2
Mg
0.5-10.9
PT
0.38
CT
2.5
15 min
23°C
CSTR
pH 9.5
   0123
             4567
             Influent Mg, nM
    Mq
8   9   10  11
FIGURE 11.   INFLUENCE OF Mg ON STEADY STATE
            CONCENTRATIONS OF Ca, Mg, CT, AND
            PT AT pH 9. 5
                      42

-------
precipitation processes.   Initially four series  of CSTR experiments were
conducted with the initial  conditions as outlined in Table 4.   The results
of these experiments, presented  in Figure 12,  show that up to. initial Cj
concentrations of 10.6 mM no significant effect  on the activity product
of TCP was  detectable.  At initial Cj levels  of  between 10.6 and 14.8 mM
the activity product of TCP increased significantly for pH values above
9.5.
                              TABLE  4

          INITIAL  CONDITIONS FOR  CSTR EXPERIMENTS  TO
              DETERMINE  EFFECT OF  CARBONATE (CT)
Exp.
No.
1
2
3
4
pH
Range
8 -11
8 -11
8 -11
8.5-11
mM
Ca
2.1-2.4
2.2
2.1-2.3
2.1-2.3
Mg
0-0.5
0.6
0-0.6
0.6
PT
0.3-0.5
0.4
0.4-0.5
0.4
CT
0
2.5-2.7
5.0-7.2
10.6-14.8
Mole Ratio
Ca/Mg
4.8-6.0
3.7
3.8-5.2
3.5-3.9
Ca/Cj
-
0.8-0.9
0.4
0.2
Ca/PT
3.9-6.5
6
4.2-6.2
3.9-6.5
The effect this increase in activity product had on phosphate residual
at pH 9.5  and  11 in systems containing calcium,  magnesium, phosphate,
and carbonate  is illustrated in  Figures 13 and 14 which are the results
of 15-min  nominal residence time CSTR experiments conducted at the  range
of initial  experimental conditions depicted in Table 5.
                              TABLE  5

          INITIAL  CONDITIONS FOR  CSTR  EXPERIMENTS TO
          DETERMINE  EFFECT OF CARBONATE  ON PHOSPHATE
                 AND OTHER COMPONENT RESIDUALS

PH

9.5
11.0
mM

Ca
2.0-2.2
2.2-2.3

Mg
0.6
0.6

PT
0.4
0.4

CT
2.6-14.4
2.5-14.8
Mole Ratio

Ca/Mg
3.4-3.7
3.7-3.8

Ca/Cj
0.1-0.9
0.1-0.9

Ca/PT
4.4-6.1
6.0-6.1
                                43

-------
25
Run
A
Q
O
O
Initial Conditions, mM x 103
Ca
2.1-2.4
2.2
2.1-2.3
2.1-2.3
Mg
0.0-0.5
0.6
0.0-0.6
0.6
PT
0.3-0.5
0.4
0.4-0.5
0.4
CT
0.0
2.5-2.7
5.0-7.2
10.6-14.8
22
                                                11
      FIGURE 12.  EFFECT OF CT ON  pA Ca3(P04)2
                      44

-------
     ro

     O)
        1.5

        1.0
       0.5  -
       0.0
        1.6
        1.4
        1.2
        1.0
        0.8
Initial Conditions, mM
Ca
2.0-2.2
Mg
0.6
PT
0.4
CT
2.6-14.4
                                   8      10

                                 INFLUENT CT, mM
12
      CSTR
      pH 9.5
      20° to 23°C
      15 min
         18.0


         16.0


         14.0

         12.0
             CM
         10.0 °
             X

          8.0 I


          6.0 ^

          4.0


          2.0

          0.0
14
16
FIGURE 13.  EFFECT OF CT ON STEADY STATE CONCENTRATIONS OF Ca, Mg,  AND PT
            AND ON AMOUNT OF PRECIPITATED  CARBONATE AT pH 9.5

-------
        4.0


        3.0



        2.0



        1.0



        0.0



        0.7
        0.6
      O)


      -a
      c
      to

      10
0.5
        0.4
        0.3
Initial Conditions, mM
Ca
2.2-2.3
Mg
0.6
PT
0.4
CT
2.5-14.8
                                                       19° to 20°C
                                                          0.9
                                                          0.7
                                                                     CM

                                                                     O
         0.5
                                                                   0.3
                                                                   0.1
                                  8       10

                                 INFLUENT CT, mM
                                          12
14
16
FIGURE 14.   EFFECT OF CT ON STEADY STATE CONCENTRATIONS OF Ca,  Mg, AND  PT

            AND ON AMOUNT OF PRECIPITATED CARBONATE AT pH  II.0

-------
At both pH values there is a very gradual increase in residual dissolved
phosphate up to  initial carbonate concentrations of 10 mM.  Thereafter an
extremely rapid  increase in residual  dissolved phosphate  occurs at both
pH values.  These data are consistent then with the effects of carbonate
on the activity  product of TCP.   Evidence exists from these data that
carbonate is incorporated into the solids increasingly with increasing
initial carbonate concentration.   There  is consistently higher carbonate
incorporation into the solids at  pH 11.0 than at pH 9.5 for all initial
carbonate concentrations.  Evidence that calcium carbonate precipitation
occurred in both of these experiments is offered by additional CSTR data
depicted in Figure 15.  Here the  ratio of Ca/PT in the precipitate is
plotted versus pH on  the same graph as the CT removed from solution.  The
curves have almost identical shapes and  support the observation that
            UJ
            UJ
            Cf-
            CL
D.



O
i—i
I—

-------
calcium carbonate precipitation occurs at pH 9.5 and increasingly  there-
after up to pH 11.0.  These data also indicate that below pH  9.0 to  9.5
in systems with typical wastewater component concentrations,  calcium
carbonate precipitation is not significant.

Examination of the amount of carbonate and calcium removed from solution
in these experiments suggests that more carbonate disappears  from  solution
than can be accounted for by the precipitation of calcium carbonate  (with
a Ca/Cj mole ratio of 1).  It must therefore be assumed,  as has been
indicated in the literature, that incorporation of the  carbonate into
the calcium phosphate solid occurs in such systems.   This phenomenon has
been reported to produce calcium phosphate solids whose solubility
increases with their carbonate content.  It is possible that  the gradual
increase of residual dissolved phosphate with increasing  initial carbonate
concentration, noted in Figures 12 and 13, and the gradual increase in  TCP
activity product, noted in Figure 11, is caused by carbonate  inclusion
in the calcium phosphate solid.

In these CSTR experiments Figures 13 and 14 demonstrate that  for the entire
range of initial carbonate concentrations and for both  pH values of  9.5
and 11 there appears to be very little if any incorporation of magnesium
into the solid.

It might be concluded then, that for the purpose of predicting dissolved
phosphate residuals, the activity product of TCP derived  in these  experi-
ments can be used satisfactorily for waters  containing  carbonate concen-
trations of below 4 mM (or alkalinities of below 400 mg CaC03/£) - values
of alkalinity that are rarely exceeded in wastewaters.
METHOD  OF PREDICTING DISSOLVED PHOSPHATE RESIDUAL
CSTR and batch experiments have shown that for the  range  of component
concentrations commonly found in wastewaters,  dissolved phosphate  resi-
duals may be predicted from the activity product of TCP as  follows:
                          [Ca2+]|s [PO;-]S2S  =  A                      (8)


where the subscript ss stands for steady state.  The mean  value  for  the
negative logarithm of the activity product was determined  to  be  23.56.
From a materials balance [Ca2+]ss can be replaced by
using the fact developed from Figure 9 that the Ca/P mole ratio of the
precipitation approximates 1.5.   Introducing a value of lo~23-56 for the
                                   48

-------
activity product and substituting Equation  (9)  into  Equation  (8) with
the elimination of [Ca2+]ss,  the logarithmic form of the equation becomes
3 log {[Ca2+]in -  1.5  [P1n
2 log
           =  -23.56
                                                                    (10)
In this equation the only unknown is  Pss,  the  steady state concentration
of total dissolved phosphate, since the value  of  PO^"  can be expressed
as a function of pH, total dissolved  phosphate, and equilibrium constants
as follows:
                                        ss
                         [H+]/k3 + [H+?/k2k3  +  [H+]3/k1k2k3]
                                                                    (11)
where kl5 k^, and k3 are the first, second,  and  third equilibrium constants
of phosphoric acid.  Equation (10) can be solved manually  for Pss as
illustrated in Appendix B.

Using this equation to predict phosphate residuals  from  the CSTR and
batch experiments in chemically-defined systems  gave excellent results
(within 25% of experimental values) for systems  with pH  between pH 8 and
10, with initial magnesium up to 2 mM, and initial  carbonate up to 4 mM
(Table 6).

At pH 11.0 an error of 68% between predicted and experimental values was
obtained for initial carbonate concentrations  up to 3 mM.  This error
increased slightly (the predicted values were  lower than experimental
values) to the neighborhood of 73% when the  initial carbonate concen-
tration was increased to 6 mM.  The increasing error in  predicted values
at both increased pH and carbonate concentration is undoubtedly caused
by calcium carbonate precipitation which reduces the effective initial
value of calcium.  These errors were reduced so  that predicted and
experimental values of dissolved phosphate residual were within 25%
when the precipitation of calcium carbonate  was  taken into consideration
(Table 7).
 CALCIUM  PHOSPHATE  PRECIPITATION AT pH
 SUSPENSION  RECYCLE
8 WITH
As previously indicated, the calcium phosphate  precipitation model of
Stumm and Leckie [18] visualized precipitate  formation  in  three steps —
nucleation of an amorphous phase, phase  transformation,  and growth of a
crystalline phase.  In the short-term calcium phosphate  precipitation
processes examined in the previous part  of this  chapter  it was postulated
that the solid present in the phase transformation step  would control the
                                   49

-------
                              TABLE  6
      COMPARISON OF  PREDICTED AND  EXPERIMENTAL  RESIDUAL
       PHOSPHATE VALUES  IN  CHEMICALLY-DEFINED SYSTEMS
PH
8.0


9.0


9.5


10.0


n.o


pA Ca3(P0lt)2
Range
High 22.92
Low 24.35
Mean 23.69
High 22.80
Low 23.68
Mean 23.49
High 22.96
Low 24.08
Mean 23.71
High 22.77
Low 24.03
Mean 23.74
High 22.44
Low 24.44
• Mean , 23.55
Dissolved Py, mM x lO1*
Experimental
2100
4200
4500
1200
620
550
280
no
34
130
230
48
94
3.3
6.5.
Predicted
1000
1300
4280
500
530
500
185
177
31
43
222
52
13.5
5.1
6.4
Error3
%
-57.2
-69.0
- 4.4
-58.4
-14.5
- 9.1
-34.0
+60.7
- 8.8
-67.0
+ 3.5
+ 8.3
-85.6
+54.6
- 1.5
% Error  =
                           x 100.
dissolved phosphate residual.   It was found that  at  residence times
typical  of the processes  this steady state solid  has the characteristics
of an amorphous tricalcium phosphate.

There is considerable evidence  [3,4,18] that the  phase transformation
step can be shortened (or even  eliminated) by the presence of preformed
crystalline material in the precipitating medium. Indeed the reported
                                 50

-------
                               TABLE  7
      COMPARISON OF  PREDICTED AND  EXPERIMENTAL  RESIDUAL
       PHOSPHATE VALUES IN  CHEMICALLY-DEFINED SYSTEMS
         CORRECTED FOR  CALCIUM CARBONATE  PRECIPITATION
pH
11.0
11.0
11.0
CT
mM
1.5
3.0
6.0
Dissolved PT, mM x TO1*
Predicted
Uncorrected
for CaC03
Precipi-
tation
5.16
5.16
5.11
Corrected
for CaC03
Precipi-
tation
3.90
14.1
13.4
Experi-
mental
>
3.23
16.1
18.7
Error9, %
Uncorrected
for CaC03
Precipi-
tation
+60
-68
-73
Corrected
for CaC03
Precipi-
tation
+22
-13
-25
     a% Error =
                P -  E
x 100.
success of the Phosphate  Extraction Process [16],  in which precipitated
solids are recycled,  is attributed to providing a  high  concentration of
solids to encourage more  rapid crystal growth of calcium phosphate.
Because of these observations, a limited investigation  was conducted on
chemically-defined systems  to determine the feasibility of obtaining low
phosphate residuals at pH 8 through the recycle of preformed precipitate.
Two CSTR experiments,  each  of about 2 weeks duration, were
a single CSTR followed by a sedimentation basin from which
pension could be recycled to the CSTR (Figure 3).
                                   conducted using
                                   settled sus-
In the first experiment  an  attempt was made to build  up  precipitate by
recycling sedimentation  basin underflow from a solution  containing the
following initial  component concentrations:  Ca,  2.2  mM; Mg, 0.5 mM;
PT, 0.3 mM; CT,  3.2 mM;  pH  8.  After a period of  13 days i,t became evident
that it would not  be possible to build up a high  precipitate level in the
reactor because  of the poor settling characteristics  of  the precipitate.
Indeed, throughout the almost two weeks of continuous operation the
reactor suspension concentration did not exceed 6 mg/£ and no significant
phosphate removal  was achieved.              '      . _t

In the second experiment an aged calcium phosphate-carbonate suspension
was preformed in the reactor by adding 3 a, 10 M  CaCl2 solution to 10 £
                                  51

-------
distilled water followed by 750 mi cone.  H3POi,. and then sufficient 50%
NaOH solution to achieve a pH of 8.  After aging the stirred suspension
for 2 days the CSTR was continuously fed  with a solution whose initial
composition was:  Ca, 2.2 mM; Mg, 0.5 mM; Pj, 0.3 mM; Cj, 3.2 mM; pH 8.2.
Settling problems were again encountered  and the initial suspension
concentration of 5900 mg/fc fell rapidly during the first 2 days  of opera-
tion to a level of approximately 1500 to  2000 mg/£.   Over the one week
period that suspension concentrations of  1500 to 2000 mg/£ were  maintained,
it was possible to remove 60% of the incoming phosphate.  Following this
period of operation, continuing difficulties in suspension separation
caused a further decrease in reactor suspension concentration to about
50 mg/x. at the termination of the experiment (Figures 16 and 17).  At
this time the dissolved phosphate removal had decreased to 20%.

At the same time that these CSTR experiments were proceeding, batch
experiments were conducted on solutions of identical composition to the
CSTR feed.  Suspension from the CSTR reactor was introduced into two 4-a
beakers containing CSTR feed solution to  achieve initial levels  of 3 and
3000 mg/i suspension, respectively.  The  batch reactors were stirred for
29 hr.  Figure 18 shows that the dissolved phosphate concentration in the
high suspension concentration reactor decreased more rapidly and to a
lower level than the dissolved phosphate  concentration in the low suspen-
sion concentration reactor.  A further illustration  that the presence of
calcium carbonate-phosphate suspension enhances phosphate removal was
provided by a series of experiments in which solutions of composition
identical to the CSTR feed were filtered  through layers of suspension
taken from the CSTR reactor that had been deposited  on the surface of
membrane filters in a 2-cm deep bed.  When the filtrate composition was
analyzed for phosphate it was revealed that one passage of CSTR  feed
solution through such a precipitate bed effected a 54% removal of phosphate;
this was increased by successive refiltrations to 56% and 60% removal.
These results are of significance because the conditions in such a
refiltration experiment might be thought  of as being somewhat similar
to those existing in the sludge blanket of an upflow clarifier.   It
might be surmised that such an expanded bed of precipitate could have
many of the important features (albeit with a lower  residence time) of a
reactor to which precipitate is recycled.

X-ray powder diffraction analysis of the  suspensions from continuous,
batch, and refiltration experiments provided an important piece  of evi-
dence in support of the postulate that the steady state solid was a
tricalcium phosphate.  The presence of Ca3(POit)2-nH20 was detected by
these analyses and by chemical analysis of the solid.  The value of n
was found to be 4.
                                                               i    i
The effect of recycled solids concentration on the activity product of
TCP is shown in Figure*19.  The close correlation between activity product
and suspended precipitate concentration serves to emphasize the  catalytic
effect of solids in the formation of more insoluble  phases.

These experiments clearly indicate that low phosphate residuals  can be
achieved at pH 8 from waters with a mineral composition typical  of many


                                   52

-------
0.5


0.4


0.3
0
0.1
       Feed
 Run 2
Ca - 2.2
Mg   0.5
Pj = 0.32
CT = 3.2
inM
                                             1	T
pH = 8     Preformed Solids
21°C
                                            ^"t*"^
                                                      6,000
                                                      5,000  g

                                                      4,000  o
                                                          3,000  o
                                                          2,000
                                                          1,000
                                                               0.
                                                               co
                                                        60
   0   1   23    4    567   8   9   10   11   12  13  14
                          TIME, days

      FIGURE 16.   EFFECT OF SUSPENSION CONCENTRATION
                  ON  EFFLUENT  PHOSPHATE CONCENTRATION
                   IN  PRECIPITATE  RECYCLE EXPERIMENTS
    01   23   4   5678   9  10  11   12   13  14
    FIGURE 17.   STEADY STATE Ca AND  CT CONCENTRATIONS
                DURING PRECIPITATE RECYCLE EXPERIMENTS
                               53

-------
en
•£»
                                        I    I    I    '    I    I    I   I
                                        Suspended Solids, 3 mg/8.
                                           Suspended Solids,  3000 mg/fc
                             pH = 8.2

                             20°C
                                        8
10  12  14  16  18  20

      TIME, nr
24   26   28 30
                FIGURE 18.  EFFECT OF SUSPENSION CONCENTRATIONS ON RATE OF CALCIUM

                             PHOSPHATE PRECIPITATION  IN  BATCH EXPERIMENTS

-------
       Run 2

Feed,  Ca, 2.2 rrfl
      Mg, 0.5
      PT, 0.32
      CT> 3.2
                      4   5   6   7   8   9   10   11   12  13   14
                                                               6,000
                                                               4,500
                                                               3,000
                                                                    Q
                                                                    LU
                                                                    a
                                                               1,500
        FIGURE 19.  EFFECT OF  SUSPENSION CONCENTRATION ON
                    pA Ca3(P04)2 IN PRECIPITATE RECYCLE
                    EXPERIMENTS
domestic wastewaters by the  use  of precipitate recycle to produce high
suspension concentrations  for calcium phosphate crystal growth.   The
experiments also emphasize the importance of precipitate separation in
calcium phosphate precipitation  processes and the absolute necessity  for
development of methods  to  enhance  the settling properties of the calcium
phosphate-carbonate suspensions  produced at pH values below 10.
                                  55

-------
                             SECTION  IX

           PRECIPITATION  OF  PHOSPHATE IN  WASTEWATER


Five series of phosphate precipitation  experiments were conducted on
wastewater in the SERL wastewater treatment  facility.  Each experimental
run lasted for at least 4 days.   In the first  four experiments the phos-
phate removal performance of each individual unit of the entire treatment
train was examined, including primary sedimentation, activated sludge,
lime precipitation, recarbonation, and  clinoptilolite sorption. In the
fifth experimental run the phosphate removal performance of a lime
precipitation unit operating on  the primary  effluent was assessed.
 OVERALL  PHOSPHATE  REMOVAL  PERFORMANCE
The phosphate removal performance over the  entire  experimental period
for the whole process train is summarized in  Figure  20.  Performance
data are given for actual phosphate removal  as well  as maximum possible
phosphate removal —i.e., the removal  of phosphate that would occur if
perfect separation of particulates were possible.

Daily experimental values derived from 24-hr  composite samples for pH,
Ca, Mg, alkalinity and dissolved and total  phosphate  are presented in
Appendix C.  Mean values of these parameters, together with the mean
values of imposed operating conditions, are  presented in Tables 8 through
12.

In experiment 1  lime slurry was added  to activated sludge effluent and
precipitation conducted in a 3-compartment mixer-flocculator with nominal
residence time of 46.5 min in a run of 8 day's duration.  Lime doses of
2.9 mM (290 mg as CaCO^) raised the pH of  the activated sludge effluent
feed from 7.2 to 10.8 and reduced the  total  phosphate from an average
influent value of 11.1 mg P/i to a residual of 0.5 mg P/s, in the effluent
from the clinoptilolite column (Table  8).   In this experiment, as was the
case in all of the wastewater runs, the dissolved  phosphate concentration
leaving the clarifier following the precipitation  unit was lower than
that following recarbonation.  In this experiment, average dissolved
phosphate levels of 0.12 mg P/a were present  in  the  settled effluent from
precipitation while the dissolved phosphate following recarbonation of
this effluent was 0.50 mg P/£.  Subsequent experiments showed that virtually
all of the particulate phosphate that  escaped sedimentation redissolved
in the lower pH  medium of the recarbonation basin  within the 5-min average
detention time of this basin.  Indeed, following recarbonation only an
insignificant fraction of the phosphate was present  in the effluent as
particulate matter.  It is this factor that accounts  for the observation
                                   57

-------
                                         TABLE 8

                   OPERATING CONDITIONS  AND RESULTS  FOR WASTEWATER
                                      EXPERIMENT  I
Sample Location
Primary Effluent
Activated Sludge
Effluent
Lime Precipitation
Effluent
Recarbonation
Effluent
Clinoptilolite
Effluent
P
Total
roM
0.37
0.36


0.017
Dissolved
raM
0.31
0.30
0.00389
0.016
0.016
Ca
Dissolved
mM
1.38
1.39
1.93
-
1.72
Mg
Dissolved
mM
0.37
0.37
0.26
0.26
0.25
CT
Dissolved
mM
-
1.97
0.67


Alkalinity
Dissolved
mM
2.65
0.85
1.50
1.75
1.76
PH
7.4
7.2
10.8
8.5
8.5
   Feed to precipitation unit:   Activated sludge  effluent; Number of compartments:  3; Nominal
residence time:   46.5 min; Total lime added:   2.9 mM; Dissolved  lime added:  0.1 mM.
                                         TABLE  9

                   OPERATING CONDITIONS AND  RESULTS  FOR WASTEWATER
                                      EXPERIMENT 2
Sample Location
Primary' Effluent
Activated Sludge
Effluent
Lime Precipitation
Effluent
Recarbonation
Effluent
Clinoptilolite
Effluent
P
Total
mM
0.39
0.35
0.053
0.056
0.056
Dissolved
mM

0.34
0.0184
0.053
0.054
Ca
Dissolved
mM
1.36
1.37
1.76
1.72
1.14
Mg
Dissolved
mM
0.42
0.41
0.36
0.36
0.37
CT
Dissolved
mM

2.10
1.95


Alkalinity
Dissolved
mM
2.48
0.84
1.77
1.79
1.78
PH
7.4
7.3
918
8.2
8.4
   Feed  to precipitation unit:  Activated sludge effluent; Number  of compartments:  3; Nominal
residence time:  46.5 min; Total lime added:  1.6 mM; Dissolved lime added:  0.1 rtM.
                                          58

-------
                                        TABLE  10

                   OPERATING CONDITIONS AND  RESULTS  FOR WASTEWATER
                                      EXPERIMENT 3
Sample Location
Primary Effluent
Activated Sludge
Effluent
L1me Precipitation
Effluent
Recarbonation
Effluent
Clinoptilolite
Effluent
P
Total
mM
0.37

0.0089
0.0076
0.0076
Dissolved
mM
0.28

0.00323
0.007
0.007
Ca
Dissolved
nfl
1.20

1.51
1.90
1.41
Mg
Dissolved
mM
0.38

0.24
0.26
0.26
CT
Dissolved
ntt
3.8

0.6
-
5.0
Alkalinity
Dissolved
mM
2.35

2.46
2.94
2.97
pH
7.4

11.0
7.7
7.8
  Feed to precipitation  unit:  Primary effluent; Number of Compartments:  3;  Nominal residence
time:   46.5 min; Total lime Added:   4.9 mM; Dissolved lime added:  0.2 mM.
                                        TABLE 11

                   OPERATING CONDITIONS  AND RESULTS  FOR WASTEWATER
                                      EXPERIMENT 4
!•
Sample Location
Primary Effluent
Activated Sludge
Effluent
Lime Precipitation
Effluent
Recarbonation
Effluent
Clinoptilolite
Effluent
P
Total
mM
0.35

0.018
0.027
0.021
Dissolved
mM
0.28
-
0.0074
0.023
0.019
Ca
Dissolved
nti
1.10
-
1.07
1.26
1.24
Mg
Dissolved
mM
0.37

0.31
0.30
0.33
CT
Dissolved
mM
3.2

1.4
4.0
4.3
Alkalinity
Dissolved
mM
-


-

PH
7.4
-
10.2


   Feed  to precipitation unit:  Primary effluent; Number of compartments:  3;  Nominal residence
time:  46.5 min; Total lime added:  2.8 mM; Dissolved lime added:  0.1 mM.
                                            59

-------
                                       TABLE  12



                    OPERATING  CONDITIONS  AND  RESULTS  FOR  WASTEWATER

                                     EXPERIMENT  5
Sample Location
Primary Effluent
Activated Sludge
Effluent
Lime Precipitation
Effluent
Recarbonation
Effluent
Clinoptilolite
Effluent
P
Total
mM
0.40
.
0.026
.

Dissolved
mM
0.33

0.019


Ca
Dissolved
nW
1.40

1.71

-
Mg
Dissolved
mM
0.42

0.36
-

CT
Dissolved
mM
3.60

3.00


Alkalinity
Dissolved
mM
2.46

3.06
-

PH
7.5

9.6
-

     Feed to precipitation unit:  Primary effluent; Number of compartments:  4; Nominal residence

  time:  42 min; Total lime added:  2.6 mM; Dissolved lime added:  0.2 mM.
   100
           Run 1
i+*
LjJ C
ee. 
-------
than no phosphate removal  was effected  by  the  clinoptilelite columns -
since removal by these columns can only be expected by filtration and
filtration is only effective in removing particles.

In experiment 2 activated sludge effluent  was  treated with  lime in a 3-
compartment mixer-flocculator with average nominal residence time of 46.5
min for an experimental  period of 5 days.   Lime doses of  1.6 mM (160 mg
as CaC03/)i) raised the pH of the activated sludge effluent  from 7.3 to
9.8 and reduced the influent total phosphate from 10.8 mg P/£  to a value
of 1.7 mg P/£ following treatment by the whole process stream  including
recarbonation and clinoptilolite columns (Table 9).  Dissolution of parti-
culate phosphate occurred during recarbonation (causing an  increase in
dissolved phosphate from 0.6 mg P/n to  1.7 mg  P/£) as the pH was reduced
from 9.8 to 8.2.

In experiment 3 primary effluent was treated in a 3-compartment mixer-
flocculator with average nominal residence time of 46.5 min for an experi-
mental period of 10 days.  Lime doses of 4.9 mM (490 mg as  CaC03/£)
raised the pH of the primary effluent to 11.0  and reduced total phosphate
from a value of 11.5 mg P/2. to a residual  of 0.23 mg P/2.  following the
process stream of recarbonation and clinoptilolite treatment  (Table 10).
Recarbonation resulted in an increase of dissolved phosphate from 0.1 mg
P/£ to about 0.2 mg P/fc as the pH was adjusted downward from 11.0 to 7.7.

In run 4 the experimental conditions were  similar to those  used in experi-
ment 3 with the exception that, over a five-day period, primary effluent
was dosed with 2.8 mM lime (280 mg as CaC03/£) to reach a precipitation
pH of 10.2.  Final process total phosphate levels averaged  0.65 mg P/a —
levels that were increased by recarbonation from the 0.56 mg P/e. achieved
following sedimentation of the precipitated effluent.

In the final wastewater precipitation experiment (experiment 5) primary
effluent was treated in a 4-compartment mixer-flocculator with an average
residence time of 42 min for an experimental period  of 4  days. At the
precipitation pH of 9.6, achieved by a lime dose of  2.6 mM  (260 mg as
CaC03/Jl), the precipitation-sedimentation  unit reduced the  total phosphate
from 12.4 mg P/£ to 0.81 mg P/i.


 COMPARISON  OF  PREDICTED  AND EXPERIMENTAL
 PHOSPHATE  RESIDUALS


Good prediction of effluent dissolved phosphate concentration  using
Equation (10) derived from the precipitation model developed from chemically-
defined solutions was obtained when based  on the concentrations of dis-
solved constituents and with corrections being made  for complex!ng of
calcium and for calcium carbonate precipitation.  This procedure was made
necessary largely by the extremely poor efficiency of  lime  dissolution
obtained in the pilot plant.  Thus Figure  21 indicates that of the lime
added to the lime dissolution unit some amount on the order of 10$ or
                                   61

-------
LU
                                                    Experiment
                                                           1
      FIGURE 21.   COMPARISON OF  LIME  SLURRY ADDITION TO
                   DISSOLVED LIME CONCENTRATION  IN
                   WASTEWATER EXPERIMENTS
 less became dissolved in the wastewater in the first compartment of the
 precipitation reactor.  Additional  complications  arose because  the lime
 continued to dissolve in the subsequent reactor compartments.   Thus
 Figure 22 shows the increase in dissolved calcium caused by lime dis-
 solution as one progresses down the precipitation reactors.    '  !

 When the dissolved calcium concentration was used the model produced
 predictions of dissolved phosphate  residuals that were within 25% of
 experimental values (Table 13).  Previous attempts to predict residual
                                62

-------
     (O
     CJ
     o
     to
     CO
 Point of
-Lime Addition
                    1         2
                        COMPARTMENT NO.
     FIGURE  22.  CHANGE IN DISSOLVED Ca AND PT
                 CONCENTRATIONS IN  REACTOR  COMPART-
                 MENTS  -WASTEWATER EXPERIMENTS
phosphate concentrations  in wastewater as a result of chemical dosing have
been in error by a factor of some 5 times [3].
                           TABLE 13

       COMPARISON OF PREDICTED AND AVERAGE  EXPERIMENTAL
               DISSOLVED PHOSPHATE RESIDUALS  FOR
                    WASTEWATER EXPERIMENTS
Experiment
No.
1
2
3
4
5
Dissolved Phosphate
Residual, mM x 10^
Predicted
47.5
208
36.2
92.2
230
Experimental
38.9
184
32.3
74.0
190
Error9
%
+22
+13
+17
+25
+21
            a0/ F     _ Predicted - Experimental   1QO
             /0 hrror       Experimental       x IUU'
                               63

-------
Figure 23 shows  the variation of the  activity product of TCP with pH in
comparison with  the values of this  constant from chemically-defined sys-
tems.  In computing the activity product of TCP for wastewaters  the
complexing of calcium was taken into  account (Appendix B).  The  line for
the wastewater points was fitted by least squares analysis.  The agreement
between the values of this constant from the two types of system is good
at pH values  below 10.5.  At higher pH  values the pA value  of  TCP in the
wastewater systems (where lime was  used as a precipitant) is significantly
lower than for chemically-defined systems where dissolved calcium salts
were used as  precipitants.
      24
    to
    o
      22
                    I

                    0
                                     Chemically-Defined
                                   _^.Systems ©
Experiment
0
B
A
^
0
1
2
3
4
5
            I
        9.0
9.5
10.0       10.5

      PH
11.0
11.5
           FIGURE 23.   DEPENDENCE OF pA Ca^PO^ ON pH
                         FOR WASTEWATER  EXPERIMENTS
 OVERALL  PHOSPHATE REMOVAL  PERFORMANCE
In the wastewater tested(which is one with an average alkalinity  of 2.4
mM and a typical Ca/Mg mole ratio of approximately 3 to 1), a phosphate
removal in excess of 80% was consistently achieved at pH 9.5 with lime
doses of, at the most, 2 mM (200 mg/2 as CaC03).  Indeed, for most of
                                  64

-------
the experimental period phosphate removals  exceeded 90%.   It should  be
noted, however, that the overall phosphate  removal  performance  (total
rather than the removal based on dissolved  phosphate concentration)  of  the
entire treatment train was dictated by the  performance  of the precipi-
tation reactor and its ensuing sedimentation basin.  It was  demonstrated
that any phosphate-containing particles that escaped sedimentation were
rapidly dissolved during recarbonation (Tables  8 through  12).   This
observation raises the important issue of the efficacy  of filtration for
improving the phosphate removal for wastewater treated  by lime  precipi-
tation.  Effluents from all such processes  will  require downward  pH
adjustment by some process such as recarbonation prior  to filtration.
As these experiments have demonstrated, however, even short-term  (5  min)
recarbonation or other low pH environments  will  cause the rapid and
complete dissolution of the particulate phosphate,  thus making  it non-
removable by filtration.  It must be concluded,  therefore, that filtration
cannot be used to reduce the phosphate residual  in  lime-treated wastewater
and that the effective phosphate residual of a lime precipitation phos-
phate removal process is determined by the  efficiency of  the sedimentation
process following precipitation.

Because of this observation, which emphasized the importance of precipitate
removal by sedimentation, some preliminary  experiments  were  conducted on
chemically-defined systems to investigate the effect of solution  composi-
tion on the behavior of calcium phosphate—carbonate suspensions during
sedimentation.


 Factors  Influencing  Separation  of Precipitates

Previous literature has attributed the inability to operate  calcium
phosphate precipitation processes efficiently below pH  values in  the
10.5 to 11 region to the production of poorly flocculated precipitates
that cannot be readily separated.  Operation of these processes at  pH
values of 10.5 or greater was thought to be beneficial  from a precipitate
separation standpoint because of the formation of CaC03 (which  supposedly
increases suspension density) and of Mg(OH)2 (which reportedly  is gela-
tinous and acts as a suspension binder).

Two series of CSTR experiments were conducted on chemically-defined  solu-
tions with the objective of determining the effect of precipitating
conditions on the settling properties of the precipitates that  were  formed.
Variables investigated included pH; initial concentrations of calcium,
magnesium, and carbonate; and the initial Ca/Mg  ratios.  Both series of
experiments employed single CSTR's at 23°C, with 15 min nominal residence
time and with mean velocity gradient values of 76 sec"1.   The settling
properties of the suspension were assessed  by quiescent sedimentation
for 30 min following the CSTR.  Initial component concentrations  for the
two series of experiments are presented in  Tables 14 and  15.

Of the variables studied, initial magnesium concentration and pH  exerted
the most influence on the separation by sedimentation of  calcium
                                    65

-------
                            TABLE  14

       INITIAL  COMPONENT  CONCENTRATION  FOR PRECIPITATE
              SEPARATION  EXPERIMENTS, SERIES I
Run No.
1
2
3
4
Ca
mM
14.2
2.2
2.2
2.2
Mg
mM
3.7
0.6
3.6
10.8
PT
mM
0.4
0.4
0.4
0.4
CT
mM
2.4
2.4
2.6
2.4
pH
(Range of
Variation)
8 -11
8 -11
8.5-10.9
8.5-11
Ca/Mg
Mole
Rati o
3.8
3.7
0.6
0.2
Ca/CT
Mole
Ratio
5.9
0.9
0.8
0.9
Ca/PT
Mole
Rati o
4.0
6.1
6.1
6.0
                            TABLE  15

      INITIAL COMPONENT  CONCENTRATION FOR PRECIPITATE
              SEPARATION  EXPERIMENTS, SERIES II
Run No.
1
2
3
4
Ca
mM
2.2
2.2
2.2
2.2
Mg
mM
0.6
0.6
0.6
0.6
mM
0.4
0.4
0.4
0.4
CT
mM
(Range of
Variation)
2.7-12.7
2.7-12.7
2.7-12.7
2.7-12.7
PH
9.0
9.5
10.0
11.0
Ca/Mg
Mole
Rati o
3.7
3.7
3.7
3.7
Ca/CT
Mole Ratio
(Range of
Variation)
0.2-0.8
0.2-0.8
0.2-0.8
0.2-0.8
Ca/PT
Mole
Ratio
6.1
6.1
6.1
6.1
phosphate-carbonate suspensions.  Examination of Figures  24  and 25 shows
that increasing the pH  of suspensions  in  general improved their settling
characteristics.  Indeed, if one were  to  select the very  modest criterion
that 90% removal of precipitate by sedimentation was the  minimum acceptable
performance for a precipitate separation  process, then only  operation at
a pH value of greater than 10.5 would  consistently achieve this objective.
In the  pH range between 8 and 10 the suspensions generally settled poorly.
The influence of magnesium concentration  on suspension settling is
                                66

-------
Run
No.
1
2
3
4
Initial Concentration
mM
PT
0.4
0.4
0.4
0.4
CT
2.4
2.4
2.6
2.4
Ca
14.2
2.2
2.2
2.2
Mg
3.7
0.6
3.6
10.8
                      CSTR (G = 76 sec'1,
                      T = 900 sec) followed by
                      30 m1n quiescent settling
FIGURE 24.
           pH

SETTLING PROPERTIES OF CALCIUM
PHOSPHATE-CARBONATE SUSPENSIONS
(SERIES I)
                 67

-------
uj z —i 0.4
i— •-'i-

-------
concentrations of a positive specie such  as  Ca2+  and uninfluenced  by  the
presence of negatively charged species  such  as  HCC>3  or CO*'.

These preliminary observations on the separation  characteristics of
calcium phosphate-carbonate suspensions point to  the importance of
suspension surface properties in the removal of phosphate  by  precipi-
tation.  They also indicate that a fruitful  area  for investigation exists
in the physical characterization of these suspensions to provide infor-
mation that will lead to effective methods of precipitate  separation  at
pH values below 10.5.  At this stage one might conclude that  high  (>90%)
removals of phosphate by calcium phosphate precipitation and  sedimenta-
tion of solids are only possible at pH values of above 10. To achieve
80% overall phosphate removal at pH values of 10 or lower, the precipitate
must be coagulated with a cationic material, and in these experiments it
appeared that a large excess calcium dose achieved this objective.  It
would be better to achieve the same results with lower concentrations of
alternative flocculants such as polyelectrolytes or alum or ferric salts
which may have both an electrostatic and bridging function.
                                   69

-------
                             SECTION X

                          ACKNOWLEDGMENTS
This investigation was supported in part by Grant No. 17080 DAR from the
Environmental Protection Agency.  The work was performed by Mr. Arnold
B. Menar'un'der the direction of Professor David Jenkins.  Mr. Warren
Schwart? served as Project Officer for the Environmental Protection
Agency.  The assistance of Miss Christine Tarr and the late Mr. Philip
Palmerlperforming analyses and conducting experimental work throughout
this study is appreciated.  The generosity of Mr. Robert Jack for pro-
viding X-ray crystallographic facilities is acknowledged.  Portions of
the experimental phase of this study were conducted with other work
performed concurrently on this project.  The assistance of Drs. John H.
Koon and Larry A. Esvelt in this regard is greatly appreciated.
                                   71

-------
                              SECTION  XI

                              REFERENCES


 1.  Nesbitt, J.  B., "Phosphate  Removal  -The  State of the Art," J. Water
     Pollution Control  Fed., 41, p.  708  (1969).

 2.  Menar, A. B., and Jenkins,  D.,  The  Fate of  Phosphorus in Sewage
     Treatment Processes.   Part  II -Mechanism  of Enhanced Phosphate
     Removal by Activated  Sludge, SERL Report  68-6, Berkeley Sanit.
     Eng. Research Lab., University  of California  (1968).

 3.  Ferguson, J. F., "The Precipitation of Calcium Phosphates from Fresh
     Waters and Wastewaters," Ph.D.  Dissertation,  Stanford University,
     Stanford, California  (1970).

 4.  Ferguson, J. F., Jenkins, D.,  and Stumm,  W.,  "Calcium Phosphate
     Precipitation in Wastewater Treatment," Water -  1970, American
     Institute of Chemical Engineers.

 5.  Jenkins, D., Ferguson, J. F.,  and Menar,  A.  B.,  "Chemical Processes
     for Phosphate Removal," Wastewater  Reclamation and  Reuse Workshop,
     Lake Tahoe, California, Proceedings, p. 147 (1970).

 6.  Rudolfs, W., "Phosphate in  Sewage and  Sludge  Treatment.  II.  Effect
     on Coagulation, Clarification  and Sludge  Volume," Sewage Works J.
     19., p. 179 (1947).

 7.  Wuhrmann, K., "Objective, Technology and  Results of Nitrogen  and
     Phosphorus Removal Processes,"  Advances in  Water Quality Improvement,
     Univ. of Texas Press, p. 21 (1968).

 8.  Mulbarger, M. C., Grossman  III, E., Dean, R.  B., Grant, 0.  L., "Lime
     Clarification Recovery, Reuse,  and  Sludge Dewatering Characteristics,"
     J. Water Pollution Control  Fed..  41, p. 2070  (1969).

 9.  Buzzel, J. C., and Sawyer,  C.  N., "Removal  of Algae Nutrient  from Raw
     Wastewater with Lime," J. Water Pollution Control Fed., 39, p. R16
     (1967).

10.  Seiden, L., and Patel, K.,  Mathematical Model of Tertiary Treatment by
     Lime Addition, Report on Contract 14-12-416  between FWQA and  General
     American Research Division  (1969).

11.  Owen, R., "Removal of Phosphate from Sewage  Plant Effluents with Lime,"
     Sewage and Ind. Wastes. 25. p.  548  (1953).
                                    73

-------
12.  Spiegel, M., and Forrest,  T.  H.,  "Phosphate  Removal:  Summary of
     Papers," J.  Sanit.  Eng.  Div., Proc.  ASCE.  95  (SA5), pp. 803-827
     (1969).

13.  Suhr, L. G., Nutrient Removal at  Lake Tahoe - Costs and Criteria,
     report by Cornell,  Howland, Hayes, & Merryfield,  Engineers and
     Planners, Corvallis,  Oregon  (n/d).

14.  Culp, R. L., Advanced Wastewater  Treatment as Practiced at South
     Tahoe, Final Report to EPA on R & D  Project  17010 ELQ (WRPD 52-01-
     67TT1971).

15.  Garland, C.  F.,  "High Density, Solids Contact," paper presented at
     FWQA sponsored seminar,  Chicago,  111., June  26-27 (1968).

16.  Albertson, 0.  E., and Sherwood, R. J., "Phosphate Extraction Process,"
     J. Water Pollution  Control Fed.,  41, p.  1467  (1969).

17.  Humenick, M. J., and  Kaufman, W.  J., "An Integrated Biological-
     Chemical Process for  Municipal Wastewater  Treatment," paper presented
     at the 5th International Water Pollution Research Conference, San
     Francisco (1970).

18.  Stumm, W., and Leckie, J.  0., "Phosphate Exchange with Sediments;
     Its Role in  the  Productivity  of Surface Waters,"  paper presented at
     the 5th International  Water Pollution Research Conference, San
     Francisco (1970).

19.  Schmidt, L.  A.,  and McKinney, R.  E., "Phosphate Removal by a Lime-
     Biological Treatment  Scheme," J.  Water Pollution  Control Fed., 41,
     p. 1259 (1969).

20.  Ferguson, J. F., Jenkins,  D., Eastman, J.  D., "Calcium Phosphate
     Precipitation at Slightly  Alkaline pH Values," (in press) J. Water
     Pollution Control Fed.

21.  Vacker, D.,  Connell,  C.  H., and Wells, W.  N., "Phosphate Removal
     Through Municipal Wastewater  Treatment at  San Antonio, Texas,"
     J. Water Pollution  Control Fed.,  39, p.  750  (1967).

22.  Ferguson, J. F., and  McCarty, P.  L., The Precipitation of Phosphates
     From Fresh and Waste  Waters,  Technical Report No. 120, Dept. of
     Civil Engineering,  Stanford University,  Stanford, Calif. (1969).

23.  Strates, B.  S.,  Neuman,  W. F., and Levinskas, G.  J., "The Solubility
     of Bone Mineral.  II. Precipitation  of Near  Neutral Solutions of
     Calcium and Phosphate,"  J. Physical  Chem., 61. p. 279 (1957).   ,

24.  MacGregor, J., and Brown,  W.  E.,  "Blood:Bone  Equilibrium in Calcium
     Homeostasis," Nature, 205, p. 359 (1965).
                                   74

-------
25.  Booth, D. H., and Coates,  R.  V.,  "The  Stability of Calcium Hydrogen
     Phosphate Precipitated from Solutions  of Calcium Nitrate and Phos-
     phoric Acid," J.  Chem. Soc..  p.  4914  (1961).

26.  Hal ton, A. G., Bodin, W. J.,  Furedi,  H., and Schwartz, A., "Nucleation
     of Calcium Phosphate from  Solution,"  Canadian  J. Chem, 45, p. 2695


27.  Van Wazer, J. R., Phosphorus  and  Its  Compounds.  Volume  I, Inter-
     science (1961).

28.  Hendricks, S. B., and Hill, W.  L.,  "The Nature of Bone and Phosphate
     Rock," Proceedings National Academy  of Science of US, 36, p. 731
     (1950).	

29.  Rootare, H. M., Deitz, V.  P., and Carpenter, F. G.,  "Solubility
     Product Phenomena in Hydroxyapatite-Water  Systems,"  J. Colloid
     Sci.. 17, p. 179  (1962).

30.  McConnell, D., "Crystal Chemistry of Hydroxyapatite  — Its Relation
     to Bone Mineral," Arch. Oral  Biol.,  10,  p.  421 (1965).

31.  Berry, E. E.,  "The Structure and Composition of Some Calcium-Deficient
     Apatites," J.  Inorg. Nucl. Chem., 29, p.  317  (1967).

32.  Eanes, E. D., Gillessen,  I. H., and  Posner, A. S.,  "Mechanism of
     Conversion of Non-Crystalline Calcium Phosphate to  Hydroxyapatite,"
     Proceedings International  Conference on  Crystal Growth,  New  York
     (1967).

33.  Stumm, W., and Morgan, J.  J., Aquatic Chemistry, Wiley-Interscience
     (1970).

34.  Newesely, H.,  "Changes in  Crystal Types  of Low Solubility Calcium
     Phosphates in the Presence of Accompanying Ions," Arch.  Oral Biol.,
     Special Supplement, 6. p.  174 (1961).

35.  Mooney, R. W., and Aia, M. A.,  "Alkaline  Earth Phosphates,"  Chem.
     Rev., p. 433 (1961).

36.  Chave, K. E., Deffeyes, K. S.,  Weyl,  P.  K., Garrels, R.  M.,  and
     Thompson, M. E.,  "Observations  on the Solubility of Skeletal Carbon-
     ates in Aqueous Solutions," Science,  137,  p. 33 (1962).

37.  Simpson, D. R., "Apatite  and Octa-Calcium  Phosphate: Effects of
     Carbon Dioxide and Halogens on  Formation,"  Science.  154, p.  1660
     (1966).

38.  Somasundaran, P., "Zeta Potential of Apatite in Aqueous  Solutions
     and Its Change During Equilibration," J.  Colloid and Interface Sci..
     27, p. 659 (1968).
                                    75

-------
39.  Parks, G. A., "The Isoelectric Points  of Solid  Oxides,  Solid Hydroxides,
     and Aqueous Hydroxo Complex Systems,"  Chem.  Rev.,  65, p.  177 (1965).

40.  LaMer, V. K., and Smellie, R.  H.,  "Flocculation, Subsidence and
     Filtration of Phosphate Slimes," J.  Colloid  Sci.,  11, p.  704 (1956).

41.  American Public Health Association,  Standard Methods for  the Examina-
     tion of Water and Wastewater,  llth  Edition  (1960)  and 10th Edition
     (1955).

42.  FWPCA Methods for Chemical  Analysis  of Water and Wastes.  Federal
     Water Quality Administration,  Cincinnati, Ohio  (1969).

43.  Jenkins, D., Analytical  Methods. Berkeley: Sanit.  Eng.  Research
     Lab., University of California (1966).

44.  Winneberger, J.  H., Austin, J. H.,  and Klett, C. A., "Membrane Filter
     Weight Determinations,"  J.  Water Pollution Control  Fed..  35. p. 807,
     (1963).

45.  Basset Jr., H.,  "Beitrage  zum  Studium  der Calciumphosphate," ~L._
     Anorg. Chemie, 59, p.  1  (1908).

46.  Bard, A. J., Chemical  Equilibrium,  Harper and Row,  New  York (1966).

47.  Sillen, L.  G., and Martell, A. E.,  Stability Constants  of Metal-Ion
     Complexes,  London Chem.  Soc.,  Special  Publication  No. 17  (1964).

48.  Somasundaran, P., and  Agar, G. E.,  "The  Zero Point  of Change of
     Calcite," J. Colloid and Interface  Sci.. 24. p. 433 (1967).
                                   76

-------
                            SECTION XII
                             GLOSSARY
 Symbol                          Definltion
[Ca2+]in      Influent calcium  ion  concentration
[Ca2+]ss      Steady state calcium  ion concentration
[CaTJ         Total dissolved calcium concentration
CSTR          Continuously stirred  tank  reactor
CT            Total carbon = H2C03  + HCOg + COg"
DCP           Dicalcium phosphate
HAP           Hydroxyapatite
hp            Horsepower
kj            First equilibrium constant of phosphoric acid
k2            Second equilibrium constant of phosphoric acid
k3            Third equilibrium constant of phosphoric acid
a             1i ter
M9(OH)2       Magnesium hydroxide
mg/a          milligram per liter
nm            nanometers
OCR           Octacalcium phosphate
P.            Influent phosphate concentration
[P0^~]        Steady state P0^~ concentration
P             Steady state phosphate concentration
PT            Total orthophosphate  = H3POl( + H2PO^ + HPOj" + POj"
PVC           Polyvinyl  chloride
                                  77

-------
 Symbol                         Definition
pA            Negative  logarithm of a solubility product
pH            pH of zero point of charge
  zp c*
sec'1         Reciprocal seconds
SERL          Sanitary  Engineering Research Laboratory
TCP           Tricalcium phosphate
zpc           Zero  point of charge
y             Mi cron
                                 78  .

-------
                          SECTION XIII

                           APPENDICES

                                                             Page

A.  Walton Technique for Determining Stoichiometry'
    of Nucleating  Solid Phase 	   81

B.  Example of Manual Computation of Residual
    Phosphate Values  	   85

C.  Daily Wastewater Experimental Data  	   89
                                  79

-------
           APPENDIX A
WALTON TECHNIQUE FOR DETERMINING
   STOICHIOMETRY OF NUCLEATING
        SOLID PHASE [26]
               81

-------
Walton et^ al. [26] state that the energy barrier to nucleation, AQ°
(nucleatiorTJ" is given by the following expression:


         AG° (nucleation) = AG° (cluster) - AG°sp (solubility)  .   (12)


The terms on the right-hand side of Equation (12) can be evaluated ,as
follows:  The standard free energy of the cluster, AG° (cluster),  depends
on th'e type and number of interacting ionic species.  In general,  if
ionic specie Aa+ interacts with ionic specie B^" to form a neutral ionic
aggregate, the metastable equilibrium is given by the equation:
                    xbAa+ + xaBb" + (Aba+Bab~)    .                 (13)
                                              /\


The condition for equilibrium is AGX  = 0, or
                                    o

                AGXQ = bxyA + axpB - xy AB - Za-jj =0               (14)


where:

    yA» ^B' an<^ yAB are ^e cnemi'ca^ potentials of species Aa , B ~, and
    (Ag+ Bb-)x, respectively,

    oij is the interfacial energy of the interface between the i and j
    phases.

After equating U/\B = ^°AB (solid). Walton defines the total standard
energy of the critical cluster AG x  as:


            r  AG°Xo (cluster) = -RT ln[A]b[B]a + z ^   .         (15) .


Also
                   AG°sp (solubility) = -RT In K^                 (16)
where KS  is the solubility product.

Now, substituting Equations (15) and (16) in Equation (12) we obtain
Equation (17):
                                  82

-------
       AG° (nucleation) = -RT ln[A]b[B]a + z -& -  RT  In  K     .     (17)
                                              x           s p

Rearranging Equation (17)


           ln[A]b[B]a = z ii + In K   - AG°  (nucjeation)     _      (1Q)
                          XKI        S p          KI


The right-hand side of Equation (18) is constant at constant  temperature
for any one critical size nucleus of constant  composition and interfacial
energy.  Thus:


                    ln[A] = - | ln[B] + constant   .                (19)


Thus, if the above criteria hold, a plot of the  logarithm of  the  critical
concentration of the cation A against that of the anion B should  yield a
straight line.  The slope of this line gives  the stoichiometric ratio of
the initial phase.

The above derivation does not depend upon the  detailed nature of  the
nucleation process and similar reasoning can  be  applied to  systems  in
which the nucleus contains more than two ionic species.

Walton and coworkers [26] visualize the formation of a critical nucleus
and the beginning of calcium phosphate precipitation as follows:


                ax HgPO^ + bx Ca(OH)2 + x (c  - 2b)  H20

             [CabH3a_2b (POJ3C H20jx  (critical nucleus)

                             precipitation                         (20)


At the onset of nucleation, the relation between the concentrations of
the species is given by Equation (20) i.e.,


           log ([Ca2+][OH"]2)b  ([H+][H2PO;])a = constant  .        (21)


Equation (21) could be rearranged


        b log ([Ca2+][OH~]2) + a log ([H+][H2PO;;] \ = constant      (22)


        log ([Ca2+][OH~]2) = - (£) log ([H+][H2PO;])+ constant     (23)
                                  83

-------
Thus  if  log ([Ca2+][OH"]2)is plotted against log ([H+][H2POi;] )for a
series of experiments, a straight line should result with  a slope of
a/b that gives the  ratio of phosphate to calcium in the nucleating or
steady state phase.  Data from Walton et_ aj_.  [26], together with data
from the chemically-defined systems of this  work, are presented in
Figure 26.  The data from the chemically-defined systems appear in the
text of  this report (Figure 9) on a somewhat more expanded scale.
      18
      16
   '=  14
   o
   (O
   O
      12
      10
                       Data from Walton et
         Ca/P =1.49
                     Ca/P =1.44
Run
0
A
°C
37
1.9-23
Time
min
1,440
15
Reactor
BATCH
CSTR
Initial Conditions
M x 103
Ca
.04-2.0
2.1-2.4
Mg
0.0
0.0-0.6
PT
1.0
0.3-0.5
°r
0.0
0.0-14.8
                n
        FIGURE 26.
13
            15
17
19
21
CRITICAL IONIC CONCENTRATIONS AS
 DETERMINED  BY WALTON et_aL  AND
 BY THIS INVESTIGATION
                                84

-------
           APPENDIX B

EXAMPLE OF MANUAL COMPUTATION OF
    RESIDUAL PHOSPHATE VALUES
                85

-------
 In this presentation the manual computation of residual dissolved phos-
 phate concentrations will be illustrated by the following example:

 Initial Conditions:  Ca, 2.19 mM; Pj, 0.38 mM; Mg, 0.59 mM; CT? 5.95 mM.

 Conditions of Precipitation:  pH 10.

 Predictive Equation:


    3 log{[Ca2+]in - 1.5 [P1f) - PS$]} + 2 log[POj~] = - 23.56      (10)


 Equation (10) is solved for Pss by trial and error.  Thus, assume

 [Pin - P$SJ = 0.9 Piri, i.e., PSS = 3.8 x 10~5M.  Now, compute


              [Ca2+]ss = (2.19 - 0.9) x 1.5 x 0.38 x 10"3

                         1.68 x 10"3M  .

Substituting this value for [Ca2"1"]-,.. in Equation (10) and solving for
                       [P0j~]ss   2.34 x 10"8M  .


The fraction that [P0ij~]ss is of Pss at any pH can be computed from
equilibrium constants of phosphoric acid and [H+].
Thus,
                          PSS = 4.9 x 10"6M
This value of PSg is not within 5% of the assumed PSs value.  Therefore,
the computation is repeated using,
                          P._ = 4.9 x 10"6M
                           j o
This yields
                              = 5.26 x 10"6M.
The difference between this value and the assumed value is again greater
than 5% so that a third iteration is necessary using,
                                  86

-------
                         PSS = 5.26 x 10~6M  .
The computation now yields


                         PSS = 5.26 x 10"6M

an identical value to that assumed.

The computation of residual dissolved phosphate from Equation (10)  can
be modified when necessary to account for the calcium that is required
to precipitate calcium carbonate.  The solubility product chosen for
calcium carbonate was


                      [Ca2+][COf ] = 2.6 x 10'6                    (24)


a value typical of a destabilized calcium carbonate formed under the
precipitation conditions encountered in wastewater precipitation
processes [10,22].

Now, setting the [Ca2+] incorporated into calcium carbonate precipitate
equal to x, we have


                  [Ca2+ - x][CO§" - x] = 2.6 x 10"6   •             (25)


This quadratic equation can be solved and the amount of calcium removed
as calcium carbonate precipitate determined.

Additional corrections for the complexation of calcium become significant
at pH values of greater than pH 10.  The various complexes of calcium
can be regarded as removing the calcium from the sphere of activity in
precipitating phosphate.  The complexes considered were:  CaOH+,
CaC03(aq), CaHC03, and CaPOi;.  Thus,
     [CaTJ = [Ca2+l + [CaOH+] + [CaC03(aq)] + [CaMCO^] + [CaPO^]   (26)


where [Cay] is the total dissolved calcium concentration.  The equili-
brium constants for these complexes are as follows:

    Ca2+ + OH" = CaOH+             KCaOH+ = 23-5

    Ca2+ + CO2' = CaC03(aq)        KCaC03(aq) = 1>59 *

    Ca2+ + HC03 = CaHC03           KCaHCO" = 18'2
    Ca2+ + POr = CaPOl            Kr,Dn- = 2.89 x 104
                                   87

-------
From the mass balance equation for total dissolved calcium, Equation
(26), and the equilibrium constants for calcium complexes the following
expression for [Ca2+] is obtained:


[Ca2+]                                                             (27)
[Ca?]
1 + KCaOH+ [0irl
+ KCaC03(aq)
[C0|-] H
h KCaHCO| [F
1C°3] + KCaPOi; Cp°'"l
The most significant terms in this equation are those for the CaC03(aq)
and CaPOi; complexes.  The CaOH+ and CaHCOlj complexes can be neglected
at pH values below 11.  Thus a simplified form of Equation (27) can be
used

                                     [CaT]
            [Ca2+] = - ! -    '    (28)

                     1 + KCaC03(aq)                     "
For the ranges of concentrations and pH conditions in the experiments
conducted here, neglecting the CaPO£ complex, would introduce an error
of 52-102 in values of [Ca2 ].  Thus Equation (28) could be further
simplified to


                                     [CaT]
                    [Ca2+] = - 1 - —    .             (29)
                                                -
                             1 + KCaC03(aq)


Computation of the initial [Ca2+] available for precipitating calcium
phosphate, therefore, consists of calculating the amount of calcium
removed by calcium carbonate precipitation and then converting the
residual dissolved calcium value to [Ca2+] by correcting for calcium
complexes by one of the Equations (27, 28, or 29).
                                   88

-------
            APPENDIX C



DAILY WASTEWATER EXPERIMENTAL DATA
                 89

-------
 8 -

 7*
S A V\ 	 S\—
> 	 ^ 	 -cj tr
_ Run 2
>-^ — -O— —
trr^fc— — gt^r5^
*- 	 ^ 	 If 	 -V
	 $—==£ §_
A ' ' i '
-ti
	 T7

	 0 i i i , ,i ,
TO
10
  I


 9


 8


 7



 9


 8
       Run 3
      Run 4
                                    _L
                                  I
 'i
T      I       T
.	.Q.     A
       Run 5
    I       I       I
'ilot Plant Streams
                       o  Primary Effluent
                       a  Activated Sludge Effluent
                       A  Lime Precipitation  Effluent
                       S7  C02 Recarbonation Effluent
                       0  ClrinoDtilelite Effluent.i ,
                             5      6

                            TIME,  days
                 8
                                                        9 /'
                                                      10
        FIGURE 27.   EFFECT OF  UNIT  PROCESSES ON pH
                               90

-------
0.5
0.4 -
0.0
„ "••» s
t! C
I 0.3
Q-
1 0.2
P "''l
1
0.0
] "^ -n 	 ^* -^^.j Pilot Plant Streams
"- o Primary Effluent
Q Activated Sludge Effluent
A Lime Precipitation Effluen.
Run 2 v C02 Recarbonation Effluent
0 Clinoptilolite Effluent
* i i i i i i i
                                                     10
                        TIME, days
     FIGURE 28.  EFFECT OF UNIT PROCESSES ON TOTAL
                PHOSPHATE RESIDUALS (RUNS  1, 2 AND 3)
                         91

-------
                                      0.4
A
LU

O
Q.
_l
1


0

0

u
o


(
.3

.2

.1
0*
Primary E
Run
^fluent -
4
"Lime Precipitated,
bonation, and
-Effluents
!h w» ^
y ^i ^
•° 1 2
TIME,


C02 Recar
Clinoptilol

?»
Jl
3
days

g
*P
4

ite
—
r
"5
1
LU
1—
D-
o
o.
^
o
5

0

0

0
0


<
.3

.2

.1
.0<
Primary
Run
Effluent -
5
Lime Precipitation"
Effluent
-
__ — • a—
r* i
1 2
TIME,

-
	 & 	 y
i *\
3 '
days
       FIGURE 29.   EFFECT OF UNIT PROCESSES  ON TOTAL
                    PHOSPHATE RESIDUALS (RUNS 4  AND 5)
o
40
co
                                                Cllnopt. Eff.-
                                                j	r    i	
        2468
         TIME, days
      Activated Sludge
          Effluent
 2468
  TIME, days
2468
 TIME, days
Lime Precipitated    C02  Recarbonated and
   Effluent      Clinoptilolite Effluents
        Activated Sludge Effluent
                                0.005-    C02 Rec.
                                           Clinoptilolite
                                                      Eff.
                                         Run 2
                                        Lime Ppt.  Eff.
                                           i      i
            234
             TIME, days
                234
                 TIME, days
      FIGURE 30.   EFFECT OF UNIT PROCESSES  ON DISSOLVED
                   PHOSPHATE  RESIDUALS (RUNS  1  AND 2)
                             92

-------
                                                 8
                                                                  10
   0.012


   0.010


   0.008^


   0.006


1  0.004
 *

a  0.002
                                  Run 3
                C02 Recarbonated Effluent
     Lime Precipitation
    "Effluent
                                                Clinoptilelite
                                                Effluent
o
GO
I-H
o
    1
0.3


0.2


0.1

0.0


0.4
                  Run 4
            Primary Effluent
        1
0.2
         Run 5
       Primary Effluent
          i       i
   1      2      3

        TIME, days
                                      0.002
                                           0.001
                                                Run 5
                                                  Lime Precipitated
                                                  Effluent
                                               1
  2      3

TIME, days
     FIGURE 31.   EFFECT OF  UNIT PROCESSES ON  DISSOLVED
                   PHOSPHATE RESIDUALS (RUNS 3, 4 AND 5)
                              93

-------
o
o

CO
I—I
o
                 Pilot  Plant Streams

                 o  Primary Effluent
                 n  Activated Sludge Effluent -
                 A  Lime Precipitation  Effluent
                 v  C02 Recarbonation Effluent-
                 0  Clinoptilolite Effluent
                         456
                             TIME, days
                               8
10
      FIGURE 32.
EFFECT OF  UNIT PROCESSES ON DISSOLVED
CALCIUM CONCENTRATION (RUNS 1, 2 AND 3)
                              94

-------
 1.8
  Pilot Plant Streams
•  o Primary EFfluent
  & Lime Precipitation Eff
  ^7 C02 Recarbonation Eff.
          234

            TIME, days
                                         Run 5
                                   2      3

                                   TIME, days
FIGURE 33.  EFFECT OF UNIT PROCESSES  ON  DISSOLVED
             CALCIUM CONCENTRATION (RUNS 4 AND 5)
                          95

-------
                    Run 1
                                               Run 2
2.0
_C02 Recarb. Eff.
   12345678
                                      12345
                                       3.0
                                           i     r
                                         Prim. Eff.
                                           Clinopt.  Eff.
C02  Rec. Eff.
     Clinipt.  Eff.
     C02 Recarb.  Eff.
     Lime Ppt.  Eff.
2.0
                             5     6      7
                             TIME, days
                                                             o
              Run  4
                                              Run 5
3.1
        ^      I       T
        Lime Ppt.  Eff.
    Prim. Eff.
    	I	I
                   I
                                       2.5 -
   12345
          TIME, days                            TIME, days    I
   FIGURE 34.   EFFECT OF  UNIT  PROCESSES ON ALKALINITY
                            96

-------
  SELECTED WATER
  RESOURCES ABSTRACTS
  INPUT TRANSACTION FORM
                                          ;. Report No.
                                                             w
       '   Calcium Phosphate Precipitation in
          Waste-water Treatment
          Menar; Arnold B.  and Jenkins, David
          California University
          Sanitary Engineering Research Laboratory
                                                             .5., KsportDate
                                                             -5     June 1972
                                                             8.  Performing Organization
                                                                             72-6
                                                              ..     ...
                                                              EPA WQO 1 7080 DAR

                                                             Is Type , f Repotl aad
                                                                Period Covered
   12. 'Srinsorir.r Organisation
                         Environmental Protection Agency report
                         number. EPA-R2-72-064,  December 1972.
      .,.-• , • \_ i
            This investigation examined the precipitation of calcium phosphate both
   from chemically-defined solutions representative of wastewater composition and from
   wastewater.   The steady state solid phase that controlled dissolved phosphate residual
   was an amorphous tricalcium phosphate.  The solubility of this phase, determined
   from chemically-defined systems, was used with success  to predict dissolved
   phosphate residuals from both chemically-defined systems and wastewaters.  Sus-
   pension recycle was found to result  in lower dissolved phosphate residuals, but poor
   suspension settling below pH 10 made this process  difficult to maintain.  Suspension
   settling was enhanced by Mg(OH)2 precipitation but not by  CaCOS precipitation.   In
   wastewater of moderate alkalinity and hardness, a  phosphate removal in excess of
   80% was consistently achieved at pH 9. 5 with lime doses of, at the most,  200 mg/
   liter as CaCO3.  The  overall phosphate  removal performance was dictated by the
   performance of the precipitation reactor and its ensuing sedimentation basin.
   Phosphate-containing  particles that  escaped sedimentation could not be  removed by
   filtration because they dissolved rapidly during the recarbonation process that
   necessarily precedes  the filtration step.
   173. Descriptors
                 * Waste Treatment — Sewage Treatment
   I7b. Identifiers
                 Calcium phosphate,  Tricalcium phosphate, Calcium carbonate,
   Calcite, Magnesium hydroxide
;;.. fOlVRR Fit-Ill
                     ,!  05D
                             
-------