PHOSPHORUS  REMOVAL
   TECHNOLOGY TRANSFER DESIGN SEMINAR
   U.S. ENVIRONMENTAL PROTECTION AGENCY
   OFFICE OF RESEARCH AND MONITORING
   NATIONAL ENVIRONMENTAL RESEARCH CENTER
   CINCINNATI, OHIO

-------
   GENERAL INFORMATION ON PHOSPHORUS REMOVAL
               Prepared for the
    U. S. Environmental Protection Agency
     Technology Transfer Design Seminar
                 Presented at
   Chicago, Illinois, November 28-30, 1972
    National Environmental Research Center
Advanced Waste Treatment Research Laboratory
      Office of Research & Monitoring
             Cincinnati, Ohio

-------

-------
                GENERAL INFORMATION ON PHOSPHORUS REMOVAL
                             I. J. Kugelman
      The key role of phosphorus in the process of eutrophication (aging
of lakes and impoundments) has been known for many years.  However,  until
recently, eutrophication of the nation's waterways was not a significant
problem.  Consequently, control of the phosphorus level in streams and lakes
was not considered an important pollution control problem.  Since the con-
clusion of World War II, however, the rate of eutrophication has increased
to the point where it is a major water quality problem,  Although o':her
                                                                  (1 2)
nucriencs pi ay a role in eutrophication, recent articles by Sawyer  '
have indicated that much of the recent increase in eutrophication rate is
linked to significant increases in phosphorus discharges to rivers and lakes.
Sawyer has shown that virtually all of the increase in phosphorus discharge
is due to the activities of man and has termed the resulting eutrophication,
cultural eutrophication.
      The major sources of phosphorus contributing to eutrophication are
domestic sewage and agricultural runoff.  Domestic sewage is the primary
source in critical areas, and only this source will be discussed here.
Phosphorus gains entrance to sewage from human body wastes (primarily urine)
and through the use of condensed inorganic phosphate compounds as builders
in detergent formulations.  Each of these sources accounts for about half
of the phosphorus in domestic sewage.  Thus, while elimination of phosphorus
from detergent formulations would be helpful, it would not be the total
answer to the eutrophication problem.  Treatment of domestic sewage to re-
move a significant portion of the phosphorus contributed by human wastes
and detergent builders would, however, have a significant effect on
eutrophication rate.

Phosphorus Removal in Conventional Treatment
      Removal of any pollutant from wastewater requires that it be converted
to either an insoluble gas or an insoluble solid.  Because none of the
chemically stable forms of phosphorus is a gas at normal temperature and

-------
                                  -  2 -
pressure, removal from wastewater is  dependent  on  formation  of  an  insoluble
solid.  Less than 10 percent of the phosphorus  discharged  to municipal  sewer-
age systems is insoluble and none of  the conventional  treatment  techniques
is particularly effective in insolubilizing this nutrient.   Thus,  phosphorus
removal in conventional treatment systems is relatively poor.  Primary  treat-
ment can remove only the 10 percent of the phosphorus  which  is  initially
insoluble.  During secondary treatment,  phosphorus removal is achieved  by
synthesis into the biomass followed by sedimentation  and sludge wasting.
However, municipal sewage contains a considerable  excess of  phosphorus  over
that required for biomass synthesis during complete utilization of the
organic carbon present; thus, removals are generally limited to 20-40
percent   .  Studies indicate that biological systems  have the  capacity for
                                                             (4)
much higher removals through the mechanism of "luxury uptake"   .   However,
attempts to implement this phenomenon in actual plants have  not been
successful   .  Indeed, a recent study has illustrated that  luxuty uptake
may actually be a chemical precipitation rather than a biologically mediated
 ,         (6)
phenomenon

Phosphorus Removal by Chemical Precipitation
      Fortunately, phosphorus forms essentially completely insoluble precipitates
with a number of substances, thus high levels of removal can be obtained when
appropriate doses of the proper chemicals are applied.  A large variety of
chemicals can be utilized for this purpose but  economic factors dictate the
use of salts of iron, salts of aluminum or lime.
      For both the salts of aluminum and iron,  the required  dose is set by
similar factors.  The major control is the stoichiometric reaction between
the metal salt and the phosphorus compound as illustrated in equation 1.

              M+3 + P0"3 ............ >MP0                             (1)
This reaction indicates a requirement of 1 mole of the metallic cation for
each mole of phosphorus in the wastewater.  However, a number of factors act
to increase the dose of metallic salt required.  As indicated by equation 2,
the metal salt reacts with the alkalinity in the wastewater to form insoluble
hydroxide .

-------
                                  - 3 -

       R+3 + 3HC03"1 ............ > M(OH)3I + 3C02                   (2)

Although this is a competing reaction with the phosphorus precipitation,  the
metal consumed is not wasted as the hydroxide formed is an excellent flocculant
of the phosphorus precipitates of equation 1.  Another factor which tends to
increase metal dosage over that theoretically required is pH.  Figure l
illustrates that aluminum phosphate is least soluble at pH 6 and iron phos-
phate is best precipitated at pH 4   .  It is usually best to use additional
metal salt rather than adjust the pH of the sewage from its normal range of
7-8.  Another factor which influences coagulant dose is the presence of sus-
pended organic solids in the wastewater.  A portion of the metal salt added
will be required for coagulation of these materials, again increasing the
dose requirement.
      The dose required to achieve any specified degree of phosphorus removal
cannot be specified with great precision because of the factors discussed
above.  It is best to conduct bench-scale jar tests on the wastewater to be
treated to establish the metal salt dose requirement .  Average of results from
                                                                        (3)
many installations which can serve as a rough guide are given in Table 1

                                TABLE 1
                  AVERAGE METAL SALT DOSE REQUIRED FOR
                  SPECIFIED PERCENT PHOSPHORUS REMOVAL
                                                 Dose Metal Salt
  %  P  Removal                                 Mole Ratio M  /P
      75                                             1.4
      85                                             1.7
      95                                             2.3
      The reaction of lime with wastewater constituents is given in equations
3 and 4:
      3HP04"2 + 5Ca+2 + 40H"1 ..... > Ca (OHMPO^) ^ + 3H 0          (3)
      Ca(OH)2 + Ca(HC03)2 ......... > 2CaC03^ + 2^0                  (4)

-------
                            -  4 -
                                                       CNJ
LJ i ',', ',_. 1.._!__ .'in ' ' '.'_.'--,.	'-M'l.1  ' -1	mil i 11
                                                                    OJ
                                                                   cs:
                                                                £'- CD
                                                                CO —•
                                                                CO

                                                                c5
                                                                   c_->
                                                                   OJ
                                                                LtJ
                                                                c~
                                                                D_
                                                    CD

                                                    CD
                31VHdSQHdOIUMO 1VROIS3H

-------
fc.O
       Ml IfUL
       p*, r* n n n< \ f n c  r n n r. /i  n r\ \ A f  i ^ f A
       KEwiuVAL FtiOlvl  iil-iW  WA
                  A  TOTAL PHOS.IUNFILT.
                  A  TOTAL PHOS.JF1LT.)
                  e  ORTHOPMOS. (UNFILT.
                  o  ORTHOPHOS. (FILT.)
                                              12
       FIGURE: 2
pll

-------
      Equation 3 represents the formation of the insoluble compound calcium
hydroxyapatite.  The kinetics of this reaction as well  as  the stoichiometry
are pH dependent.  Elevated pH values 9.5-11.5 are required to ensure good
phosphorus removal by precipitation and liquid solids separation  techniques.
                                   (8)
Figure 2 illustrates this pH effect   .  Thus, while pH was a secondary factor
in establishing the dose requirement  with iron or aluminum as the precipitant,
it is the primary factor when calcium is utilized.
      In order to reach the pH range  which will ensure good phosphorus removals,
the alkalinity demand of the wastewater must be met, as illustrated in equation 4.
Because the bicarbonate concentration in wastewater is  much higher than the
phosphorus concentration (several hundred mg/1 vs. =s 10 mg/1), therefore,  most
of the lime dose required to achieve  effective phosphorus  removal is used  to
meet the alkalinity demand.  Here a  situation exists similar to  that occurring
when metal salts are used for phosphorus removal; i.e., the product of the
cation alkalinity reaction serves as  a flocculant of the phosphorus pre-
cipitate.  Thus, the reaction of the  lime with the alkalinity does not
completely waste the lime.
      When reaction 4 is complete, the pH of the wastewater is in the range
of 9.5-10.0.  For wastewater with moderate to high levels  of alkalinity
(> 200 mg/1), sufficient calcium carbonate is formed under these  conditions
to effectively bring down the hydroxyapatite.  After sedimentation, the pH
is reduced to < 8 by the addition of CO .  This process is referred to as
single-stage lime precipitation.  If the alkalinity of the wastewater is low
> 150 mg/1, however, insufficient calcium carbonate is formed.  For this
situation, the pH must be raised above 11 which will ensure precipitation
of magnesium hydroxide according to equation 5:

          Ca(OH)2 + Mg+2	> Ca+2 + MgtOH)^             (5)

Magnesium hydroxide is a gelatinous precipitate which will effectively
scavenge hydroxyapatite from solution.  Adding excess lime to raise the pH
to high levels results in a wastewater containing excess calcium.  The calcium
is reduced by carbonation with CO  .  As illustrated in equation 6, this re-
duces the pH and results in a second precipitation of calcium carbonate:

-------
                                  - 7 -

        C02 + Ca+2 + 2 OH" 	> CaC03^'+ ^0               (6)

The second precipitation is carried out in a sedimentation tank separate from
that in which hydroxyapatite, calcium carbonate and magnesium hydroxide pre-
cipitation takes place.  The recarbonation reaction is complete at pH 9.5-10.0,
which is the same end point pH as in the single stage process.  A second dose
of CO  is then applied to reduce the pH below 8.  The final pH reduction in
the two-stage and single-stage processes is utilized to prevent post precipi-
tation of CaCO., in other treatment units.  Flow diagrams of the single-stage
and two-stage lime precipitation processes are given in Figure 3.
      Table 2 gives a list of chemicals which are useful for phosphorus
precipitation.  Small doses of organic polymer may also be required to aid
flocculation.  In selecting the chemical for use at any particular site, the
factors listed in Table 3 should be taken into account.  It is clear that
the cost of the chemical is only one factor among many which bear on the
ultimate cost of phosphorus removal.  Of prime importance is dewatering and
ultimate disposal of sludges.  Here, lime has an advantage because lime
sludges are more easily dewatered than iron or aluminum sludges.  In addition,
lime sludge can be recalcined to reusable lime in an incinerator.  However,
large capital expenditures such as an incinerator can only be justified at
large treatment plants.
      Phosphorus removal is achieved by precipitation followed by liquid
solids separation.  For the most part, the usual liquid-solids separation
equipment in a treatment plant can be utilized for phosphorus removal.  This
results in a considerable savings in capital as well as integration of
phosphorus removal into conventional treatment plant operation.  In addition,
it has been found that the use of chemical precipitants in conventional treat-
ment can markedly upgrade performance of the treatment plant.  This results
from coagulation of organic suspended and colloidal solids by the chemicals
added to precipitate phosphorus.

-------
             SINGLE STAGE LIME  TREATMENT

            LIME  SLURRY
WASTE
WATER
STE
            ph-VARIABLE
           SLUDGE
        F r n/i F n A I
         i: Kill ML
                 FILTER
                    r- T A n* r  & ?
                    biAut  i.I
             T: r' r, "rnn r r *,~
              i '-'if ;' ) i* (••] «
              t t. L, i 1 i ', ^ t V i
LIME
            SLURRY
CARBON
           V
           DIOXIDE
     SLUDGE ph   11
          1
       [ill    10
           ¥
         SLUDGE
           u
         i:/iT|\fi!:M
         t./i t ivi L. l\i
          ¥
       SLUDGE
   SECOND  STAGE
CHEMICAL  TREATMENT
                              CLARIFIED
                                WATER
                                              CLAR1F1'
                                               WAT Hi
                                            FILTER
                          FIGURE: 3

-------
                     - 9  -




                   TABLE 2




        CHEMICALS FOR PHOSPHORUS REMOVAL
Ferric Chloride



Ferric Sulfate



Ferrous Chloride



Ferrous Sulfate



Alum



Sodium Aluminate



Steel Mill Pickling Liquor



Lime
Fed,
FeSO,
NaAlO,
FeCl. + FeSO.
    2


Ca(OH),
                    TABLE 3




       FACTORS AFFECTING CHOICE OF CHEMICAL


              FOR PHOSPHORUS REMOVAL




   Influent Phosphorus Level



   Wastewater Suspended Solids and Alkalinity



   Chemical Cost Including Transportation



   Reliability of Chemical Supply



   Sludge Handling Facilities



   Ultimate Disposal Methods



   Compatibility with Other Treatment Processes in Plant



   Potential Adverse Environmental Effects

-------
- io -
                                    CO
                                    n>
                                    C£

                                    O
                                    :r
                                    Q_
                                    CO
                                    O
                                    IT
                                    D.

                                    U_

                                    O

                                    _J

                                    O
                                 <*•

                                 UJ
                                    <
                                    U-J

                                    IT
                                    O
                                    LO-

                                    CO
                                    O
                                    u
                                    O

-------
                                - 11 -
      Figure 4 illustrates a conventional treatment plant with the three
general sections in which phosphorus removal can be carried out.   Chemical
can be added either:
      a.  Just before the primary tank with removal taking place  in the primary.
      b.  In the secondary (biological) section of the plant with removal  in
          the secondary sedimentation tank.
      c.  In a tertiary stage as was discussed in the section on  suspended
          solids removal.
      Table 4 illustrates typical results obtained with phosphorus removal
in the primary, secondary, or tertiary.  As can be seen, good removals are
obtained in all sections, however, the lowest levels of phosphorus remaining
are achieved in the tertiary addition.  One reason for this is that a filter
is usually included in the tertiary plant, thus better removal of fine preci-
pitates is achieved.  In addition, when the flow reaches this section of the
plant all the complex phosphorus forms which are more difficult to precipitate
have been hydrolyzed to orthophosphate which is the easiest to precipitate.
      In plants where removal in the primary was practiced a major effect  to
note is the significant increase in BOD and suspended solids removal achieved
over the removal usually obtained in the primary tank.  This may  be important
in helping to meet water quality standards for BOD and suspended  solids if the
treatment plant is overloaded.
      In plants where the chemical is added in the secondary section, it has
been observed that much more stable operation of the activated sludge is
obtained than before chemical addition.  The effect of the chemical is to
weigh the sludge down, preventing its loss when a filamentous or dispersed
growth predominates.  Even in plants which have historically exhibited
excellent performance, chemical addition has improved performance by helping
maintain a higher concentration of activated sludge in the aeration tank.
This is illustrated by the data in Table 5 from parallel operation for one
                              ( 9 )
year at Penn State University     .  It has been found best to add the
chemical between the boilogical reactor and the final sedimentation tank
rather than at the head end of the secondary tank.
      Table 6 summarizes the advantages and disadvantages of carrying out
phosphorus removal in the various sections of a treatment plant.

-------
                                                     -12  -
        to
        W
        c_>
        53
        w
a
        to

        g
        o
        co
        o
Q>
V
C
C)
CJ


V}
•


Q
O












rH
O

C)
Pi
rH
M
O
£ j
CJ

e>

rH
 co
H
O
-H
rH
CJ
cj
•H
C
O
u
f—l

G)
CS| rH
rH H -H
' CJ

60
•H C_>
rj •
o o
•H O
H -H
c: rd
o o
60 •(->
PRIMARY
Place
C 60 -
TH C M
rH -H O
>x X *J
rj tn rt
M cj 0)
0 ti «



CM n 
d c
•« CJ ^ O
gj co tn
Q CO to JJ
^3 C to rt
O O rj ,c
o n c o
U O q -H
to fi< H oi













































><
TERTIAR
                                                                                                     in
oo
o
in

o

rH
                                                                                                      V

                                                                                                      G
              CJ


             3
.
^
£
4J
C
a
Q
C-J

D
P3
W
tn
w
p^
M-l
•H
rH
O

A
0)
O
.C
CO
H

CJ
^t
03
H-3


O
-H
r*
0

*k
C
o
C
o
rQ
t)
I-J

-------
                 TABLE  5

COMPARISON OF ACTIVATICD SLUDGE PERFOR1-IANCE

      WITH AND WITHOUT ALUM ADDITION
Parameter
S.S.
BOD
COD
Soluble P
Total P
Influent
mg/1
110
71
172
6.7
10.0
                                         Effluent,  mg/1
                                      Alum	Normal

                                      22            26

                                       9            13

                                      55            68

                                      0.36          6.7

                                      1.4           7.3

-------
                                                                      14
vD
H

CM

^
O

CO

cj


rH
CO
AJ
0)
S

4J
R
a)
•H
o
oj
U-l
U-l
w

4->
CO
o
CD
.u
cO
t-l
4J
CO
r:
o
t-
0)
«

&
a)
J>
o
o
a)
t4

a)
6
•H
 O
T.1
 R
 C)
 O
 o
CO
                                            o
                                             CO
                                             0)
                                             ti
                                             o
                                            ,H
                                                  fsl   ro
              O
              CO
              o
              «

              rH
              CO
              O  ^
              •H  )4
              i:  tj
              ai  n
                                                  U
                                                      P-i
-H   a)
r-l   M>
•H  13
.n   3
 CO  rH
4->  to
CO
    13
'O   d
 a   i->
 >   rt
 o   >
                                                            0)
                                                            60
                                                            CO
                                                            CJ
CO
tj
r\\
*u
tH
CX
rH
rH
cO

O
•-J 4J
It
rj a)
(. r-l
•r 1 .0
>4 CVJ
n< o
vl
rH
fv,
(X,
<
0)
*[ 1
L>
c:
r I
Vi'
p,
Ul
»-j
C/J

O
t-l
n ai
d ,'•
f'1 O
i:
'v.1 0
o i ;
to
Cj HI
C) 'vl
>< -, i
C) . 1
t: o
M CO
KJ
r-l
r.<
^U
4J
J'i
**-<
U(
O

01
0)
)^
to
C)
(-1

4)
10
c>
J<
0
K-l
                                                                      tr
                                                                      a
                                                                     .u
                                                                     o
                                                                      M
                                                                      d)
                                                             rH
                                                              o
                                                             p-l































CO
w
o
<;
f-l
^
<2
>
§
to
1 — f
Q























4-)
w
o
o
rH
CO
ft 5
co ft,
•H CO
4J U
• .
M
C) 4->
H «
cu
•§>
•H
W
•
rH

4J
W
60

O
4J

rH
O
M
•U
d
o
. 0
>J
t-i a
w <%
•u
C tH
0 D
C.I M-4
t) QJ
to >-i
CO
O

rH
R
O
•H
•U
w
N
-H
t — (
•H
*J
t>
4J
R
8)
-T-l
f»J 0
It tl
W lu
{.' «w
s 1 10
n
fXi
4J
Dl
rO
C)
I-J
60
CO
^i
CO
cu
r^l
rH
CO
l-i
o
S
4J
CO
a)
•a
*H
r— <
rM
•
cs















rH
*^-^
tJ}
E

rH



PM

















tH
10
il
C)
^•*

144
0
















O
3
CO
0

>,
CO
B

rH
cO
JJ >.
O 4-1
S T-l
u
l*-4 -rl
0 X
o
(U EH
to
o :ri
13 fX
V-J
CD ;*
> 0
O i-l

CM






M
O
>4-l
•xi
a)
X
•H
;> t;
cv o
Ql -il
cv; 41
rJ
Vl r-l
ci :i
I. u
:•• ci
r ^ Cl
o .-i
fX! (X
















ti_i
HH
o

cu
CO
3
CO
CJ
a)
P3

CJ
fi
-H
rJ

C)
to
p

4J
O
R
R
d
O

CO



o
4_)
4-J
rH
y
CI
_ _j
^~1
in
144
1-1
f'J
a>
jj
o
-';
CI
to
•1.1
: i
t i
to






























p^
CU

o
>
•H
to
CO
V
o
X
W


a>
N)
•VI
;)
tH
to
>,
t-J
«
I :
•r-l
M
Pi

I*
rt
'4 j
(•J
CJ
4 1
C\)
t%
C)
W
                                            r-l    CNl

-------
                                 -15 -
      The cost of adding phosphorus removal to existing treatment  plants
is given in Table 7.  The lowest cost option is addition in the secondary
because capital expenditure for chemical feed equipment and extra  sludge
handling equipment are least in this mode.   However,  removals below 1  mg/1  P
are difficult.  The most costly option is tertiary because of the  high capital
cost.  However, an excellent quality effluent is obtained which is very low in
phosphorus, BOD and suspended solids.  Note the higher tertiary cost here than
for the upgrading system.  This is due to higher chemical doses, and additional
sludge handling difficulties.  Table 8 lists a number of plants where phosphorus
removal is being conducted or is in the planning stages.

-------
                                 - 16  -
                            V.
                              TABLE  7


                    ECONOMICS OF PHOSPHORUS REMOVAL
                     Cost Cents per 1000 Gallons
Plant Size,
      iGD

Chemical Addition
    Point	

Primary

Secondary

Tertiary
            10
   100
6-8 .
6-7
28 '
3-5
3.5-4.0
12
3-4
3.0-3
6

.5

                               TABLE  8
          INSTALLATIONS  1MIACTICING  OR PLANNING PHOSPHORUS RKMOVAL
     Lor at 3 on
  Chicago, Illinois
  Seatt -1 e, Washington
  Pomona,  Ca J jfornia
  S.  Lake  Talioc,  California
  PiscaLtaway >  Mary.l and
  Colorado SprJngs, Colorado
  Richardson, Texas
  Blue P.laJnr.
  El  Laju), Texas
  Hal field l\vp., Pa.
Capacity, MOD
    30
    20
     2
     7.5
     5
     2
     1
   300
    0.5
    5.0
Status
Design
Design
In Operation
In Operat ion
Construct ion
In Operation
In Operation
In Operat. ion
Construe Lion
In Operation

-------
                                 - 17 -
                               BIBLIOGRAPHY
 1)  Sawyer, C. N. "ABC's of Cultural Eutrophication Part I Cultural  Changes,"
     Water & Sewage Works 120, 278, Sept. (1971)

 2)  Sawyer, C. N. "ABC's of Cultural Eutrophication Part II  Wastewaters ,"
     Water & Sewage Works 120, 322 (Oct. 1971)

 3)  Process Design Manual for Phosphorus Removal,  EPA Technology Transfer,
     Washington, DC (Oct. 1971)

 4)  Levin, G. V., Shapiro, J. "Metabolic Uptake  of Phosphorus by Wastewater
     Organisms"  Journal WPCF, 37, 800 (1965)

 5)  Mulbarger, M. C., Shifflett, D. G., Murphy,  M. C., and Huffman,  D.  0.,
     "Phosphorus Removal by Luxury Uptake,"  Journal WPCF 43, 1617, (1971)

 6)  Menar, A. B., Jenkins, D. "Fate of Phosphorus  in Waste Treatment
     Processes; Enhanced Removal of Phosphate  by  Activated Sludge."
     Environmental Science and Technology, 4,  1115  (1970)

 7)  Recht , H. L., Ghassemi, M., "Kinetics and  Mechanism of Precipitation
     and Nature of Precipitate Obtained in Phosphate Removal from Waste  Water
     Using Al III and Fe III Salts," EPA Report 17010 EKI (1970)

 8)  Stamberg, J. B0, Bishop, D0 F., Warner, H. P., Griggs, S. H., Lime
     Precipitation in Municipal Wastewaters.  Chemical Engineering Symposium
     Series Water 1970 67, 310.

 9)  Op. Cit. (3) pages 7-13

10)  Kreissl, F. F.,  "Phosphorus Removal Today,"  presented at  the Sanitary
     Engineering Institute, University of Wisconsin, Madison,  Wisconsin,
     March 9-10, 1971.

11)  Bishop, D. F., O'Farrell, T. P., Stamberg, J.  B. "Physical  Chemical
     Treatment of Minicipal Wastewater." Journal  WPCF 44, 361, (1972)

12)  Mdddleton, F. M., Convery, J. J., "Municipal Pollution Control Technology
     in the United States of America." EPA-NERC,  Cincinnati, Ohio (1971)

13)  Mulbarger, M. C., Shifflett, D. G., "Combined  Biological  and Chemical
     Treatment for Phosphorus Removal."   Chemical Engineering Progress
     Symposium Series, 67, 107 (1970)

14)  Laughlin, J. E. , "Modification of a Trickling  Filter Plant  to Allow
     Chemical Precipitation"  presented at EPA Technology Transfer Seminar,
     Dallas, Texas.  July (1971)

-------
                                  - 18 -
15)  Gulp, R. L., Gulp, G. L. Advanced Wastewater Treatment, Van Nostrand
     Reinhold Co.  New York (1971)

16)  Progress Reports EPA - Nassau County Advanced Waste Treatment Project
     WPRD 61-01-67

17)  Cohen, J. M.3 "Nutrient Removal from Wastewater by Physical-Chemical
     Processes."  EPA report, NERC, Cincinnati, Ohio, March  (1971)
                 ^TT
-------

-------