f FA
A TECHNICAL SEMINAR
NUTRIENT REMOVAL
and
ADVANCED WASTE
TREATMENT
Lloyd Center Auditorium Portland, Oregon
February 5-6, 1969
sponsored by
FEDERAL WATER POLLUTION CONTROL ADMINISTRATION
Northwest Region
U.S. Department of the Interior
-------
FORMS AND MEASUREMENT
OF
HITROGEN AND PHOSPHORUS
B. F. Earth
ADVANCED WASTE TREATMENT SEMIMR
Portland, Oregon
February 5 end 6, 1969
-------
This topic will consider the general chemistry of nitrogen and
phosphorus from a treatment and control standpoint.
Nitrogen and phosphorus form many similar type compounds, but
there are important differences.
Because of the reagents and procedural operations Involved in the
determination of pnosphorus careful analytical control is necessary.
The reagents and test condition for various nitrogen forms are usually
very specific. However, because of the complex biological transforma-
tions of nitrogen, nil forms must be measured.
The most comprehensive review of phosphorus determinations in water
Is:
3igurd Olsen, "Determination of Orthophosphate in Water."
Proceeding of I.B.P. Symposium Chemical Environment JDn The
Aquatic Habitat. Published by N.V. Noord-Hollandsche
Uitgevers MaatschapplJ, Amsterdam, 1967*
3-J
-------
The following Is a list of slides that will be used to present
the -topic.
1. Periodic Table
2. Phospnorus and Biological Treatment
3. Determination of Various Phosphorus Forms
J*. Fractions of Various Phosphorus Forms
5» Compounds of Phosphorus
6. Solubilities of Phosphorus Compounds
7. Effect of Acid Concentration on Molybdate Color
8. Biological Transformations of Nitrogen
9. Nitrogen Removal Efficiency
1O. Van Slyke Reaction
3-2
-------
PERIODIC TABLE OF THE ELEMENTS
SUBSKELLS BEING COMPLETED
d
1
H
1 COBO
IICHT
1 A
7
1
1
ft
1
3
8
8
1
7
B
8
7
e
18
18
B
t
IB
33
18
B
3
Li
6940
11
No
33991
19
K
39 100
37
Rb
85 48
55
Cs
13291
87
Fr
17731
1 2
He
4003
MfTAlS
II A
7
7
7
fl
7
8
B
7
4
Be
901]
12
Mg
74 37
20
Co
40 oa
'
38
18 Sr
a «'
1
4
18
18
e
j
13
32
IB
B
8763
f 3
Ba
I) 36
88
Rn
77601
iKArf^iiivjn ncHTT m
III t l\ B \ B MB Ml B ^
J 21
5 Sc
44 96
8 39
18 y
9 '
1 88 97
57-71
&»
S«,.t,
89-100
s..
Aclimd*
S»"«i
2
8
10
7
7
8
IB
10
7
2
8
IB
11
10
2
22
Ti
4790
40
Zr
91 72
72
Hf
17B 50
7
ft
1 1
7
7
II
17
i
7
B
IB
U
1 1
7
23
V
5095
41
Nb
9791
73
To
18095
7
it
13
1
7
B
IB
13
I
'
3?
I?
24
Cr
i70l
42
Mo
9595
74
W
18386
3
B
13
7
7
U
57
La
131 fl
69
Ac
1171
.1 58
2 Ce
* 1
1 1 40 1 3
?, 90
",| Th
7
7
B
15
7
7
8
IS
ts
1
7
B
IB
37
'3
7
^•^ it ii B
27
Co
58 '4
45
Rh
IOJ 91
77
Ir
197
;
59 i
" n ''
• Pr »
7 U09J 1
; 91 ;
. ?a !?
7
7
8
16
7
7
18
IB
7
fl
IB
17
'7,
28
Ni
58?'
46
Pd
ICW.
78
Pt
.11 17
7
a
IB
a
IB
IB
\
7
a
IR
37
IB
1
29
Cu
•» 54
47
Ag
i-.- BIO
79
Au
107 0
I 30
',' 2"
45 :
n
I 48
;: c,
3 .,7
J
41
I 30
15 Hg
" 700
60
Nd
144 ]7
92
U
33«0?
,. 61
VJM.1!
•93
,. ij--'i
, 62
Sm
1)0 Ji
'94
'. -Pu
, U4?;
, 63
> En
i95
3 Am
' '""
61
III A
7
3
3
8
7
B
18
3
5
18
18
3
7
0
1 B
37
5
B
n A t A M i
1087
13
Al
7698
31
Ga
6-»:
J
49
In
1 1
0
•'
1
n
"1-.4
"'
f
-•»—
.1 64 ;
13
Gd ;!
117)6 .
96 ;V
Cm ^
! ,„, ,
1C
'
ii
4
e
18
6
c
17011
14
Si
7B09
32
Ge
77 6P
i! so
is r
ID' J"
' [i 1 "3
"7" 32
;, Pb
'MJ(-,.
2
5
7
8
5
j
B
'I
5
a
18
18
5
J
a
37
10
5
7
N
14 008
15
P
30975
33
As
74 91
51-
Sb
171 76
83
Bi
7CJ
7
7
•>
6
7
1
IB
6
7
H
IS
19
6
^
n
i *
i
•NEPT |
GASES
, — ^ — ,j
k Ml A
8
0
16000
16
5
37066
34
Se
71
96
52
Te
17;
"
84
Po
710
65
Tb
97
Bk
66
' Oy
'»: J.
• 98
;; cf
(
H
1 *
It
J
67
Ho
141 94
99
Es
6b
Er
100
Fm
'""
2
7
7
8
7
A
18
7
7
18
'8
9
F
1900
17
Cl
35 <
57
35
Br
79 916
f
' -74
J
8
1 a
37
IH
J
91
85
1
7
7
8
3
g
8
7
R
IB
II
1
2
He
«OC3
10
Ne
?0 IP3
18
Ar
39 «4i
16
Kr
83 s?
54
'" *e
1 I • I i3
;' 86
At J R"
17!
.|69
• Tm -I
* 114! »< }
ihoi
•»'
Ii Md
j| .».
V
7
]
70
Yb
17304
102
No
, 71
Jjlu
!
'»• 1 ,
-------
PHOSPHORUS IS DIFFERENT
ELEMENT
PHOSPHORUS
BIOLOGICAL
REACTION
RESULT OF TREATMENT
NET LOSS
OXYGEN
CA**O1i I OXIDATION
\ AND/OR
HYDROSEN ? REDUCTION
NITROCEN }
SULFUR /
CO 2 CM4
H2 H2O
N2
V
NO OXIDATION
NO REDUCTION
NO LOSS
ONSANIC PHOSPHORUS
11
INONSANIC PHOSPHORUS
-------
TABLE I
Determination of Various Rio«phonM farm*
Savple
Mllllpore
OASu Filter
Wo ' Acid Acid +
Treatment
/
' a)
Soluble:
Ortho
Hydrolysis Persulfate QL|
(2)
Soluble:
Ortbo *
POly *
Some Org.
(3)
Soluble:
Total
No Acid Acid *
Treatment Hydrolysis Persulfate Dig.
(M
To tad:
Ortbo
(5)
Total:
Ortho +
Poly ^
90M Org,
(6)
Total
-------
TABLE II
Phosphorus
Pita Treatment Components Determined
A. Soluble Forms
1 1) ortho
2 2) ortho + poly 4 sow org'a.
3 3) total
2-1 k) poly f some org's.
3-2 5) Boat org's.
B. Total Forms
k 6) ortho
5 7) ortho t> poly «• aoaw org's,
6 S) total
$ - k 9) poly • «OM org1!.
6-5 10) lott org's.
C. Ptrticulttt rorma
% . 1 11) ortho
. (2-1) 12) poly + MM org's.
. (5.2) 13) •Bit org's,
6 - 3 1^ total
-------
TABLE III
Forms and Keaaureaents of Phosphorus
c Anhytixide P^O^ (usually calculated a*
Orthophosphates Pound in Water
j'nosphoric acid ^^k belov PH 2'2
Jihydrogen phosphate ion (aonovalent) H2P01» fro* pH 2.2 to 7«2
Monohydroften phosphate ion (divalent) KPO^* fro* pH 7.2 to 12.^
Phosphate ion (trivalent) PO^" abore pH 12.k
Dlcalclum phosphate CaHPOj^ from pH 6 to 8
(Trlcalcijn phosphate, CA^PO^)^ doee not form in water)
Hydroxynpatite C*5CH(POU)3 abore pH 7-8
Fluorapatite Ca5F(POu)3 abcnre pR 7
AliBlnioi hydroxide adsorbs HPO. froa pH 5 to 10
Almlnate ion Al(OH)," forma above pH 10
Ferric hydroxide adsorbs KPO^" abore pR k
hydroxide flocculates colloidal phosphates abort pi 10-11
Polyphosphates
All polypnosphates hyirolyxe slowly to orthophosphates
Sodium TrimetApnospnate Ra3(P03)3 — a ring compound
Sodlisi Polymetapnosphate (K*P03)n — a Long chain
(there is no Justification for the naae "aexMMtApnosphate")
Sodlia Pyrophosphate MAi»P2°7
Sodlisi Tri Polyphospnate Ba5P3°10
Orsjanio
,
Esters and anhydrides BOPOj
___ *
Phoephacens RXPOo
Reference
(1) Van Wa*er, J. R. Phosphorus and Its Compounds. Interscienc* 1958-
-------
tABU IV
PHOSPHORUS OCMPOUHDS CLASSIFIED R CHBQCAL
AMD SuUIBII.nT RELATIOBS*1* _
Pom
1. Ortoo phosphates
(PO.)
-3
Mater Soluble
Contained with nonovalent
cations such as R, Ma, K,
Insoluble
Coablned with «ultl
valent o
.. CA*2*
2. Poly phosphate*
v-3
and others depending upon
the degree of dehydration.
(a) M la 1 (above), but
solubLLlty decreases as
•olecular Might Increases
(a) as la 1 (above)
(b) sultl P poly-
pbospoates (high
•ol. wt.) in-
cluding the
"glassy" phos-
phates
3. Organic phosphorus
R-P, R-P-R (3)
(unusually varied nature)
(a) certain ebe*leals
(b) degradation products
(c) cniyme ?
(d) phosphorylated nutrients
!a) certain enemies
b) cell s»ss, mmy
chealcala
be
colloidal or
(c) soluble P sorted
by insoluble
residues
U. Mineral phosphorus
(a) as in 1 (above), but
solubility decreases for
foraa
(a) as In 1 (above)
(b) geological P
pboslllcates
(c) certain Mineral
coaplexes
(1) Table drawn from a lecture, "Detsrmlnatlon of Phosphorus In the Aqueous
by P. Luduhck, PWPCA Trainlac Activities, Ohio Basis Region.
(2) Used to Indicate usual stats under eoMne conditions.
(3) R represents an organic radical, P represents P, POj,, ox Its derivatives.
-------
TABLE V
Effect of Varies Sulfuric Acid - AaMonlua Molybdate
Ratio on Heteropoly Blue Color Development
(Staunou* chloride In glycerine reductant)
Acid MOlybdate Reagent
Composition
AJMsonlua Molybdate
gsVlOO al
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
Concentrated
Sulfuric Acid
•1/100 Bl
5
10
ii>
20
2&
30
3b
»*0
X)
60
Rora»llty*
of acid In
final 100 id
as»ay volume
0.072
0.144
0.216
0.266
0.360
OA32
o.x*
0.^76
0.720
0.664
Absorbancy (O.D. )
Distilled
Water
> 2.0
o.bo;>
0.260
0.036
O.OO4
0.000
O.OOO
0.000
0.000
O.OOO
0.05 •£ P
Standard
-> 2.0
1.00
0.5*»
0.47*
0.41^
0-36b
0->b
0.2^2
0.060
0.012
•Calculated aorvality of HjflO, In the 100 •! final vol
4.0 al of the particular Ida •olybd*t« aalutloa.
produced by
-------
-------
TABLI VI
•Itrotra Tozmm la 1 1 Mixture of
Blfb-Rkt* and nitrified Effluent*
•Itrogen
Pen
Orpulc-B
MMonli-I
-£
9m
•rriuant
9
10
-
15
•Itrlfled
(mum
2
0
1 13
12
IS
Denitrified
Vflueot
3
5
0.5
9
(teldlied nitrogen
Reoov&l
^
Over-All
Reaond
**
FIGURE
VAJT 3LYKB REACTIOM
NO,
ROH
* K I +
-------
PRESENT AND PROJECTED PHOSPHORUS REMOVAL
IN CONVENTIONAL TREATMENT
E. F. Earth
ADVANCED WASTE TREATMENT SEMINAR
Portland, Oregon
February 5 and 6, 1969
-------
PRESENT AMD PROJECTED PHOSPHORUS REMOVAL
IH CONVENTIONAL TREATMENT
Portland, Oregon
February 5 and 6, 1969
This discussion will Indicate the magnitude of the phosphorus
problem, some of the shortcomings of our treatment and collection
systems, and the biological cycle of phosphorus In nature.
Many investigators have considered various means of controlling
phosphorus; these will be discussed and the three approaches
currently In practice will be briefly described.
The parameters thought necessary for biological removal of phosphorus
will be presented, and the flow sheet for a treatment plant specifically
designed for biological, uptake vill be shown.
Substantial improvement in conventional treatment could be obtained
by breaking the cycle fca* digester supernatant. Progress along
this •>•*"• vill be presented.
The following slides vill be used, and vhere available, a reproduction
of the slide is attached:
Slide Bo.
1 Population vith Wast* Treatment System* 19&
2 Treatment Classification by Population 19&2
4-1
-------
-2-
3 Treatment Classification by Facilities 1962
k phosphorus Removal by Conventional Treatment
Systems
5 Estimate of Total Amount of Phosphorus
Discharged to Water System
6 Cycle of Phosphorus in Nature
7 Various Approaches to Phosphorus Control.
ti Three Main Approaches
9 Flov Diagram of Approaches
10 Parameters Deemed Necessary for Biological
Removal of Phosphorus
11 Design of a Phosphorus Removal Biological
Treatment Plant
12, Characteristics of Digester Supernatant
13 Digester Supernatant Process
4-2
-------
POPULATION WITH WASTE SYSTEMS FOR YEAR 1962
186 x 10 Total population
118 x 106 Total sewered (63%)
104 x 10 Receive treatment
837, of sewered
567. of total
4-3
-------
TREATMENT CLASSIFICATION BY POPULATION FOR YEAR 1962
Type of 1 of % of
Population treatment treated of total
38 x 10 Primary-intermediate 37 20
33 x 10 Activated sludge 32 18
23 x 106 Trickling filter 22 12
10 x 106 Other 9 6
4-4
-------
TREATMENT CLASSIFICATION BY FACILITIES FOR YEAR 1962
Number Type % of tota'l
11,655
2,277
9,378
2,794
6,584
800
3,506
1.348
Total
Raw discharge
Treated discharge
Primary -intermediate
Secondary
Activated sluc^e
(127. of secondary)
Trickling filters
(537. of secondary)
Stabilization ponds
(207. of secondary)
100
19.5
80.5
24.0
56.5
7.0
30.0
12.0
4-5
-------
PHOSPHATE REMOVAL BY WASTEWATER TREATMENT PLANTS
Type T» removal
Primary sedimentation 5-15
Extended aeration (sludge wasting) 8-15
Trickling filter 20-30
Activated sludge 30-50
4-6
-------
ESTIMATE OF TOTAL AMOUNT OF PHOSPHORUS DISCHARGED
TO WATER SYSTEM
Per capita contribution of phosphorus is 4 pounds
per person per year
Population x per capita x efficiency Pounds discharged
Not served by any treatment
70 xlO6 x 4 x 0.00 280 xlO6
Served by,primary treatment only
40 xlO6 x 4 x 0.10 154 xlO6
Served by trickling filter or
extended aeration
30 xlO6 x 4 x 0.25 80 at 108
Served by activated sludge
40 xlO6 x4 x 0.40 96xl06
Total pounds of phosphorus discharged
to water system per year 610 x 10
4-7
-------
BALANCE BETWEEN' ITETErtOTROPIJIC AND
AUTOTROPHIC GROWTH (Stumm & Tenney. 1963)
Treatment
Plant
Receiving
Stream
Wasiewaler contains organic and inorganic
C-H-G-N-P. By heterotrophic activity this
is converted to CO2> H2O, NOg, PO4- A
large portion of oxygen demand is satisfied
by conversion to CO , but N and P are
solubilized and enter receivin stream.
CQ2J N
1 P04 + light energy ^ >154
Autotrophic activity
The Cycle
1 # Phosphorus ^-75 # of algae ^ 160 "# Og
A steady state exists in the overall environment
of the earth. There is always excess O in the
i
atmosphere.
In localized environments pollution results
from an imbalance.
Excess heterotrophic activity ^ DO depletion
caused by excess organic matter.
Excess autotrophic activity = large algal mass
produced caused by excess inorganic matter.
4-8
-------
FIGURE I
Phoiphorus Removal by Waatewater
Treatment (Excluding Lagoons and Percolation)
PHOSPHORUS REMOVAL DEPENDENT UPON*.
Mineral Compoaition ^Interaction Biological Synthesic
Addition of Supplement
to Primary Effluent
Guggenheim (1)*
* E. A. Thomas (2)
Separate Unit Process
on Biological Effluent
Rohlich (3)
Culp (4)
Yee (5)
Stumm and
Tenney (6)
Control of Conventional
Operational Parameters
Levin and Shapiro (7)
Vacher et al. (8)
Primary Settling Efliminated
to Increase C;N:P
Finstein (9)
Algae Culture
Gates (11)
Eliminate Digester
Supernatant Feedback
Wectberg (10)
} t•arate* iitetwtwre reference.
-------
*•
t-'
o
TABLE HI, APPROACHES TO PHOSPHORUS REMOVAL
CONTROL OF BIOLOGICAL UPTAKE
CHEMICAL TREATMENT OF BIOLOGICAL EFFLUENT OR
PROCESS STREAM TO FORM SLIGHTLY SOLUBLE
PHOSPHORUS COMPOUNDS
COMBINATION OF THESE TWO ABOVE APPROACHES
-------
BIOLOGICAL
CONTROL
Digester
Supernatant
Sludge
CHEMICAL
CONTROL
f *| PRIMARY
Oi<3«tcr I
Supernatant I
BIOLOGICAL
PROCESS
I Recycle
(R»f.
Sludge
BIOLOGICAL
PROCESS
L Recycle
COMBINED
CHEMICAL- BIOLOGICAL CHEMICAL
ADDITIVE
Digester
Supernotont
Sludge
BIOLOGICAL
PROCESS
Recycle
Sludge
CHEMICAL
IDDITIVE
FINAL
Sludge
Sludge
(Rat SLECHTA AND
GULP, 195 7)
FINAL
(Raf. BARTM AND
ETTINGEft, S67)
Sludge
FIGURE I. PHOSPHORUS CONTROL BASIC SYSTEMS
4-11
-------
PARAMETERS DEEMED NECESSARY'FOR EFFICIENT
BIOLOGI-CA L iR-EIvlO VA L OF 1JHOSPI1ORUS
1. Eliminate.Digester Supernatant Recycle
2. Aeration Time and Rate of Air Supply
3. Concentration of Mixed Liquor Suspended Solids
4. Concentration of Dissolved Oxygen in Aerator
.5. Period of Sludge Retention in Secondary Clarifier
6. Additional Benefit Possible by Stripping of Phosphorus
from Return Sludge
a) Anaerobic Stripping
b) Acid Stripping
4-12
-------
Grtat r NNmMMt 3«nlt ry District
1 m.g.d.
NO PRIMARY SEDIMENTATION
TREATED DIGESTER SUPERNATANT
WASTEWATER
PRIMARY
AERATION
(TURBINE)
FLOTATION
UNIT
RECYCLE
SECONDARY
AERATION
(TURBINE)
M CLARIFIER —*-
RECYCLE
WASTE
TREATED
SUPERNATANT
SUPERNANT
SLUDGES
WASTE
DIGESTERS
SECONDARY
SAND BEDS
-------
TABLt II
Sunnsry of Supernatant Data (6)
Analysis 80% Confidence Median ttean
C-icarbonate Alkalinity 850 - 2,950 ,580 1681
(ng/1 as CaC03)
5-Day COO (rig/1) 500 - 1,260 910 1401
Total Phosphate (ng/1 as P04) 46 - 330 132 211
(mg/1 as P) 15 - 124 43 69
Amnonta Nitrogen (mg/1) 100 - 710 405 413
pll 6V6 - 8.0 7.3 7.3
(6) Hasselli, Joseph !/. ct.al. "The Effect of Industrial Wastes on Sewage
Treatment", flow Cnoland I/ater Pollution Control Commission, Dos ton»
Massachusetts
4-14
-------
CO
DIGESTER SUPERNATANT
AND RELATED PROCESS
ST AMS TEMPERATURE
AT 35°C
NH,
SUBMERGED
COMBUSTION
HEATER
A
TEMP
LIME
DIGESTER GAS
OR
NATURAL GAS
•*-SCRUB BED AIR
MAKE-UP ACID
AMMONIA
SCRUBBER
LIME SLAKER-
MIXEH
AMMONIA
STRIPPER
CONCENTRATED
, SOLUi !CN
AIR
TREATED
SUPERNATANT
GRAVITY SETTLER
SETTLED,WASTE SLUDGE
FIGURE-I
(JjD y M C CORPORATION
hi
CT
em
_
M
•UKI
«...
TREATMENT SYSTEM
DIGESTER SUPERNATANT
0.1 1 It >*•<•
?1
-------
REFERENCES
1. Guggenheim Process. Described by: Kiker, J.E., "Waste Disposal
for'Dairy Plants," Sanitarian (Los Angeles) 16. 11-17 (1953).
Florida Engng. Exp. Station. _7, Leaflet Series No. 51 (1953).
Mo6;re, R. B. , "Biochemical Treatment for Anderson, Ind."
Water Works and Sewage. Nov. 1938..
2. Thomas, E'. A., "Phosphat —Elimination in der Belebtechlemannlage
von Mannedorf und Phosphat-Fixation in See-und Klarschlamm,"
Vierteljahrsschrift der Naturforschenden Gesellschaft in Zurich.
Jahrgang. 110. Schlussheft, S. 419-434 (Dec. 1965).
3. Rohlich, G. A., "Chemical Methods for the Removal of Nitrogen and.
Phosphorus from Sewage Plant Effluent." Conference on Algae and
Metropolitan Wastes. Robert A. Taft Sanitary Eng-. Center,
Cincinnati, Ohio. U.S.F.H.S. (1961).
4. Culp, G., "Nitrogen Removal from Sewage." Final Progress Report
U.S.P.U.S. Demonstration Grant 86-01, (Feb. 1966).
5. Yee, W. C., "Selective Removal of Mixed Phosphates by Activated
Alumina." J. Am. Water Wks. Assoc.. 58. 239-247 (1966).
6. Stumm, W., and Tenney, M. W. , "Chemical Flocculation of Microorganisms
in Biological Waste Treatment." Proc. of the 19th Ind. Waste
Conf. Purdue Univ. Eng. Ext. Ser. No. 117, Part 2, 518 (1964).
7. Levin, G. V. and Shapiro, J., "Metabolic Uptake of Phosphorus
by Wastewater Organisms." J. Water Poll. Cont. Fed.. 37. 800-821 (1965)
8. Vacher, D., Connell, C. H. , Wells, Wm. N., "Phosphate Removed
Through Municipal Wastewater Treatment at San Antonio, Texas."
Presented at the Annual Short School o.f the Texas Water and
Sewage Works Assn., March 9* 1966. Texas A&M University.
9. Finstein, M. S., "Nitrogen and Phosphorus Removal from Combined
Sewage Components by Mlcrobial Activity." Appl. Microbiol.,' 14
679 (1966).
10. Westberg, N., "Nutrient Salt Reduction at Sewage Treatment Works."
Vattenbygien. 19. 2-10 (1963). (In English).
11. Moore, W. A., et al.t "Effects of Chromium on the Activated Sludge
Process," J. Water Poll. Cont. Fed.. J33:54 (1961).
12. Mulbarger, M. C. and Castelli, J. A., "A Versatile Activated'Sludge
Pilot Plant," 21st Annual Purdue Industrial Waste Conferences,
Purdue University, 1966.
13. Standard Methods for the Examination of Water and Wastewater.
12th Edition. Am. Public Health Assn., New York 1965.
14. Lee, F.G., et al., "Studies on the Analysis of Phosphates in Algal
Cultures," Int. J. Air and Water Poll.. _9, 715 (1965).
4-16
-------
REFERENCES
Asimov, I., Discovery of the Elements, Fawcett World Library, Nev York. 1962.
Advanced Waste Treatment Research Summary Report, January I960-June 196U,
PHS Publ. No. 999-WP-2U. 1965. 1^2 pp.
Earth, E. F.f and Ettinger, M. B., 'Mineral Controlled Phosphorus Removal
in the Activated Sludge Process," J. Water Pollution Control
Federation,39!1362-1368 (Aug. 1967).
Brunner, C. A., "Pilot-Plant Experiences in Demineralization of Secondary
Effluent Using Electrodialysis," J. Water Pollution Control
federation, 39:2, R1-R15 (Oct. 1967).
Buszel, J. C., and Sawyer, C. N., "Removal of Algal Nutrients from Raw
Wastewater with Lime," J. Water Pollution Control Federation,
39:2, R16-R-24 (Oct. 1967).
Dorr-Oliver, Inc., Phosphate Removal Process, Envir. Sci. and Tech., !_:
G59 (Oct. 1967).
Eberhardt, W. A., and Nesbitt, J. B., "Chemical Precipitation of Phosphate
Within a High Rate Bio-Oxidation System," Proc. 22nd Ind. Waste Conf.,
Purdue Ur.iv ., May 1967.
Eliassen, R., "Chemical Processing of Wastewater for Nutrient Removal,"
liOth Ann. Conf. Water Pollution Control Federation, New York, N. Y. ,
Oct. 1967.
4-17
-------
Foyn, E . , "Removal of Sewage Nutrients by Electrolytic Treatment, "
Intern, Ver. Theoret. Angew. Limnol., Verhandl., 1£: 569-579 (1962).
Lee, G. F., and Fruth, E. G., '*The Aging of lakes," Industrial Water
Engineering, 3j26-30 (Feb. 1966).
Middleton, F. M., "New Treatment Techniques for Municipal Waste Waters,"
Consulting Engineer (Mar. 1967)-
O'Connor, B., Dobbs, R. A., Vllliers, R. V., and Dean, R. B., "Laboratory
Distillation of Municipal Waste Effluents," J. Water Pollution Control
Federation, 3£:2, R25-R31, (Oct. 1967).
Rand, M. C., and Memerov, K. L., "Removal of Algal Nutrients from Domestic
Wastewater. Part II. Laboratory Studies," New York State Dept.
of Health (Mar. 1965).
Slechta, A. F., and Gulp, G. L., 'Water Reclamation Studies at the South
Tahoe Public Utility District," J. Water Pollution Control Federation,
22:787 -81** (May 1967).
Tenney, M. W., and Stumm, W., '^Chemical Flocculation of Microorganisms in
Biological Waste Treatment, " J. Water Pollution Control Federation,
37:1370-1388 (Oct. 1965).
4-18
-------
Thomas, E. A., "Phosphate Elimination in the Activate::, ".luufje ?:_a:it Mannedorf
and Phosphate Fixation in Late and Sewage Sludge, " u;.erteljahrsschriff
der Haturforschenden Gesellschaft in Zurich I'LO'.klk-k^h (Dec. 1965),
(in German).
Vacher, D., Connell, C. H., and Wells, W. N., "Phosphate Removal through
Municipal Wastevater Treatment at San Antonio, Texas," J. Vater
Pollution Control Federation, 39:750-771 (May 1967).
Yee, W. C., "Selective Removal of Mixed Phosphates "by Actlvatpc? Alumina,"
J. Am. Water Works Assoc.f 58:239-2^9 (1966).
4-19
-------
PHOSPHORUS RB40VAL BT TERTIARY TREATMENT WITH LINE AMD AL1M
by
Robert B. Dean
Chief, Ultiaate Disposal Research Program
Advanced Waste Treatment Research Laboratory
First Technical Session
11:45 A.N.
Wednesday, February 5, 1969
Symposium on Nutrient Removal and Advanced Waste Treataent
Portland, Oregon
-------
I. LIME TREATMENT OF SECONDARY EFFLUENT
A. Chemistry — in secondary effluent
1. a. Ca(OH)2 + HOD - . > CaOO J,+ OH" + HgO (l)
Slaked lime + alkalinity —^ calcium carbonate + hydroxyl ion
b. Ca"" + HOO ' + OH'—* CaOO + HgO (2)
Calcium in vater + alkalinity » calcium carbonate
c. 5 Ca++ + k OH" + 3 HPOJ —* Ca5 OH(K)4)3 ^ * 3 HgO (3)
Orthophosphate ••^ hydroxyapatite
d. Mg++ + 20H'_4 Mg(OH)2| (U)
Magnesium in vater •••A magnesium hydroxide
2. a. Equations 1 and 2 are complete above pH 9.5.
b. Equation 3 forming hydroxyapatite starts above pH 7 but is very
slow below pH 9. Hydroxyapatite is the only stable phase in
alkaline vater.
c. Some phosphate may be colloidal 1y dispersed as hydroxyapatite
and escape sedimentation.
d. Mg(OH)2 equation is a gelatinous precipitate that collects
particles.
e. Phosphate removal improves as pH is raised.
B. Lime Requirements
1. Lime dose is primarily related to alkalinity.
a. Figure 1. Lime as CaO to reach pH 11.
b. Attempts to calculate lime dose from alkalinity,
NHY, etc. alvays come out too lav.
5-1
-------
FIGgRE 1
LIME REQUIREMENT FORpH^ll.O AS A FUNCTION
OP THE WASTEWATER ALKALINITY
o
CO
o
00
500
Al
o:
O
400
Q
C> 300
LU
cc
200|-
QE:
o
8
DATA FROM THIS INVESTIGATION
DATA FROM OTHERS WITH REFERENCE
100 2CO 300 400 500
WASTEWATuR ALKALINITY mg/L - CciC03
S-i
-------
C. Systems
1. Single Stage
a. Add lime slurry to pH 9.5, settle.
b. Precipitate calcium carbonate, calcium phosphate (hydroxy-
apatite) and organics.
c. Calcium phosphate may be difficult to collect. May need
coagulant aids. Iron salts, activated silica or polymers
and filtration.
2. Two Stage, figure 2, System A.
a. Add lime slurry to pH 11+, settle. Recarbonate to pH 9.5.
Clarify.
b. First stage precipitates calcium carbonate, calcium
phosphate, magnesium hydroxide and organics.
c. High pH is suitable for ammonia removal by air stripping.
d. Sludge may be recarbonated with flue gasses to redissolve
magnesium hydroxide and aid filtration. Figure 2,
Systems B and D.
e. Second stage precipitates only CaCO? (100 mg/l) if COg
is used.
f. Split treatment uses alkalinity in part of. the stream to
adjust pH to 9.5, Systems C and D. Precipitate contains
calcium phosphate and some organics as veil as calcium
carbonate.
5-3
-------
FIGURE 2.
EXPERIMENTAL SYSTEMS
("C"AND"D")
MIX C02
FLOC
SETTLE -
GRAVITY
THICKENER
REFERS TO
FLOW STREAM.
SEE TABLE 2
SYSTEM "A"
TOTAL TREATMENT
SYSTEM "B"
TOTAL TREATMENT WITH SLUDGE CARBONATION
SYSTEM"C"
SPLIT TREATMENT (0.5Q to 2nd STAGE)
SYSTEM "D"
IT TREATMENT 'VITH SLUDGE CARBONATION
NOTE LIME DOSE rOR SYSTEMS "C" AND "D" IS H OF
THE L VIE DOSE USED FOR SYSTEMS "A" AND "B"
5-4
-------
D. Equipment
1. Slake lime as CaO in slurry or paste slaker to make a slurry
of Ca(OH)2 in saturated water.
2. Rapid mix tank or zone for precipitation.
3. Clarifier to remove solids.
a. Circular. South Lake Tahoe, California. Tables 1 and 2.
k. Sludge blanket systems precipitates in presence of preformed
solids.
a. Lebanon, Ohio. Figure 3. 75 gpm.
o
b. Upflov rate 1 gpm/ft , 110 minutes detention.
c. Sludge sucked up from bottom, mixed with new lime and
secondary effluent.
d. Water rises through sludge blanket. Top of blanket is
organic and colloidal, larger calcium carbonate settles
out.
e. Remove sludge as 2-3$ slurry and gravity thicken to 1O£.
f. Poor performance when biological plant is upset and high
colloids are coming over.
g. Coagulant aids not needed during normal operation.
5. Lime Solution System
a. Requires high volume ratio of lime water to effluent.
About 1/4 of total flow.
b. Does not recycle insolubles from recovered lime.
6. Filters to remove residual turbidity and colloidal phosphates.
a. Multi-media — coal on sand, etc., or rapid sand filter.
b. Effluent needs pH adjustment to avoid calcium carbonate
precipitation on media.
5-5
-------
SOUTH TAHOfi PUD
PROCESS EQUTPMENT
Primary and Secondary Treatment. Process "".'jipment used at this plant
for primary and secondary treatment is quite conventional in design and
design loadings for various units may be termed conventional. Both rectangular
and circular primary clarifiers are employed; each exhibit approximately equal
efficiencies. Both conventional spiral roll and completely mixed activated
sludge systems are used; oxygen transfer to the spiral roll system is by means
of coarse bubble diffused air and to the completely mixed system by means of
sparged turbine aerators. As expected, the completely nixed activated sludge
system exhibits much more stable operation and load handling ability than docs
the spiral roll system. Two types of circular, sludge suction type, secondary
clarifiers are employed. Performance data from these secondary clarifiers to
date does not indicate significant difference in the efficiency or performame
of the two un i ts.
Wasting of activated sludge is accomplished by means of a mass density
system which provides automated control of sludge wasting. The system employs
a nuclear density meter to determine percent solids contained in the waste
sludge. Waste sludge flow is measured by a magnetic flow meter. The combined
outputs of the density meter and the flow meter are electrically multiplied
in an indicator-totalizer and used to position the waste sludge valve by
comparison of the multiplied output signal with a preselected wasting rate.
Waste sludge is thus recorded in terms of the pounds of dry solids wasted
per day (kg/day).
Waste sludge is thickened by means of a flotation type thickener. Polymers
are added to enhance thickening and solids capture. Operation of the thickener
has been very successful with essentially 100 percent solids capture and float
production containing 5 percent solids easily obtained. Polymer costs in the
range of 3 to 5 dollars per ton ($3.30 to $5-50/metric ton) are normal. Cationic
polymers appear most effective with this sludge.
Chemical Coagulation^ Chemical coagulation and phosphate removal equip-
ment are also conventional with flash mix, slow mix (flocculation) and a
conventional clarifier being employed. Lime dosage to the flash mix is on
the order of ^00 mg/l as CaO. The addition of approximately 0.25 mg/l poly-
electrolyte after flocculation has proven to be very beneficial in reducing
effluent turbidity from the clarifier. Turbidity in the chemical clarifier
effluent commonly ranges 3 SJU or less. Operating expense to date indicates
that conventional clarifiers rather than upflow type inechani s ims ore definitely
preferable for this type of service.
Coagulant Recovery. Spent lime sludge is removed from the clarifier by
pumping, thickened in a conventional gravity thickener, dewatered by centrifuge
and recalcined for reuse by means of a multiple hearth furnace. Pneumatic
conveying equipment is utilized for transporting both recalcined and new lime
to overhead storage bins. Gravimetric ]ime feeders and pug mill type slakers
are used to feed Ca(OH)2 slurry to the flash mix bases and other process
application points. Lime feed rates are controlled by signal from continuous
pH measurement.
5-6
-------
TABLE 1
MAJOR DESIGN CRITERIA FOR
THE SOUTH TAHOE PUD
WATER RECLAMATION PLANT
Parameter
Design Criteria
Flow
BOD cone.
BOD amount
SS cone.
SS amount
Lime dosage as CaO
Recalciner capacity
Sludge incinerator capacity
Carbon furnace capacity
7.5 mgd
350 mg/1
22,000 Ibs/day
200 mg/1
12,500 Ibs/day
<«00 mg/1
39,600 Ibs/day dry
21 ,600 Ibs/day dry
5,000 Ibs/day dry
solids
solids
spent carbon
Note mgd x 3785 = cu m/day
Ibs/day x 0.1*5** - kg/day
TABLE 2
CHEMICAL AND PHYSICAL CHARACTERISTICS
RAW SEWAGE AT SOUTH LAKE TAHOE PUD
Waste Characteristic
COD
Turbidity
Phosphate
II i trogen
Organic
Ammon i a
as P
Concentration in Raw Sewage
as
as
N
N
N02 and NCK as
1*00 - 60U mg/1
Htoh 100 SJU
10 - 15 nwj/1
10 - 15 mg/1
25 - 35 mg/1
Trace
5-7
-------
LIME
SECONDARY EFFLUENT
75 gpm
^m i
CLARIFIER
PRODUCT
SETTLER
MIXING ZONE
FLOCCULATING ZONE
SLUDGE
COLLECTING
RING
FIGURE 3. LIME CLARIFIER
-------
E. Performance
Single Stage Clarifier (Lebanon). Figure k.
a. Effluent: Ca 100 mg/1
Mg 30 mg/1
PO^- P 10 mg/1
Alkalinity kOO mg/1
b. 90% removal above pH 9.7
Single Stage plus Dual Media Filters. Figure 5.
a. Media. 18" - 0.75am anthracite
6" - O.J*6mm sand
b. ^Operation. 2 gpm/ft2 to "f" Hg pressure loss in 1*8 hours
c. 97% Removal of PO^ at pH 10, 8O% at pH 9.O.
d. Approximately 250 mg/1 of Ca(OH)2 (l9Q',mg/l as CaO)
required _f or gQfft removal.
3. Two stapo.
d1. Tahoe
FINAL EFFLHENT-qUALm
South Lake Tahoe
September 1968
Median Maximum
MBAS 0.14 0.19
OOD 9. 13.
Turbidity-JU 0.1 O.U
Phosphate-P 0.11 0.1Y
Nitrogen-N
NH &+ .N03~ 11.U 13.1
b. Rapid mix before polymer addition.
c. Recarbonate to pH 7.5.
d. 110 mg/1 fresh lime and 227 mg/1 recalcined lime (CaO).
Predicted kOO mg/£l in lab, studies (Slechta '67, Suhr '68)
e. Washington, B.C. 70 gpm. 90% in clarifier, 95% plus filter.
Whole plant, 97%.
5-9
-------
8
M
U.
o
9.0
t5
10.0
10.5
11.0
CLARiFIER PH
11.5
FIGURE 4. EFFECT OF pH ON PHOSPHATE CONCENTRATION OF
EFFLUENT FROM LIME CLARIRER
-------
6
fe
o
- 3
ft
O
85
9.0
9.5 10.0
CLARIFIER pH
IQ5
11.0
11.5
FIGURE 5- EFFECT OF pH ON PHOSPHATE CONCENTRATION OF
EFFLUENT FROM FILTERS FOLLOWING
LIME CLARIFIER
-------
Line Recovery Studies. Two Stage, 75 1 batch (Mulbarger).
Figure 2.
a. Hard water (3*10 mg/l) and relatively soft water (175 mg/1 as CaOO )
b. Composition did not affect removal efficiencies. 97$ SS,
99% FO removal.
c. Split treatment reduced PO. removal to
P. Cost of Lime Treatment
1. Without recarbonation or post filtration. Table III
a. 300 mg/1 Ca(OH)2 (225 ngA CaO)
b. 0.87^/1000 gal. for coagulant aid equivalent to 50 mg/1
c. k-1/2% for 25 years.
d. Cost of recalcined lime is close to cost of fresh lime.
e. Savings are in reduced disposal of lime sludge.
2. Lake Tahoe. 7.5 agd plant (4.5 mgd operation)
a. 9^/1000 gal. (Suhr) versus 1O.4 estimated (Slechta).
Lime costs $2U.50/ton because of high freights.
5-12
-------
TABLE III
TOTAL COST OF
(Cents per
Capital amortization
Land amortization
Operating and maintenance
Cost of chemicals
Lime
Iron salt
Cost of sludge disposal by hauling
(to lend .-Till (25-mile one-way trip)
TOTAL
Savings if sludge can be
recalclned
TOTAL (with recalcining)
PHOSPHATE REMOVAL
1000 gallons)
Size of
1 ogd 10 mgd
•97 .79
.09 .09
.Ui .lU
1.75 1.75
.87 .87
.67 .67
It. 76 lv.31
-.67 -.67
U.09 3.6^
Plant
100 ngd
.65
.09
.08
1.75
.87
.67
U.ll
-.67
3.UU
250 rngd
•59
.09
.06
1.75
.87
.67
U.03
-.67
3.36
Source: IWPCA, Lake Michigan Water Pollution Enforcement Conference
5-13
-------
II. TREATMENT OF SECONDARY EFFLUENT WITH ALUMINUM IONS
A. Chemistry
1. Low pH. Alum. Al (SOjL • X HgO > Al+++
a. Al + ^^ - * Al PO^ + 2 H
2. Above pH 5-6 also get aluminum hydroxide precipitation
a. Al""'* -i- 3 OH" 7 - » Al(OH), I and
^ j 4-
Al(OH) + H POj - > Al • PO^ + 2 OH" + HgO
b. Final product approximates Al(OH) • Al PO, ,
but mole ratio depends on pH. Figures 9 and 10.
c. Alumina te is an alternative source of aluminum hydroxide.
+ H+ - > A1(OH)
3
d. Natural alkalinity can supply H* or OH" within limits.
3 H00~ + Al+ + 3 H20 - > A1(OH)3
H00~ + Al(OH)j^ - > A1(OH)3 + CX>*
e. Alum and aluminate can be combined:
3 AI(OH)J; > if Ai(oH)3
f. Polyphosphates are harder to remove than ortho (Stunm) and
may reduce removal percentage.
B. Dose Requirements
1. Basic requirement 2 Al to 1 P.
a. Substantially greater than for clarification. Figure 7.
b. At Lebanon: 250 mg/1 alum for 90$ removal
At Tahoe (Slechta): 300 mg/1 for 90% removal
C. Systems and Equipment
1. Similar to Mf**
2. Lover rise rates in clarifier 0.5 - 0.75 mg/1*
3. Post filtration needed only to exceed 80$ removal.
U. Sludge Disposal is a major unsolved problem.
5-14
-------
D. Performance
1. Lebanon. Table 3 and Table k.
a. 90$ removal of PO^ without filtration using, activated
silica as a coagulant aid.
b. Filtration increased P removal to $6% and significantly
improved solids removal.
E.
10 mgd plant requiring 300 mg/1 alum plus 2.5 mg/1 silica.
a. darlfier only. 80-900 removal of P. Cost 3.9^/1000 gal.
b. Filter after Clarlfier. 96$ removal of P. Cost 7.^/1000 gal.
Does not include sludge disposal.
a. Freezing especially natural.
b. Aluminate recovery. See later papers.
5-15
-------
In- Depth Filtration
TAHLE 1. OPTIIV.fJM OPERATING CONDITIONS FOR TERTIARY
CI^IUFICATION USING ALUM COAGULATION
f^ILTRATION AT LEBANON, OHIO
V.in.ibic-
(1) Over!low Kate
GPD/FTa
(2) Alum Dose
mg/1
(a) Clarih_ation
(b) -Phosphate removal
(3) Type of Media
(a) Sand
(b) Coal
(c) Sau
-------
In-Depth Filtration
TABLE 2. TYPICAL RESULTS UNDER OPTIMUM
OPERATING CONDITIONS
Process Stream
Suspended
Solids, mg/1
Acid Hydrolyzable -
Phosphate, mg/1 PO
Turbidity
J.T.U.
Secondary Effluent
Chemically Treated
Settled Effluent
Filtered Product
45.6
11.0(76%)*
1.3(97%)
22.4
2.2(90%)
0.9(96%)
12.2
1.5(88%)
0. 5(96%)
Operating Conditions:
Overflow Rate:
Alum Dose:
Aluminate Dose:
Silica Dose:
700 GPD/FT
82 mg/1
68 mg/1
3 mg/1
* Denotes Removal Efficiency as %
The clarifier is operated at a rise rate
of 1 gpm/ft . Dual media filters;
consisting of IS inches of 0.75 mm coal
over 6 inches of 0.45 mm sand polish
the clarifier effluent. The filters are
operated at 2 gpm/ft to a terminal
headloss of 9 feet of water. Some
preliminary operating results are shown
in Table 3.
The suspended solids concentration of
the settled sludge ranged from 1.5 to
2.5% solids by weight and was further
concentrated to 12-15% solids by weight
in a gravity thickener and additional
drying on sand beds prior to disposal
to a land fill.
C Jar Test Results
When effluent quality standards require the
removal by phosphate in addition to clari-
fication, significantly higher coagulant
doses can be anticipated. A rough
stoichiometry of 2 moles of A L+T+ to 1
mole of P has been found empirically, to
exist in tertiary treatment applications.
Filtration Rate: 3GPM/FT
Effective Size of Coal: 1.325 mm
Length of Run: 25 hours
Sludge Concentration: 1.2% by weight
Solids Concentration in
Filter Backwash Water: 475 mg/1
In addition,, as shown in Figure 1 all the
primary coagulants are equally effective
in removing phosphate. These results were
obtained from jar test studies on' a
pilot plam activated sludge effluent.
VI INDUSTRIA L APPLICATIONS12
A Food Industry
1 Meat
Grit removal, flocculation and sedi-
mentation will remove 50% of the total
volatile solids and BOD and 75% of the
suspended solids in a manure-bearing
wastewater. Alum coagulation removes
an additional 2. 5 mg/1 of BOD and
1.5 mg/1 of grease per mg/1 of alum
added. The chemical costs for 50%
and 78. 5% increased removal efficiency
is $0.64 to $2.05 per 100 Ib. BOD
removal, respectively.
2 Poultry process wastes
Plain sedimentation can remove 17-28%
of the BOD and 30-65% of the S.S.
4-10
5-17
-------
TABLE 4. TYPICAL RESULTS UNDER OPTIMUM
OPERATING CONDITIONS
Process Stream
Suspended
Solids, mg/1
Acid Hydrolyzable_
Phosphate, mg/1 PO
Turbidity
J.T.U.
Secondary Effluent
Chemically Treated
Settled Effluent
Filtered Product
45.6
11.0(767.)"
1.3(97%)
22.4
2.2(90%)
0.9(96%)
12.2
1.5(88%)
0.5(96%)
Operating Conditions:
700 gpd/ft2
82 mg/1
68 mg/1
Overflow Rate:
Alum Dose:
Aluminate Dose:
Silica Dose:
3 mg/1
* Denotes Removal Efficiency as %
Filtration Rate: 3 gpm/ft
Effective Size of Coal: 1.325 mm
Length of Run: 25 hours
Sludge Concentration: 1.2% by weight
Solids Concentration in
Filter Backwash WaLer: 475 mg/1
5-18
-------
50
CJl
vc
III
o
a.
in
Ml
O
a.
O
O
O
CO
Lbl
ae
30
20
10
I
o 100
FIGURE 7
INITIAL MINIMUM
TURBIDITY 16.0 j.u. 3.6 J.u. AT 300 mg/1
ALKALINITY 404 mg/1 191 mg/1
AS CaC03
pH
7.7
6.5
I
I
I
200 300 400
ALUM. CONCENTRATION, fflg/1 AS AL 2 (S
-------
SELECTED REFERENCES
1. Buzzell, J. C., and Sawyer, C. M., "Removal of Algal Nutrients
from Raw Wastewater with Lime," Journal Water Pollution Control
Federation. 39, R 16 (October, 1967).
2. Albortson, 0. E., and Sherwood, R. J., "Phosphate Extraction
Process," presented at Pacific Northwest Section Meeting of the
WPCF at Yakima, Washington (October, 1967).
3. Machis, A., "What Third-Stage Sewage Treatment Means," American
City. 110 (September, 1967).
4. Malhotra, S. K., Lee, G. F. , and Rohlich, G. A., "Nutrient Removal
from Secondary Effluent by Alum Flocculation and Lime Precipitation,"
Int. Journal Air and Water Pollution. 8_, 487 (1964).
5. Owen, R., "Removal of Phosphorus from Sewage Plant Effluent with
Lime," Sewage and Industrial Wastes. 25, 548 (1953),
6. Slechta, A. F., and Culp, G. L., "Water Reclamation Studies at
the South Tahoe Public Utility District," Journal Water Pollution
Control Federation. 39_, 5, 787 (May, 1967).
7. Stumrn, V., Discussion on: Rohlich, G. A., "Methods for the Removal
of Phosphorus and Nitrogen from Sewage Plant Effluents," Advances in
Water Pollution Research, Vol. 2, Pergamon Press, The MacMillan
Company, New York, N. Y. (1964).
8. Mulbargor, M. C., Grossman, E., and Dean, R. B., "Lime Clarification,
Recovery and Reuse for Wastewater Treatment," (in preparation for
publication).
9. Smith, R., "Cost of Removing Phosphorus from Wastewater," Seminar
on Phosphorus Removal, Chicago, Illinois (May 1 & 2, 1968).
10. Suhr, L. G. and Culp, R. L., "Design and Operating Data for a 7.5 NOD
Nutrient Removal Plant." Presented to the Water Pollution Control
Federation Annual Meeting Chicago, 1968.
5-20
-------
PHOSPHORUS REMOVAL
BI
MINERAL ADDITION
X. F. BBTth
ADVANCED WASTE TREATMENT SEMINAR
Portland, Oregon
February 5 and 6, 1969
-------
PHOSPHORUS REMOVAL BY MINERAL ADDITION
Portland, Oregon
February 5 «nd 6, 1969
combined «^«»'H'lgaT~T?lolLO|gL^**3- Approach 'too phosphorus ^**"^y*f^ offers
several advantages.
Screening data shovs that effluent phosphorus concentrations are re-
lated to the cations present In the waatewater.
Various flov patterns are used in the mineral addition process. Results
from dosing a 2 mgd activated sludge plant will be presented.* Comparison
of alum versus eliminate and observed operational problems vill be die-
cussed.
Logistics of mineral addition versus tertiary coagulation vlll be
considered*
results of dosing *\mit{nmii compounds directly onto tri/**^ tns filter
rocks will be shown.
Brief mention will be made of straight g*««^«t<*«7 processes for phosphorus
control of effluents along with a listing of advanced processes under
development.
The following slides vill be used and, vhere available, a reproduction
of the slide is attached:
Slide Ho. _ Title
1 Advantages of Direct Addition
2 Screening Data
3 Flov Diagram of Chemical-Biological Processes
k- Viant Diagram- of Pomona Plant
6-1
-------
5 View of Chemical Dosing at Pomona
6 Aeration Chamber - Pomona
7 Final Data - Pomona
fl ftH'Tf" Versus AluodLnate
9 pH and Turbidity
36 Mineral Addition Versus Tertiary
11 Falrborn Plant
12 Chemical Dosing — Falrborn
13 C^^^lcfl p»mp . Falrborn
1* Dosing at Weir crest
15 Filter Spray
15 Pilot Settler
17 Final Effluent * Falrborn
Ifi Results - Falrborn
19 Straight Chpmlcftl Approaches
20 Advanced Processes
6-2
-------
COMBINED BIOLOGICAL AND CHEMICAL PROCESSES
Add Mineral Supplement Directly to Aerator
I. Dual Use of Aerator
a) Mixing by Diffusion Apparatus
b) Residence Time for Reactions
II. Tremendous Surface Area of Floe
Available
III. Uses Existing Final Settler
6-3
-------
TABLE II
Results of Screening for Mineral Supplement
to Activated Sludge Process
Direct Dosage to Aerator
No Supernatant Recycle, No Primary Settling
Mineral Addition
None (base line)
Ca. 150 mg/1
Ca, 150 mg/1
plus
F, 6 mg/1
Mg, 20 mfi/1
Fe. 15 mg/1
Ai, 20 mg/1
Al, 30 mg/1
plus
Ca, 20 mg/1
Introduced as
-
CaO
CaO
NaF
MgS04
FcCl3
A12(S04)3
A12(SOA)3
CaO
To Form
-
Hydroxyapatite
Apatite
MgNH4P04
FeP04
Al?04
A1P04
Overall Rnuoval
40X
641
75%
50%
73Z*
70**
901
* Turbid effluents
6-4
-------
CHEMICAL
ADDITIVE
J
PRIMARY
I
BIOLOGICAL
PROCESS
;
f flecycte
FINAL I
JlfW Tl
i /
(Rof THOttAS. 1935)
I
Sludge
PRIMARY
T
Sludge
FLASH
IMIX-
CHEMICAL
ADDITIVE
BIOLOGICAL
PROCESS
tEBERKARO AND
NCSDITT ISC7)
Sludge
D/IICAL
ADDITIVE
HIGH RATE
BIOLOGICAL
PROCESS
FLOCULATION
FINAL
Sludge
I
CF
JtRaf. TENNY AND
f STUMU.BS5)
f Recycle
FIGURE 3. COMBINED CHEMICAL- BOLOGICAL, VARIATIONS
OF BASIC SYSTEMS
6-5
-------
A.F.iT. *;*
I J
t*
Aft BIT ia*0
I
I I "
I t I
MB* .HCt -I-— '-
»UCUtfM ! ' I
i
I ':
,._.._
i 1
i |
: i
i * i
i •
p|
'""I?
r it
*
: •
111
j
;
,
i
•
•
I
1
1
/
1^ J— itLa^. _L
'L
i]». : ,*^i-»*«- ; --•»
r- ~ [^ ', ''.' -•'-.-••
'• : i i ! i - !
Mr!-'
/ i • ; '
I ~- ' U3 -
?" jT: 71 ]
;t" c j curETilziJ
•VbUbflS . ...'.......
N
v
\
..A
I
1
*
9
I"
i
- i
-
-
-------
.9
I
V"
I
//i
99
.•9
s ' JS
or 74to$pftor:r/s fe
! I
4U/A
I !
/.4t
f
! i
I.Si
«f»!-/.if
/tf-*V
u,
! i
/i«
&8
/Jiff
/A 8
/£
I i
I
»L/Ulp
• i
iKfr W.-9
tu
».¥
-t
-------
mBK
w3£SX
. i • : i
±izlirn
-i-rn* i;--;.
76.—\——J*S. :
-------
-------
ir
•i
; ; i...(., j
iM-
-
I
j.
JflWff
m^
i>!••"•'iliH-ji-ij-- i-i '*ttj -H^ij+;ir];-»4-rJ ~^!Ji •'; j t1ia.
•]. :;:'^tm:f;^W! f'"S^^Sffl^g'n"'^ ^ ^<^ffi^:1l'i ilTi !l»'
j • -f:lft: |l-||- -| H§t ttHllrT i: ll'ff "^R^SfflEl^ffiK^
'I Ti'i. riilYHMi;Tff^i-VTw1::-+lTffr1?t:1^^
U^U-U-i-U-U..
-------
100
I
_J
<
>'
O
S
UJ
o:
O
I
DL
i/>
O
I
Q.
90
O
85
MINERAL
ADDITION
COMPARISON OF PHOSPHORUS REMOVAL
BY MINERAL ADDITION AND TERTIARY TREATMENT USING ALUM
(POMONA, L.A.C.S.D.)
80
I
I
1.0
1.2
1.4
1.8 2.0
AhP RATIO - tb./lb.
2.2
2.4
2.6
2.8
-------
FA1RBORN. OHIO TRICKLING FILTER
(Percent Removals)
Unit efficiency - Primary effluent to final effluent
Dosed
Filter
Control
Filter
Total
Phosphorus
63
13
pH
COD S. S. Mode Al : P
74 51 7.9 1:1
75 51 7.8
Overall efficiency
Dosed
Filte,r
Control
Filter
Entire
Plant
Total
Phosphorus
64
17
37
- Raw waste water to final effluent
pH
COD S. S. Mode Al : P
83 84 7.9 1:1
83 85 7.8
84 92 7. 6 0.3:1
6-12
-------
_
AOO IT tVf
Sludge
(Ref. RAND AND MEMEROW. 1965)
CHEMICAL
ADDITIVE
\
PRIMARY
Sludge
BIOLOGICAL
PROCESS
Sludge
B
(Ref BUZZEL AND
SAWYER. 19671
CHEMICAL
ADDITIVE
SOLIDS
CONTACTOR
1
BIOLOGICAL
PROCESS
t Recvde
FINAL
INCINERATOR
"[Stoker
.ASH
( Ref. DORR-OLIVER. I«i7)
IN A.8.0RC WHEN
USING LIME RECARBON-
NATIQN MAY BE CON-
SIDERED.
FIGURE 2.CHEMICAL TREATMENT VARIATIONS OF BASIC SYSTEM
6-13
-------
TABLE Ttf. ADVANCED PROCESSES FOR
NUTRIENT CONTROL
ION EXCHANGE
ELECTRODIALYSIS
REVERSE OSMOSIS
SORPTION
ELECTROCHEMICAL
DISTILLATION
6-14
-------
REUSE AND DISPOSAL OF LIME AND ALUM SLUDGES
by
Robert B. Dean
Chief, Ultimate Disposal Research Program
Advanced Waste Treatment Research Laboratory
Second Technical Session
2:45 P.M.
Wednesday, February 5, 1969
Symposium on Nutrient Removal and Advanced Waste Treatment
Portland, Oregon
-------
I. Removal of phosphates by lime leads to sludge that must be disposed of.
A* Chemical equations representing removal of phosphate by calcium hydroxide
and aluminum sulf ate
Treatment
60T + 3 1^0 (l)
Equations 2 and 3, which are the lime softening reactions, also occur:
Ca(HC03)2 + Ca(OH)2 — ) 2 Ca003 J + 2 HgO (2)
Mg(HC03)2 + 2 Ca(GH)^ - » 2 Ca3 ^ + Hg(OH)2 ^ + 2 HgO (3)
B. Ca_(OH)(PO. )_ (hydroxyapatite) is the most stable form at pH's greater than
7-7.5, but precipitation is slow unless there is a large excess of OH". The
precipitates Indicated by "^" settle in the sludge together with other
insoluble material.
Co Lime recovery
CaCO - » CaO + CO2 (4)
Hg(OH)2 - » MgO + HgO (5)
CaO + H20 - > Ca(OH)2 (6)
MgO does not dissolve in vater at normal temperatures.
7-1
-------
D. Treatment of sludges
1. Thicken. Recycle supernatant.
a. Conventional circular clarlflers are effective either upflow or
downflov with scrapers. Lime sludge settles well but Is retarded
by large quantities of Mg(OH)g.
b. Lagoon.
2. Recarbonate. Recarbonation with C0g to pH 9.5 redlssolves Mg(OH)2,
makes filtration and centrifugation easier. Mg(HCOo)2 stream cannot
be recycled and could cause problems at some locations.
3* Dewater. Recycle supernatant.
a. Centrifuge—as at Lake Tahoe. Figure 1. Suhr '68.
b. Filter—rotary vacuum filter or pressure filter.
4. Recalcine. Capture fly ash and odors. C00 to atmosphere. See
Crow '60, Wertz '60. 2
a. Multiple hearth furnace—Lake Tahoe.
b. Fluidized bed incinerator.
c. Horizontal rotary kiln—large scale. May have trouble with dirty
lime.
5. Slake. CaO + HgO j» Ca(OH>2 + heat. Insolubles (ash) serve as
a filter aid and sludge conditioner.""**11
a. Slurry. Cheapest for clean lime.
b. Paste—easier to adjust to dirty lime.
c. Dissolve Ca(OH)p. Solubility l£00 mg/1 at room temperature.
Discard ash. PEP process. Dorr-Oliver. Figure 2.
6. Discard excess.
a. Land fill. Ca5
-------
Phosphate Removal at South Lake Tahoe,
with Lime Recalcination and Reuse
Lime feed
INFLUENT
N
l
Effluent to
Ammonia Stripping,
Recarbonation,
Filtration,
Carbon Treating,
and Chlormation
Lime feed to rapid mix basin
-------
Solids Handling and Lime Recovery in PEP System
WASTE
PRIMARY
SLUDGE
DILUTION WATER
RECYCLE
MAKEUP LIME
GRIT REMOVAL
WASTE ACTIVATED SLUDGE
(CENTRATE)
TH.CKENER |(SLUDGE)>
(RECYCLE
LIME SOLUTION)
CENTRIFUGE
(SLUDGE)
(GASES AND
CALCINED SOLIDS)
WASHER
T
OFF-GAS
EFFLUENT
WATER
ASH (TO WASTE)
7-f-
-------
K
i
CO
0
i
UJ
Q
_J
Ui
cc
111
FIGURE 9
FILTER YIELD RATE VS.
LIME RECYCLE
MEAN VALVES
IMMERSION FRACTION- 0.25
VACUUM = 15" Hg
LIME DOSE tngACa
TO ACHIEVE A pHlL
450
300
225
2
LIME RECYCLE
-------
I. Removal of phosphates by aluminum hydroxide. See Lea '54, Stumm '68.
A. Chemical equations.
Influence of pH on Al ^ form:
A1(OH)3^
p 3 PH 7 PH 12
Phosphate removal:
2A1(OH) + HPO^= . a, AlPO^'Al(OH) A + 20H" + HgO (8)
1. Phosphate removal is a function of pH and other constituents of
waste stream. Optimum near pH 7«
2. Alumino phosphate shown is n\>t a pure ccnipound and may contain
calcium. Approximately 2 moles of Al are required per mole of P.
Figure 3.
B. Treatment of sludges.
1. Thicken. Neubauer '68.
a. Lagoon. Ineffective. Never solidifies. Maximum of about
8-lo£ solids.
b. Freezing. Very effective if done slowly. Doe '67 and
Christensen. Expensive if artificial freezing is required.
125 kwh/1000 gal.
c. Sand beds.
d. Precoat vacuum filter.
e. Pressure filter with lime.
f. Centrifuge. Ineffective.
2. Filtrate or supernatant can be recycled. If lime is used, this may
contain useful aluminate.
3. Solids can be used for land fill if 20$ or greater.
7-6
-------
REMOVAL OF PHOSPHATE FROM WASTEWATER BY ALUMINUM HYDROXIDE
1.0
•vj
I
-J
o:
ui
CL
cn
o
u.
O
0.8
0.6
0.4
0.2
0123
FIGURE 3
MOLAR RATIO OF Al ADDED TO PHOSPHORUS INITIALLY PRESENT
-------
k. Recovery with alkali. See Lea '5^ Slechta '67.
Solubilization:
5Ca(OH)2 + 60H" (9)
6Al(OH)i,."
Loss of aluminum ions as calcium aluminate precipitate:
Ca**" + 2Al(OH)jf -
a. Aluminate may be reused in process.
b. May need heat. Requires large vessels. Process works best
at high dilutions avoiding precipitation of calcium aluminate.
c. Up to 75# recovery as aluminate balances acidity from makeup'
alum. Higher recoveries of Al are unlikely.
d. Heat treatment of sludge improves filterability.
Acid recovery does not separate alum from phosphate and recycle
is not practical.
7-8
-------
Pertinent Literature
Chrifitensen, W., Manager, Municipal Water Works of the City of Copenhagen,
see Doe's report (p. 37)
Crow, W. B., "Calcination Techniques," Jour. AWA, pp. 322-326 (Mar. I960)
Doe, P. W., A report on "The Disposal of Sludge from Water Treataunt Plants,"
Jubilee Traveling Fellowship, 1966^7, pub. The British Waterworks Assn.
Lea, W. L., Rohlich, G. A., and Kats, W. J., "Removal of Phosphates from
Treated Sewage," Sewage and Industrial Wastes, 26, No. 3, 261-275 (Mar. 195*0
Mulbarger, M. C., Grossman, E., Ill, and Dean, R. B., "Lias Clarification,
Recovery, and Reuse for Wastewater Treataent," U. S. Dept. Interior,
FWPCA, Cincinnati Water Research Laboratory, Cincinnati, Ohio (Jan. 1968),
in preparation for publication
Neubauer, W. K., "Waste Alum Sludge Treatment," J. Am. Wat. Works Assn.,
60, No. 7, 819-826 (July 1968).
Slechta, A. F., and Gulp, G. L., "Water Reclamation Studies at the South
Tahoe Public Utility District," Journal Water Pollution Control
Federation, £g, No. 5, 787 (May 1967).
Stumn, W., and O'Melia, C. R., "Stolchiometry of Coagulation," J. Am. Wat.
Works Assn., 60, No. 5, 51^539 (May 1968)
Suhr, L. G., and Gulp, R. L., "Design and Operating Data for a 7.5 MGD
Nutrient Removal Plant." Presented to the Water Pollution Control
Federation Annual Meeting, Chicago, 1968.
Young, E. p., "Water Trea-fament Plant Sludge Disposal Practices in the
United Kingdom," J. Am. Wat. Works Assn., 60, No. 6, 717-732 (June 1968)
Webster, J. A., Daer Alum Recovery Plant, J. Instn. Water Engrs., 20, No. 2.
167-98 (1966) ' —' '
Wertz, C. F., "Miami Llae Recovery Plant," Jour. AWA, pp. 326-332 (Mar. I960)
7-9
-------
ALTERNATIVE METHODS OF PHOSPHORUS REMOVAL
Jesse M. Cohen, Chief
Physical and Chemical Treatment Research Program
Advanced Waste Treatment Research Laboratory
Cincinnati, Ohio
Technical Seminar
Nutrient Removal and Advanced
Waste Treatment
Portland,Oregon
February 5 and 6, 1969
-------
ALTERNATIVE METHODS OF PHOSPHORUS REMOVAL
J. M. Cohen
INTRODUCTION
It would be remiss if this conference were to end with the
impression that removal of phosphorus can be obtained only by reagents
consisting of iron, aluminum or lime. At the present time, these are
the reagents of choice; they have been shown to be effective in precipitating
phosphate, the technology and the engineering design' have been fairly well
advanced from small pilot-scale operation to relatively large application.
The costs have been shown to be reasonable. The purpose of my talk is
to point out that there are alternative methods for removing phosphates.
At the present state of development, these processes are clearly not
competitive. The research for some of these methods is just getting
started. It is entirely possible that some of these alternative processes
will never be competitive for the large waste treatment plant that can
afford the capitalization; the skilled operation and required manpower
to operate a removal plant based, say, on lime precipitation and calcination.
Special circumstances - perhaps the relatively small treatment plant may
dictate the selection of a process that has a higher unit cost for
removing phosphates but which requires less attention.
8-1
-------
ACTIVATED ALUMINA
One alternative process that has good prospects is the removal
of phosphates by activated alumina. Activated alumina is a synthetic
material consisting largely of aluminum oxide which has been activated
at high temperature. It is available in a variety of forms ranging
from fine powders to granules. A particular property is its high specific
surface area ranging from 200 - 400 square meters per gram. In this
respect it is similar to activated carbon which has surface areas of
= 800 - 1000 m2/gram.
The ability of activated alumina to remove phosphates and some
data to support this contention was published only as recently as 1966
by Yee working at the Oak Ridge National Laboratory. Not only could
phosphate be sorbed from solution by activated alumina but, most importantly,
the sorbed phosphate can be quantitatively desorbed by appropriate
regeneration chemicals. In this respect, alumina behaves like an ion-
exchange material but with some significant differences, as I will point
out later. The probable mode of operation would be column contactors,
much like those in current use for activated carbon. These require, as
you know, reduced capital costs and a minimum of direct labor costs.
The principal cost is that of the alumina itself, and especially the cost
of the regenerating chemicals and their final disposal.
There are a great, variety of activated aluminas available, varying
greatly in unit cost and capacity to sorb phosphates. Yee, tested three
commercially available types and this information is shown in Figure 1.
8-2
-------
Evaluation of Three Commercial Aluminas*
Item
Cost— i/lb
Surface area — sg m/g
Alumina sorption capacity
for mixed phosphates
(as PO43~)— mg/g
Mixed phosphates(as POi3")
eluted from alumina in
regeneration — ntg/g
Regenerant waste — B V's
Alumina dissolved in re-
generation — per cent
Type of Alumina
A
12
250
19
17
12
8
B
100
250
17
10
16
5
C
40
400
>27
28
12
35
* Performances were evaluated in upflow through
fluidized beds.
-------
The aluminas were packed in columns and were then fed solutions containing
phosphorus in a variety of forms such as or'thophosphate, tripolyphosphate,
pyrophosphate and hexametaphosphate. Flow rates were of the order of
0.25 to 1.25 gallons per minute/sq ft, about the rates normally used
for an ion-exchange process.
It is clear from this table that of the three-aluminas that were
tested only one, type A, would have commercial value for phosphate removal,
based on its lowest cost, 12^/lb and more importantly'on the low solubility
during regeneration. Also note that regeneration is almost complete -
17 rag/gram of phosphate was recovered from a loading of 19 rag/gram. These
data, incidentally, were obtained from a pilot-plant operation rather
than the bench-scale work previously referred to.
Some breakthrough curves are shown in the next slide, Figure 2.
The curves represent removal patterns for various forms of phosphate by
plotting bed volumes versus a ratio of influent to effluent phosphate
concentration. The smaller the value the greater the removal. With a
feed concentration of 25 ppm of phosphate, which is about what most
wastewaters will contain, phosphate ion could not be detected until more
than 1000 bed-volumes had passed - providing more than 99 percent removal.
No process that I know of obtains such complete phosphate removal. Also
note that alumina removes polyphosphates even better than orthophosphate,
with breakthrough curves indicating «* 1800 bed volumes before- breakthrough.
As a "polishing" process to remove the last trace of phosphate after
conventional treatment had reduced the concentration to say 1 ppm,
alumina may be very appropriate. Under these coaditions better than
8-4
-------
DO
3
O
JC
ra
-------
99 percent removal would be obtained while processing 20,000 bed volumes
of treated effluent and producing only about 10 bed volumes of regenerant,
a volume reduction factor of about 2000.
Time will permit only a brief listing of the additional advantages
of alumina.
1. Alumina is highly selective for phosphate and removes
polyphosphates even better than the orthophosphates.
2. In contrast to alum, iron or calcium salts, no additional
inorganic ions are added.
3. Treatment with alumina does not affect pH.
4. Variations in feedwater composition and quality have little
or no effect on removal efficiency.
5. Removal efficiencies are so high that extremely low residuals
are achieved, of the order of 0.05 ppm of phosphate.
Based on Yee's work, chemical costs can be estimated at
3.9 cent/1000 gallons for a feed containing 14 mg/1 of phosphate ion
and 6.4 cents/1000 gallons for a feed of 23 mg/1. These estimates assume
an 8 percent loss of alumina for each regeneration cycle, when in fact,
cycles after the second amounted to only 5 percent.
Activated aluminas have a wide spectrum of structural and sorptive
characteristics. Additional research should reveal the effect of these
properties on phosphate removal capacity. There is some evidence that
capacity can, in fact, be increased. The current regeneration system
using sodium hydroxide followed by reacidification with nitric acid may
not be best, and alternate systems based perhaps on ammonium hydroxide
which would enable regenerant recovery and reuse should be investigated.
Efforts are currently being made in this direction.
8-6
-------
Precipitation by Lanthanum
In the search for reagents which produce insoluble salts with
phosphate compounds, it was natural to consider very early the common
precipitants based on aluminum, iron and calcium. Another metal which
we are considering is lanthanum. Although the use of this salt may appear
to make the process prohibitively expensive, the possible greater efficiency
of the phosphate removal coupled with the use of a regeneration step could
make this approach attractive. The solubilities of the metal phosphates
are shown in the next slide (Slide ). The solubilities in terms of
mg/1 of phosphorus were calculated from reported solubility products.
It is clear that lanthanum phosphate is about as insoluble as iron and
aluminum phosphate. In practice, however, these theoretical solubilities
are not obtained with iron or aluminum even though large excesses of
precipitants are added. The reasons apparently are that competing
hydrolysis reactions reduce the concentration of metal ions available for
phosphate precipitation. This basis has been used to explain the' lower
solubility of aluminum phosphate despite the fact that its solubility
product is greater than that of ferric phosphate. Non-hydrolyzing calcium
added to wastewater in dosages several times greater than those of aluminum
or iron reduces residual phosphate to about the same level although the
theoretical solubility of calcium hydroxyapatite, the form of the
precipitate, is 6-7 orders of magnitude greater than the iron and
aluminum phosphates. In short, solubility product is not an accurate
guide to the practical usefulness of a precipitant. It should be
remembered that solubility product represents equilibrium conditions
and does not consider
kinetics. Thus, rate of precipitation is also an important consideration.
8-7
-------
SOLUBILITY OF PHOSPHATES
SALT K CALCULATED SOLUBLE P
SP MG/L
FE P04 10~23 9 X 10"8
LA P04 3.75 X 10~23 2 X 10"7
AL P04 10'21 9 X 10~7
(MG)3(P04)2 2 X 10"27 0.1
10-9° 0.3
-------
The extent of removal of phosphate by hydrolyzing metal ions like
iron and aluminum really is a function of both solubility product, rate
of reaction and importantly, of the competing hydrolysis reactions.
Laboratory studies have shown that both aluminum and iron form soluble
complexes in the presence of large excesses of either the metal or
phosphate ions. Moreover, under certain conditions a stable colloidal
precipitate which was difficult to remove from suspension-was formed.
These studies showed that lanthanum is, 1) only slightly hydrolyzed, and
2) shows less tendency to form soluble complexes with phosphate. These
factors led us to consider the technical feasibility of using lanthanum
as a phosphate precipitant.
To be successful, certain fundamental criteria would have to be
met: 1) the phosphate should be readily reduced below 0.1 mg/1 of
+3
phosphorus; 2) the residual La in solution should be negligible and
3) recovery and reuse should be inexpensive.
A quick test in our laboratories produced the data shown in the
next slide. (Slide ). The sample of wastewater used in -this test was
collected after a day of rain and hence showed a low phosphorus content.
A dosage of 23 mg/1 of lanthanum added as lanthanum nitrate reduced the
phosphorus residual to just detectable concentration - 0.1 mg/1 of P.
Also note that turbidity was reduced to 4 units.
When we first considered lanthanum as a precipitant, our enthusiasm
was curbed because the then current price was $3.00 a pound of lanthanum
oxide. A more recent price quotation of 60 cents a pound for a lanthanum-
rich mixture of rare earths is certainly more encouraging. Clearly,
8-9
-------
PRECIPITATION OF PHOSPHATE WITH LANTHANUM
SAND-FILTERED SAND-FILTERED
MG/L LA TURBIDITY ORTHO-P
0
23
46
69
92
139
17
4
0.6
0.4
0.9
1.5
1.2
0.1
< 0.1
< 0.1
< 0.1
<0.1
-------
it is not yet competitive with 3 cents a pound alum. To become competitive:
it will be necessary to use virtually stoichiometric dosages of lanthanum
to maintain low residuals of lanthanum in the effluent. But since the
phosphate concentration in the wastewater is constantly changing, some
means will have to be worked out to adjust lanthanum additions to phosphate
changes.
And, finally, methods will have to be developed to recover the
lanthanum with high efficiency. From desk-top considerations an alkali
regeneration process appears to be promising. Use of lanthanum for
phosphate removal will be explored for technical feasibility.
Ion Exchange
Ion exchange materials, both natural and synthetic, have potential
for removing phosphates from wastewater, although the primary function is
the removal of inorganic ions in general. But modifications can be made
on natural ion-exchange materials to obtain a degree of selectivity for
phosphate ion, in which case, ion-exchange as a process for phosphate
removal becomes more attractive.
A typical ion-exchange material can be looked upon as a micro-
porous matrix containing many ionic sites, either all positive or all
negative as shown in the next slide (Slide ). In order to maintain
electro-neutrality, each fixed ionic site must have associated with it
and ion of opposite charge or counter ion. This counter ion must be
mobile enough so that it can be replaced by, or exchanged for, another
ion of the same charge upon contact with a solution of this new ion.
8-11
-------
TYPICAL EXCHANGE MATERAL
-------
When all of the original counter ion is replaced, the ion-exchange
material is exhausted. A concentrated solution, 4 to 10 percent,
of the original counter ion must be used to regenerate the ion exchanger.
Depending on the counter-ion the exchange material can be either a
cation or anion exchanger.
The most important group of natural ion-exchange materials are the
zeolites and clay minerals. These materials are largely alumina silicates
which generally carry a net negative charge and hence are dominantly cation
exchange materials. But certain of the clay minerals are capable of
exchanging anions as well. For example, exchange of hydroxyl ion for
chloride, sulfate or phosphate ions has been observed in montmorillonite,
kaolinite and certain feldspars and micas. The problem with these
naturally occurring ion-exchangers is that anion capacity is low -
generally less than 10 percent of the cation exchange capacity.
It is technically possible, however, to chemically treat the natural
mineral so that replacement of certain ions in the clay structure occurs
with iron or aluminum ions. These replacement ions can now react with
phosphate ions. Depending on the degree of replacement obtainable by
chemical modification one can expect increasing capacity to react with
phosphate. It should be possible to regenerate the clay mineral by
stripping the phosphate off with some regenerating reagent.
Currently, this concept is in its barest infancy. We know that
these minerals have some natural phosphate exchange capacity but it is
too small to be useful. It's been shown that this capacity can be
increased by appropriate chemical modification. How much increase
remains to be determined. The process could become attractive economically
8-13
-------
since the raw materials, clay minerals, are abundantly available and
cheap - just cents a pound. If chemical modification costs are modest
and if the increase in exchange capacity is reasonable then the process
could find use for phosphate removal. At the present time a modest
research program is being directed toward developing this treatment concept.
The most important ion-exchange materials are the synthetic organic
types with which all of you are undoubtedly familiar. Synthetic anion
exchangers have adequate capacities to remove anions from the water. The
problem is that there is competition among the various anions for the
exchange sites; thus the selectivity of the resin for phosphate ion
becomes a critical factor. The competition is not random but is dictated
by considerations of ionic size and ion valence. This selectivity can be
illustrated by the breakthrough patterns shown in the next slide.(Slide ).
In alkaline wastewater, the dominant phosphate species is the HP04
double negative. The resin selectively exchanges this ion in preference
to Cl , NC>2> or NO^. Unfortunately, the resin has about the same
preference for sulfate ion as for phosphate, thus the anion exchange
capacity must be shared with the two anions. The problem here is that
wastewater generally contains 2 to 5 times as much sulfate as phosphate
ion, thus effectively reducing the capacity to remove phosphate ion.
In order for synthetic ion-exchange materials to become a
competitive process, they must become highly selective for phosphate ion.
Such selective resins have been developed for certain ions such as
potassium, fluoride and borate. None, to my knowledge, has been developed
for phosphate ion, and until such materials are made, synthetic ion
exchange will not be economically useful for removing phosphate.
8-14
-------
o
c?
PHOSPHATE
SULFATE
100 200 300
TMROUGHPl
-------
REMOVAL IN SOIL SYSTEMS
One of the processes for removal of phosphates that has
generally been overlooked is use of naturally occurring soil. Soil
systems as treatment devices, rather than as means of disposal, are
receiving more attention as a potentially cheap method of treating wastes.
Soil is a composite medium that may contain relatively inert rock,
gravel and sand as well as reactive clays and minerals; and additionally,
varying amounts of organic materials such as humus, living and dead
vegetable and animal matter. Many of the components possess useful
properties to accomplish effective waste treatment. Treatment mechanisms
include biological oxidation, chemical precipitation, ion-exchange,
adsorption, chemical oxidation and plant assimilation. The sum total of
these effects is capable of providing very effective treatment of
wastewater.
A major effect is the removal of phosphates by the soil system.
The potential capacity of soil to remove phosphate is enormous. Actually,
hydraulic failure is more likely to occur well before the phosphate removal
capacity has been taxed. Data we have obtained from field and lysimeter
studies at Westby, Wisconsin, and Loveland, Ohio, are shown in the next
slide. (Slide ). These data were collected over a number of years.
In almost every instance, better than 90 percent of the phosphorus was
removed by a six-foot depth of soil. Phosphate removal was not greatly
affected by seasonal variation, vegetation, or anaerobic soil conditions.
Other data we have obtained show that most of the phosphate, approximately
75 percent, is actually removed by the upper 1-foot layer of soil.
8-16
-------
EXPERIMENTAL RESULTS
PHOSPHATE REMOVAL IN SILT LOAM
INFLUENT P 5 FT PERCOLATE PERCENT
MG/L MG/L REMOVAL
WESTBY
SOIL & VEGETATION 9.7 2.4 75ll)
SOIL & VEGETATION 9.7 0.7 93
LQVJ1AJU1
BARE SOIL, WINTER 6.3 0.3 95
BARE SOIL, SUMMER 11.5 O.i: 86
SOIL 8 VEGETATION, SUMMER 11.5 0.15 99
BARE SOIL, ANAEROBIC 13 1.0 92
(1)1 FT PERCOLATE
-------
More important to this discussion is the fact that these systems
had by no means begun to-exhaust the potential for phosphate removal.
The next slide (Slide ) shows the phosphate loadings in our studies
as compared to reported capacities. In the 6 years of field operation
at Westby, Wisconsin, 1.6 tons of phosphorus had been fixed in the upper
6 inches of an acre of soil. Capacities for some coastal sandy soils
range from 11.2 to 40 tons of phosphorus per acre half-'foot. Some coastal
sandy soils have capacities ranging upward to 205 tons per acre half-foot.
It is obvious that only a very small portion of the potential capacity
had been used at either Westby or Loveland. Both of these operations were
operated at about 1 gallon per day/sq ft. At this rate the phosphorus
capacity would not be exhausted for upwards of a century.
The mechanisms for phosphate fixation in the soil are complex and
can be attributed to many factors. Among the more important are: type
of soil, particle size, pH, reduction potential, temperature, organic
content and reaction time. All of these factors contribute to the mechanisms
of anion exchange and adsorption, chemical precipitation and assimilation
into organic matter.
The potential of soil to treat wastes and to remove phosphates
should mot be ignored. When land is available, this method should be
considered.
8-18
-------
EXPERIMENTAL RESULTS
PHOSPHATE REMOVAL
WESTBY SYSTEM (SILT LOAM)
LOVELAND SYSTEM (SILT LOAM)
SASSAFRAS SOIL
.H 6.5
DURATION
6 YEARS
1.5 YEARS
COLLINGTON SOIL pH 6.5
DUTCHESS SOIL pH 6.5
(1) TONS PER ACRE-HALF-FOOT
(2) BAILEY, FROM TOTH & BEAR, 1947
PHOSPHATE-P
REMOVAL
1.6
0.4
CAPACITIES (2)
11.2
40
35
- THESE ARE SANDY SOILS.
-------
REVERSE OSMOSIS
Phosphates can be removed'by the process of"reverse osmosis,
although .:the process is being developed principally as a device to
remove contaminancs in general.
A simple description of r.everse osmosis', is shown in the next
slide. .(Slide ). When two solutions of unequal salt concentration
are separated by a semi-permeable membrane, there is a tendency for pure
water to., flow from the more dilute to the more concentrated solution. The
driving force is the salt concentration gtadient. The semi-permeable
membrane is simply an imperfect barrier between the two liquids "which
selectively permits passage of some component"of the solution, either
solvent or solute.
In the case of osmosis - diagram on the left - (Slide )•'water
passes through the membrane and the inorganic ions are rejected.' Under
the impetus of the concentration gradient, the water flows through the
membrane, and a higher pressure is created in the receiving vessel
(called the osmotic pressure); the salt concentrations in the two
vessels, thus tend to become equal. Now, if additional pressure is
applied to the solution already under greater pressure, the water flow
will be.-reversed; that is, water will,now -flow in the reverse direction,
hence the name reverse, osmosis or ultrafiltration.
Reference to f-iltration. is appropriate, since water flows from
the more -oncentrated to tfce more dilute side whil« salt ions are rejected
or filtered. The magnitude*ot the tlow and the degree or salt rejection
are dependent upon the applied pressure, t'he characteristics of the membrane,
among other factors.
8-20
-------
OSMOSIS AND REVERSE OSMOSIS
OsmCSiS
.*""*N
/V-e'
(I © ?
% * ^-^ S
SL *o e«« ,
to tf*
j u/a
1 ^
f
c * *
0 » • e °
X C 0 ° e
o e (C9\ o c
00 e c
o ° o • c
mrmbrane
o o
0
Process
OSMOSIS
REVERSE OSMOSIS
Driving Force
o£ Process
concentration
gradient
pressure
C1>C2
Primary
Flow
solvent
(water)
solvent
(water)
Principal
Result
the concentrations are equalized
pollutants are removed from the water
-------
Membranes can be made with selectivity for certain ionic or
nonionic materials, but current research is directed toward development
of membranes which remove all contaminating substances. However, as a
general rule, ion removal is directly proportional to ion valence, thus
multivalent phosphate and sulfate are more readily removed than monovalent
chloride.
Large scale tests of reverse osmosis have shown that virtually
100 percent of the phosphate ion is removed.by "tight" membranes while
94 percent was removed in the "looser" membranes.. 'A reverse osmosis.
unit currently available from a manufacturer has been designed to'remove
organic and particulate matter and to operate at the relatively low pressure
of less than 100 'pounds per square inch. Ninety-five percent removal of
the phosphate ion is claimed. The manufacturer puts the cost of the
total treatment, which includes phosphate removal, ;at about 30 cents.per
1000 gallons.
Thus, if reverse osmosis can be developed to an economically useful
process for partial demineralization, it is useful-to know that phosphate
will be almost completely removed.
8-22
-------
Sludge Blanket Process
In the final portion of this discussion I would like to comment
on the merits of the sludge-blanket or up-flow clarifier. This, of course,
is not an alternative method of removing phosphate but rather is an
alternative to the widely used horizontal-flow flocculators and
sedimentation basins. There are some significant advantages to up-flow
clarification. Capital costs for equivalent flows are lower; since the
up-flow system requires a single vertically oriented tank with a much
reduced detention time, construction costs and land requirements are
reduced. Several unit processes take place in a single tank - coagulation,
flocculation, sedimentation and sludge removal.
Because of the extremely complicated hydraulics and the multi-
functional character of the device, there has evolved a great number of
individual designs. For those of you who are not familiar with this type
of equipment, the next slide (Slide ) illustrates the usual components
of a sludge-blanket clarifier.
Incoming wastewater dosed with coagulant enters the unit at the
central well and then flows downward to the bottom of the twin tanks
which contain a "blanket" of coagulant sludge. The flocculated waste-
water is thus caused to flow through the blanket. Clarified supernatant
is withdrawn at the top and settled sludge is withdrawn at the bottom.
Upflow clarifiers are operated at rates of 10-15 gal/hr/sq ft which
provides detention times of about 1 hour. Sludge volume generally amounts
to less than 5 percent of the throughput. The ability of the sludge-
blanket device to remove pollutants such as phosphates, suspended solids,
color and organic solids is equal to or greater than a corresponding
horizontal unit. A major advantage is the intimate contact obtained
in the sludge blanket thereby enhancing chemical and physical reactions.
8-23
-------
Sludqo
CROSS SECTION OF A CIRCULAR
UPFLOW REACTOR
(AFTER AITKEN)
-------
Preliminary results in our laboratory with an experimental unit
showed phosphate removals of 80-97 percent with blanket depths of 4-8 feet,
using alum, ferric sulfate and lime.
We have currently under way a study of this device for application
to wastewater treatment.
Concluding Remarks
I hope that these brief remarks on alternative methods for removal
of phosphates have made the point that our program of research on phosphate
removal is considering methods other than the ones based on alum or lime.
I want to emphasize that none of the methods I've discussed is now ready
for application. Hopefully, research will make some of them useful for
application in specialized instances.
8-25
-------
REFERENCES
1. Engelbrecht, R. S., and Morgan, J. J. ; "Studies on the Occurrence and
Degradation of Condensed Phosphate in Surface Waters," Sew. and Ind.
Wastes, 31, 458 (1959).
2. Klotter, H. E., "Investigations on the Removal of Orthophosphates and
Condensed Phosphates from Sewage Water." Vom Wasser 31. 181 (1964).
3. Yee, W. C., "Selective Removal of Mixed Phosphates from Water Streams
by Activated Alumina." Jour. AWWA, 58, 239 (1966).
4. Matijevic, E., and Milie I., "On Precipitation Effects of Aluminum,
Lanthanum, Iron(III) and Thorium Phosphates." Kolloid Z., 188, 129
(1963).
5. Eliassen, 'Rolf, and Tchobanoglous, G. , "Removal of Nitrogen and Phosphorus."
23rd Purdue Industrial Waste Conference, Purdue Univ., Lafayette,
Indiana, May 8, 1968.
6. Eliassen, Rolf, Wycoff, Bruce M., and Tonkin, Charles-D., "Ion Exchange
for Reclamation of Reusable Supplies." Jour. AWWA, 57. 1113 (1965).
7. Garland, Chesley F., "Wastewater Renovation and Reuse." Infilco
National Seminars, pp. 15-25, Ifilco, Tucson, Arizona (1966).
8. Greenberg, A. E., and McGauhey, P. H., "Chemical Changes in Sewage
During Reclamation by Spreading." Soil Science, 79. 33, (1955).
9. Russell, G. C., and Low, P. F., "Reaction of Phosphate with Kaolinite
in Dilute Solution." Soil Science Society Proc., 18. 22 (1954).
10. Bailey, G. W., "Role of Soils and Sediment in Water Pollution Control."
U.S. Dept. of the Interior, FWPCA, Southeast Water Laboratory, March
1968.
11. Pennypacker, S. P., Sopper, W. W., and Kardos, L. T., "Renovation
of Wastewater Effluent by Irrigation of Forest Land." Jour. WPCF,
3_9, 285 (1967).
12. Preul, H. C., "Contaminants in Groundwater Near Waste-Stabilization
Ponds." Jour. WPCF, 40, 659 (1968).
13. Bendixen, T. W., Hill, R. D., Schwartz, W. A., and Robeck, G. G.,
"Ridge-and-Furrow Liquid Waste Disposal in a Northern Latitude."
Jour. San. Engr. Div., Proc. ASCE, 94, 147 (1968).
8-26
-------
14. Anon., "Reverse Osmosis as a Treatment for Waste Water." Aerojet-
General Corp.
15. Foyn, E. , "Removal of Sewage Nutrients by Electrolytic Treatment."
Inter. Ver. Theoret Angew Limnol., Verhandl., 15, (1962), Published
in English (1964).
16. Merten and Bray, "Reverse Osmosis for Water Reclamation." 3rd Int'1
Conf. on Water Pollution Research.
8-27
-------
COMBINED TREATMENT FOR REMOVAL
OF HITROGEN AHD PK)3PHORUS
E. 7. Barth
ADVANCED WASTE TREATMENT SEMINAR
Portland, Oregon
February 5 and 6, 1969
-------
COMBINED TREATMENT FOR REMOVAL
OP NITROGEN AHD PHOSPHORUS
Portland; Oregon
February 5 and 6, 1969
The Importance of nitrogen removal is much more broadly based than
ita ability to serve as an algal nutrient In receiving waters.
The biological oxygen demand of most effluents is largely due to
Plant efficiencies of COD or suspended solids removal have no
correlation with nitrogen removal.
By accident of design some plants are efficient for nitrogen
removal*
Effective control requires a design that considers the cycle of
nitrogen. A three-stage denitrification process has been developed.
The mineral addition process for phosphorus control can be united
with the biological denitrification process. Details of this
process will be presented, along with tentative cost estimates.
Dentriflcation can also be accomplished on carbon or sand columns.
A brief outline of the denitrification columns at Pomona will be given.
Because most nitrogen compounds are soluble there are only a limited
nonber of strict chemical approaches to nitrogen control.
9-1
-------
Bie following slides will "be used and, where available, a repro-
duction of the slide is attached:'
Slide Ho. _ Title _
1 Importance of nitrogen
2 Carbon Oxidation Versus Aononia Oxidation.
3 ELant Efficiencies
k Flov Diagram - Archibald
5 Utrogen cycle
6 Biree-Stage Denltrlf loatlon
7 Objectives of Combined Rocess
8 Flow Sheet of Combined Krocess
9 Changes in Cooposition
10 Mass Balance - Efaosphorus
11 Mass Balance - Nitrogen
32 Mixed Liquor Characteristics
13 Bff ect of A'*'"'"'fir""» on Sludges
Cost Data
V!iev of Pomona
16 Bltrate Concentration Versus Column Time
17 General Data - Banana
18 QuiiirtCHl Approaches to Bitrogen Removal
9-2
-------
IMPORTANCE OF NITROGEN
NH, in effluents can cause DO sag in receiving water
NH is corrosive to copper fittings
1 NH requires 7 plus C12 for breakpoint
NO causes high C12 demand
NH influences C12 contact time
Nitrogen compounds are nutrients
NO can be health hazard
9-3
-------
RELATIONSHIP OF CARBON OXIDATION
AND NITROGEN OXIDATION
Organic Matter
Oxygen Demand
NHo
Oxygen Demand
Total oxygen
JjpKffunA
Oxygen Demand
Due to HE*
ag/1
Wastewater
250
375
25
112
W7
22*
Final Effluent
25
37
2O
90
127
71*
9-4
-------
(T) L'
(nut of tliu .luh Mum .lni UN..i, W.\-jni I'ui.u.'iia.v C'OMI.OL
I'rinii-fl in t S A
Vol. 38(7), 1203-19 (1966)
TAULE III. -KtHeicncy ol COD Romoial, Fnc-Daj Averages
7-oc.itiun anJ
ItalinCi. I'cnuJ
Archbold, Ohio
I
II.uiiilli.il, Ohio
I
II
Lebanon, Ohio
I
II
Love-hud, Ohio
I
Milfuid, Ohio
I
II
Influent
(me/I)
Oil
ol.S
-tot
120
332
32G
513
Col
Pnnmr) Kf!Ju< at
('HB/D
5--JS
270
205
100
127
32G
SSI
31-1
Itf not Jl
(?i)
43
55
:io
21
02
0
20
52
I'mil Fil'iiriit
(M.E/I)
Cl
02
5S
10
C9
12S
151
10
IVnunal
C~J
S'J
77
SO
51
•15
Cl
GO
-IS
Oicrill
Reno^al*
(r<)
94
00
S7
Gl
70
Cl
GO
7-1
1 f!cfoie
\sitli any
\\.isto.
TABLE II.—Elneiencj of Suspended Sol
-------
© Cwi>)-rii;lii us imrt of the .hilv |!HM .|ui KX.M. WATKM POLLUTION CDXTKOL
Wuhliiiiulun, J). O. 20010
I'rininl ID U S. A.
Vol. 38(7), 1208-19 (1966)
TABLE V.—Nitrogen Removal Efficiencies
Finn: ami
H.il.mce
An lihnM
I
Ihiiiiilton
I
II
i
ii
Lowland
I
Milford
• I
II
Nitro^tn \rriiuntrd
for
<'*)
Primary
Treat-
ment
102
US
93
102
lOo
101
108
07
Sccund-
ar>
Treat-
me n i
G2
'.15
08
81
70
74
SG
100
I'lVmcno of Niiru;in Kvmotal
IVt)
Priin.iry
nii-iii
0
-27
2
5
2S
0
17
10
Sfcond-
ar>
Tn.it-
nunt
57
31
20
8
42
30
25
8
OviTjII
Ri : toval
Gl
1C
22
13
58
30
•
38
25
NitriruMtion IXua
(ti.i-ij on nitroL-i-n rnli-iins fccoii'lary
(rrati.-.i-nt)
('it
Nitrogen
KlMliiM I
bi Dint-
tnfu-jlion
3S
5
2
19
24
2G
14
0
\|i| .I'Cllt
.Niir--.rn
Ox«!.zi 1
22
0
2
4S
2G
10
1
1
Total
Nil.'i >,«•'»
GO
5
4
G7
50
3G
15
1
Nitroscn
In Kinal
.KAIticnt
50
0
8
CO
ai
14
5
3
9-6
-------
SUPERNATANT
INFLUENT
2,010
SECONDARY
DIGESTER
482
FILTRATEn
DIGESTED
SLUDGE
CAKE TO
LAND FILL
PRIMARY
DIGESTER
PRE-
AERATIDN
PRIMARY
SLUDGE
1,416
EFFLUENT
maaat
1,904
PRIMARY
SETTLER
423
WASTE
SLUDGE
1 mgd
RECYCLE
FINAL
SETTLER
FINAL EFFLUENT
•••
777
ESTIMATED RESERVOIRS OF NITROGEN
AERATION UNITS 600 POUNDS!
DIGESTERS 2750 POUNDS
0
r
<
«
REAERATION
A -41
AERATION
A= 0
REAERATION
AERATION
dw
NT
A
Srf^
( )* indicates sum at this point
t multiply by 0.454 to obtain Kg
Figure 1. FIVE-DAY NITROGEN BALANCE: ARCHBOLO. OHIO. ALL VALUES IN POUNDS
-------
Cellular
Synthesis
Influent _L excess „ -
Bitrogen ^^3 ^ N°2
,1 I .
plus HO.
Primary A.r.inc plus Carbon
FIGURE 1
9-8
-------
FIGURE 5
IMARY
HIGH RATE
C — C02
RETURN
SLUDGE
NITRIFICATION
— N03
DENITRIFjCATION
/N
RETURN
SLUDGE
-------
OBJECTIVE
DEVISE TECHNIQUES FOR EFFICIENTLY REMOVING
NITROGEN AND PHOSPHORUS FROM WASTEWATER
COMPATIBLE WITH ACTIVATED SLUDGE TREATMENT.
APPROACH
1. UTILIZE THREE-SEPARATE SLUDGE SYSTEMS TO
ISOLATE SPECIFIC BIOLOGICAL ACTIVITIES,
CARBON OXIDATION AND SYNTHESIS,
NITRIFICATION, AND DENITRIFICATION.
2. REMOVE BULK .OF CARBON AHEAD OF NITROGEN
CONTROL IN-.A HIGH-SYNTHESIS, SHQRT
DURATION MICROBIAL UNIT.
3. CONTROL NITRIFICATION AND DENITRIFICATION
WITH ENRICHED CULTURES OF MICROORGANISMS.
4. SUPERIMPOSE CHEMICAL PRECIPITATION OF
PHOSPHORUS ON THE THREE-STAGE NITROGEN
CONTROL SYSTEM.
9-10
-------
PROCESS FLOW DIAGRAM
MAJOR
PROCESS
FUNCTIONS
CHEMICAL
ADDITIVES
REMOVE | REMOVE BULK OF
SETTLEABLE SOLIDS) SOLUBLE ORGANICS
AND PARTICULATE
COD
NONE
AND PRECIPITATE
PHOSPHORUS
SODIUM
ALUMINATE
CONVERT
NH3 TO NO3
NONE
CONVERT NO3 TO
NITROGEN GAS AND CO2
AND FURTHER REDUCE
PHOSPHORUS
SODIUM ALUMINATE
AND
METHYL ALCOHOL
RETURN
Unit
Primary
High-Rate
Nitrification
Denitrification
-------
CHANGE IN COMPOSITION OF PROCESS STREAM DURING TREATMENT
PROCESS
STREAM
RAW
WASTE WATER
PRIMARY
EFFLUENT
HIGH-RATE
EFFLUENT
NITRIFIED
EFFLUENT
FINAL
(DENITRIFIED)
EFFLUENT
PERCENT (%)
REMOVAL
mg/1
IB TOTAL
s. s.. 1 P
320
218
64
43
44
86
^^H^AM^BMBBBB^Hia^v
157
90
9
7
7
95
12.6
11.9
2.8
2.6
1.5
88
ORG« NH- N00 NO, TOTAL
N FT N^ NJ N
10.3 + 11.3 4- / / — 21-6
5.9 4- 13.7 4- / XX *" 19>B
0.8 -f. 7.7 4- 1.1 4. 4.3 ea 13.9
0.4 + 0.8 4- 0.3 4- 11.5 « 12.8
0.4 4- 0.3 4- 0-3 + 0.9 «- 1.9
91
vo
ro
-------
MASS ACCOUNTING
OF PHOSPHORUS
A+B+C+D=
TOTAL INPUT
1.
2.
LEGEND
PHOSPHORUS PATHWAYS
THROUGH PLANT
LIQUID REGIME
SLUDGES
-------
T
A
L
U
T
N
I
T
5
?
MASS ACCOUNTING
OF NITROGEN
i
•*•
% P
T P
• S
" 5
U 5
O
1OO
80
^^^^m
6O
_
4O
20
0
f ft
i*..~tfb!
W
A
S
T
E
R
A
W
~7
X
L
U
D
G
E
W
A
S
T
E
W
A
T
E
R
^
A+B+C+D+E=
TOTAL INPUT
NITROGEN PATHWAYS
THROUGH PLANT
1. LIQUID REGIME
2. - SLUDGES
3. DENITRIFICATION
-------
MIXED LIQUOR CHARACTERISTICS AFTER ADDITION OF ALUMINUM
High-Rate
Mixed Liquor
Nitrification
Mixed Liquor
Denitrificatior
Mixed Liquor
Sludge Age
days
(Approx. )
2
22
38
MLSS
mg/1
3,385
1,195
2,010
MLVSS
mg/1
1,740
770
1,385
%
Volatile
51
64
69
SDI
1. 12
1. 15
1.46
7c
Nitrogen
V. S. Basis
8.3
10.6
9.7
%
Phosphorus
V. S. Basis
12.3
8.0
8.7
°/c
Aluminum
V. S. Basis
11.0
6.6
6.0
1C
in
-------
EFFECT OF ALUMINUM ON SLUDGES AND SUPERNATANT
1. HIGH-RATE MKED LIQUOR CHARACTERISTICS
BEFORE
A1W
AFTER
Al*"1"1"
MLSS
mg/1
1,450
3,385
MLVSS
mg/1
900
1,740
%
VOLATILE
62
51
SDI
0.44
1.12
ORGANIC LOAD
Ib COD/ day/
. Ib MLVSS
2.6
1.8
2. PERCENT PHOSPHORUS IN SLUDGES - TOTAL SOLIDS BASIS
BEFORE
Ai*"
AFTER
A1W
PRIMARY
1.1
1.1
WASTE
ACTIVATED
2.1
6.2
MIXED
PRIMARY
PLUS
W. ACTIVATED
1.6
4.5
DIGESTED
2.2
5.2
3. TOTAL PHOSPHORUS IN SUPERNATANT - mg/1
BEFORE
AFTER
Ai1
170
60
9-16
-------
CHEMICAL REQUIREMENTS AND COSTS
Phosphorus Concentration
of Primary Effluent
rhg/1 as P
Al to P ratio
Chemical Cost of
Phosphorus Removal
c/1000 gal
Nitrate Concentration of
Nitrified Effluent
mg/1 as N
MeOH to NO -N ratio
Chemical Cost of
Nitrogen Removal
c/1000 gal
Total Chemical Cost
c/ 1000 gal^
Pilot Plant
Data
12.2
1.5
6.4
13.0
5.8
2.3
8.7
Design
Guideline
10. 0
1.2
4.3
20.0
4.0
2.7
7.0
J-17
-------
CHEMICAL APPROACHES TO
NITROGEN REMOVAL
PH ADJUSTMENT OF EFFLUENT AND STRIPPING OF AMMONIA
FORMATION OF MgNH4 P04
o) EFFLUENT
b) DIGESTER SUPERNATANT
ADVANCED METHODS- ION EXCHANGE, ECT.
9-18
-------
REF£KENCES
1. Johnson, W. K. und Schroepfer, G L., "Nitrogen Removal by
Nitrification ..nd Denitrification," J. Water Pollution
Control Federation, 36, 1015 (196U).
2. Wuhrmann, K., "Nitrogen Removal in Sewage Treatment Processes,"
Verh. Internat. Verein. Lljnnol., 1£, 580 (1961*).
3- Earth, E. F. and Ettinger, M. B., 'Managing Continuous Flow
Biological Denitrification," Proceedings of the Seventh
Industrial Water and Waste Conference, University of Texas,
(June 1967).
IK Haltrich, W., "Elimination of Nitrate from an Industrial Waste,"
Proc. 22nd Ind. Waste Conf., Purdue Univ., May 1967.
5- Thomas, E. A., "Phosphat-Elimination in der Belebtechlemannlage
von Mannedorf und Phosphat-Fixation in Secund KLarschlamm."
Vierteljahrsschrift der Naturforschenden Gesellschaft in
Zurich, Jahrgang (Switzerland), 110, M9 (1965).
6. Tenney, M. W. and Stumm, W., "Chemical Flocculation of
Microorganisms in Biological Waste Treatment, " J. Water
Pollution Control Federation, 37, 1370 (1965).
7. Earth, E- F. and Ettinger, M. B , "Mineral Controlled Phosphorus
Removal in the Activated Sludge Process," J. Water Pollution
Control Federation, 39, 1361 (1967).
8. Eberhardt, W. A. and Nesbitt, J. B., "Chemical Precipitation
of Phosphate Within a High Rate Bio-oxidation System,"
Proc. 22nd Ind. Waste Conf., Purdue Univ., May 1967.
9-19
-------
9. Downing, A. L., Painter, H. A. and Khowles, G., "Nitrification In the
Activated Sludge Process," J. Institute Sewage Purification,
Part 2, 130 (196U).
10. Earth, E. F., Mulbarger, M., Salotto, B. V. and Ettlnger, M. B.,
"Removal of Nitrogen by Municipal Waatewater Treatment Plants,"
J. Water Pollution Control Federation, 38, 1208 (1966).
11. "Royal Commission on Sewage Disposal." Fifth Report, Appendix V,
Section k (1908). (London, England).
12. Ludzack, F. J. and Ettinger, M. B., "Controlling Operation to
Minimize Activated Sludge Effluent Nitrogen," J. Water Pollution
Control Federation, 3_U, 920 (l°£2).
13. Christiansen, C. W., Rex, E. H., Webster, W.. M. and Vigil, F. A.,
"Reduction of Nitrate Nitrogen by Modified Activated Sludge,"
U. S. Atomic Energy Commission, TID-7517 (Pt. la), 26U (1956).
14. Placak, 0. R. and Ruchhoft, C. C., "Studies of Sevage Purification.
Part XVII, The Utilization of Organic Substrates by Activated
Sludge." Public Health Reports, 62, No. 20, 697 (19^7).
15* McKinney, R. E. and Conway, R. A., "Chemical Oxygen in Biological
Waste Treatment," Sevage and Industrial Wastes, 2£, 1097 (1957).
16. "Standard Methods for the Examination of Water and Wastevater."
12th Ed. Amer. Pub. Health Assn.,. Nev York (1965.).
17* Weatherburn, M. W., "Phenol-Hypochlorite Reaction for Determination
of Ammonia," Analytical Chem., 39, 971 (1967).
18. Kamphake, L, J., Hannah, S. A. and Cohen,, J. M., "Automated
Analysis for Nitrate by Hydrazine Reduction," Water Research,
1, 205 (1967).
9-20
-------
19- Lee, G. F-, Clesceri, N. L. and Fitzgerald, G. P., "Studies
on the Analysis of Phosphates in Algae Cultures," Air and
Water Pollution International Jour., 9, 715 (1965).
20. Packham, R. E., 'The Absorptiometric Determination of Aluminum
in Water," Proc. Soc. Water Treatment and Examination,
7, 102 (1958).
21. Instrument Manual. Per kin-Elmer Model 303, Revised Nov. 1966.
22. Antczak, K. and Piotrowskl, "Methanol Determination in
Expired Air and in Urine in Acute Poisonings," Medical
Pharmacy, lU, 321 (1963). (Chem. Abs.t 62, 3307(6) (1965)0
23. Oglnsky, E. L. and Unbreit, W. W., An Introduction to Bacterial
Physiology, 2nd Ed. W. H. Freeman and Co., San Francisco (1959).
2U. Coombs, J. A., "Activated Sludge Process of Sewage and Other
Impure Liquid Purification." U. S. Patent l,90U,9ljS (1933).
25' Slraner, F-, "Process for Biological Purification of Wastevater. "
U. S. Patent 1,751.^59 (1930).
9-21
-------
PHYSICAL-CHEMICAL JBTHODS FOR NITROGElf REMOVAL
by
Robert B. Dean
Chief, Ultimate Disposal Research Program
Advanced Waste Treatment Research Laboratory
Third Technical Session
9:^5 A.M.
Thursday, February 6, 1969
Symposium on Nutrient Removal and Advanced Waste Treatment
Portland, Oregon
-------
I. Nitrogen forms in secondary effluents. Neale '6U, Barth '69.
A. Ammonia-ammonium ions NH_ + H . > NH^
1. Av. 14, Range 0 to 36 mg/1 as N.
2. Derived from urea and other organic nitrogen in raw sewage.
3. Suggested limit, 1 mg/1 as N.
B. Nitrite ions, NO,,'.
1. Av. 0.4, Range 0 to 10 mg/1 as N.
2. Produced from ammonia and C0_ by Nitrosomonas bacteria when organic
food is low*
3. Suggested limit, 1 mg/1 as N.
C. Nitrate ions, NO^".
1. Av. 2.7, Range 0 to 25 mg/1 as N.
2. Produced from nitrite and C02 by Nitrobacter.
3. PHS limit 10 mg/1 as N (45 mg/1 as NO^").
D. Ammonia-Nitrate are an eithcr-or situation. Hare to get substantial
quantities of both. Ammonia in high rate systems, nitrate with
extended aeration.
10-1
-------
H. Ammonia removal
A. Physical. Evaporation at high pH..
1. NHj/ + OH" E2J
as NH3 at pH 10.75, 99S& at pH 11.75-
2. Solubility. Very great compared with all ordinary gases.,
a.
COMPARISON OF SOLUBILITY OF
CARBON DIOXIDE AND1 NHi IN WATER
TEMP.
HENRY'S LAW CONSTANT (H)
20
40
60
WHERE pa = Hx£
co2
l',420
2,330
3,410
NH3
0.7
1.8
3.8
Pa - VAPOR PRESSURE OF
DISSOLVED SUBSTANCE
"a" (atm)
x = MOLE FRACTION OF "a"
H - HENRY'S LAW CONSTANT
b. 14 mg/1 = .001 mole/55 moles water or .000018 mole fraction,
Therefore P = 0.7 x .000018 Atm or .01 mm of Hg-
10-2
-------
Air requirements ccaqparecl to CO
a- MINIMUM AIR REQUIREMENT FOR
STRIPPING NHi AND COp FROM WATER AT 20°C
IN A COUNTERCURRENT OPERATION
Stripping NH3 out
of basic solution*
NHi in water NH-^ in air
20 mg/1 A8.7xlO-6 g/g
2 mg/1
0
Stripping CC>2 out
of acidic solution*
CO? in water
20 mg/1
2 mg/1
V
CO? in air
A17.6xlO-3
0
Air Required: 225 cu ft/gal
O.LI cu ft/gal
*Entering water stream and exiting air stream are assumed to be
in equilibrium
b. Actual performance requires even more uir—300 cu.ft./^al,
Tor 90$ removal determined at Tahoe. fJlechta '67.
c. Temperature throughout most of the tower is vr:t bulb.
This can be below zero decrees C in the winter and the
tower will freeze. Henry's law constant is lower at low
temperatures.
d. Packed towers have prohibitive pressure rlrops.
38 inchcjs of water in Intalox packing.
Fraction of an inch in cooling tower.
10-3
-------
B. Ion exchange as NH. + at pH's belov 8
1. Adsorption., R" is a cation exchange material.
a» R"Na+ + NH"1" > R-MH"1" + Na+
b. R2-Ca + 2
2. Regeneration
a. Reverse of 1 using excess salt or
b. R'NH^* + Na+OH" _ ^ R"Na+ + NHg + HgO
2R"NH^+ + Ca(OH) - > R2~Ca^ + 2NH3 * ^2°
Displacement of NH- with base.
3. Specificity
a. Ion exchange resins favor Ca** over MH^"1".
Hardness of vater \ises up available capacity.
bo Uome zeolites favor NH^+ or Ca+lf.
Clinoptilolite. Battelle-Northwest. Ames '68. Figures 1 and 2.
k. Adsorption. Conventional ion exchange equipment.
Packed towers up or downflow.
5. Regeneration
a. Conventional resins with brine.
b. Specific zeolites with base. Lime can precipitate
calcium carbonate in zeolite*
6. Disposal
a. To ocean. .Brines.
b. To air by lime treatment and blowing.
10-4
-------
1.0
~ U.8 -
S
rsi
c
o
_o
"o
rrs
O.b -
0.4
0.2
0.10
NH. Concentration, eq/l
FIGURE 1. The 23°C ion Exchange Isotherms for the
Reaction of NHj with Na+, K*. Mg+2 and
Ca42 for Hector Clinoptilolite and the Reaction
NHJ with Ca+2 for a Polystyrene Sulfonate
ion Exchange Resin. Total Solution Normality
was Constant at 0.1 for all Systems.
o
o
o
0.8 -
0.4 -
100 200 300
Column Volumes
FIGURE 2. Effect of Flow Rate on Uptake of
400
500
Column
Feed
C/Co
35 cc of 25 x 50 Mesh Clinoptilolite
Simulated Clarified Secondary Effluent
Effluent NHd Concentration
InfluentNHfl Concentration
10-5
-------
C. Chemical destruction
1. Chlorine
a. 3C12 + 2NH3 _ > Ng •»• 6HC1
b. NH3 6l2 > NH2C1 C12 > KHClg C312> KC13
c. Requires from 6.25 up to 10 nag dp per mg ammonia N.
2. Ozone
OH-
Not efficient at low concentrations.
Converts NHo limit 1 mg/1 into NO-" limit 10 mg/1.
III. Nitrate removal
A. Ion exchange. R NO ~
1. Conventional. Competition from Cl", HCO-", and especially
and
2. Specific' resins* Research being carried out.
a. Most resins show some specificity for NO ~ but lov selectivity
versus SO. ~. ^
B. Chemical reduction. Rocketdyne
5Fe Fe2
2. . 2NO
a. Reactions 1 and 2 can go in about equal proportions.
b. FeSO, *7H 0 "copperas" is cheap but there is a lot of precipitate
of black^iron hydroxides and sulfates.
c. Utility limited to availability of , ferrous salts at low cost
and convenient dump for precipitates.
d; Useful1 in 'situations where neither denitrif ication nor ammonia
stripping can be used. Toxic chemical wastes and low tempera-
ture part '-time operations.
10-6
-------
REMOVAL OF IONS BY
ION-EXCHANGE RESINS
Cation-Exchange Resin
Na+
Ma*
Na+
+ 6NH4+
Anion-Exchange Resin
cr
CI
Cl
cr
2Na+
4NO
2CI
10-7
-------
IV. Cost of reducing nitrogen from 20 to 2 mg/1. j£/1000 gal. Slechta.
10 MBD basis.
A. As aDononia
1. Stripping at high pH.
a. 2.3j£ at pH 11 adjusted for cold weather operation.
b. ^»9p including cost of lime to raise pH.
Line cost can be shared with phosphate removal*
c. Cost is not sensitive to fluctuation in actual'nitrogen
quantities.
2. Chlorine
a. 5-3^
b. Cost directly related to actual NH-, quantity.
3* Ion exchange
a. Non-specific resin 8.1/1. Eliassen.
b. Zeolite. Unknown.
c. Costs strongly dependent on actual quantity treated.
B. As nitrate. No.cpsts assigned to excess aeration required to achieve
nitrification.
1. Ion exchange
a. Non-specific resin l6.0j£.
2. Chemical reduction by ferrous sulfate
a. 8,6ff not including sludge disposal. Unpublished.
3. Biological reduction
a. 2.7j£* Barth.
10-8
-------
LIST OF PERTINENT REFERENCES
1. Slechta, A. F., and Gulp, G. L., J. Water Poll. Control Fed.,
_39.(5), 787-814 (1967), "Water Reclamation Studies at the
South Tahoe Public Utility District."
2. Eliassen, R., and Bennett, G. E., J. Water Poll. Control Fed..
.39(10), (Part 2), R81-R-90 (1967), "Anion Exchange and
Filtration Techniques for Wastewater Renovation."
3. Perry, R. H., Chilton, C. H., Kirkpatrick, S. D., "Chemical
Engineers' Handbook," 4th Ed., McGraw Hill, N. Y. (1963).
4. Ames, L. L., Jr., "Zeolitic Removal of Ammonium Ions from
Agricultural and Other Wastewaters," in Proc. 13th Pacific
Northwest Indust. Waste Conference, April 6-7, 1967, pub.
by Tech. Extn. Services, Washington State University,
Pullman, Washington.
5. Farrell, J. B., Stern, G., and Dean, R. B., "Removal of Nitrogen
from Wastewaters," Internal FWPCA Report, Cincinnati Water Research
Laboratory, 4676 Columbia Parkway, Cincinnati, Ohio, May 1968.
6. Neale, J. H., "Advanced Waste Treatment by Distillation,"
PHS Pub. No. 999-WP-9, AHTR-7, March '64.
7. Earth, E. J.- Unpublished.
10-9
-------
SOLIDS REMOVAL PROCESSES
by
Jesse M. Cohen, Chief
Physical and Chemical Treatment Research Program
Advanced Waste Treatment Research Laboratory
Cincinnati, Ohio
Technical Seminar
Nutrient Removal and Advanced
Waste Treatment
Portland,Oregon
February 5 and 6, 1969
-------
Tampa, FJoi ji
November 13-1A,
Solids Removal
J. J. Convery
I. Screening Devices*
A. Functional Description
Microscrccniug is a form of simple filtration. These mechanical
filters consist of a rotary drum which revolves on a horizontal
axis. The peripheral surface of the drum is covered with a
stainless steel filter fabric. The effectiveness of the woven
nifsh screen for retaining fine pair tides is dependent on the
size of openings in the screen and on the pattern of the weave
employ I'd. Influent enters the open end or the drum and is
filtered through the fabric with the intercepted solids being
retained on the inside surface of tlic fabric. As the drum
rotates, the solids are transported and continuously removed at
the top of the drum by pumping strained effluent, under pressure,
through a scries of spray nozzles which extend the length of the
drum. The solids and wash water are collected in a central
trough within the drum and discharged through a hollow axle.
Some mechanism such as intermittent chlorlnation (every 3 to 8
days), or continuous ultraviolet lighting must be. employed to
prevent the growth of biological slimes on the fabric.
B. Physical Description
A typical micros training unit is shown in Figure 1.
^'Comments and operating experience refer to the patented Micro-Strainer,
a product of Glenfield and Kennedy, Ltd., London, England. Mention of
a proprietary name is not an endorsement by IWCA.
11-1
-------
'iii. I Tvnieal Micr"~:trainer Unit
AOt Sho\vn
Wasiewater Hopper
L'ltra-violet Lam:
\\ ash -u n•;«->!• Pump
Drive Unit
Rotating Driur1
W^sh-wau "• Jets
M.cro-fabric
Influent Chamber
j_;;iuent Chamber
Eifluen' V\ eir
-------
o
1. UniL Sizes and Capacity
Drum Sizes Approx.
in Feet Capacity
Diameter Width
5 1 0.05 - 0.5
5 3 0.3 - 1.5
7% 5 0.8-4
10 10 3 - 10
2. Fabric Details
Sec Table 1
C. Design Basis
The basic design criteria for the proper sizing of a micro-
screening unit is the filterability Index, I. A device called
the filtrameter is available from Glenfield and Kennedy, Inc.
for the field measurement of the filterability index I. This
index is an expression of the volume of water obtained per unit
headless when passed at a standard rate through a unit area of
standard filler fabric. The Index is used to calculate the re-
quired area to treat a given flow of a specific wastewater ac-
cording to the following expression:
H - mQC{
e
nlQ/S
where:
H «= headless, in.
Q *= rate of flow, gpm
G£ - Initial fabric resistance, sec.
3
A = effective submerged filter area, ft
I - filterability index, ft"1
S = speed of rotation, area per unit time
entering the water, ft^/mln.
11-3
-------
The unit conversion factors m and n have values of 0.0267 and
0.1337, respectively. Cf is 1.8 sec. for the Mark 0 fabric and
1.0 sec. for the Mark I fabric.
TABLE 1
CHARACTER IS TICS -iOF THREE FILTER FABRICS
Apertures:
No. per sq. in.:
No. per lineal in.
Warp:
Weft:
Mark 0
1/1000
23 "microns
160,000
400
120 mesh
2 wires .0018
600 mesh
1 wire .0016
Mark I
1.5/1000
35 microns
80,000
280
76 mesh
2 wires .0024
500 mesh
1 wire .0022
Mark II
2.5/1000
60 microns
60,000
240
60 mesh-
2 wires .0040
420 mesh
1 wire .0028
D. Operating Parameters
1. Head loss
Microscreening devices should be installed at an elevation per-
mitting gravity flow thereby eliminating the possible fractiona-
tion of flocculant particles. Experience has demonstrated that
the removal efficiency of particulates is reduced when flocculant
material is pumped. The maximum operating head loss across the
screen is normally limited to six (6) Inches to prevent rupture
of the delicate filter fabric.
2. Drum Speed
Some flexibility of operation can be achieved by varying the speed
of rotation, S. Peripheral velocities normally range from 25 to
135 ft/min.
3. Wash Water Requirements
^ function of backwash-ing is- ;tO;':thoroughly clean the filter
fabric during each revolution .with ;a minimum" 'o'f wash water.
Pressure appears to be more critical than the quantity of wash
water to effect the cleaning. Wash water pressures range from
20 to 55 psi and water requirements are less than 57. of the
11-4
-------
throughput. The wash water is normally recirculated to the
head end of the plant. The suspended solids concentration
of the wash water approach 750 mg/1.
E. Performance in a Tertiary Treatment Application
o
1. Removal Efficiency and Flow Rate
The effect of aperture size on removal efficiencies and flow
rate are shown in the following table:
Removal Efficiency Flow
Fabric Solids BOD Gals . /Ilr
Mark 0 (23 microns) 70-807. 60-707. 400
Mark I (35 microns) 50-607. 40-507. 600
2. Operating Problems
The Federal Water Pollution Control Administration has conducted
a study J of microscreening at Lebanon, Ohio. The secondary
effluent from the municipal activated sludge treatment plant
had an average suspended solids concentration of 17 mg/1 while
the Mark 0 fabric was tested and 27 mg/1 during tests of the
Mark I fabric. The filterability index, coincidentally, was
17.0 during both tests.
One of the major problems associated with operation of the
microscreen was the unit's inability to handle wide fluctuations
in solids loadings. On one occasion the solids in the feed in-
creased 8 fold from 25 to 200 mg/1. Within minutes the through-
put of the unit was reduced from 60 to 13 gpm (21.77. of rated
capacity). Figure 2 shows a frequency distribution of suspended
solids values in the Lebanon secondary effluent.
Suspended solids concentrations in excess of 50 mg/1 occurred
107. of the time. If the microscreening unit can produce 257. of
its rated capacity during periods of high solids loading, (> 50
mg/1) 757. of the design flow must come from storage 10% of the
time. Therefore, the plant must have an excess design capacity
of 8.33%. A recent design improvement has been the coordination
of drum speed and wash water pressure with changes in headless
across .the fabric.
11-5
-------
1000
500
200
100
O>
E
§
8
S
Q
50
20
10
I T
I I I I I I I
I I
I
RESULTS OF 384
OBSERVATIONS
I I I I I I
I I I
I I
I I I
12 5 10 20 30 40 50 60 70 80 90 95 9 99.5 99.8
PERCENT EQUAL TO OR LESS THAN
Figured. Frequency distribution of suspended solids in secondary effluent.
11-6
-------
3. Limitations
Fine fabric microscrecning is not recommended for the treatment
of waters with a chloride content greater than 500 mg/1 because
of potential corrosion problems.
F. Capital and Operating Costs
The capital and operating costs associated with micros training
of secondary effluent have been prepared by Smith ^ and are
shown in Figure 3.
II. Coarse Media Upflow Filtration9
The work on upflow filtration, which I have summarized, was a 12-
month study performed by Truesdale and Birkbeck aL Letchworth,
England in 1967.
A. Description
The upflo\7 coarse media filter is somewhat similar in appearance
to the conventional rapid sand filter except that the direction
of flow is upward. Backwashing in an upward direction grades
the filter media with the coarser media on the bottom and finer
media on the top. During filtration the effluent contacts pro-
gressively finer media and the intercepted solids are distributed
throughout the filter bed rather than being concentrated near the
surface. This uniform distribution of headless and successively
finer filtration are the reasons for the longer filter runs and
improved product quality normally associated with in-dcpLh filtra-
tion.
Upflow coarse media filters consist of 5 feet of 1-2 mm sand
supported on progressively coarser layers of gravel and a dis-
tribution system. Total height is approximately 12 feet. A
metal grid, consisting of parallel bars spaced 4 in. apart
placed 2 in. below the surface of the bed, is used Lo prevent
expansion of the filter media at flow rates as high as 6.7
gpm/ft . The backwash rate of 11.7 gpm/ft2 is sufficient to
expand the bed to permjt cleaning. A wash cycle of 15 minutes
was found adequate to clean the filter.
B. Performance
The sol'ids concentration of the activated sludge effluent varied
from 10 to 34 nig/1. The effluent from the upflow coarse media
11-7
-------
Figure 3
MICROSTRAINING OF SECONDARY EFFLUENT
Capital Cost, Operating & Maintenance Cost, Debt Service
vs.
Design Capacity
10.0
GO
HP
n
ti
0)
0
.1.0
1
o
*
0.10
: Cost Adjusted to June,
rrnTrTrrrr! i"
1.0
CQ
0.10
03
c
o
CO
•p
•H
P«
CO
u
0.01
« 7 8 9 10
1.0 10.0
Design Capacity, millions of gallons per day
100.
C = Capital Cost, millions of dollars
A = Debt Service, cents per 1000 gallons(U 1/2% - 25 yr.)
0 & M = Operating and Maintenance Cost, cents per 1000 gallons
T = Total Treatment Cost, cents per 1000 gallons
11-8
-------
filter contained from 3.2 to 10 mg/1 of suspended solids. The
coarse media filter removed on the average 60% of the suspended
solids and 537. of the BOD. The removal efficiency of a filter
is a function of the hydraulic loading as shown in the following
table:
Filtration Rate Solids Removal Efficiency
gpm/ft2 %
2.9 77
A.I 58
5.9 38
The longest length of run was 19.2 hours at a -filtration rate of
3.3 gpm/fL2. Wash water consumption ranged from 4.6 to 6.7 percent
of the filter throughput. The authors did not mention what term-
inal headless was used to initiate the automatic backwash mechanism
but they do indicate it was set too low. Therefore, extended filter
runs and lower wash water consumption may be obtainable.
A micros training unit with a Mark 0 fabric was operated in parallel
with the upflow filter during this study. Results indicated that
the filter effluent was superior to the strainer effluent at a
filtration rate of 3.33 gpm/ft2 or less. Filter performance was
inferior to strainer performance at a filtration rate of 6.67
gpm/ft2 or more. Performance was comparable at the intermediate
filtration rates of 3.33 to 6.67 gpm/ft2. The microstrainer was
operated at A.3 gpm/ft2.
One advantage of coarse media filtration is the availability of
some floe storage capacity within the interstices of the filter
media to permit partial phosphorus removal by chemical pretreatment.
The practice of chemical pretreatment prior to microstraining has
generally not been successful. The fragile floe breaks up on the
screen.
The capital and operating costs for filtration through sand or
graded media at 4 gpm/ft^ has been estimated by Smith*0 and are
shown in Figure 4.
11-9
-------
Figure 4
FILIATION THROUGH SAND OR GRADED MEDIA - UGFM/SQ FT1
Capital Cost, Operating & Maintenance Cost, Debt Service
vs.
Design Capacity
10.
I
gl.O
o
0.1
U 1 Li . J L . LI 4.;:; .<_. I _•
nth ' i ' I i HI 71.T i 1 . .
Cost Adjusted to June, 196? "
1.0
u
IS
0.1
co
o
•H
H
H
•a
o
co
s
I
4 5 678910
.01
1.0
10.0
Design Capacity, millions of gallons per day
100.
C = Capital Cost, millions of dollars
A = Debt Service, cents per 1000 gallons(U 1/2% - 25 yr.)
0 & M = Operating and Maintenance Cost, cents per 1000 gallons
T = Total Treatment Cost, cents per 1000 gallons
11-10
-------
III. Moving Bed Filter Technique
A. Description
The Moving Bed Filter (MBF) is a new filtering technique that
is currently being developed by Johns-Manville Products
Corporation.* Evaluation of a pilot unit is currently being
supported under contract by the FWPCA. A schematic diagram
of the Moving Bed Filter is shown in Figure 5.
This unit is basically a sand filter. Particulate matter in
the water is removed as the water passes through the sand
(.6 - .8 mm). As the filter surface becomes clogged, the
filter media is moved forward by means of a mechanical dia-
phragm. The clogged filter surface is guillotined off thereby
exposing a clean filter surface. The sand and'accumulated
sludge is washed and the sand is returned to the base of the
filter. The unit is thus a form of countercurrent extraction
device feeding sand countercurrent to the water being filtered.
The Moving Bed Filter has a renewable filter surface analogous
to the microstrainer and the advantage of depth filtration
comparable to the coarse media filter. The unit does not have
to be taken off suream for backwashing. In theory, 1% of the
filter is being backwashed 100% of the Lime compared to the con-
ventional practice of backwashing 10078 of the filter 1% of the
time.
B. Performance
Several pilot MBF units have been built to date and used to treat
settled and unsettled trickling filter effluent and primary efflu-
ent. The system lends itself well to the use of chemical aids
ahead of filtration because of designed flexibility to handle
high solids loadings.
*Mention of a proprietary device does not indicate endorsement by
the FWPCA.
11-11
-------
BASIC CONCEPT OF MOVING BED FILTER
INFLUENT1
CHEMICALS
(OPTIONAL)
FEED
HOPPER
DIAPHRAGM
HYDRAULIC
SYSTEM
,.SAND
RECYCLE
WASH.
WATER
EDUCTOR
SAND
WASHING
UDGE
WASTE
SLUDGE
VJT.
-------
Preliminary results obtained while treating unsettled trickling
filter effluent are as follows:
ALUM
DOSE
mR/1
200
200
200
100
COAG.
AID
mg/1
0
0.5
1.0
0.5
TUMIDITY
JTU
Inf.
41
40
22
28
Ff f
fcl1' Red.
10.0 76
6.3 85
9.2 57
8.4 64
BOD
me/1
Inf
40
49
46
64
. Eff.
8.8
9.7
10.0
9.7
%
Red.
78
80
79
79
COD
mfi/1
Inf.
120
143
111
172
Eff.
40
39
42
43
7.
Red.
66
73
62
65
TOTAL PHOSPHATE
mg/1
Inf. Eff.
36 1.5
36 2.5
40 1.8
30 1.7
*/.
Red.
9G
93
96
94
These very preliminary results look encouraging. Of particular
interest is the excellent phosphate removal obtained with 100 ing/l
of alum. This represents an AL/P ratio of about 1. We are follow-
ing this relationship very closely.
It is too premature to even suggest a treatment cost for the Moving
Bed Filter. Design and performance information developed from the
Johns-Manville-FWPCA contract will be analysed to project capital
and OSM costs for comparison with alternate treatment processes.
11-1 3
-------
REFERENCES
1. "Mici-ostraining," Bulletin of Glenfield and Kennedy, Inc., Box 191,
King of Prussia, Pa., No. 41, 6th Ed. (1961).
2. Evans, G. R. , "Microstraining Tests on Trickling Filter Effluents
in the Clear Creek Watershed Area, Texas," Public Works. (October,
1965).
3. Bodien, D. G. and Stenburg, R. L. , "Microscreening of Effluent from
a Municipal Activated Sludge Treatment Plant," Water and Wastes
Engineering. (September, 1966).
4. Evans, S. C., "Ten Years Operation and Development at Luton, Sewage
Treatment Works," Water and Sewage Works Journal. Vol. 104,, n5,
pp. 214-219, .(May, 1957).
5. "Microstraining of Sewage Effluent," Report by the Ontario Water
Resources Commission (1960).
6. Truesdale, G. A. , Birkbeck, A. E. and Shaw, D. , "A Critical Examina-
tion of Some Methods of Further Treatment of Effluents from Percolat-
ing Filters," Conference Paper No. 4, Water Pollution Research
Laboratory, Stevcnage, England (July, 1963).
7. "Micros training, of Sewage Effluent," Report by the Ontario-" Itoter
Resources Commission, (1960).
8. Diaper, E. W., "Microstraining and Ozonation. of Sewage Effluents,"
Presented at-the 41st Annual Conference of the WPCF, Chicago-,
Illinois (September, 1968).
9. Truesdale, G. A. arid Birkbeck, A. E., "Tertiary Treatment of
Activated Sludge Effluent," Reprint No. 520, Water Pollution
Research Laboratory, Stevenage, Herts, England (1967).
10. Smith, R., "Cost of Conventional and Advanced Treatment of Waste-
waters," Journal Wafer Pollution Control Fcdsration. 40, 9, 1546-
1574 (September, 1968).
11-14
-------
ORGANIC RESIDUE REMOVAL
Jesse M. Cohen, Chief
Physical and Chemical Treatment Research Program
Advanced Waste Treatment Research Laboratory
Cincinnati, Ohio
Technical Seminar
Nutrient Removal and Advanced
Waste Treatment
Portland, Oregon
February 5 and 6, 1969
-------
USE OK GKANULAR ACTIVATED CAHI3ON
1 IN'i HODUCT iOi\
A Firjurc 1 ,sh->ws the predicted \\ati t use and
watc-i supply picture: through I he year 2000
This irf a national piclure -- in some areas,
demand aln-:idy exceeds supply and popula-
tion, indusli i.il nnd cc onomic growth are
limited. Tlu i gi apli 'loos not allow for
reuse so the situation is not as dark ns
indicated.
B This Nation's usable walrr supply must be
increased -- potential solutions are de-
salting of sua water, weuthei control, build-
ing of reserve irs, and wastewater renovation.
All appro.ichcs should be investigated
thoroughly because ol immense but den that
water shortages can place on well-being
of citizens.
C Before wastewatci can he reused lur potable
purposes, we must remove objectionable
contaminants added hy use. fablr 1 lists
the organic materials added to a municipal
water Supply by one usu and not removed
by conventional (activated sludge) treatment.
U These refractory oigamr materials must
be removed In i grcatci degree than is
done now whether the water is directly re-
used or rctuiiii-d to a stream. Kigure 2
shows the projected input of organic mate-
rials as measured by chemical oxygen
demand (COD) thi ough the year 2000 Most
of these organics are dissolved rather than
suspended solids Since these COD mate-
rials are, for the most part, rfsislnnt to
biological ti i ntmunt, tlicir concentration
will not be aficcted by widespread applica-
tion of the activated sludge process (solid
lino). An efficient organic removal process
is needed to reduce the input to that shown
by the clotted line. These materials, if put
into a stream, promote algal growth, con-
tribulr- to fish kills and tainting of lish flesh
and to taste and odor ol w.itei supplies with-
drawn from tin- river. In addition, there is
a possibility that they wiJl have cumulative
harmful physiological clfcMs
E Processes thrit have been and arc being
considered for removal of these oiganic
materials include distillation, freezing,
reverse osmosis and adsorption on activated
carbon, both granular and powdered.
!•' Ad-.oiption on granular cnibon is the
f.iitlicist advanced ot these processes
Ali i-ady in commrieial use foi waste-
water treatment, the technology is ri-.idy
lor use on a large scale This outline
considers the "State oi tin- Art" of c;u boil
use.
II TU1-: MANUKACTUmNC; Ol ACTIVATED
CAKHON
A Soui ce materials include < oal, wood,
sawdust, pc-at, lignite ami pulp mill char
B In the lirst step ol m.iiuil 'cturinp, the raw
material is carbonized in lln abscm e of
air, usually below 600°c This step re-
moves the- hulk of the volatile materials
and yields a product with slightly enhanced
ad.sorptive i apacity.
1 Carbonization is sometimes made more
efficient by the addition of metallu
chlorides (zinc < hlorul«- is common) to
the carbonaceous matci lal before
heating. The heating is continued until
zinc chloride vapors come off, after
which the char is (.ooled and washed to
recover the /.me salt lor reuse
C Activation is controlled oxidation at
elevated temperatures Activation ga»es
arc usually steam or carbon dioxide nnd
the temperature is held bi.tween 800 jnri
1000°C. The tinn- of activation vat ic.s
between 30 minutesa and 24 hours, depending
on the oxidizing conditions and on the quality
of active carbon dcsn nd
The ma]or manuf.ictut ITS of activated
carbon are listed below, in alphabetical
order:
1 Granular
Inc.
Inc
American Norit Company,
Atlas Chemical Industr K-S,
National Carbon Company
Pittsburgh Activated C':irbon C'o
(Calgon Corporation)
West Virginia I'ulp and Paper C'o.
Witco Chemical Company, Inc.
SK.TT.pp 6. II 07
12-1
-------
G
CO
ffi
700
M
o 2
Z o
< =5
I I
ec o
I1
ESTIMATED TOTAL DEVELOPABLE SUPPLY
GROSS NATIONAL
WATER REUSE
REQUIREMENTS
INDUSTRIAL
ESTIMATED
DEPENDABLE
SUPPLY
AGRICULTURAL
MUNICIPAL
300 —
200
100
o
1
i
I
re
a
n
w
cr
o
YEARS
1900
1920
1940
1960
1980
2000
Figure 1. WATER USE AXD WATER SUPPLY IN THE
UNITED STATES (1900-2000)
-------
Use of Clranul.'u Activated Cai bon
Table I
EXAMPLES OF ORGANIC MATERIALS REMAINING
IN SECONDARY EFFLUENT
TANNINS ALKYLBENZENESULFONATE
LIGNINS SUBSTITUTED BENZENE INSECTICIDES
ETHERS PROTEINACEOUS MATERIAL
END PRODUCTS OF BIOLOGICAL OXIDATION
2 Powdered
American Noiil Company, Inc.
Atlas Chemic ill Industries, Inc.
West Virginia I'ulp & I'.iper Company
II PRINCIPLES OK ADSORPTION ON
GRANULAR CAHBON IN COLUMNS
A Adsorption is a phenomenon by which solutes
in a solution are attracted to and adhere to
the surface of solid mntei ial.s Activated
carbon is a particularly good adsorbent be-
cause it has an extremely large surface area
per unit of volume -- areas of 1000 m^/gm
are not uncommon. The bulk of the area is
on the walls of pores in the granule. The
driving force lor- adsorption i:-. a function ol
concentration of the mater lal to be adsorbed,
active surface ai-ea available, type of
ad.sorbate and, to a lesser- degree, pll and
temperature.
B Various theci ics have been proposed for the
adsorption phenomenon but, in general, the
process mechanisms arc still largely
unexplained. One empirical formula, the
Kreundlich adsorption isotherm, states that
the quantity of material adstuhcd per unit
weight of adsorbent is proportional to the
concentration of solute in equilibrium with
the adsorbent. This equation ean be written
as follows:
X .
~5vT
= kC
1/n
C
wliet c- \= unit1* of ni.iJcnal rulsoi bed
M = weight of adsorbent
C = equilibrium concentration of
materials remaining unadsnrbcd
in solut ion
k&n = constants which have different
values for each solute and1
adsorbent. They are also
tcmperatm e and pH dependent
This formula doi-s not lake mli> .id mint
the fact that the adsorbent will i ea( h .1
saturation point at which further mcr i .isc--.
in r-oncentration will not increase the Ifading.
Figui r 3 shows isotherm datj taken by
adding different amounts of adsoi bi nl 1<>
given quantities of wastewater and mi ism -
ing residua] chemical oxygen demands
after equilibrium has been reached. In
the ease of granulni rarbons, this may l.ikc
several weeks so the earbon is pulvon/cd
for the toftt. The lines ean be extrapolate I
to the feed concentrations (CQ) to dctrimmu
the adsorption capacity of the i arbon Im
the organic material in a given wa&tcwnfiM
D
Some generalities relating to Ihe type
material"? adsorbable by eurbon are:
of
1 Weak electrolytes aie absorbed bettei
than strong electrolytes
2 The mure ionic a material is, the mot e
difficult it is to adsorb.
12-3
-------
JO
CO
O
o
MUNICIPAL REFRACTORY ORGANICS (million Ibs./yr.)
DISCHARGES TO U.S. STREAMS
o
o
00
o
o
ro
O
o
(*)
o
o
CO
o>
o
o
I
o
§
8
d
VI
n
o
I
H-'
P
o
ft
n
c.
n
V
S-
o
3
Figure 2: ESTIMATED-LOAD OF MUNICIPAL REFRACTORY ORGANIC
MATERIALS TO U. S. STREAMS
-------
FIGURE 3: COD /SO7HERMS USING VIRGIN CARBON AND DIFFER EN J
SECONDARY SEWAGE EFFLUENTS
NS
I
Q
kU
CO
oe
O
Q
<
Q
O
U
d>
E
2
O
CO
oe
U
0)
E
(•027
o.oi
0.1
10 100
(Cj RESIDUAL COD CONC. (ppm)
CO CO CO
100.0
u
-------
Use of Grannliir Activated Carbon
soluble matt-rials ait1
generally ud&orbed better than highly
soluble materials.
High-molecular-weight materials may
be adsorbed better than those of low
moleculai weight. There arc exceptions:
ABS, for instance, is a very soluble
material and yet is absorbed well on
carbon and other adsorbents.
In industrial and municipal waste, you
have* mixed materials that complicate
the problem. When adsorption is applied
to two materials such as phenol and ABS,
each inhibits the other's adsorption. In
general, however, the net effect of mixed
material': is not detrimental, and you
may gel an enhancement of the total
weight of materials adsorbed from
mixtures
IV LABORATOKY STUDIES
A A detailed laboratory study of the perform-
ance of activated carbon at, an adsorbent for
organic materials was conducted by
Drs. Morris and Weber u( Harvard Univer-
sity under sponsorship of the Federal Water
Pollution Control Administration.
B Figure 4, from this sludy, shows the elfect
of the molecular weight of tin- adsorbate
molecule on the rate of adsorption. In
addition to size, the molocular configuration
has an effect -- extensive branching tends
to reduce the rale of adsorption".
C As mentioned befort-, si?r i eduction of the
carbon reduces the time necessary for
reaching equilibrium. Figure 5 shows the
effect of particle size on adsorption rate.
If the mechanism of adsorption is one of
uptake on external sites, the rate should
vary as the reciprocal of the first power
of the diameter If the mechanism is one
of diffusion within the particle, the rate
should vary with thtr rec iprocal of the
square root This figiin- ••.hows that mtra-
particle difiusiun is the rate-controlling
mechanism
D Decreasing tc-mpprnturc and pH act to
increase both the capacity of the carbon
and the i ate of adsorption It seems
evident at this lime, however, that the
cost of controll.ng cither the temperature
or pH far exceeds any benefit that could be
obtained by this action
E An effort was made to determine a
relationship between the efficiency of a
carbon for removing COD from waste-
water and one or more of the classical
methods for evaluating carbons. These
classical methods include 1) surface area
by nitrogen adsorption, 2) adsorption
capacity for phenol and 3) adsorption
capacity for methylene blue. None of
these indices were useful in rating
activated carbons for the service intended.
The adsorption capacity for alkyl benzene
sulfonate (ABS) came closest to being
meaningful but still was not reliable
enough to be used
F At this time, our recommendation to a
potential user would be an empirical
test in which the waste to be purified is
contacted with granular carbon in small-
scale columns. Publication No. 8 in the
selected reading list at the end of this
outline presents a detailed procedure for
arriving at a system design by this method.
V PILOT PLANT STUDIES AND RESULTS
A Carbon adsorption was tested by setting
up small columns in series at a sewage
plant in the Pittsburgh area. Figure 6
is a schematic drawing of the experimental
apparatus.
B The principal conclusions drawn from these
studies, from similar ones at a pilot plant
area in Pomona, California (a joint project
of the FWPCA and the Los Angeles County
Sanitation Districts) and from others are
as follows:
1 The COD in a well-treated secondary
eifluent can be reduced to less than
3 ppm by adsorption on carbon.
2 At feed rates between 4 and 10 gpm/ft2,
the product quality is a function only
of column length The actual velocity
of the water past the carbon granule
(in the range mentioned) has no effect
so diffusion of the solute molecule to
the sm face of the carbon granule is
not controlling. Figure 7 illustrates
this point
3 Recent indications are that, in sugar
decolorizing at rates much below 4
, gpm/ft2, diffusion in the liquid becomes
controlling This is an extremely
important point and is being checked
out on wa.stewater.
12-6
-------
ro
i
220
200
180
^ 160
140
120
100
80
I I
I
180 200 220 240 260 280 300 320 340 360 380
MOLECULAR WEIGHT, g
C
v.
n
c
OJ
«-t-
n>
a
O
Figure 4. EFFECT OF MOLECULAR WEIGHT ON RATE OF ADSORPTION
o
-------
M
I
00
i
160
20
(D
O
">
O
*t
I
h^
P
n
o
1
cr
c
120
l/(diam) , (mm)
-2
Figure 5. EFFECT OF PARTICLE SIZE ON RATE OF ADSORPTION
-------
SECONDARY
EFFLUENT
>
NJ
&:•;
BACKWASH
TO DRAIN
SAND
FILTER
55 gal
RESERVOIR
-tx>-
2" I.D. PYREX
PIPE COLUMN
G-PRESSURE GAGE
S-50 MESH
SS SCREEN
P-SAMPLE PORT
TO DRAIN
ZENITH METERING GEAR PUMPS
G
en
(B
p
Figure 6. Apparatus for granular carbon column
tests on secondary effluent
a
S"
a
0
B
3-
3
-------
Use of Granular Activated Carbon
Figure 7. Effect of Mass Vi.-locity on Carbon Column Performance
20
10
8
o>
£
2
z
o
-J
e
1
0.8
0.6
MASS VELOCITY:
o 4 gpm/ft^
A 10 gpm/ft2
o -
15
20
25
30
35 40
RESIDENCE TIME ( MIN )
Granular carbon can be regenerated
chemically and reused but does suffer
a significant loss in capacity with each
regeneration.
Recent studies at Pomona indicate that
the nitrate ion can be removed from a
nitrified secondary effluent m a cat bon
column. This is accomplished biologi-
cally and requires the addition of a
supplementary food source (e. g. -
methanol). The denitrifying bacteria
utilize the oxygen from the nitrate ion
to metabolize ihe methanol. The same
results have been obtained in a sand
media. Figure 8 shows recent results
from Pomona.
C A 200 gpm pilot plant for treating municipal
secondary effluent was built and is being
operated at Pomona. Figure 9 is a sche-
matic of this pilot plant
1 Contactor No 1 i_s operated at
200 gpm (7 gpm/ft2) and the effluent
is discharged to waste When only
25% of the dissolved COD is being
- removed, the c.u bon is removed from
the contactor, regenerated and returned
to service.
2 Contactor Nos 2 through 5 are operated
in series. When the dissolved COD from
the last contactor in series reaches 12
ppm, the catbon in the first contactor
12-10
-------
Use of Granular Activated Carbon
40
_ 30
e
Z
1 2O
10
O
Z
10
I
CONTACT TIME
o o 5.5 minutes
I 1. 0 minutes
A A 16. 5 minutes
A- A 22. 0 minutes
INFLUENT
2" CARBON COLUMN
7 gpm/ft2
I I i i I I L
J L
024 6 8 IO 12 14 16 18 20 22 24 26 28 30
JUNE- 1967
Figure 8. Deni tr if ication on granular act ivated carbon
in series is removed, regenerated
thermally and returned to the same
contactor. The piping is then re-
arranged so that the contactor contain-
ing the freshly-regenerated carbon is
last in the series . By this mode of
operation, the carbon in the first con-
tactor in series is nearly in equilibrium
with the feed (and therefore contains the
maximum possible loading of organics)
and the freshest carbon contacts the
water just before it leaves the system as
product.
The performance of the four contactors
in series over the first year of operation
is shown in Table 2. After two years of
operation, the carbon loading (in #COD/
100 Icarbon) is still about 58. The
carbon dosage to maintain a maximum
COD of 12 ppm is now about 350-400
#/mg.
Table 3 gives data on the quality of the
feed and product over the first year.
There has been no significant change
over the second years operation. About
80% of the dissolved COD has been re-
moved from the secondary effluent.
D
5 As of September 1967, the carbon in
each of the four contactors in serie.s
had been regenerated twice. The
carbon in the accelerated contactor,
however, has been regenerated eight
times.
6 A column of carbon never becomes
completely exhausted. Dissolved
COD removal levels off at about 15% -
this is evidently a result of biological
activity on the carbon.
7 Very few BOD analyses have been run
at Pomona. The BOD in both the
influent and effluent are only about
5 ppm, which is below the level of
reliability of the test.
A 2. 5 mgd granular carbon adsorption
pilot plant has been in operation at Lake
Tahoe, California for over two year.s.
To protect the water of Lake Tahoe from
increasing domestic wastes, it will soon
be necessary to export the wastewater out
of the Lake Tahoe Basin and export quality
requirements cannot be met by secondary
treatment alone. A 7. 5 mgd plant is being
built to meet this need. Figure 10
shows the flow sheet used in this plant.
12-11
-------
Use of Granular Activated Carbon
FIGURE 9
CARBON ADSORPTION PILOT PLANT
SCHEMATIC FLOW DIAGRAM
SECONDARY
CLARIFIER
i 1 i
I 1 * -*-
CHLORINATION
TO
PRIMARY
CLARIFtERS
SECONDARY
EFFLUENT
STORAGE
BACKWASH
TANK
BACKWASH WATER
Px.
CARBON CONTACTOR
REGENERAT [D
FURNACE
TO
SAN JOSE
CREEK
PRODUCT
WATER
STORAGE
TO
SAN JOSE
CREEK
QUENCH
TANK
DRAIN
BIN
12-12
-------
TABLE 2
ro
i-1
OJ
MAIN CARBON COLUMN PERFORMANCE
VOLUME TREATED (MG)
DAYS ON STREAM
DAYS IN POS'N. "A"
CARBON DOSAGE
(Ibs/MG)
WT. OF ORGANICS
REMOVED
COD Ibs/IOOIbs of carbon
TOC " " " " "
ADC ii M it ii ii
IOA
24
86
86
280
T
73
28
D
46
—
5.3
3LOA,B
59
211
125
220
T
122
32
0
66
16
5.1
nroA,B,c
81
288
77
250
T
106
27
D
57
18
4.9
EOA%B,C,D
100
365
77
270
T
87
24
D
58
17
2.8
T « TOTAL
D - DISSOLVED
c
s
c.
1
n
0
-------
TABLE 3
AVERAGE WATER QUALITY CHARACTERISTICS
OF MAIN CARBON COLUMN
10
JUNE 1965 TO AUGUST 1966
PARAMFTFR COLUMN
PARAMETER INFLUENT
SUSPENDED SOLIDS, mg/l
COD,
DISSOLVED COD,
TOC,
NITRATE, as N,
TURBIDITY,
COLOR,
ODOR,
CCE,
mg/l
mg/l
mg/l
mg/l
JTU
mg/l
10
47
31
13
6.7
10.3
30
12
—
COLUMN
EFFLUENT
<|
9.5
7
2.5
3.7
1.6
3
1
0.014
r.
-------
Use of Granular Activated Carbon
CHLORINE
COAGULANT AND
POLYELECTROLYTE
THERMAL CARBON
REGENERATION
SYSTEM
FIGURE 10 FLOW THROUGH ORIGINAL TERTIARY PLANT
1 Biological Oxygen Demand (BOD) is
reduced from 200-400 ppm in raw
sewage to <1 ppm in final effluent.
COD is reduced from 400-600 ppm to
3-25 ppm. These data indicate that
BOD is preferentially removed by
carbon. A report recently issued by
the South Tahoe Public Utility Districts
claims 6!)% COD removal by the carbon
columns.
2 Flow through the carbon columns at
Tahoe is upward, in contrast to the
down-flow at Pomona.
E A 30 gpm plant has been run at Nassau
County, New York, to provide treatment to
allow injection of treated wastewater.
Secondary effluent was pre-treated by alum
coagulation, sedimentation and filtration
before carbon adsorption. A 400 gpm plant
is presently under construction. The product
water will be injected into the ground to
prevent sea water intrusion and to
replenish the groundwater supply.
VI REGENERATION
A Two types of regeneration have been
studied - chemical and thermal.
1 Chemical regeneration, even by the
strongest oxidants, was uneconomical.
Hydrogen peroxide, the most effective
regenerant tried, showed recoveries
of 70% of the adsorption capacity on
the first regeneration, 50% on the
second and 20% on the third.
2 In the accelerated contactor at Pomona,
the carbon has been regenerated eight
times. Figurf 11 shows the effect of
these regenerations on the capacity of
the carbon.
12-15
-------
320
I
I—
0
Q
O
O
280
_
O
m
CE
^240
O
O
:200
z260
cr
I
o 80
40
/
/
/
/
/ I
•NUMBER OF REGENERATIONS
20 40 60 80 100 120 140 160
CUMULATIVE VOLUME TREATED (10* ga I Ions)
180
200
220
o
Ml
C
B
n
PI
cr
5
Figure II. Effect of regeneration on carbon performance
-------
Use of Granular Activated Carbon
B In a small plant in which installation of
a regeneration facility could not be justified,
the spent carbon can be burned as a fuel
to produce power or boat. It may be
economical to consider chemical regenera-
tion (low capital cost) to prolong the life of
the carbon.
C In a larger plant, a furnace can be used to
obtain almost complete regeneration. All
present installations use a multiple-hearth
type furnace with rabble arms to move the
carbon from hearth to hearth.
D Pollution of air from the furnace stack is
a potential problem. At Pomona, we arc
investigating cyclone separators and
afterburners. Lake Tahoe is using a water-
scrubbing column on the stack gases.
VII COSTS
A Table 4 shows a brt-akdown of the total
costs for a 10 mgd plant. Note that the
amortization is for 15 years at 4%. This
plant was designed to be identical in flow
path and operating conditions to the Pomona
Plant - that is, four pressure contactors in
series, no pre-treatment, average product
dissolved COD of 7 ppm. The capital cost
is lf» 7£/gallon/day and the total cost is
$83. 00/mg
B Lake Tahoe estimates that their 7. 5 mgd
plant can be operated for $30/mg. They
predict 65% COD removal with a contact
time about 1 /3 of that in Pomona. In
addition, Tahoe amortizes over 20 years
at 4%. Similar amortization at Pomona
would reduce the cost to $76/mg.
C Nassau County amortizes over 30 years
at 3-1/2% and gets a cost of about $40/mg
with a contact time of 24 minutes.
Amortization over 30 years at 3-1/2%
would reduce the Pomona costs to $67/mg.
D Figure 12 shows the estimated cost data
for plants up to 100 mgd. These estimates
may change as larger plants are built and
more cost data are accumulated.
E The overall system needs to be optimized
for costs. Some: of the major factors that
need to be considered are listed below.
1 Pressure vs gravity flow. Pressure
tanks cost more but can be run at a
higher rate and take up less area.
Backwash requirements will be less
for pressure systems. Gravity systems
may require suspended solids removal
before the rarbon
True countercurrent vs batch counter-
current. True countercurrent flow of
carbon and water will utilize the
maximum capacity of the carbon. We
need to determine the cost of equipment
for obtaining true countercurrent flows
on a large scale.
Contact time. Contact time is inversely
related to the rate at which the carbon
is regenerated to maintain a given
quality. Pomona could get the same
quality product in three contactors if
they regenerated at a higher rate.
VIII FUTURE PLANS
A In addition to the 7. 5 mgd plant at Lake
Tahoe and the 400 gpm plant at Nassau
County, a 5 mgd plant is being planned
at Piscataway, Maryland in the Washington,
D. C. area; a 750, 000 gpd plant in East
Chicago, Indiana; and a 50, 000 gpd plant
in Santee, California.
B A contract has been initiated to study the
effect on costs of types of contacting
system used. Another study is in progress
on the effect of carbon regeneration condi-
tions on recovery of adsorptive capacity.
C Studies are in effect or being planned to
determine the effects of carbon granule
size, velocity of water past the carbon
and type of carbon.
IX USE IN INDUSTRY
A Activated carbon is used in industry to
improve the quality of wastewater but
such use is not wide-spread. The main
reason for this is that discharge standards
do not yet require removals over and above
those which can be obtained by conventional
procedures such as settling, coagulation
and biological treatment. It is clear that
effluent quality standards will become more
stringent and activated carbon is expected
to find wide use in industrial waste
treatment. Some of the present uses in
industry are listed below.
1 Oil and phenolic compounds are
removed from the effluent from a
20, 000 barrel /day refinery in Bronte,
Ontario at Trafalgar Refinery.
12-17
-------
d
«g
n
o
m
TABLE 4 §
COST ESTIMATE
FOR 10 MGD ACTIVATED CARBON TREATMENT PLANT
00
CAPITAL ($1,670,000; 15 YRS.® 4%) $41.00
POWER (K/KWH) 8.50
LABOR ( 4 MEN ) I 5.00
MAINTENANCE 5.00
CARBON REGENERATION:
POWER, GAS a WATER 2.50
MAKEUP CARBON (10 % LOSS ) I 1.00
TOTAL $ 83.00
n
a.
9
ITEM COST/ MG
-------
Use of Granular Activated Carbon
100
PLANT SIZE, milliongallons per day
Figure 12. Cost estimates for granular carbon adsorption
12-19
-------
Use of Granular Activated Carbon
Treatment level is about 5 ppm of carbon
and phenol is reduced from 1. 5 ppm to
15 ppb
Caprolactam monomer is removed to
beds of granular carbon in southeastern
U S The monomer is recovered by a
series of hot water extractions and the
carbon is regenerated with steam. It is
believed the process pays for itself
through recovery of the monomer.
Dye wastes are being successfully
treated using up to 2fa5 ppm of powdered
activated carbon. This dosage, followed
by alum coagulation, removes 86% of the
color.
The Wyandotte Chemical Company in
Washington, New Jersey is removing
mixed organic wastes from their effluent
in a 20 ft. deep upflow column. They
remove and regenerate 10-15% of the
carbon every 14 hours. Most of their
problems have been in transporting the
carbon to and from regeneration. A
carbon loss of 2% per cycle has been
reported.
REFERENCES
1 Hassler, J. W. Active Carbon
Chemical Publishing Company, Inc.
1951
2 Adsorption of Biochemically Resistant
Materials from Solution 1. AWTR-9
Public Health Service Publication
No. 999-WP-ll.
3 Adsorption of Biochemically Resistant
Materials from Solution. 2 AWTR-16.
Public Health Service Publication No.
999-WP-:i3.
4 Feasibility of Granular Activated-Carbon
Adsorption for Wastewater Renovation
AWTR-10. Publjc Health Service
Publication No. 900-WP-12.
5 Feasibility of Granular, Activated-Carbon
Adsorption for Waste-water Renovation.
2. AWTR-15. Public Health Service
Publication No. 999-WP-28.
6 Evaluation of the Use of Activated Carbons
and Chemical Regenerants in Treatment
of Waste-water. AWTR-11. Public
Health Service Publication No. 999-WP-13
7 Levendusky, J. A., et. al. Developing
and Using Basic Adsorption Data for
Realistic Plant Evaluation and Design.
Graver Water Conditioning Company,
Technical Reprint T-191.
8 Fornwalt, H. J. and R. A. Hutchins.
Purifying Liquids with Activated
Carbon. Reprint No. D-101.
Atlas Chemical Industries, Inc.
9 Dobtal, K. A. , et. al. Development of
Optimization Models for Carbon Bed
Design. Journal American Water
Works Association, 58. 9, Sept. 1966.
10 Parkhurst, J. D., et. al. Pomona 0. 3 mgd
Activated Carbon Pilot Plant. Submitted
for publication to the Journal Water
Pollution Control Federation.
11 Hager, D. G. and M. J. Flentje. Removal
of Organic Contaminants by Granular-
Carbon Filtration. Journal American
Water Works Association, 57, 11,
Nov. 1965.
12 Bishop, D. F., et. al. Studies on Activated
Carbon Treatment. Journal Water
Pollution Control Federation. Feb. 1967.
This outline was prepared by Arthur N. Masse,
Chemical Engineer, Project Analysis Activities,
Division of Research, Cincinnati Water Research
Laboratory, FWPCA, SEC.
12!-20
-------
USE OF POWDERED CARBON IN WASTEWATER TULA'I WENT
Powdered activated i.arbon is a lughly adsorp-
tive material and can be used to treat waste -
water. Carbon can be produced with particle
diameters of a few microns. The small par-
ticle size and the highly porous structure make
it an ideal adsorbant. Powdered carbon has
been used in the United States to remove taste
and odor from water supplies since 1930.
Currently, 40-45 million pounds of powdered
carbon are used each year for water treatment.
The successful use of powdered carbon has
led to the investigation of its use for the treat-
ment of wastewater.
I CONTACTING SYSTEMS
A Single Stage
carbon
D Two-stage
water
water
carbon
B Two-stage, Split Carbon Feed
carbon carbon
water
water
carbon
carbon
C Two-stage Crossflow, Split Wastewater
Feed
carbon
water
water
water
i current Flow
water -__ water
carbon
cuibon
II PILOT PLAN'I STUDIES
Pilot plant studies luvc; been made of the single
stage and two-stagi: counlcrcurrent flo* systems.
A Single Stage - Tucson, Arizona
An Infilco Acti. La.oi unit was usei. lo
investigate various aspects of waste-water
renovation at the Tucson Municipal Sewage
Treatment Plant (Figure 1). Bcebi, and
Stevens reported wdti-r with a COD ut
14-23.5 mg/1 produced from a pretreated
secondary efflucni with a COD of 36-43 mg/1.
Turbidity was reduced from 15-18.4 to
values ranging from 1-6.9 units.
B Twojstage countcrcurrwnt flow - Lebanon,
Ohio
Research has been conducted at the Lebanon,
Ohio pilot plant facilities by the federal
Water Pollution Control Administration u< ing
a two-stage counU.-rcurreni flow system.
The pilot plant design (Figure 2) wa»i based
on contract work done- by Dnrr-OJiver-2.
Each stage .-on.sisis of a contact tank, a
flocculation lank and a settler.
The influent i > thi- uj:it is pumped into a fust
stage contadoi vvliore it is imxuJ with
partially spent po vdercJ caiiun fiom the
second stage. The carbon water mixture
overflows to Iho lloccul.itiori tank. Di.rmg
the passage 'i n.j conUi I tank to flocculatoi.
an organic i>ol>i luctrulyte is adde - lu
agglomerate tin- < .IMKJII and facilitate its
settling. Tin- IKK L^ uinlt up n ihr floccula-
tion t;mk, fj on- w,hjt ii ihe wa-i-i ovi'rflows
SE. TT.pp 7. 1) (.7
12-21
-------
Use of Powdered Carbon in Wastewater Treatment
AGITATOR
COAGULATION ZONE
CHEMICALS
INFLUENT
^lf-"w*W WW 1
INLET DUCT
V
X
\
rr
LAUNDER
SLOWDOWN
CONCENTRATOR
V
PRIMARY MIXING ZONE
Figure 1. SINGLE STAGE TREATMENT UNIT
Reprinted with permission from January 1!)67 issue of "Water and Wastes Engineering. "
to the settling tank. The carbon settled out
at this stage has contacted the wastewater
in both stages and will require re-generation
for further use. The settler overflow is
pumped into the second stage contact tank.
A virgin carbon slurry is added to the
partially treated water in the second stage
contact tank. The water carbon mixture
again has a poly electrolyte added, is floc-
culated and settled. The settled carbon is
pumped to the first stage contact tank. The
settler overflow is filtered through a
dual-media filter (sand and anthracite). A
tank is required to collect the filter back-
wash water so that carbon retained by the
filters will not be lost from the system.
The effluent obtained from this system is
characterized by the data presented in
Figures 3 and 4. Figure 3 shows that
effluent turbidity is 1 or less Jackson Units
until the influent to the system exceeds
approximately 20 Jackson Units of turbidity.
Figure 4 shows the relationship between the
TOC of the influent and the TOC of the
effluent. Assuming a linear relationship.
12-22
-------
Use of Powdered Carbon in Wastewater Treatment
FIRST STAG1
TREATMENT
0
CARBON CONCENTRATION
IOO TO
10 TO iO gpm fltO
SECONDARY EFFLUENT
FROM CONVFNTIONAl
NT PLANT
SPf NT CARBON TO *
REGENERATION FURNACE I
SECOND STAGE
TREATMENT
CARBON CONCENTRATION
IOO TOISOO
Figu,e2. POWDERED CARBON ADSORPTION PILOT PLANT
Lebanon, Ohio
the curve of best fit has the formula
Y * 1.23 + O.Oli X, where Y = the
TOC (mg/1) <-l the effluent and X =* the
TOC (mg 1) of the influent. This line
appears to be representative until the
TOC of the influent exceeds 25 mg/1, and
shows a TOC reduction of over- 90% of the
TOC exceeding 1.2 mg/1. A TOC of 1.2
mg/1 appears to be the lowest consistent
value obtainable under the stated conditions.
These results were obtained using Lebanon
secondary effluent and the given parameters
The parameters required to obtain a given
effluent, quality must be determined for eacl
particular wastewater and situation.
Work has begun to determine the adaptability
of this .system for the treatment of primary
effluent. The data obtained during two 120
hour runs is shown in Table 1.
12-23
-------
Use of Powdered Carbon in Wastewater Treatment
SJO
40
30
g 20
10
Influrnl -prondary effluent
Kiow 10 spin
Carbon 200 rng/l
Polymer First Stage - 1) 5 mg/1
Second Stage - I 0 ing/1
Each puinl represents SO sdmpli-s ami 120 hours of operation
10 15 20
Influent rurbulil>
-------
Use of Powdered Carbon in Wastewater Treatment
TABLE I
Influent: Primary Effluent
Flow: 5 gpm
Carbon: 200 mg/l(based on 5 gpm)
Polymer: 1st stage - 1.0 mg/1
2nd stage - 1.5 rng/1
June 13 - 17, 1967
TOC (ppm) Feed
High
Low
Avg.
216.0
21.0
69.0
Turbidity (JU) Feed
High 90.0
Low 12.5
Avg. 41.7
July 11-14, 1967
TOC (ppm) Feed
High 67.0
Low 14.6
Avg. 41.7
Turbidity (JU) Feed
High 38.0
Low 9.0
Avg. 23.4
Product
32.5
4.4
10.2
Product
10.0
0.83
3.3
Product
5.8
1.6
3.7
Product
2.8
0.03
1.0
in HANDLING POWDER ED ACTIVATED
CARBON t3'
Care must be used in the handling of powdered
activated carbon. As with most finely pulver-
ized material, there will be a dust problem if
proper precautions are not taken. Dusting
problems are minimized if the carbon is
handled as a slurry whenever practical.
Materials which come in contact with a carbon
slurry should be corrosion resistant as
activated carbon in water permits galvanic
corrosion.
A Kraft Paper Bags
1 Until a few years ago, all powdered
activated carbons were supplied in Kraft
paper bags. The use of these bags in-
volved transferring the bags to storage,
moving the bags to the feed room, emptying
the bags, and final disposal of the bags.
12-25
-------
Use of Powdered Carbon in Wastewater Treatment
Storing the carbon in slurry for m
involves only emptying the bags and
final disposal of the bags. This reduces
the number of points at which dusting
problems r.iay arise.
2 The greatest amount of dusting occurs
when the bags are emptied. The fol-
lowing precautions should be observed
in order to minimize the dusting
a The carbon bag should be opened
only at the place it is to be emptied.
b The seam at the top of the bag should
be removed by cutting it with a knife.
c The carbon should be allowed to flow
out slow ly during the emptying
operation and never allowed to fall
any appreciable distance. Cutting a
hole in the opposite end of the bag
will prevent a vacuum from develop-
ing within the bag and permit the
carbon to discharge easily.
d The empty bags should not be shaken,
since that will create dust without
recovej ing a significant amount of
caibon.
e The empty bags should be stored in
a drum or barrel until such time as
they are incinerated.
B Bulk Handling
Treatment facilities of a million gallons per
day or larger should investigate the econom-
ics of bulk shipments.
1 Shipment
Powdered carbon can be shipped by
railroad hopper cars and by Air slide
railroad cars 'and Airslido trailers.
Airslide railroad cars and trailers are
specifically designed for bulk transporta-
tion of powdered commodities.
2 Handling
Bulk deliveries can be dumped directly
into underground tanks where the carbon
is wetted with water and stored until
needed. The concentration most widely
utilized is one pound of carbon per
gallon of slurry. Higher concentrations
have resulted in difficulties. The
addition of one pound of powdered carbon
will expand the volume of the resulting
slurry about 10%. Slurry tanks should
therefore have a minimum usable space
of 40, 000 gallons for Airslide cars,
and 30. 000 gallons for C&O hopper cars
and bulk trucks.
Activated carbon is insoluble, and has a
specific gravity greater than 1, therefore,
agitation must be provided in storage
tanks to maintain a uniform slurry.
C Transfer Pumps
Both Moyno and Duriron pumps have been
used to transfer carbon slurries success-
fully. Other makes of pumps would also
be satisfactory. Parts of the pumps that
are in contact with the slurry should be
corrosion resistant. A pump that incor-
porates ball check values could be trouble-
some at carbon concentrations of 10% or
greater.
D Metering
Various metering pumps are suitable for
feeding carbon. The literature3 shows
the Omega Rotodip feeder to be widely used
in water treatment.
IV DESIGN
Only pilot plant facilities have used powdered
carbon for wastewater treatment. The folio-*--
ing information relates to pilot studies that
have not been optimized.
A Single-stage Contactor - Infilco Accelator
The pilot model accelator has a nominal
flow rate of 5 gpi'i.
Tank: 5 feet in diameter, 3 feet deep
Liquid Depth. 3b inches
Liquid Volume »00 gallons, 260 gallons of
which are in the clarified
liquid zone
Settler: Area = 17.2 sq.ft.
Rise ran.- - 0.29 gal/min/.iq. ft.
at n f^als/min flow
12-26
-------
Use of Powdered Carbon in Wasle.t-aier Trv.itmi'iit
Pressure Sand Filter: 42 inch diameter
Chemicals: Powdered carbon, ferric
sulfate and polymer
B Two-stage Countercurrent Flow
The Federal Water Pollution Control
Administration's pilot plant unit at
Lebanon, Ohio, has a flow of 10 gpm.
1 Contact tank (2)
Dimensions: 26. 75 in. diameter with a
liquid depth of 31.5 in.
Volume: 75 gallons
Detention Time: 7.5 minutes
Mixer: 1750 rpm with two 3.3"
diameter super-pitch propellers
Baffles: four 2" baffles attached to the
wall
2 Flocculation tank (2)
Dimensions: 35 in. diameter with a
liquid depth of 36 in.
Volume: 150 gallons
Detention time: 15 minutes
Mixing: gate-type flocculation paddle,
8 rpm
3 Settler (2) - gravity, conical bottom
Dimensions: tank - 53 in. diameter with
a liquid depth of 49 in.
Volume: 515 gallons
Detention time: 51.5 minutes
Surface Area: 12.5sq. ft.
Rise Rate: 1.25gpm/sq. ft.
Bottom Scraper: 5 rph
Weir: peripheral, V-notch
4 Sand Filter: support - 1/4 in. to 3/4 in.
stones
- 4 in. coarse sand
dual media - 9 in. of a 0. 6 mm
sand
-22 in. of Anthrafilt.
effective size of
0. 8 mm
V REGENERATION OF POWDERED CARBON
A To make the powdered carbon adsorption
process economically attractive, a method
for re-establishing the original adsorptive
capacity to the spent powdered carbon must
be developed. Davies and Kaplan working
with the 1-ederal Water Pollution Control
Administration conducted suflicient labora-
tory studies to indicate that a dewatered,
spent powdered carbon can be regenerated
to its original capacity with losses in the
range of 1-5% by weight and an ash buildup
of 2% by weight per re-generation cycle.
Regeneration was cm ried out in a laboratory
externally honied screw conveyor at a tem-
perature of 7i>00F with a 20 minute detention
time.
B It was recognized Hi it serious heat transfer
problems exist in this type ol direct thermal
regeneration and that a unit designed on this
principle would br extremely large and would
have great difficulty in r'-aulung the required
regeneration lempeialure. Accordingly,
several methods are being studied to accom-
plish regeneration.
1 Microwave heating - Door-Oliver4
(Figure 5)
Spent powdered carbon at about 10%
solids will be dried in a rotary drum
filter to about 30% solids. The filter
cake shall then be dried in a Thermal
Screw Dryer to approximately 75%
solids. The dried cake will be processed
in the Microwave Heating Unit.
2 Reaction lurnacc - FMC5 (Figure 6)
Spent powdered carbon at about 10%
solids will be pumped to a furnace.
The carbon is rapidly heated to 1500-
1600 F in the furnace by a burner which
admits combustion products and excess
air to pyrolyze and consume the pollution
loading on the spent carbon. The steam
regeneration carbon is then cooled to
150°F in a contactor such as a venturi
scrubber
3 Fluidized-bed - Battellu6 (Figure 7)
The application of fluidization methods
for regeneration of spent powdered
carbon requires development of operating
methods for proper fluidization ol the fine
powder. A fluidized bed of coarse, inert
particles will be used as a constant-
temperature zone. The spent, dried,
12-27
-------
JO
n>
•»»
3
Q.
O
8
3
n
0)
*1
I
2
»^
0>
30% Solids Spent Carbon
Filter or Centrifuge Cake
Wet
Carbon
Feeder
(Star Valve Type)
Dryer Off Gas
(To scrubber or
after burner)
Microwave Heater off gas
to Scrubber or Afterburning
Thermal Screw Dryer
I
K>
I
M
CO
Oil or
Steam Heating
System
Heat
Regenerated
Carbon
Collector
Regenerated
Carbon
Reslurry
Tank
Microwave
Power
Supply
Water
Regenerated Carbon
Slurry Return to
Adsorption Process
TJ
o
o.
n
1
n
a
3-
o
3
V
CO
!-»•
ID
0)
n
B
rt
Figure 5. Proposed Powdered Carbon Regeneration-Process Utilizing Microwave Heating
-------
Use of Powdered Carbon in Wastnwater Treatment
SPENT
SLURRY
FROM
PROCESS
Figure 6. FLOW SHEET FOR PROPOSED BENCH SCALE AND PILOT PLANT
A (Patent applied for by FMC Corporation)
1
SPENT
SLURRY
SURGE
HOPPER
OFF. GAS
FEED
PUMP
FURNACE
BURNER
INCINERATOR
T T
GAS AIR
a
AIR
VENTURI
SCRUBBER
GAS
AIR
I COLO
SLURRY
SCRUBBING
I TANK
MANUAL
CONTROL
• COLD WATER .N
WATER COOLED INDIRECT
HEAT EXCHANGER
COOLING WATER OUT
RECYCLE
PUMP
powdered carbon would be fed into the
bottom of the coarse, inert bed and
carried through the inert bed by the
action of the fluidizing gas. This method
of fluidized-hed operation should offer
control of retention time- of the fine
carbon powder and good heat-transfer
characteristics. The finely divided
carbon would be recovered with cyclone
collectors or some other collection
device.
There are other methods of regeneration,
Atomized Suspension Technique for example,
but the above methods arc those currently
suppoi ted by the Federal Water Pollution
Control Administration.
VI COST ESTIMATE FOR USE OF POWDERED
CARBON
The cost estimates arn based on work that
has been completed anu no effort has been
made to optimize the design. The cost
estimates are believed to be conservative.
A Some lactoj-s whn.h would lend to reduce
the estimated cost are pivcn below
1 Studies of regeneration thai are in
proci &.= . The estimated cost for
12-29
-------
Usr of Powdered Carbon in WnstowaU-r Treatment
FILLER TUBE
BED THERMOCOUPLE
TRANSITS TOP
JACKET THERMOCOUPLE
CYCLONE
DUST COLLECTOR
OUST RECEIVER
-' ALUMINA TUBE 4" 1.0.
FINE CARBON
COARSE INERT
BED MATERIAL
DRY SPENT CARBON
METERED AIR SUPPLY
METERED NATURAL
GAS SUPPLY
METERED STEAM SUPPLY
-GLOBAR HEATING ELEMENT
CAST ALUMINA BASE AND GAS INLET
OVERFLOW DISCHARGE FOR PINE CARBON
FIGURE 7 EXPERIMENTAL FLUIDIZED-BED REGENERATION APPARATUS
Reproduced courtesy of Battelle Memorial Institute.
12-30
-------
Use of Powdered Carbon in Wastewaler Treatment
regeneration is bat>ed on a process
(wet oxidation) that produces a powdered
carbon that is not as active as some
produced by other regeneration methods.
The cost figure for carbon make-up can
also be considered conservative because
it is quite possible that llvere will be a
smaller percentage of loss than the
assumed 15%.
2 Contacting methods other than the
method tested at Lebanon (a two-stage
countercurrent process) are being
considered. A contacting scheme
utilizing a sludge-blanket type of settler
or possibly pipeline contact may be
efiective. Improvements in the system
could lower costs appreciably.
3 Research is underway to evaluate the
properties of powdered carbons that
would make them particularly suitable
for treating wastewater. This would
help to determine (he best carbon for a
particular waste and could result in a
savings in material costs.
B Assumptions made in estimating the
elements of cost ai e listed below:
1 A contact or rapid mix time of 7-1/2
minutes was assumed. The tanks were
assumed to be 5 it. deep.
2 The slurry nux tanks were assumed to
be used to mix up a 10% slurry of carbon
in water with u 2-hour mixing time.
There were two stainless steel mixing
tanks used.
3 Estimated use ol powdered carbon 1.66
Ib. per 1, 000 gallons (200 mg/1 in the
feed stream).
4 Cost oi powdered carbon delivered is
Q$ per pound. With no i (^generation the
cost of carbon alone is ISC' per 1, 000
gallons.
5 Virgin carbon sluriy is 10% solids.
Spent carbon slurry is 10% solids.
6 Cost of wet oxidation taken as 1.99$.
equivalent to aboul J.24 per ton of dry
solids.
7 A 15% overall loss or make-up require-
is assumed.
8 It is assumed that the operating cost is
equal to a conventional secondary
treatment plant.
$/mgd = 5,000 +21. 500 (l/mgd)'b>J
9 The total detention tirno for the plant
is 2-3/4 hours.
10 With 200 mg/1 carbon feed. 1.910 Ib. of
carbon is in the 10 mgd plant.
11 Assumed overflow rate for settlers =
700 gal/day/sq.ft.
C The following list of cost elements is based
on the preceding assumptions.
COST ANALYSIS FOR POWDERED
CARBON ADSORPTION PROCESS
CAPITAL COST (10 mgd)
2 Settlers 360.000
(700 gpd/sq.ft. )
2 Flocculators (15 mm.) 94,400
2 Rapid Mix Tanks (2 hr.) 60. 800
Filtration and Chlorination 450. 000
Air Blowers (mixing) 16, 000
3 Slurry Pumps 12, 000
3 Main pumps 22, 500
$1,022.700
Multiply by 1.29 to correct for con-
struction cost index • $1,319,283.
TOTAL COST POWDERED CARBON
PROCESS
(cents/I, 000 gallons)
10 mgd
Amortization Cost 2. 6
(20 yr. . at 4%)
Carbon Regeneration 2. 0
Carbon Make-up 2. 3
Operating and Maintenance 2.7
Organic Coagulant 1. 3
Total Cost
10. 9
12-31
-------
Uae of Powdered Carbon in Wastewater Treatment
Vll POWDERUD CARBON VS. GRANULAR
CARRON
A Both processes have the capability to reduce
organic contamination well brlow any other
proven unit, process The processes, with
regeneration included, do not create addi-
tional disposal or pollution problems.
B The powdered carbon process has further
merit because:
1 It is readily adaptable to large scale
treatment.
2 The quality of the effluent can be adjusted
by changing the carbon concentration.
3 Regeneration can be incorporated into a
continuous system.
4 The powdered carh
-------
CHEMICAL OXIDATION
I INTRODUCTION
Chemical oxidation has been used to a small
extent in treatment of industrial wastewaters (1).
Air oxidation has. for example, been applied
to certain petroleum refining wastes for
destruction of ammonia, hydrogen sulfide and
mercaptans. Chlorine is sometimes used for
destruction of cyanides and has been suggested
along with ozone for destruction,of phenols in
coke manufacturing wastes. Generally,
however, chemical oxidation has been found
too expensive for industrial wastewater
treatment. Although the process has not been
applied on plant scale to municipal wastewater.
cost estimates based partly on the results of
laboratory studies with secondary effluent
suggest a similar result (2) (3). The aim of
work on secondary effluent was to remove
the remaining organic contaminants. There
are a number of oxidizing agents that are
cheap enough to be practical if the amount of
oxidizable materials is not excessive. The
problem lies in applying the oxidizing agents
in such a way that large excesses and long
contact times can be avoided. The fact that
there is a chance for oxidation to be practical
makes it sufficiently interesting to consider
further study.
II THEORY
Although there undoubtedly are inorganic
materials in industrial wastes that might be
affected by chemical oxidation, those materials
to be considered here are the organics. The
following discussion refers specifically to
municipal secondary effluents, but should
apply also to many industrial wastes. There
are four groups of oxidants that can be
considered for organic oxidation. These are
active oxygen species such as ozone, hydrogen
peroxide, and hydroxyl free radicals from
other sources; elemental oxygen with or
without a catalyst; chlorine and derivatives;
and oxy acids or salts such as potassium
permanganate. Electrochemical treatment
can also be used, but this treatment is
actually a method of producing oxidizing
materials such as chlorine or active oxygen
species in situ. It is not a cheap process
for organic removal. Of the available
oxidants those that have the most chance of
being useful appear to be the active oxygen
species and elemental oxygen. Chlorine may
be useful also. The following discussion
pertains primarily to the active oxygen forms.
The theory of oxidation in wastewater is not
well known. It is complicated by lack of
knowledge of the complete composition of the
organic materials. One type of reaction that
appears to be important is attack by free
radicals, including chlorine and especially
hydroxyl free radicals, on hydrogen atoms
attached to a carbon atom. The hydroxyl
radicals can be produced from chemicals
such as hydrogen peroxide or by means such
as ultraviolet or nuclear radiation. Reactions
of the type RH + 'OH >R' + HgO can be
postulated where R represents an organic
molecule. The organic radical might com-
bine with a hydroxyl radical forming an
alcohol. The organic radical also might
react in the presence of oxygen to form a
peroxy radical, R' + ©„ > ROO* . The
peroxy radical can react further with another
organic molecule to give another free radical
and a hydroperoxide,
ROO- + RH > ROOH + R'
The production of a new free radical demon-
strates autoxidation. If these radical
producing reactions could be made to continue
a large number of times, the result would be
a very cheap oxidizing method.
Although these free radical reactions and
others that can be written result in oxidation
of the organic molecule, they do not explain
a decrease in the concentration of organic
carbon. To remove carbon from the water
requires rupture of carbon-to-carbon bonds
with formation of carbon dioxide. A number
of types of compounds resulting from prior
oxidative attack can undergo decarboxylation
SE.TT.pp.8. 11.67
12-33
-------
Chemical Oxidation
to carbon dioxide. Two compounds and their
reactions are.
R-CH.-CO-CH -CO.H > R-CH -CO-CH, + CO,
6 6 £ 2 J *•
H02C-CH2-C02H s CH3-C02H + CO2
These reactions and. no doubt, many others
then result in a decrease of 'the organic
content of wastewater.
orgamcs irom synthetic and actual waste-
water samples. Electric power requirements
were imp radically high. For destruction
of the orgamcs in secondary effluent
power costs could be more than $1.00/1000
gal. A limited study of oxidation by
nuclear radiation from a cobalt-60 source
also indicated high costs. At 100/cune
an optimistic estimate of treatment costs
is 50(1/1000 gal.
Ill STUDIES WITH MUNICIPAL SECONDARY
EFFLUENT
A Catalyzed Hydrogen Peroxide Oxidation
A laboratory study (2) was undertaken to
determine the technical feasibility of
combining hydroxyl free radical oxidation
and molecular oxygen oxidation or aut-
oxidation for destruction of the orgamcs
remaining in secondary effluent. Hydrogen
peroxide at pH of 3 to 4 and in the presence
of ferrous or ferric iron was used as a
source of free radicals. Both forms of
iron catalyst were found to give appreciable
removal, of the organics in secondary
effluent. Ferrous iron could be used at
room temperature, but ferric iron required
elevated temperatures. From 36 to 65
percent of the chemical oxygen demanding
materials were oxidized to carbon dioxide
and water. When the available oxygen
from hydrogen peroxide was considerably
less than the chemical oxygen demand of
the water, some autoxidation was found lo
occur in the presence of pure oxygen. The
degree of autoxidation was not sufficient
to make the process practical. That free
radicals are produced was demonstrated
by carrying out a polymerization reaction
that is free-radical initiated.
B Oxidation with Other Fret: Radical Sources
Although free-radical oxidation had been
demonstrated, no cheap source of radicals
was available to make the process practical.
Two possible sources are corona electi L^
discharge and nuclear radiation. A
laboratory study ol treatment by corona
discharge gave very small removals of
IV PLANNED STUDIES OF MUNICIPAL
WASTEWATER
Oxidizing agents that may be economical for
secondary effluent treatment, if technical
feasibility can be demonstrated, are atmos-
pheric oxygen or pure oxygen, ozone, and
chlorine. A study oi catalytic oxidation with
air is planned. Promising results from this
work would be of great importance for organic
removal. Investigation of ozone treatment
of secondary effluent is also planned to
determine the efficiency of oxidation.
Preliminary work with chlorine in the presence
of ultraviolet radiation indicates that the
rate of oxidation ot some organics is
accelerated. Light catalyzed oxidation will,
therefore, bo explored further.
V APPLICATION TO IN DUST RIA L WASTES
Development oi a catalyzed-oxygen oxidation
process could be of fjrcat importance for
industrial waste It i-atmt nt. Atmospheric
oxygen and even pure oxygen can be supplied
cheaply enough that fairl> concentrated
wastes could be treated Light catalyzed
chlorine oxidation may have some applications,
but these would be more limited. The
inherently high cost ol o/.one limits its
application to very dilute solutions.- Ther«
may be, however, uses that have been ovei-
looked.
REFERENCES
1 Gurnham, C. K., Editor, "industrial
Wastewater Control", Academic Hress,
New York iiml London, 19G5.
12-34
-------
Chemical Oxidation
Bishop, D. F., Stern, G.. Fleischman. M., AWTR 14. Public Health Service
Marshall, L.S., "Hydrogen Peroxide Publication No. 999-WP-24.
Catalytic Oxidation of Refractory
Organics in Municipal Wastewaters".
Presented at the 148th National Meeting
of the American Chemical Society, This outline was prepared by Dr. Carl A.
Chicago, Illinois, Sept. 1964. Brunner. Chemical Engineer, Project
Analysis Activities, Division of Research.
"Summary Report, Jan. 1962-June 1964, Cincinnati Water Research Laboratory,
Advanced Waste Treatment Research", FWPCA, SEC.
12-35
-------
NEW DEVELOPMENTS IN SLUDGE HANDLING AND DISPOSAL
by
Robert B. Dean
Chief, Ultimate Disposal Research Program
Advanced Haste Treatment Research Laboratory
Fourth Technical Session
2:30 PJI.
Thursday, February 6, 1969
Symposium on Nutrient Removal and Advanced Waste Treatment
Portland, Oregon
-------
NEW DEVELOPMENTS IN SLUDCE HANDLING AND DISPOSAL
I INTRODUCTION
A Pollutants removed from waste-waters must
be treated in such a way that they will not
pollute the environment.
1 A pollutant is a substance that interferes
with the intended use of the environment.
2 Incineration to reduce volume of organic
wastes must not lead to air pollution.
3 Likewise, the effluent from a scrubber
used to control air pollution from a
furnace must be treated to prevent
water pollution.
4 Only three places to put polluting
substances. Air. land, ocean. Not
in surface waters.
5 Pollutant substances must be rendered
innocuous either by dilution below
background level or by locking up.
B Ocean Disposal
1 Barge to deep water and sink to bottom.
Lock up on bottom.
2 Ocean pipeline to deep water and sink
solids. Dilute and lock up. West Coast.
3 Ocean diffuser into well mixed area.
Dilute, Gulf and East Coast Continental
Shelf.
4 Food chains may concentrate poisons
killing larger species.
5 Long pipelines contemplated to collect
sludges to bring them to disposal areas.
C Land Disposal. Fill.
1 Not suitable for soluble substances such
as salts.
2 Needs dewatering to produce solid that
will bear a load. Useful for insoluble
inorganic wastes.
3 Organics will putrefy and decay and
may produce foul seepage and sub-
sidence of the surface.
D Land Disposal. Surface.
1 Not suitable for solubles except nuti u-nls
in quantities utilized by plants.
2 Low cost dewatering, can handle liquid
sludges.
3 Low cost oxidation of organic matter.
4 Most elements locked up on soil
minerals.
5 Improves soil for agriculture and
forestry.
6 Chicago planning long pipeline to carry
sludge to waste lands.
7 Morgantown, West Virginia, planning
to recover mine spoil with organic
sludge.
E Disposal into Wells.
1 Wells into porous formations are
unsuitable for sludges or liquids con-
taining filterable solids.
2 Useful for salt disposal into saline
aquifers.
3 May leak out and contaminate other
waters.
4 Large volumes may produce earthquakes
and land movement.
F Disposal into Underground Cavities
1 Natural salt domes in oil fields.
Suitable for salts, acids, alkalis and
sludges.
2 Limestone caverns. Danger of leakage
if underground water is present.
AWT. UD. 1.12.67
13-1
-------
New Developments in Sludge Hand,] ing «nd Disposal
3 Artificial caves, nuclear cavities. More
suitable for radioactive wastes than
sludges. High cost relative to other
methods.
C, Consider pollutants on an element-by-
clement basis. Fortunately do not have
101 problems as most elements will not
lie pollutants.
II ORGANIC SUBSTANCES
A Carbon, Hydrogen, Oxygen, Nitrogen,
Phosphorus, Sulfur, Ash.
II Principal problem in disposal of organic
sludges is water. Twenty to fifty times
as much water as all other substances in
waste.
C1
Carbon ami hydrogen in organic compounds
can be oxidized to CO., and H^O which do
not pollute the atmospnere.
by proper furnace operation.
Avoid CO
D Heat Production:
1 Oxidation of 1 Ib. of organic sludges is
sufficient to evaporate about 2 Ib. of
water. Up to 3 Ib. for oily sludge since
combined O as in carbohydrates reduces
heating value.
2 Fuel value of sludge is determined by
bomb calorimeter. Can be estimated
from volatile solids content. Sludge
is about half as good as coal on a dry
weight basis.
.) High temperature oxidation uses all the
heat ol combustion to evaporate water
left in tlie sludge and usually requires
excess fuel.
4 economics of incineration are therefore
closely tied to dewatcring by sedimen-
latioi,, filtration, and drying.
II Treahm-nt oi wet shulgo to aid further
1 Anaerobic digestion. Reduces solids
about 50% by hydrolysis and fermen-
tation to methane gas which is burned
to CO and water. Produces foul
supernatant liquor which returns
organics and nutrients to the plant for
recycle.
2 Sludge cooking at 37OOF. Improves
filterability of solids. Returns 10-20%
of the BOD and 60-80% of the nitrogen.
3 Wet oxidation - Zimpro at 350QK
removes 15% of COD by oxidation;
dissolves 25% of solids and 90% of
nitrogen. Higher temperatures destroy
more solids. Improves filtration,
produces a foul supernatant liquor.
4 Aerobic digestion. Aerate for 1-30
days. Stabilizes solids, retains
nutrients but does not aid dewatering.
F Oxidation and dewatering on land surfaces.
An "old-fashioned" process.
1 Low cost dewatering if sludge can be
sprayed on land.
2 Organics are oxidized by soil bacteria.
3 Nutrients and other pollutants are fixed
to a significant extent and kept out of
water supplies.
4 Soil is improved for agriculture or
forestry. Low economic value as
fertilizer but significant in some areas.
5 Must control putrefaction and spread
of pathogens by pretreatment.
13-2
-------
III. CHEMICAL SLUDGES
A. Lime
] . CaCOp CaijOilfFO^K, Mg(OH)2, CaSO^, Ca{l)H)2Dewater in lagoon
or vacuum filter.'
2. Land fill
a. Alkaline due to .gxcess Ca(OH)g
"b. Non-leaciiable FO]J
c. SOP leaches out, l'}00 mg/1
^. (Recovery of CaO from CaOO-, by burning.
B. Iron
1. Fe(OH),, Fe+2. fc Ve*^ (OHJtj, FeK>^, CaCO^, CaSO^ --
Dewate^iiiR more difficult.
2. Land fill.
3. Iron recovery rarely worthwiii IP .
C. Aluminum
1. Al(OH)^, Al PO^, CaCO^.
2. Very difficult to dewater; Freezing.
j^. Alumjnate recovery may be practical.
13-3
-------
IV. SPECIAL DEMURRING TECHNIQUES
A. Freezing — Destroys colloidal structure; gives good filtration
1. Needs slow freezing over several hours after -thawing.
2. Complete freezing and minimal mechanical agitation.
3. Most promise for Activated Sludge and Aluminum Hydroxide.
4. Does not increase strength of filtrate.
5. Does not kill organisms.
B. Radiation
1. Selectively attacks high polymers and may reduce water holding
capacity.
2. 'Does not disinfect. May kill 90# of microorganisms.
3. Releases organics to filtrate.
k. Requires costly shielding.
C. Pressure Cooking — 250° C (kQO° F)
1. Gives good filterabillty.
2. Strong "soup" needs further treatment.
3. Disinfects.
D. Cooking with SO2 catalyst
1. Good filterability at lover temperatures and pressures.
2. Strong "soup" may have food value for livestock.
3. Disinfects.
E. Cooking with Oxygen — Zimpro
1. High temperature to destroy organic solids (^50° C)
2. Low temperature to get good filterability (250° C)
3. Produces "soup" that is difficult to treat.
U. Disinfects.
5. Power recovery not economical.
13-4
-------
LIST OF PERTINENT REFERENCES
1. Dean, R. B., "Ultimate Disposal of Waste Water Concentrates to the Environ-
ment," Env. Sci. and Tech. 2(12), 1079-1086 (Dec. 1968).
2. Dalton, F. E«, Stein, J. £., and Lynam, B. T., "Land Reclamation— A Complete
Solution of the Sludge and Solids Waste Disposal Problem," J. Water Poll.
Control Federation to(5), 789-80^ (1968).
3. Warner, D. L., "Deep-Well Injection of Liquid Waste," PHS Pub. No. 999-WP-21
(1965).
U. Healy, J. H., Rubey, W. W., Griggs, D. T., and Raleigh, C. B., "The Denver
Earthquakes," Science l6l (38^8), 1301-1310 (Sept. 27, 1968).
5. Owen, M. B., "Sludge Incineration," J. San. Eng. Div. Am. Soc. Civil Eng. 83,
Paper 1172 (Feb. 1957).
6. Evans, S. C., and Roberts, F. W., "Heat Treatment of Sewage Sludge," The
Surveyor, 9-12 (Jan. 3,
7. Teletzke, G. H., "Wet-Air Oxidation," Chem. Eng. Progr. 6p_(l), 33-38
8. Slechta, A. F., and Culp, G. L., "Water Reclamation Studies at the South
Tahoe Public Utility District," J. Water Poll. Control Federation 22(5),
787-813 (1967).
9. Doe, P. W., Benn, D., and Bays, L. R.. "The Disposal of Wasnvater Sludge
by Freezing," J. Inst. Water Eng. lg(4), 251-291 (June 1965)-
13-5
-------
DEMINERALIZATION OF WASTEWATERS
Jesse M. Cohen, Chief
Physical and Chemical Treatment Research Program
Advanced Waste Treatment Research Laboratory
Cincinnati, Ohio
Technical Seminar
Nutrient Removal and Advanced
Waste Treatment
Portland,Oregon
February 5 and 6, 1969
-------
ELECTRODIA LYSIS
I INTRODUCTION
Usually when water is used, there are added
varying amounts of inorganic materials. In
the case of domestic wastes, the increment
amounts to about 300 to 400 mg/1. Most of
the inorganic material is composed of ions
common to natural waters. Also added are
phosphate and nitrogen as ammonia or nitrate.
Removal of the added material is obviously
necessary to maintain the quality of the
water. With industrial wastes, especially
from industries dealing with metals, the
inorganic materials added are heavy metal
ions that may be very toxic. Their removal
is even more important than the removal of
the ions usually found in natural waters.
Electrodialysis is useful for the partial
demineralization of fairly dilute waters. For
practical application usually the total dissolved
solids (TDS) concentration is about 2000 mg/1
or less. Since municipal wastewaters nearly
always have a TOS less than 1000 mg/1 they
are well within the range where the process
should be economically feasible. Many
industrial wastes also undoubtedly fall into
the practical concentration range. With
electrodialysis sufficient demineralization is
usually carried out to reduce the TDS to
500 mg/1 or slightly less. Further deminer-
alization results in appreciable added cost.
II PRINCIPLES
A direct electric voltage impressed across a
cell containing mineralized water will cause
positively charged ions or cations to migrate
to the negative electrode and negatively charged
ions or anions to migrate to the positive
electrode. If cation- and amon-permeable
membranes are placed alternately between the
electrodes as shown in Figure 1, alternate
compartments become more concentrated
while the intervening compartments become
diluted. Many membranes can be placed
between the electrodes forming many dilute
and many concentrate compartments.
Manifolds can then be added so that the
membrane stack can be fed continuously and
demmeralized product and waste concentrate
streams can be removed continuously.
The cation(+) and anion(-) permeable mem-
branes that make electrodialysis possible are
composed of ion exchange materials. Cation
membranes are likely to be sulfonated
styrene with divmylbenzene added for cross
linking and strengthening of the polymer.
Anion membranes are also often styrene
base with various armne groups being used to
give the ion exchange property. The
sulfonate and amme groups make the mem-
branes very hydrophilic. Membranes absorb
appreciable water when in contact with
aqueous solutions.
Although the structure and function of the
membranes is complicated, a rough
quantitative explanation of the reason for
their ability to transport ions of one charge
can be obtained from consideration of the
Donnan principle. This principle, in an
incompletely rigorous form, states that the
ion concentration product of a compound will
be equal for water in the membrane and for
the surrounding solution. A cation membrane
in contact with very dilute sodium chloride
can be considered as an example. The
product of the concentrations of the sodium
and chloride ions will be very small in dilute
solution. In the structure of the membrane
there will be a high concentration of fixed
negative charges which, to maintain
electro-neutrality, will attract an equal number
of positive charges or in this case sodium ions.
The concentration of sodium ions associated
with the water in the membrane will then be
very high in relation to the original solution
concentration. By the Donnan principle the
chloride ion concentration in the membrane
must be very small. If a voltage is applied
across the membrane, sodium and chloride
ions will move in opposite directions through
the membrane. Because of the preponderance
of sodium ions, however, most of the current
will be carried by that ion. The cation
WE.TT.pp.9. 11.67
14-1
-------
I ,!'•< I I <
©
ANION CATION A -ANION PERMEABLE MEMBRANE
C-CATION PERMEABLE MEMBRANE
i; t Klectrudu lysis Hruic iplu
Pii-i'il)! .me- will, ihei I'fore, havu :i
seJci livity lot sodium and for cations >n
grin i i. A hiinilar arjjurnenl can be US.L-I! to
sh'1 v •. liy .iriion meinhranes aru selective to
ariK " .. As the Domi.i'i pi incijile null' ,itc.->,
boll ttiviiy decri-asi-s &s solution concc:r. ra-
tion 11-1.1
On- ol th
-------
Electrodialysis
required voltage is more difficult to deter-
mine. Classical thermodynamics allows the
minimum voltage to be calculated (1) for a
membrane separating solutions of different
concentrations. In the usual electrodialysis
application this minimum voltage, which only
applies strictly at a zero current, is a small
fraction of the total. The usual approach is
to consider the stack as a series of resistances
for which the voltage can be calculated from
Ohm's law. Resistances that should be
included are diluting and concentrating liquid
streams, membranes, and the thin stagnant
liquid films adjacent to the membranes.
Usually, however, the resistances other than
those of liquid streams are simply combined
a£ a ringle value which must be determined
experimentally. This empirical resistance
can then be used to estimate the approximate
voltage for an electrodialysis stack using the
same kind of membranes. To calculate the
totax power requirements the voltage require-
ments of the electrodes and electrode com-
partment .streams must be added to the
voltage for the remainder of the stack.
Ill OPERATING PROBLEMS
The stagnant liquid films that occur at the
membranes are responsible for significant
problems during operation. Serious con-
centration gradients can develop across these
films because of the higher electrical trans-
ference numbers in the membranes compared
to transference numbers in solution. Asa
result there is a decrease in concentration
from bulk dilute stream to the membrane
surface and an increase in concentration from
bulk concentrate stream to the membrane
surface. The high concentration at the face
of the membranes on the concentrating side
leads to scaling from precipitation of com-
pounds with low solubility. The low mineral
ion concentration at the dilute aide of the
membranes causes transfer of hydrogen and
hydroxyl ions from the water through the
membranes. Polarization is the term used
to describe operation when water decom-
position becomes significant. The
decomposition of water into ions wastes
electric power. A more important effect,
however, is the increased scaling potential
at the concentrate side of the anion mem-
branes. Hydroxylion increases the
precipitation of magnesium hydroxide and
calcium carbonate.
An effect that is common to electrodialysis
is membrane fouling. This phenomenon is
different from scaling since it involves
deposition of materials on the dilute side of
the membranes. Although there can be a
number of causes in wastewater treatment
fouling appears to be due to electrophorctic
movement of negative;ly charged colloidal
particles to the anion membranes. Because
of the size of the particles they cannot move
through the membrane. Instead they form a
layer at the membrane surface that interferes
with demineralization.
IV APPLICATION TO MUNICIPAL
WASTEWATERS
A Laboratory Investigation
Bench scale study (2) of electrodialysis
for treatment of municipal secondary
effluent indicated that good removal of
soluble and insoluble organic materials
was necessary for s-itisfactory operation.
Cellulose cartridge filters were used for
removal of the suspended organic solids.
These filters, however, would not be
practical on a large scale and would have
to be replaced by another form of sus-
pended solids removal. Granular activated
carbon was used for soluble organic
removal. Using adequate pretreatment
the cost of electrodialysis, exclusive of
pretreatment and waste concentrate
disposal, was estimated at less than
10 cents /1000 gal.
B Pilot Plant Investigation
As a result of encouraging rnsults from
the laboratory study a pilot scale investi-
gation was undertaken (3). Equipment was
installed at the Lebanon, Ohio Sewage
Treatment Plant. Secondary effluent was
the feed to the system. Based upon the
experience gained from the laboratory
work, clarification and soluble organic
14-3
-------
Electrodialysis
removal were considered necessary
pretreatment for the electrodialysis stack
feed. Diatomaceous earth filtration, using
for the most part an alum treated water
grade filter aid, was chosen as the method
of clarification and granular carbon in
fixed beds was chosen for adsorption of
soluble organics. The pilot system is
shown in Figure 2. The electrodialysis
stack is an Ionics Mark III capable of
holding up to 150 cell pairs. It is designed
to remove about 40 percent of the TDS from
the feed. Membranes for this stack are
18 X40 in. and have about 3000 m2
active area. Spacer material and edge
gasket are combined in one piece. Spacer
thickness is 0.032 in. Using all 150 pairs
the nominal capacity of the stack is 50 gpm
product water. As a result of difficulties
with the diatomaceous earth filter, it was
necessary to reduce the feed water rate.
The number of cell pairs was reduced to
125 and the product rate to 42 gpm. The
concentrate waste rate at reduced feed
rate has been 4 gpm. Intert.al recirculation
of concentrate at a rate about equal to the
product fate is necessary, however, to
maintain approximately equal pressures
on both sides of the membranes. The j>H
of the concentrate stream i:-, held at 5 or
less to prevent calcium cailxmate scale
formatior-.
Operation of the pilot plant has resulted
in only one serious problem fouling or
anion membranes. Membrane fouling
increases stack resistance and necessitates
higher voltage to maintain the desired
degree of demineralization. Figure 3
shows how demineralization at constant
voltage was affected by fouling. Fouling
rate depends to some extent upon feed
turbidity as might be expecU-d. The
unexpectedly low foulinp, rate for the run
made after the stack was acid rinsed at a
low enough pH to kill organisms suggests
biological growth on the membranes as a
contributing factor. Fortunately toulmg
has not been permanent. Shutdown over a
weekend has always restored the fraction
demineralization to near normal. Even
with some fouling, operation is considered
practical since excess stack capacity can
be provided to take care of down time for
self-cleaning of membranes. Control of
feed turbidity is necessary for practical
run lengths before shutdown is required.
Scaling of anion membranes, a problem
that is serious in many electrodialysis
installations, has not been particularly
troublesome in treatment of municipal
wastewater. Calcium carbonate scale
formation has occurred on occasions,
but has not done serious damage to mem-
branes. Failure to control the pH to 5 or
less can cause serious scaling that ruins
the anion membranes. In the pilot woik
concentrate blowdown has been held at
9.5 percent of the product. Scaling
potential would be reduced if the amount
of blowdown were increased. Such an
increase would, however, increase the
problem of brine disposal. From a
pollution control standpoint, therefore,
the concentrated waste must be kept as
low in volume as possible.
Permanent damage to membranes during
pilot runs over a period of 1500 hours has
not been great. Membrane pair resistance
at the beginning ol runs, for example,
only increased about 5 percent even though
fouling conditions during some runs were
severe. Current efficiency, that is the
fraction of the current that results in
demineralization, has not changed
significantly. Serious membrane damage
would result in decreased current efficiency.
To obtain better control of turbidity and
more dependable operation than was
possible with the diatomaceous earth
filter, a chemical clarification system
has recently been installed before the
carbon columns. This system consists
of an upflow clanfier and dual media
filters. Lime has been chosen for use
in the clarifiei because of the softening,
phosphate removal, and alkalinity removal
that this material offers. This change
in pretreatment has not eliminated mem-
brane fouling. The fouling rate is roughly
the same as was obtainable with the
diatomaceous earth filter when it was
producing water of less than 0. 1 JTU water.
14-4
-------
60TO 70 G.PM
I/I 0#/S.F FILTER
SECONDARY EFFUJENTl QQ ,.
FROM CONVENTIONAL]50 ~100 mg/l
TREATMENT
PLANT
DIATOMACEOUS EARTH FILTRATION UNIT
LOWERS pH TO 4 IN CONC. STREAM
40-5% G.PM
ELECTRODIALYSIS
UNIT
40% DISSOLVED SALTS
REMOVAL
FINAL PRODUCT WATER
WASTE CONCENTRATE
-------
Electrodialysis
0.45
6.40
0,35
o:
UJ
z
UJ
o
0.30
025
02O
FEED TURBIDITY = 0.03 v>TU
FEED TURBIDITY = 1.3 JTU
STACK ACID RINSED
FEED TURBIDITY-0.2 JTU
IO 15 20
RUN TIME (hrs)
25 30
FIGURE 3. EFFECT OF TURBIDITY ON
DEAMNEftALIZATfON .
14-6
-------
Electrodialysis
Further work with chemical clarification
will be carried out.
C Other Investigations
Ionics Incorporated has operated a small
pilot electrodialysis stack at Orange
County on trickling filter effluent that had
been alum clarified, chlorinated .and treated
with granular activated carbon. In addition
the stack had before it 10 and 3 micron
cartridge filters and an ultraviolet light
for sterilization. The fouling rate
apparently was considerably lower than
observed during operation of the Lebanon
stack. Operation was halted after more
than 800 hours of run time when a serious
stack resistance increases occurred.
More recently this same electrodialysis
system, including cartridge filters and
ultraviolet light, has been operated on
carbon treated, chlorinated secondary
effluent at Pomona California. Here
carbon treatment is carried out without
any clarification. Frequent replacement
of the cartridge filters is necessary.
Preliminary results suggest that fouling
will be at a lower rate than observed on the
Lebanon stack. Further work is required
to verify the early results and to determine
the effect of the cartridge filters and
ultraviolet light.
V APPLICATION TO BRACKISH WATER
The principal use for electrodialysis has
been in treatment of brackish waters. These
are inland waters containing enough salts to
prevent or limit direct use. The mineral
content is considerably less than for sea
water, often about 2000 mg/1, and their
compositions can vary widely from one
location to another. Electrodialysis has
been found practical for demineralizing these
waters to 500 mg/1 TDS or slightly less.
There are problems associated with brackish
water treatment, some of which are different
from those observed in wastewater treatment.
Fouling is sometimes troublesome, and is
often not of organic origin. Instead inorganic
materials such as traces of iron and man-
ganese can cause the difficulty. It is usually
considered necessary to remove essentially
all the iron and manganese from ground
water to be treated by electrodialysis.
Scaling is often a problem because of high
concentration of materials of low solubility
such as calcium sulfate and calcium carbonate.-.
To avoid scale formation the amount of con-
centrate waste is kept considerably higher
than is possible in wastewater treatment.
Often the volume of concentrated waste is
from 33 to 100 percent of the volume of
treated water.
There is an increasing number of electro-
dialysis installations using brackish water
in the United States. Two of these are
particularly well known. They are at
Webster, South Dakota and Buckeye, Arizona.
The Webster plant began operation in 1962
as a Demonstration Plant by the Office of
Saline water and has a nominal capacity of
250,000 gpd. Electrodialysis equipment was
supplied by the Asahi Chemical Industry
Company of Japan. The feed water is not
unusually high in TDS for brackish water
since it contains only about 1700 mg/1. It
is, however, a very difficult water to treat.
It is high in calcium, sulfato, and bicarbonate
so that scale formation is a serious problem.
The water also contains significant amounts
of iron and manganese that must be removed
by using potassium permanganate injection
and filtration through manganese zeolites
before electrodialysis treatment. One of the
wells that was to provide part of the feed
water contains organic materials that cause
a rapid increase in stack resistance. Many
problems have arisen, therefore, during
operation of this plant. These have been
described in detail in a number of Office
of Saline Water publications and summarized
in a 1965 paper (4). As a. result of the prob-
lems the Webster site has been used to test
a number of pretreatment methods, both for
iron and manganese removal and softening.
An interesting feature of the plant as now
operated is the use of periodic polarity
reversal to reduce scale formation. Improved
operation is obtained at a slight loss in
current efficiency. Successful use of polarity
reversal could represent a significant break-
H-7
-------
Electro dialysis
through in the application of electrodialysis
to high-hardness, badly-scaling waters.
Results at Buckeye. Arizona have been more
optimistic than results at Webster. The
water, although higher in TDS, has a relatively
large amount of sodium and chloride ions.
Scaling is, therefore, much less likely to be
a problem. The water contains no manganese
and little iron. The equipment installed at
Buckeye was supplied by Ionics, Inc. and
has a design capacity of 650, 000 gpd. Over
the period October 1962, through June 1965,
it is reported (5) that there were no major
malfunctions. During that time the average
load factor was only 32 percent. This
unexpectedly low figure resulted partly from
the recirculation of water through evaporative
coolers. Recirculation was not possible with
the more highly mineralized brackish water
available before installation of the electro-
dialysis equipment and, therefore, the daily
water consumption was higher. The low
load factor has resulted in a higher than
expected water cost when amortization of
equipment is included, but the installation is,
nevertheless, considered a success.
VI INDUSTRIAL APPLICATIONS
Use of electrodialysis for treatment of
industrial wastes has been very limited. Two
possible applications that have a pollution
control aspect are treatment of pickle liquor
and treatment of spent sulfite liquor. The
first application is actually more nearly
electrolysis than electrodialysis and results
in recovery of iron and sulfuric acid from
the waste (6). There is more interest in this
process in Europe than in the United States.
The socond process was developed by the
Sullite Pulp Manufacturers Research League
(7) and results in recovery of cations from
the ^pcnt liquor. The principle of this
modified electrodialysis process is shown
in Figure 4. Spent liquor enters compartments
3 ,md 7 where cntions and organics such as
sm.ill Hfino-sulf cmate ionR are removed. New
liquor is made up LJI compartments 4 and 8.
Compartments 5 and 9 serve js sources for
sulfito ion. The spent liquor \iith low mole-
cuLif weight 01 games removed may have a
numl.i.-r of uses. The process operates at
high current efficiencies and high current
densities. Because of the arrangement of
membranes, the anion membranes are
protected from fouling by organics in the
spent liquor.
Although electrodialysis has not yet been
used extensively for treatment of industrial
wastes, it may have wider application in the
future. Its practicability is improved if
useful materials can be recovered from the
treatment. With the present cost picture
for electrodialysis, recovery is probably a
necessity. It is accomplished in the two
examples cited above. There are undoubtedly
many industrial waste streams where chemi-
cal recovery is a possibility.
In its basic form of alternating dilute and
concentrate streams the process would
ordinarily only be applicable to streams free
of fouling soluble organics and suspended
materials. By using special membrane
arrangements such as in the sulfite liquor
treatment, fouling problems can be reduced.
It should he recognized that in applications
where useful materials are recovered the
added maintenance resulting from membrane
fouling may not be prohibitive in cost.
The maximum concentration that can be
economically treated would be limited to the
brackish wuter range if by-product recovery
were not possible. Electric power consump-
tion for removing a large concentration of
ions, ordinarily becomes prohibitive. Where
useful materials arc recovered the maximum
concentration, could be increased until the
cost for- electrodialysis exceeds the cost of
other processes such as distillation. A
detailed study of each particular application
would have to be made to determine whether
electrodialysis is practical.
There .in- (wo limitations to electrodialysis
that may pruv^nt its use in some applications.
One limitation is the inability to practically
remove ions below, a concentration of several
hundred mg/1. Whore waste are very dilute,
ion exchange should be more appropriate.
The other limitation is the- inability for the
process to or very selective for any one
matenM in a mixture. At present a valuable
metal ion could not in general be removed
14-8
-------
°9
T c . " . c
N
4-
3
NH* —
-LSOj
-SUGAR
^URGE
3
4
z <
V
5
H
6
X
Z <_
«--
7
-LS03
-SUGAR
^ARGE
LO^^M
8
Z <
I
H.
1
so;
C-CATION SELECTfVE MEMBRANE
A-ANION SELECTIVE MEMBRANE
N-NON-SELECJIVE MEMBRANE
Figure 4. Electrodialysis Process for Treating Spent Sulfite Liquor
K
H^
n
o
o
a
-------
Electrodialysis
with high selectivity from common ions such
as sodium and calcium. Relatively little
work has been done to develop selective
systems, however, and improvements are
likely if a strong need arises. Again ion
exchange may be a more practical form of
treatment.
VII ULTIMATE DISPOSAL OF CONCENTRATED
WASTE
Experience in the use of electrodialysiu en
municipal wastewater indicates that the volume
of the concentrated waste will be 5 to 10
percent of the product volume. To prevent
pollution of surface waters by this highly
mineralized waste it must eithet be trans-
ported to the sea or in some other way be
treated for disposal. Disposal, no matter
what the method, will not be insignificant in
cost except near the sea. For brackish
water the potential disposal problem is also
great. Presently the treatment of brackish
water is limited enough that disposal of brine
has not been a serious problem. Industrial
w.iste treatment.-would not prevent a
significant disposal problem if recovery of
by-products from the concentrated waste
were practiced. In that case the waste
would be further treated anyway. If recovery
were not possible, disposal or treatment of
the concentrate to make it innocuous would be
necessary. The volume of concentrate would
vary widely depending upon the particular
waste. II it is low in membrane scaling
materials .the volume may be only a few
percent of the product. Each case 'must be
considered separately.
VIII EQUIPMENT DESIGN
Since electrodialysis is a recently developed
process, design of equipment is not yet highly
sophisticated. The major American maim*
factuter, for example, has available a limited
number of stack configurations to be used
primarily for brackish water. Minor
modifications are made in the equipment to
adapt it to other uses. It would be a fortunate
coincidence if obtainable equipment were the
optimum for any given application. Because
of the limited market, manufacturers have
not been able to justify tht- extensive research
and development wotk that is required for
'optimized design.
Several discussions of equipment design have
been published (8, 9). Application of the
rtiethods were also demonstrated for the case
of municipal wastewater (2). The reader is
referred to these publications for details.
In principle, these methods depend upon Ohm's
Law and Faraday's Law and material balances.
The resulting equations allow membrane area
and power requirements to be calculated.
Before making design calculations, it is
necessary to know momljianu resistances
and limiting current density. These are
empirical quantities, but manufacturers
should have the necessary data available for
their particular slack configurations and for
brackish water feeds. Experimental data
might be required for design of equipment to
treat an industrial waste.
One serious lault with published design
methods is that they do not emphasize the
design oi membrane spacers. After the
designer chooses an available spacer and
has the necessary empirical data it is
relatively easy to size the equipment and
determine power i equirements. Optimum
design of u spacer i.s a much more difficult
problem. Piesenll.y, spacers range in form
from rather open types that give low velocity
and pressure drop to tortuous path types that
give high velocity and pressure drop. To
reduce production <_osts, manufacturers tend
to pick one type and use it for all applications.
The Oil ice of y a line Water is presently
sponsoring work on spacer design. Prelim-
inary results suggest that eVen the brackish
water spacers may be improved significantly.
Further work is obviously justified.
At the present time, the user of electro-
dialysis equipment usually depends heavily
upon the limited number of manufacturers for
the ultimate design ol a system. The number
of types of stack components available is
small. Use of the equipment is not yet so
widespread that consulting engineers are
experienced enough to undertake design them-
selves. If the market increases substantially.
14-10
-------
Electrodialysis
the situation will change. Not only will a
greater variety of components become
available, resulting in more flexibility of
design, but the designing itself will be done
by a greater number of people.
IX ELECTRODIA LYSIS COSTS
A Capital Costs
When discussing the capital cost of
electrodialysis, distinction must be made
between present and projected costs. The
market for this equipment has not expanded
to the point where mass production of
components is possible. The costs of
membranes and spacers especially are
very high. Capital costs do not represent,
therefore, the minimum that would be
possible just from an increased market
and the competition that should develop.
It is difficult to make capital cost estimates
without consulting a manufacturer. There
is not sufficient information available in
published form to make the precise
estimates possible. Ionics, Inc. has at
various times made available in company
literature plant costs for their equipment
as a function of capacity and feed concen-
tration assuming the feed to b'e brackish
water and product to contain 500 mg/1 TOS.
Table 1 shows some typical costs. These
figures do not include pretreatment beyond
a protective cartridge filter. Cost is
strongly affected by both feed concentration
and capacity. An independent cost
estimate has been made for treatment of
secondary effluent based partly upon
experience from pilot operation. For a
10-mgd plant and using Ionics equipment,
the estimated installed cost is $0.34/gpd.
This is significantly higher than the Ionics
estimate for 900 mg/1 brackish water to
which it should most closely compare.
Much of the difference can be accounted
for in the conservative feed rate per cell
pair that was chosen for the wastewater
estimate. A small amount was contributed
by the inclusion of pH control equipment
on the concentrate streams. This equip-
ment is not usually supplied for brackish
water treatment, but is desirable on
wastewater treatment where there can be
frequent changes in alkalinity.
The Office of Saline Water has, in the
past, supported work on a procedure for
estimating the cost of electrodialysis
applied to brackish water. More recently,
work was done on optimization of the
process. This work will appear in
publications from that organization.
B Operating Cost
Long term operating costs for electro-
Table 1
Investment Cost for Electrodialysis Plants
Plant Capacity
(tngd)
Feed Concentration
(mg/1)
Installed Plant Cost
($/Kpd)
2
2
10
10
100
100
900
3,000
900
3,000
900
3,000
0.33
0.69
0.19
0.39
0.11
0.25
14-11
-------
Electrodialysis
dialysis are available from only a few
brackish water installations. Estimates
are usually based1 upon experience combined
with, assumptions about long term main-
tenance requirements. Membrane life is
probably the most questionable factor.
Usually a 20 percent replacement per year
is assumed. Pilot experience on waste-
water suggests that membrane life is
greatly affected by the care that is-taken
in handling the membranes and in operating
the equipment. Although definite proof is
not yet available, there is reason to believe
that an annual 20 percent replacement may
not be necessary. Ionics has estimated the
cost of electrodialysis operation over a
wide range of capacities, and feed water
concentrations. Some examples are given
in Table 2. The estimates are for brackish
water and assume 90 percent load factor
and 20 percent annual membrane replace-
ment. They include amortization of
equipment. 'An independent estimate for
the treatment of 10 mgd of wastewater was
16c7l. 000 gal.- This is somewhat higher
than the 120/1, 000 gal for the nearly
comparable 900 mg/1 brackish water. The
difference can be accounted for in the
conservatively high capital cost estimate
for the wastewater plant.
The effect of load factor on operating cost
is significant. At the Buckeye, Arizona,
brackish water plant, for example, it is
estimated that at full load, the total cost
of producing water should be 32$/1, 000
gal. The plant was actually designed to
operate at a 48 percent load factor and
produce water for 52$/1, 000. gal. For a
number of reasons the load factor declined
below the design value. For the first 32
months of operation it was only 32 percent.
The result was a water cos. of about
70$/1, 000 gal. This is.an extreme
example, but it points out the need to
consider ways for minimizing fluctuations
in flow. For industrial wastes, it may be
possible to make extensive use of relatively
cheap water storage to allow operation, at
a high load factor.
REFERENCES
1 Wilson, J.R.,, "Demmeralization by
Electrodialysis", Butterworths
Scientific Publications. London. 1960.
2 Smith, J. D., Eisenmann, J.L.,
"Electrodialysis in Advanced V/aste
Treatment. " Water Pollution Control
Research Series Publication No.
WP-20-AWTR-13.
3 Brunner, C.A., -"Pilot Plant Experiences
in Demineralization of Secondary Effluent
Using Electrodialysis". presented at
the 39th Annual Conference of the Water
Pollution Control Federation, Kansas
City, Missouri. Sept, 1966.
Table 2
Operating Cost for Electrodialysis
Plant Capacity
(mud)
Feed Concentration
(IDK/I)
Total- Operating Coat
(S/1.000 gal)
2
2
10
10
100
100
900
3,000
900
3,000
900
3,000
16
39
12
29
8
21
14-12
-------
Electrodialysis
4 Calvit, B.W., Sloan, J.J., "Operation
Experience of the Webster, South Dakota,
Electrodialysis Plant", presented at
the First International Symposium on
Water Desalination, Washington, D. C..
Oct. 1965.
5 Gillilard, E.R., "The Current Economics
of Electrodialysis", presented at the
First International Symposium on Water
Desalination, Washington, D.C..
Oct. 1965.
6 Farrell, J.B.. Smith, R. N.. Ind. Eng.
Chem., 54, No. 6, 29-35(1962).
7 Dubey, G.A., et al., TAPP1, 48, 95(1965).
8 Mason, E.A., Kirkham, T.A., Chem.
Eng. Prog. Symposium Series, 55,
No. 24, 173 (1959).
9 Mintz. M.S., Ind. Eng. Chem., 55, No. 6,
19 (1963).
This outline was prepared by Dr. Carl A.
Brunner, Chemical Engineer, Project
Analysis Activities, Division of Research.
Cincinnati Water Research Laboratory,
FWPCA, SEC.
14-13
-------
REVERSE OSMOSIS
I INTRODUCTION
A The phenomenon of osmosis was first
observed in the mid-18th century when
studies were being made on biological
membranes.
B It was noted that when two solutions of
different concentration were separated by
a s^mi-permeable membrane, the solvent
would flow from the dilute to the concen-
trated side. A scini-purmuabli; nn>ml>ranc
is defined as one thai will permit passage
of some materials (usually a solvent)
while rejecting others.
C The fn st experiments with artifically-
prepared membranes were conducted in
the mid- 19th century when a membrane
was made by precipitating copper ferro-
cyanide in the pores of porcelain.
O By 1920, interest in osmosis wanud and,
except for its biological importance, it
was considered a laboratory curiosity.
E Developments in the last ten years by
C. E. Reid and others of the University of
Florida and S. Loeb and others at UCLA
showed that the process can be reversed
by applying pressure to the higher con-
centration side and that it has potential
as a process for recovering reuseable
water from a contaminated or saline
source.
II THEORY
A Figure 1 illustrates the principle of
osmosis and reverse osmosis. At equil-
ibrium, the pressure on the more
concentrated solution is known as the
osmotic pressure. This pressure is
dependent entirely on the difference in
concentration between the two solutions
and is not a function of the type of mem-
brane, provided it is semi-permeable.
It is well known that the "activity" or
"chemical potential" of a solvent decreases
when in a solution. Thus, the activity
of water in distilled water is greater than
the activity of waU r in a salt solution.
This greater activity is thought to be the
driving force that, in normal osmosis,
cause? water to flow through the membrane.
In brackish and sea water, this driving
forct; is about 1 psi lor every 100 ppm
salt in solution.
The actual mrchamsm of transfer ol
water through a reverse osmosis mem-
brane is not completely understood. The
most genurally-acceptcd tho-ory is thai
water passes through tin.' mombranc by
successive transfer from one adsorption
site to the next. In the case of cellulose
acetate, the most common membrane
material, the adsorption foi C-PS arise from
hydrogen bonding.
The rate of transfer of water through a
membrane in reverse osmosis is directly
proportional to the applied pressure minus
the osmotic (back) pressure and inversely
proportional to the thickness of the
membrane. Increasing temperature,
since it raises the activity of the water,
will result in a greater water flux through
the membrane.
As stated before, the flux of water is
directly proportional to the pressure
driving force. The flux of salt, however,
is a function only oi the salt concentration.
Raising the applied pressure, therefore,
reduces the salt content of the product.
Ill MEMBRANES AND THEIR MANUFACTURE
A Reverse osmosis was first made practical
by the discovery (by Dr. Locb and co-
workers at UCLA) of a process for making
a cellulose acetate membrane with a very
thin active layer. These membrane are
about 100 microns in total thickness. The
active, dense layer that transmits the
SE.TT.pp. 10.11.67
14-14
-------
50
1
3
0>
O
tn
3
Nornal Osnosis
I
tn
freah ^ waste
water 1 water
0«Botlc Equilibrium
Rererse Oanotij
freah waste
water water
THE PRINCIPLE DF REVERSE OSMOSIS
(WASTE WATER)
Figure 1
-------
[lt»vers»e Osmosis
water and rejects the salt is <>nny 0.25
microns thick - the rrsi is a s>poiifjy back-
up material that has little, li any,
desalination properties.
B A typical formulation for making cellulose
acetate membranes is shown in Table* 1.
TABLE 1
CASTING SOLUTION FOIl CELLULOSE
ACETATE REVERSE OSMOSIS MKMBItANES
Material
Cellulose Acetate
Magnesium Perchlorute
A cetone
Water
\\_oight
22.2
1.1
(iC.7
10.0
This casting solution is poured on a glass
plate and its thickness is adjusted by a
doctor blade. The plate is then immersed
in ice water for 30 to CO minutes during
which essentially all of the magnesium
perchlorate is removed from the membranes.
The membrane it. then immersed in hot
water for 3-10 minutes. The v.atrr tem-
perature is critical - Irom 55O C on up,
the salt rejection properties are improved
at the expense >)f flux through trie mem-
brane. Continuous casting procedures
have been developed but thuy ai c
proprietary and nothing hai, h«_un published
on them.
C The degree of acetjlation ol tin- r.cllulose
is an important factoi. 1 he optimum
membrane is prudm-cd irom c<:llulosr,-
acetate containing about 2.5 ac -Sate groups
per cellulose muleculu. The iii Acetate
is produced and it is hydroly/.ed to the
2.5 level. Further hydrolysis degrades the
membrane - this is controlled in &ervice
by maintaining thi: j>M ul th • lee-i water
at about 5.5.
Compression i,; tin. sporitjy under- layer
at the elevj..'.i pressures necessary
during operation results 11. Jiux decline.
Efforts are underway 10 produce a thin
(0.25 micron) donhu layor that could be
put on a porous, non-i.o1 .iressihlc
substructiii e.
Consicicrable cflorl h.,s Loon and is being
expended on lhf> (k-vi'lopmeni of new
mjmbrane ni.n <.n Ji<> As yet, none o!
tlio.su iiivebti^jted l.a^ |)invun as effectiv.'
ns cullulose nt-titatu. s> mi of the moro
promising m.itc-riiila sliii-iii.d to date
ini.luiU poly (virA l':-ni •.iibonatf)J
polyvmylpyrolliJone .uul graphitic oxide.
A rucynl dovcloj in-'i t in this fielo has been
miule at the Oak Ku.^u National Laliora-
torics of the A lontic f.nei-gy Commission.
Dr. Kraus and co-worl ers have devclopod
a dynamically -lor mul mumbiane having
permeation i atet. of 100-TOO gallons/
day it (gfd) (.oruparod to a typical rate
for cullulosu dcetatc1 ol 2(1- iO gfti.
1
The rejrctni(f Ja>n s aif; formed by
pumping wau-r (.ontjiiung a hydrous
oxiao {^inconiuii'i 01 thorium) or a
polyelectrnlytr | puly (vinylbenzyl-
tr imc-tliy l-.iniinoniuiti t hlorirlc)] over
a jjoroub body.
2 The poious Moily c;vn liavt- a pore size
up to 5 microns in diameter.
3 Salts containing ,ifi[_/wltjnt counter-
ions (Myf++ ni SU ) a?-e poorly rejected
and destroy thu i-'.-jrrlton capacity of
the mcmbiMin.'S lor mono-valent salts.
IV ENGINEERING Ol-
UNITS
It LV EH SK OSMOSIS
Several diflcruiil dobmnt. li,- /o been
proposed *o iiii-f i thi d< -iyn objeclivc of
producing then:. \it ium .unount ol quality
product per unit i.osl
14-16
-------
Reverse Osmosis
B
C
D
1 The "flat plate" membrane configuration
is similar to a plate and frame filter
press. These units have been tested
extensively on brackish- and sea-water
but little work has been done on waste-
waters.
2 The spiral-wound membrane is illustrated
in Figures 2 and 3. The objective of
this design is to increase the surface
area to volume ratio. This configuration
has been evaluated on brackish- and
sea-water and is being evaluated on
wastewater at the FWPCA Pomona.
California pilot plant.
3 The tubular unit is shown in Figure 4.
Another form this can take is that of a
metal tube with weep-holes drilled for
delivery of the product water.
4 The newest configuration consists of
hollow fibers with water passing from
the outside to the insido ol the fibers.
Figure 5 shows a schematic of duch a
unit.
Two differc nt concepts !ia *• iu-en employed
in the design of these uiutb. The* lust
three described above depend on relatively
high permeition rates (2'J Rfd). The
hollow fiber units rely me. rn on very high
surface area per unit vl • PC with low
permeation rates (U. lf> ,»O.
Equipment must be designed to minimize
the "boundary layer" effect - that is,
the existence of a high concentration layer
at the membrane surJace. This problem
is related to the- permea1 ion rate, becoming
more severe as the ratt- increases.
Rapid membrane replacement and
inexpensive pressure construction are
also goals of the design M s, in this field.
Membrane life is stilJ .1.1 unknown quantity,
particularly in wastewater treatment, so
configurations in which membrane replace-
ment is costly could b--- impractical.
LABORATORY INVESTIGATIONS ON
WASTEWATER
Aerojet-General conducted a 1-year
study on the application of reverse osmosis
to wastewater. The equipment consisted
of two 0. 5 ft membranes in series (flat-
plate). The concentrate stream was
returned to the feed tank to determine
the effect of increasing concentration.
The principal findings are listed below.
1 The product quality was excellent.
Table 2 shows the product quality from
two test runs.
2 Water flux through the membrane
decreased rapidly. Acid addition to
the feed to maintain a pH of 5 minimized
this problem but flux reduction with
time was still severe.
3 Aft.'r a period of several hours, the
flux at ifiOO psi feed pressure was
about the same as that obtained at
750 psi.
4 Thi; I lux through "loose" membranes,
though high originally, rapidly declined
to the same level as the "tight" (low
flu:., high salt rejection) membranes.
5 Removal of organics from the feed by
carbon adsorption was helpful in
reducing the rate of flux decline.
b Ttjst results showed that biological
.itt'u'k of *he membrane could be a
seven i problem.
The General Atomic Division of General
Dynamics ran several spiral-wound
modules on wastewater at the Pomona
Water Renovation Plant of the Los Angeles
Couniy Sanitation Districts. This was a
continuous flow-through operation
recovery about 5% of the feed as product.
1 Wat.-r quality was again excellent as
c^n he seen in Table 3. The "A"
moduli's contained "tight" membranes,
the B - "loose" membranes and the "C" -
intermediate.
14-1Z
-------
Reverse Osmosis
ROLL TO
ASSEMBLE
BRINE-SIDE
SEPARATOR .
SCREEN x-->
BRINE FLOW
PRODUCT-WATER FLOW
(AFTER PASSAGE
THROUGH MEMBRANE)
PRODUCT WATER
PRODUCT-WATER-SIDE BACKING^
MATERIAL WITH MEMBRANE ON
EACH SIDE, GLUED AROUND EDGES \
AND TO CENTER TUBE \
MEMBRANE
PRODUCT-WATER-
SIDE BACKING
MATERIAL
MEMBRANE
BRINE-SIDE SPACER
Figure 2. A Spiral-Wound Reverse Osmosis Module
Reprinted with permission from General Dynamics.General Atomic Division
14-18
-------
QMNBRAi. DYNAMIC*
Gonorml Atomic Division
100,000 GALLONS PER DAY WATER TREATMENT PLANT
JO
I
D
I
0
DO
PRETREATMENT
EQUIPMENT
AIR COMPRESSOR
FEED
MAIN PUMP
CONTROL CONSOLE
PRESSURE
VESSEL SKID
PRESSURE
VESSELS
SWITCH
RACK
CONCRETE
PAD
Figure 3
5005-MD-60
-------
Reverse Osmosis
OUT
(1) FIBERGLASS TUBE
(2) OSMOTIC MEMBRANE
(3) END FITTING
(4) P\/C SHROUD
to collect product water
(5) PRODUCT WATER
(6) FEED SOLUTION
(7) EFFLUENT
IN
Figure 4. A Tubular Reverse Osmosis Unit
Reprinted with permission of Havens Industries
2 As bt-ore, the "loose" (high flux-low
rejection) membranes plugged fastest
and were soon delivering no more than
the "tight" membranes.
3 Flux decrease was still a problem,
ranging from 25% in 1500 hours for the
tight membranes to 86% in 290 hours
for the loose membranes.
Similar work was done on a spiral unit by
the New Jersey Department of Health and
others on a sewage plant effluent and on
water from the Hackensack River. Results
were similar. Bacteriological removals
were determined during this study and the
MPN reduction/ range was from 80 to 99.9%.
VI PILOT INVESTIGATION ON WASTEWATER
A The only pilot investigation on wastewater
of reasonable duration has been conducted
at the Pomona pilot plant on a 5000 gpd unit
utilizing the spiral membrane configuration.
B The first year of operation was plagued
by operating difficulties and no encouraging
results were obtained. The more signif-
icant problems are listed below.
1 Some of the feed by-passed the mem-
brane module, going between the outside
of the module and the pressure casing.
This caused a flow decrease through
the feed channel, thus accentuating the
boundary-layer effect.
14-20
-------
n
a
so
1
ID
£>
ro
•d
it
go
n
&
I
a
(B
3
o
i
H
B
CL
n
I
REJECT
FEED
PERMEATE
Figure 5. A Hollow-Fiber Reverse Osmosis Unit
-------
Reverse Osmosis
TMd Water
m
TDS 550
IBS 4.5
COD 95
PH 51
fa* 85
K** 40
IE* 25
Ca** 125
*** 50
Ite** I. ».
a" 65
*>" 2
2100^ 260
GO* 0
S0_ 200
ttQ£ 30
POj (total) 25
TABLE 2
ft H 1 IRHATi AlAlrTSXS
Product
(V»at 18)
15
0.1
2
6.3
4.6
4.9
0
0
0.7
0
20.3
0
8.1
0
1.7
2.9
0
Product
(itat 28)
28
0.1
6.0
5.5
6.1
2.7
3.9
2.6
0
0
22.1
0
3.5
0
3.3
2.9
0
•.ft. - not dctomLud
lpB aaJMto* dviac tut
14-22
-------
Reverse Osmosis
TABLE 3. PRODUCT QUALITY - REVERSE OSMOSIS TREATMENT OF WASTEWATER
Product From
Feed Module 1-6-3
Water* SA1 **
Product from
Module 3-14-1
3B2 **
^Chlorinated and diatomaceous earth filtered secondary effluent.
**Preasure
Temperature
Time in Use
205 psi
700F
1464 hours for 1-^-3 (5A1)
160 hours for 3-24-1 (3B2)
975 hours for 1-22-1 (5C2)
Product From
Module 1-22-1
5C2 **
Alkalinity, ppm CaCOg
Ammonia Nitrogen/: ppm N
Total Nitrogen, ppm N
Specific Conductance, /Jmhos/cm
Chloride, ppm Cl
Hardness Total, ppm CaCO
Phosphate Total, ppm PO
Potassium, ppm K
Sodium, ppm Na
Calcium, ppm CaCOg
Magnesium, ppm CaCO
0
Dissolved Solids, ppm
Sulfate. ppm SO,
Fe, ppm Fe
ABS, ppm
COD, ppm O
PH
213
11.2
12.6
933
92
195
5.6
17.8
107.2
138
58
531
71.4
0.10'
3.8
41
7.2
< 25
2.8
2.8
60
4
3.8
0.2
1.2
7.5
2.8
1.4
28
5.3
0
<0. 1
4
6.6
88
7.0
7.4
399
68
27.4
2.2
10.4
63.0
20.3
7.1
224
2.1
0
1.1
8
7.1
37
3.5
3.5
215
32
6.1
0.4
1.0
18. •
1.8
2.3
118
3.9
0
<0.1
4
8.7
2 The modules tended to "telescope" in
the direction of flow due to pressure
drop through the feed channels.
3 pH control was erratic and the control
set-up was such that the feed pump
would shut down when pH control was
lost. The pressure surge caused by this
sudden loss of pressure accentuated the
telescoping and ripped apart some of
the taped joints holding the module
together.
4 Solids build-up in the feed channels
raised the prmsure drop through the
unit. Calcium sulfate an'd biological
solids were, the principal source* of
trouble.
C From late June until late September of
1967. the unit hs* been operating satis-
factorily. The f»*d water was carbon-
treated secondary effluent rather than
untreated secondary effluent as before.
14-23
-------
KeviTsi1
In addition1, pH control was switched from
sulfunc to hydrochloric acid and 5 ppm
chlorine was added .to the feed. The
results have been most satisfactory, as
listed below.
1 We are operating at a feed rate of About
4800 gallons per day and recovering
83% of the feed as product. The flux
rate is 5 gfd at a pressure of 400 psi
and there has been no significant decline
in flux in 2000 hours of operation.
2 A special anti-telescoping device
installed by General Atomic has been
successful in overcoming this problem.
3 Product quality has remained high as
shown by Table 4. Table 5 shows
gradual degradation of product quality
D
between sla.it-up cind ciflfr 2000 hum s
uii-slrcam. Th«>si increases ari
probably due to slight pin-hole: lujlc.
developing in the niembrnni-.s.
4 Daily flushing with an air-tap water
mixture' has been suc.c<>ss!ul in removing
any particular mate-rial that acuuinu-
latos or forms in the morluli-s.
The study program lor the fului» at
Pomona calls lor 1) another attempt at
feeding secondary effluent instead ol
carbc>n-trealt;d water, 2) using sulfunc
rather than hydrochlutic acid lor pll
control, and 3) increasing the operating
pressure to obtain 90% recovery of feed.
Table 4 WATER QUALITY DATA FROM GENLRAI, ATOMIC KTVI I(SI OSnlOSlS UNIT
Tulie No
Feed
1A
IB
2C
3C
4C
5C
6C
7C
8C
Total Product
Brine
% Reduction
PC
Ave
30
0
0
1
1
0
0
0
0
, 0
0
177
'J8
9
22
16
4
1
43
48
41
'17
77
S7
2
>4 (me 1)
Ranee
21 6 - 37
09-
05-
1 3 - 1
.07- 3
07- 1
08-
08- 1
04- 1
07- 1
13- 1
7'i -240
COD (mil 1) NHi
8
5J
15
5
4
I
<)0
0
H
H
2
Ave
10 B
1 2
1 1
2 4
1 6
1 8
2 0
1 6
2 .i
i 4
1 7
43 8
84
Range
6 7
0 0
0 0
0 4
0 0
0 G
0 0
0 0
0 0
0 0
n o
2b 6
-14 4
- 2 3
2 6
- 6 2
- t 6
- 4 0
-50
- T 6
- b 1
- •) 0
- '< 4
-bO 4
Ave
•J
1
1
2
1
1
1
1
2.
3
1
U4
112
t
3
2
H
H
7
7
8
6
•i
7
-N (mu
1)
Raiiec-
2 S
0 6
0 f>
0 7
1 0
0 'J
1 1
1 0
1 i
1 8
0 y
11
-'2B
- 2
- 2
- 7
- 2
- 2
- ••
- A
- 4
- -i
1
-240
0
0
2
J
6
6
6
1
6
t
4
N0r
Avu
2 4
0 4
0 5
0 ')
0 7
0 'J
0 ')
1 0
1 0
1 fa
0 II
7 ->
67
\ (niji 1)
If an£
0 !' -
0 (1 -
0 II -
0 4 •
0 (I -
U 4 -
0 4 -
07-
n ', -
II li -
Cl •! -
t H -1
[<-•
S (1
0 »
0 H
1 7
1 b
1 b
1 .
1 b
i 4
2 b
1 2
1 i 7
Avf
021
•n •.
S
Ran^c
-,!0 -
0
'1 -
-
11 -
21 -
.
I'M,
il -i
I...I-I
14-24'-
-------
Reverse Osmosis
TA BLE 5. PRODUCT WATER CONDUCTIVITY -
GENERAL ATOMIC REVBRSE OSMOSIS UNIT
Tube No.
1A
IB
2C
3C
4C
5C
6C
7C
8C
Initial
Cond. (iimhos/cm)
50
50
100*
70
70
75
85
90
110
Present
Cond. (iimhos/cm)
80
80
140
200
110
110
120
220
250
Feed Conductivity: Before acidification: 800
After acidification: 950
Brine Conductivity: 4000
*High due to use of several modules with slight leaks
VII COSTS
Reverse Osmosis is in an early stage of
development and, as could be expected in
this stage, there is a wide range on
estimated costs. Different manufacturers
have estimated costs for a 10 mgd plant
ranging from $0. 25 to over $1. OO/ 1000
gallons. Some of the factors that will
determine the cost of a reverse osmosis
system for wastewater renovation are
listed below:
1 Flux through the membranes is the
prime factor affecting costs. Most
cost estimates have been based on an
optimistic flux of 20-30 gfd.
2 Membrane life is also an important
cost consideration. If the membrane
life is short, the reverse osmosis unit
will have to be designed so that mem-
brane replacement is rapid and
inexpensive. This is an approach
being taken by some manufacturers.
The degree of pro-treatment necessary
prior to entering the-reverse osmosis
cell will influence application of the
process. If orfanic materials .have to
be removed before the membranes.
then reverse osmosis is simply a
demineralization process and must
compete economically with electro-
dialysis and ion exchange.
The product to waste ratio obtainable
with reverse osmosis Is an important
consideration. If 90% of the feed is
recovered as product, the cost of
disposing of the concentrate stream by
injection into deep wells (providing the
geology is appropriate for this) would
be about 2$/1000 gal. of water produced.
If only 80% of the feed can be recovered,
the cost of disposal will be 5-6C/1000
gallons.
The production rate of membrane
materials will affect the cost of the
14*25
-------
Reverse Osmosis
operation. The raw material costs are
almost nc»gU^:ble b.it the manufacturing
costs are IU^P. M-.mbrane consumption
must be high enough to support continuous.
automated production facilities.
Despite all these considerations, reverse
osmosis could hi vc- widespread application
in the future in ". jcutions where demineral-
ization of waste*at»r is required for
reuse or pollution Control purposes. It
must be remembered that reverse osmosis
is still an "infa.u ' when compared to
processes such : 3 Jisrillation and carbon
adsorption. The fidld is wide open and
significant devt-ljpments are occurring
rapidly.
VIII FUTURE PLANS
A The FWPCA ha =3 plans for continuing
support of deve -.prriert work in the reverse
osmosis field, both in testing of new
membrane mate-, lals and in improvements
in engineering
P. A major undertaking under consideration
is the funding oJ work to determine the
minimum pretreatment necessary to
prepare a raw sewage for treatment by
reverse osmosis. It seems most unlikely
that conventional primary and secondary
treatment provide the ideal conditioning
of the wastewater for the membrane
process.
C A major study funded jointly by the FWPCA
and the Eastern Mun-?ipal Water District
of Hemet, California is just getting under-
way. Starting with secondary effluent.
several different tyr, es of reverse osmosis
units will be run in parallel. Pretreat-
ment of the secondary effluent by any
combination of clarifj nation, filtration
and carbon adsorption will be possible.
-X INDUSTRIA L APPLICATIONS OF REVERSE
OSMOSIS
A The Pulp Manufacturers Research League
in Appleton. Wisconsin has investigated
reverse osmosis ap .. method for solving
both their water pollution and water
supply problems. They hope to operate
a 50. 000 gpd plant that could be evaluated
on several waste streams at a number of
pulp mills. Laboratory investigations
have been limited to short-term batch
runs but the results appear promising.
B The Aerojet-General Corporation has
received a grant for a study of the appli-
cation of reverse osmosis to various
industrial waste streams at the Odessa,
Texas petrochemical complex. A 560
square foot unit will oe mounted on a
trailer that will also include a clarifier
and diatomaceous earth filter to serve as
pretreatment units. The objectives of
the study will be to investigate the follow-
ing factors:
1 Reliability of reverse osmosis as an
industrial waste treatment process.
2 Sensitivity of the membrane performance
to input water quality.
3 Membrane life.
4 Design data for large-scale units. ,
5 Economics of the process
Table 6 presents characteristics of some
of the waste streams that will be fed to
the unit.
REFERENCES
1 "Desalination by Reverse Osmosis. "
Ulrich Merten, editor. The M.I. T.
Press. Cambridge, Massachusetts,
1966.
2 Wilford, J. and Perkins. F. K. "Test of
G.A. Reverse Osmosis Unit in
New Jersey, 1965." New Jersey State
Department of Health and New Jersey
Department of Conservation and
Economic Development. January, 1966.
3 Loeh. S. and Johnson, J.S. "Fouling
Problems Encountered in a Reverse
Osmosis Desalination Pilot Plant. "
14-26
-------
INS
TABLE 6. ODESSA INDUSTRIAL COMPLEX
WASTE WATER STREAMS
Stream
Total Hardness
Ca
Mg
Fe
NH
3
Total Dissolved Solids
pH
r
(COD)
Temperature. F
ABC
80-200 10-100 0-30
(130) (40) (15)
21 25 0-8
6 3 0-2
2.5 1 3-5
(4)
2-15 1-15 40-125
(60)
2250-2500 700-1150 4400
8.5-9.5 8.5-11.5 9-12
(9.6) (10)
2300 770 2440-3715
100-120 96-130 110-135
(110)
D
130-200
(150)
45
3
7
0-50
(10)
10,800
6.5-7.5
140
68-85
E F
150-320 0-625
(200)
109
8
2 1
0-13 4
(3)
1550-1810 2300
6.5-8.0 8.5-12
(7.3)
45-80 1500-7500
70-95 75-110
(85)
Combined
Waste
60-1700
(ISO)
20-150
2-30
<1.0-2.5
1-25
(10)
3. 500-25. 000
(5000)
8.5-10.0
(9.5)
650-800
85-110
(90)
90
•
H
•
n
0
B
3
3
*
-------
Reverse Osmosis
Preprint 21A - Presented at the
Symposium on Desalination: Part II.
Sixtieth National Meeting, American
Institute of Chemical Engineers.
September 1966.
4 Marcinkowsky, A.E. et. al. Hyperfiltration
Studies - IV. Salt Rejection by
Dynamically-Formed Hydrous Oxide
Membranes, Journal American
Chemical Society, 88, 5744-46. 1966.
5 Kraus. K.A. et. al. "Hyperfiltration
Studies - VI. Salt Rejection by
Dynamically-Formed Polyelectrolyte
Membranes." Desalination, 1, 225-230,
1966.
6 Bray. D. T. et. al. "Reverse Osmosis
for Water Reclamation, " Report No.
GA-6337. General Atomic Division,
General Dynamics Corporation, San
Diego, California.
7 Okey, R.W. and Stavenger, P. L.
"Industrial Waste Treatment with
Ultrafiltration Processes." Dorr-
Oliver, Incorporated, Stamford, Conn.
8 Wiley, A.J. et. al. "Application of
Reverse Osmosis to Processing of
Spent Liquors from the Pulp and Paper
Industry" Tappi, 50, 9, 455-60.
September 1967.
This outline was prepared by Arthur N. Masse,
Chemical Engineer, Project Analysis
Activities, Division of Research, Cincinnati
Water Research Laboratory, FWPCA, SEC.
14-28
-------
ION EXCHANGE
I INTRODUCTION
Ion exchange is a well known method for
softening water and foi producing deionized
water. It may also be practical for removing
the inorganic materials added to water during
use.' Although ion exchange equipment is
very different in form from electrodialysis
equipment, the two ptocesses have many
similarities. Both have application over the
same range of feed concentrations and have
a number of similar operating problems. Ion
exchange has greater capability for selective
removal of particular ions down to very low
concentrations. It is also capable of total
demineralization. Chemical regeneration of
ion exchange resins is less desirable than the
continuous electrical regeneration that occurs
in electrodialysis.
II PRINCIPLES
A Mechanism of Ion Exchange
Ion exchange materials are available as
solids and liquids. Since the bulk of ion
exchange applications presently use the
solid ion exchange materials, these will
be discussed here. Solid ion exchangers
are adsorbents that can he thought of as
having many ionic sites in their structure.
To maintain electroneutrahty each ionic
site has associated with it an ion of opposite
charge or counter ion. If the ion exchange
material is contacted with a solution of
ions different from the original counter
ions, the latter can be replaced. The
replacement of ions of one type by another
is ion exchange. When all the original
counter ions hnve been replaced the resin
is exhausted. A concentrated solution of
the original counter ion must be used to
regenerate the ion exchanger.
Many kinds of m.itenals, both synthetic
anu naturally occurring, have the capability
to exchange ions. Synthetic organic resins,
however, have replaced nalui al materials
in some applications and will probably
continue to do so in the future. The
synthetic resins can be tailored better to
meet the requirements of a specific
application. There are four major types
of ion exchange resins. These are strong
and weak acid and strong and weak base.
The names Jicid and base come from the
fact that the counter ions associated with
those resins may be hydrogen and hydroxyl
respectively. Other names for acid and
base resins are cation and anion resins.
For many applications these last terms
are the more appropriate.
Strong acid resins are commonly composed
of sulfonated styrene and divinyl benzene.
In the hydrogen form they can be used to
remove other cations from solution. This
is called salt splitting and is shown in the
following equation:
RSO- H + Na Cl
RSO Na
HC1
Regeneration would be the reverse of this
reaction, probably using sulfuric acid.
Acid in excess of the stoichiometric amount
is required for good regeneration.
Softening can also be carried out with these
resins as follows:
2RSO,, Na + Ca
+ 2 Na
After softening, sodium chloride is usually
used for regeneration.
Weak acid resins often have carboxyl
groups in their structure. These resins in
the hydrogen form will not split salts, but
will remove cations from solutions of weak
bases as shown:
RCOOH
RCOO~Na+ + HO + CO
Ci ft
In the hydrogen form these resins are not
ionized. Less excess regenerant is re-
quired for these resins than for the strong
acid resins.
Strong base resins have quaternary
ammonium groups in their structure. They
will split salts by exchanging hydroxyl ion
for other anions as follows
RN(CH3)3OH"+ NaCl * RN(CH,
Na OH
SU.TT.pp. 11. 11.67
14-29
-------
Ion Exchange
This is not a common use, however, since
salt splitting can more cheaply be carried
out with strong acid resins. These resins
are useful for removing the anions of weak
acids formed after cation exchange.
Regeneration is carried out with a strong
base. Considerable excess regenerant is
required.
Weak base resins contain various amine
groups. In the base form these are not
ionic and do not remove anions by ion
exchange. They react with acids in
solution, howe/er, and in this way accom-
plish anion removal. A common use is in
removal of the acids formed after cation
exchange:
RNH, + HC1 —* RNH^Cl"
Z <3
These resins will not remove most weak
acids; a strong base resin after the weak
base resin is required. Weak base resins
can be regenerated with weak bases such
as ammonia. Only a slight excess is re-
quired.
For an excellent discussion of resin types
and detailed consideration of ion exchange
rates and equilibria the reader is referred
to the book by Helfferich (1).
B Conventional Method of Application
Most ion exchange operations use fixed
columns of resin granules or beads.
Usually the solution to be treated is fed
down-flow through the column. As opera-
tion continues the resin becomes exhausted.
There is no sharp distinction, however,
between fresh and exhausted resin. A
schematic diagram of the degree of
exhaustion along the column is shown in
Figure 1 for exchange of sodium ion for
hydrogen ion. It is assumed the resin was
initially all in the hydrogen form. Before
all the hydrogen ion is exchanged there
will be a breakthrough of sodium ion.
Regeneration must then be carried out.
The resin is first backwashed to remove
suspended material that may have collected.
The exhaustion profile is disturbed by the
backwash because of fluidization and mixing
of the bed. Regeneration is then begun.
usually in down-flow. If sufficient acid
were used the resin could be restored
completely to the hydrogen form. This
is not possible, however, without using
an impractical excess of regenerant.
Ordinarily the resin would be regenerated
until nearly all the resin ax the top of the
column was m the hydrogen form. Further
down the column there would be an increasing
fraction of the sodium form that leads to
leakage of sodium ion when ion exchange
is resumed. The amount of excess
regenerant necessary depends upon the
amount of leakage that can be tolerated.
After regeneration the column is rinsed
for thorough removal of regenerant from
the interstitial liquid before being put
back into use.
C Countercurrent Regeneration and
Continuous Operation
Although the conventional method of opera-
tion can be made to give good results, it
is not particularly economical. Regenerant
requirements can be reduced substantially
if countercurrent regeneration can be used.
Consider, for example, what would happen
in Figure 1 if regenerant were added in
the opposite direction from the feed. The
least exhausted resin would contact the
most concentrated regenerant giving a
greater concentration driving force than
with cocurrent regeneration. As the
regenerant continues through the bed it
contacts resin that is more nearly exhausted.
Although the regunerant concentration
driving force at the feed end of the column
is less than in thu case of cocurrent
regeneration, the driving force for the
column as a whole is greater. The most
highly regenerated resin occurs at the
outlet of the column rather than the
entrance as in cocurrent regeneration.
This resin acts as a polish to reduce
leakage of unwanted ions. Conventional
operation would require excess regenerant
to accomplish the same degree of leakage
control.
If carried out in a single column, counter-
current regeneration requires that feed
and regenerant enter the column at opposite
14-30
-------
Ion Kxc!iaii"(
INITIAL CONDITION
1.0
UJ
Ot
PARTIAL EXHAUSTION
FLOW DIRECTION
NoCI SOLUTION
DISTANCE "ALONG COLUMN
I' igure 1. Effect of passing NaCl Solution through
Cation Resin in the Hydrogen Form.
i-nds. By breaking the column into two
or more sections in series it is possible
to get so.nr measure of countercurrent
action even il downflow is used for both
streams. It would be necessary, however,
to regenerate the last section of column
first and reuse the partially spend regen -
erant in each preceding column.
In some respects the most efficient type
of operation is that obtained from continuous
countercurrent treatment. In such a
system, resin and raw water move
continuously in opposite directions through
a column with exhausted resin existing at
the water entrance and regenerated resin
entering at the water exit. The cxhau.-it'-d
i i.-sin continuously enters a regeneration
chamber where it is couiitercurrontly
contacted with regeneranl. This system
iiiakes best use of resin capacity, gives
minimum bed size and minimum regenernnt
requirements. Equipment costs are,
however, higher than for I'ixed bed con-
tractors. Mechanical problems can arise,
in the? resin-moving system. Several
nearly continuous systems are in use, one
of these being the Higgins Contactor (2).
D Mixed Bed Ion Exchange
When a cation exchanger in the hydrogen
form is used to remove mineral cations
from solution, there is a gradual buildup
of hydrogen ion in the water down through
the column. This tends to regenerate the
resin and slows down the remov.il of the
mineral cations. In the case an anion
resin in the hydroxyl form the solution
tends to become alkaline and again there
is a slowing of exchange rate. If particles
of the two types of resin arc well mixed.
14-31
-------
Ion Exchange
however, demineralization will take place
without significant pH shifts. The mixed
bed type of operation is quite effective when
highly deminerahzed water is required.
There is a problem in regeneration, how-
ever, since the two types of resin must
first be separated. By making the resins
of different density or particle size
separation can be achieved by backwashing
at a high enough rate to fluidize the bed.
Ill PARTICULAR ION EXCHANGE SYSTEMS
A Strong Acid - Weak Base System
A simple system for good removal of cations
and most anions is treatment by a strong
acid resin in the hydrogen form followed
by a weak base resin in the free base form.
Cations are effectively removed giving an
acid solution of the anions. Anions forming
strong acids are then removed on the weak
base resin. Silicate and bicarbonate are
not removed. Bicarbonate can, be removed.
however, by air stripping. At the low pH
it is removed as carbon dioxide. If it is
necessary to remove silica, a strong
base resin must be added at the end of the
system. The cation resin is regenerated
by a strong acid and the anion resin by
either a weak or strong base.
The regenerant requirements depend upon
the ion concentration to be removed from
the feed water. In the usual application
of this system the cost of regenerants
would be rather high. A method of recover-
ing the chemicals for reuse might result
in a sizable saving. A system using
nitric acid and ammonia has been proposed
for carrying out recovery. After regen-
eration there would result a waste solution
of nitrate salts from the cation exchange
column and a solution of ammonium salts
from the anion exchange column. By
evaporation of the nitrate solution and
steam calcination of the residue the nitric
acid would be recovered. The mixture of
hydroxides left after calcination would
then be used in a lime still to recover
ammonia. The degree of recovery of
nitric acid is important to process
economics.
Investigation has shown that nitrate salts
of calcium, magnesium; aluminum, and
iron calcine to give a good yield of acid.
Sodium and potassium nitrates jdo not
calcine well. The economics of the
process are not well defined, but would
depend upon the nature of the feed water.
B DESAL Process
Although regenerant recovery would
appear to offer the lowest eventual ion
exchange costs, systems that use a
minimum of low cost regenerant would
represent a substantial savings over the
usual methods of ion exchange operation.
A system that does use a minimum of
regenerants has been proposed by the
Rohm and Haas Company (3). This system
is shown in Figure 2. It uses weak acid
and base resins that can be regenerated
with almost stoichoimetric amounts of
chemicals. It can be used to treat waters
of widely varying composition. Feed
water enters a column of anion resin,
IRA-68. in the bicarbonate form where
anions other than bicarbonate are exchanged.
The bicarbonate solution then enters a
column of acid resin, IRC-84. in the
hydrogen form where cations are exchanged.
Because the solution is alkaline this
exchange will take place. The carbon
dioxide solution that results enters another
anion exchange column in the free base
form. The carbon dioxide converts this
column to the bicarbonate form. When
the resins are exhausted the first column
in the series is regenerated to the free
base form with .\mmonium hydroxide
solution and the cation resin is regenerated
to the hydrogen form. If the direction of
flow is reversed the exhaustion cycle can
be repeated since the third column is
already in the bicarbonate form. The
ammonia regenerant can be recovered by
reacting with lime if desired. Columns
in the bicarbonate form must be kept under
a pressure of several atmospheres.
The resin IRA-68 has a high capacity for
organic materials. A sample of the resin
has been used to treat sand filtered
secondary effluent through a number of
exhaustion cycles. Roughly 50 percent
14-32
-------
Ion Hxchange
,NaCI
n
WEAK
BASE
RESIN
HCOg
FORM
WE
AC
RES
H
FOI
1
WEAK
BASE
RESIN
OH'-
FORM
NaHCC>j
CO2
Figure 2. DESAL Process
14-33
-------
Ion Exchange
of the organics measured by organic carbon
is removed. Although the long term effect
of the organics can only be determined
through extended use, available results
indicate that the organic contaminants are
effectively removed during regeneration.
The ability,not only to resist fouling by
organics, but to adsorb a significant
fraction of them is an example of the
advances that have been made in resin
technology.
IV OPERATING PROBLEMS
A number of problems common to electro-
dialysis also occur in ion exchange.
Fouling of an ion resins by organics and
some inorganics is well known. Some of
the newer resins are designed to over-
come fouling. These may be used for
removal of traces of organics from water.
Generally, however, it is considered
desirable to have the feed water as free
from suspended matter and soluble organics
as possible to lengthen resin life. Scaling
may occur upon regeneration of cation
resins with sulfuric acid when there are
calcium ions on the resin ion. If too
strong an acid solution is used, c alcium
sulfate will precipitate. Precipitation
does require enough tintu that it '.an be
avoided if regeneration is carried out
properly.
V APPLICATION TO WASTEWATERS
A Municipal
There are two applications to municipal
wastewater that may be useful. One
application is the removal of the increment
of minerals added during u-,f. This
could be carried out b> completely
demmeralizing a fraction ol the water.
often about 40 percent, and mixing the
demuie rah zed watpr with untreated
water. It might also b-j i arried out by
partial demmeralizntion ot the whole
stream. The latter method would be most
appropriate with continuous counter current
equipment. Obviously both cation and
anion exchange would be required.
Pretreatment would probably be deter-
mined primarily by the use to be made
of the treated water. With the older types
of resin clarification and some degree of
organic removal would probably be nec-
essary. Less pretreatment should be
required for the newer low-fouling resins.
The second use is for selective removal
of certain ions, especially ammonium,
nitrate, and phosphate. Removal of these
nutrient materials is becoming increasingly
necessary for control of algae growth in
lakes and streams. Eliassen and co-
workers (4) have investigated the use of
anion exchange with sodium chloride
regeneration for anion nutrient removal.
This is not selective ion exchange in a
strict sense since all anions except
chloride would tend to be removed. In
this work sand filtered secondary effluent
was fed to the ion exchanger. Removals
of phosphorus and nitrate were 84 and 77
percent respectively.
Little work has been done on truly selective
removal of these materials, but a study
is underway on ammonium ion removal.
Certain natural zeolites have a high
affinity for ammonium compared to most
other cations in wastewater. An attempt
will be made to develop a workable process
from one of the zeolites. Hopefully,
regeneration of the ion exchanger with
lime will be possible. No highly selective
material for nitrate removal is known.
There is no reason, however, to believe
that such a resin cannot be prepared. It
may be of considerable value in the future.
Phosphate is generally preferentially
removed by anion exchange, but the
situation with this material is not as
critical as for the other nutrients because
of the availability of other proven removal
methods.
B Industrial
The use of ion exchange in industrial
wastewater treatment has been very
limited. With the growing need for
industrial pollution control, applications
14-34
-------
Ion I7xi.
VI
may increase. Although electrodialysis
and other processes could be strong
competitors, there are certain instances
where ion exchange should'be most suited.
These are situations'in which small
amounts of valuable or very harmful
malt-rials must be removed to very low
levels. These may be the only contami-
nants present or there may be a mixture.
The amount of regenerant required should
be small. If selective removal from a
mixture is required, ion exchange, offers
the most promise through development
of selective resins or appropriate regen-
eration procedures. In the metal plating
industry there are a number of instances
in which ion exchange is presently used to
remove small amounts of heavy metals
from wastewaters. Examples are treat-
ment of the rinse waters from copper,
nickel, and chromium plating. At the
same time as metal ions are recovered
that would otherwise be lost, the quality
of the water is improved. Often the
product water is reused in the plating
system. A rattier recent application for
ion exchange treatment of wastewater is fop
removal of radioactive ions from nuclear
wastes. The method can be made to give
•A high degree of decontamination.
Although there are no known instances in
which ion exchange is used for the general
removal of an increment of total dissolved
solids from industrial wastes as would be
required for treatment of domestic wastes,
such treatment may be necessary in some
locations in the future. A study of each
particular case will be required to deter-
mine whether ion exchange is the most
economical treatment method.
ULTIMATE DISPOSAL OF CONCENTRATED
WASTES
Ion exchange produces a significant amount
of waste in the form ol backwash water, spent
reg«.i»ierant, and rinse water. In most
applications of the process little thought is
given to minimizing the volume ol waste.
Where pollution control is of primary concern
evoiything possible must be u'one to decrease
this volume. The minimum volume of waste
that might result from ,\ ,ir«pt ••!} rU-.,i f
and opt-r.Hfd u.isLi « jtrr Ir-Miinrjiil system
cannot lx- u< cui aicly estim..leu becjus* <>i
lack of data. IJsniy mdmifacliu ore1
recommend.; I ions thi aim unit is quiti largi-.
For a coini'tctily used en lion ''usin tin1 „
manuiactni vr suggests a I least CiO gal/ft
bed croKE-st-«.tifi.'i,.l area lur li;;rkwnsh. It
the ion i xi han:gn &j, stum is j»i redvd b} a
solids-removal treatment, the backwash could
be recycled hack through (hat, treatment'.
Using the usual concf tit rations of acid regi-n -
erant, the spent repenei a-.t volum« would li
at least one percent of the fei-il for i minei,-!
concentration lypii.nl ot municipal wnstt-wdl. ••.
Naturally, more conroi-truted waters would
require more reeenerant. 1-or i egeneranl
rinse 150 gal/ft rcjsiii i^ rfuggested. Again
the volume of rinse dt.-fx.nJs upon the watd
concentralion sin^f this ilfeets the resin
volume. Foi a mvimcipal wastewater the
rinse water would bu 5 to JO percent of the
feed. Obviously, lor i system ol cation an'
anion exchangers the volumo of wasie woulu
be lar'gtj. Koduuuoi! would nnc/oubh dly bu
possible il rcsti ictions wi-re plat-cd upon
waste disposal.
VII OPERATING COSTS
In the past ion exch^n^e h.ib not been n clienp
process to operate. Costs wcru often in the
same range as tlistulaiiou, more than
$1.00/1000 gal. Tor wastewater treatment
in many casc-s sncl, o cost rloc-s not appear
competitivi' with thi. jirnjecl-1'! rosts ol olhe.-
processoh that ,n-i lining developed. Tliroui>h
improved resm.s, improved design of cqui[>
ment, and possihly reyc.-:ier.-ril r«-fovcry it
is believed, now vi-r. Mint ilir i'ost ol ln«-
proccss can \»c; • ulisMniially induced. \Voi li
on nuti lent ri-mov;il hv union exchange, for
example, re-nill"d n- an estimated cost oi
17-19^/1000 (i-tJ. Mxpi-rimental work presenily
being conducted should provide further cost
information, f-'oi srJi-cMive removal ol a
specific ion, ion .^cli-ir-...- jicm-rnllj has nn
cOnnpetitors. In ihesf crises t t.-covery of a
valuable malarial is usunlly possible.
Relatively high treairn^ni cost may, therefore,
be tolerable.
14-35
-------
Ion Exchange
REFERENCES 4 Eliassen, R.. Bennett. G.E., "Anion
Exchange and Filtration Techniques
1 Helfferich, F.. "ion Exchange. " McGraw- for Wastewater Renovation", presented
Hill, New York, 1962. 'at the 39th Annual Conference of the '
Water Pollution Control Federation.
2 Higgins. I.R.. Ind. Eng. Chem. 53, 635 Kansas City. Missouri, Sept. 1966.
(1961).
This outline was prepared by Dr. Carl A.
3 See "AMBER-HI-LITES", published bi- Brunner. Chemical Engineer, Project
monthly by Rohm and Haas Company. Analysis Activities. Division of Research,
Philadelphia, Pennsylvania. Cincinnati Water Research Laboratory.
FWPCA. SEC.
14-36
------- |