-------
-------
EPA-600/8-77-005
SE2
^^^^^^
\
UJ
Manual of
Treatment Techniques
for Meeting the Interim
Primary Drinking Water
Regulations
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
WATER SUPPLY RESEARCH DIVISION
Cincinnati, Ohio
First Printing May 1977
Revised April 1978
-------
NOTICE. The mention of trade names of commercial products in this publication
is for illustration purposes, and does not constitute endorsement or recommenda-
tion for use by the U.S. Environmental Protection Agency.
-------
Contents
Page
Introduction 1
Treatment Techniques for the Removal of Inorganic Contaminants from Drinking
Water 2
Arsenic (As) 7
As+5 Removal 7
As+3 Removal 7
Arsenic Removal (General) 10
References 10
Barium (Ba) 11
Removal 11
References 11
Cadmium (Cd) 13
Removal 13
References 14
Chromium (Cr) 16
Cr+3 Removal 16
Cr+6 Removal 17
References 18
Fluoride (F) 20
Removal 20
References 21
Lead (Pb) 21
Removal 21
References 21
Mercury (Hg) 24
Inorganic Mercury Removal 24
Organic Mercury Removal 24
References 25
in
-------
Nitrate (NO3) 28
Removal 28
References 29
Selenium (Se) 29
Se+4 Removal 29
Se+6 Removal 29
References 30
Silver (Ag) 32
Removal 32
References 32
Treatment Costs for the Removal of Inorganic Contaminants 34
Cost of Modifications or Operational Changes at Existing Treatment
Plants ' 34
Cost of New Treatment Facilities 34
References 36
Treatment Techniques for the Removal of Turbidity from Drinking Water 37
Treatment Techniques 37
Granular Media Filtration 38
Diatomaceous Earth Filtration 38
Disposal of Filter Plant Sludge 39
Treatment Costs for Turbidity Removal 39
References 40
Treatment Techniques for the Removal of Coliform Organisms from Drinking
Water 44
Maximum Contaminant Levels (MCL's) 44
Disinfection of Water 45
Turbidity 45
Disinfection Byproducts 45
Chlorination 45
Ozone 45
Chlorine Dioxide 50
Cost of Water Disinfection 57
Summary 51
References 52
IV
-------
Treatment Techniques for the Removal of Organic Contaminants from Drinking Water 53
Occurrence of Pesticides in Water Supplies 53
Endrin 53
Lindane 54
Toxaphene 55
2,4-D 55
2,4,5-TP (Silvex) 56
Methoxychlor 56
Summary of Treatment Techniques 56
Estimating Cost for Reducing Trace Organics (Pesticides) Below MCL 59
Adsorption With PAC 59
Adsorption With GAC 59
References 59
Bibliography-Occurrence and Fate of Pesticides in the Environment 61
Treatment Techniques for the Removal of Radioactive Contaminants from Drinking
Water 62
Alpha Emitters 62
Maximum Contaminant Levels 62
Radium in Water Supplies 63
Removal of Radium from Water 63
Lime or Lime-Soda Softening 63
Ion Exchange Softening 63
Reverse Osmosis 63
Disposal of Treatment Waste 65
Methods for Lime Sludge Disposal 65
Methods for Lime Softening Backwash Disposal 66
Ion Exchange Brine Disposal 66
Disposal of Reverse Osmosis Wastes 66
Treatment and Disposal Costs to Remove Alpha Emitters 67
Manmade Radionuclides, or Beta and Photon Emitters 67
Maximum Contaminant Levels 67
Monitoring for Beta and Photon Emitters , 71
Removal of Manmade Radioactivity 71
Glossary 72
References 72
-------
-------
Introduction
Following the passage on December 16, 1974, of
Public Law 93-523, the Safe Drinking Water Act, the
National Interim Primary Drinking Water Regulations
were promulgated on December 24, 1975, to take
effect June 24,1977. These regulations set maximum
contaminant levels (MCL's) for 1) 10 inorganic
constituents, 2} turbidity, 3} coliform organisms, 4) 6
pesticides, and 5) radionuclides.
Public Law 93-523 stated that the Primary Drink-
ing Water Regulations should consist of MCL's and
identify treatment technology that could be used to
achieve them. This document provides the latter
information in five sections as related to the fore-
going five groups of Interim Primary Drinking Water
Regulations. It is based on the literature and the
research being conducted by the Water Supply
Research Division, and is not meant to stifle inno-
vative treatment technology. It attempts to state
technology known at the date of effectiveness of the
Interim Primary Drinking Water Regulations that will
allow utilities, with assistance from their consulting
engineers, to apply whatever treatment might be
necessary to improve their drinking water quality
such that it meets the Interim Primary Drinking
Water Regulations.
One difficulty encountered in preparing this docu-
ment was the lack of information on treatment
technology applicable to the small water utilities
serving 1,000 consumers or less. Research is now
underway in an attempt to fill that void; because the
research has not been completed, this document
does not contain the information. Cost data were
another difficulty. It is impossible to prepare treat-
ment cost information that is universally applicable
to all utilities. The authors, therefore, recognize that
the costs contained in this document may not apply
to all situations.
The authors hope that this document will be
helpful to consulting engineers and to water utilities.
They anticipate, however, that it will need to be
updated as new information on treatment technology
becomes available through research and development
and experience at the many treatment plants now in
operation.
A list of references concerning each contaminant
follows the relevant discussion.
-------
Treatment
Techniques
for the Removal
of Inorganic
Contaminants
from
Drinking Water
by THOMAS J. SORG
Water Supply Research Laboratory, MERL
U.S. Environmental Protection Agency
Cincinnati, Ohio
The National Interim Primary Drin king Water
Regulations established maximum contaminant levels
(MCL's) for 10 inorganic chemicals: arsenic, barium,
cadmium, chromium, fluoride, lead, mercury, nitrate,
selenium, and silver (tables 1 and 2). Except for
nitrate, the MCL's for all of these inorganic chemicals
are applicable to community water supply systems.
The nitrate level applies to both community and
noncommunity systems.1
Most of the treatment information and data
available for the removal of inorganic chemicals are
on conventional coagulation or lime softening treat-
ment. These treatment methods are commonly used
by large water supplies. As stated in the Introduction,
research is underway to fill the need for information
on treatment technology applicable to small water
systems.
There have been many studies on the removal of
some of the inorganic contaminants from municipal
and industrial wastewater, but few have been con-
ducted on the removal of these contaminants from
drinking water sources that would generally contain
much lower concentrations. A review of the literature
indicates that much of the available information on
drinking water is the result of laboratory and pilot
plant studies conducted by the Water Supply Re-
search Division, U.S. Environmental Protection
Agency, in Cincinnati, Ohio. The EPA research pro-
gram centered on conventional coagulation and lime
softening treatment methods. Only when these meth-
ods were found to be ineffective were other methods
studied, such as reverse osmosis and ion exchange.
The studies by EPA and others have shown that no
National Interim Primary Drinking Water Regula-
tions for definitions of community and noncommunity
systems.
one treatment technique is effective for ail contam-
inants. A summary of the best treatment methods for
the inorganic contaminants is presented in table 3.
Most of the methods listed are conventional coagu-
lation and lime softening. Other treatment tech-
niques—such as ion exchange, reverse osmosis, distilla-
tion, and electrodialysis—may be equally effective.
These methods are generally more expensive and,
except for ion exchange, they are not commonly used
for treating drinking water. They may have practical
applications in special cases, however, particularly for
small communities, and should not be ruled out
entirely. Because data are lacking on their effective-
ness to remove certain contaminants from drinking
water, these systems are not discussed in detail.
The studies on conventional coagulation treatment
and lime softening showed that removal results
frequently depend on the pH of the treated water,
the type and dose of the coagulant, and initial
concentration of the contaminant. Of these variables,
the most important was found to be the pH of the
treated water (figs. 1, 2, and 3). This finding is logical
because the solubility limits for metal hydroxides,
carbonates, and so forth, are normally pH dependent.
Another important factor in the removal of a
contaminant is its valence. Several contaminants, such
as arsenic, chromium, and selenium, may be found in
water in more than one valence state. Mercury may
be found in either the organic or inorganic form.
Studies of these substances have shown significant
differences in removals between forms of the contam-
inants. For example, the oxidized state of arsenic
(As+5) is easily removed by conventional coagulation
treatment, whereas the reduced state (As"1"3} is not
(figs. 1, 2, and 3). The valence of the contaminant is
an important consideration, therefore, in selecting the
proper treatment technique.
-------
TABLE 1. National Interim Primary Drinking
Water Regulations: Maximum Contaminant
Levels (MCL's) for Inorganic Contaminants
Except Fluoride3
Contaminant MCL, mg/l
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Nitrate (as N)
Selenium
Silver
0.05
1.
0.010
0.05
0.05
0.002
10
0.01
0.05
aThe MCL for fluoride is determined by the annual
average of the maximum daily air temperature for
the location in which the community water system is
situated (see table 2).
TABLE 2. National Interim Primary Drinking
Water Regulations: Maximum Contaminant
Level (MCL}a for Fluoride
Temperature
MCL, mg/l
53. 7 and below
53.8 to 58.3
58.4 to 63.8
63.9 to 70.6
70.7 to 79.2
79.2 to 90.5
1 2.0 and below
12.1 to 14.6
14.7 to 17.6
17.7 to 21. 4
21 .5 to 26.2
26.3 to 32.5
2.4
2.2
2.0
1.8
1.6
1.4
determined by the annual average of the maxi-
mum daily air temperature for the location in which
the community water system is situated.
TABLE 3. Most Effective Treatment Methods for Inorganic Contaminant Removal
Contaminant
Most effective methods
Contaminant
Most effective methods
Arsenic:
As+3
As+5
Barium
Chromium:
Cr+3
Cr+6
Ferric sulfate coagulation, pH 6-8
Alum coagulation, pH 6-7
Excess lime softening
Oxidation before treatment
required
Ferric sulfate coagulation, pH 6-8
Alum coagulation, pH 6-7
Excess lime softening
Lime softening, pH 10-11
Ion exchange
Ferric sulfate coagulation, above
pH8
Lime softening
Excess lime softening
Ferric sulfate coagulation, pH 6-9
Alum coagulation, pH 7-9
Excess lime softening
Ferrous sulfate coagulation, pH 7-
9.5
Fluoride
Lead
Mercury:
Inorganic
Organic
Nitrate
Selenium:
Se+4
Se+6
Silver
Ion exchange with activated alu-
mina or bone char media
Ferric sulfate coagulation, pH 6-9
Alum coagulation, pH 6-9
Lime softening
Excess lime softening
Ferric sulfate coagulation, pH 7-8
Granular activated carbon
Ion exchange
Ferric sulfate coagulation, pH 6-7
Ion exchange
Reverse osmosis
Ion exchange
Reverse osmosis
Ferric sulfate coagulation, pH 7-9
Alum coagulation, pH 6-8
Lime softening
Excess lime softening
-------
o
H
O
z
O
CO
O
u
u
o
Di
o
u.
o
o
s
LJJ
LU
Ql
3
O
LL.
-------
CO
-------
V
V
\
u
\
8*
!l
!l
h
1 !
I
\
LU
H
LL.
o
t/1
LU
CD
Q
UJ <
x y
°- Z
U-
o
o
s
LU
Di
(T)
LU
O
O
o
oo
o
(N
U-
lN3Di!3d
-------
Standard analytical detection methods (atomic
adsorption) do not distinguish between valence states,
but measure only the total concentration. Such
measurements are adequate for normal routine moni-
toring, but if treatment must be provided to lower
the concentration, additional analytical testing is
recommended to identify the form of the contami-
nant to determine the proper treatment method.
Information on the chemistry and occurrence of
inorganic substances in water indicates that normally
the reduced form of a contaminant is found in
ground water, and the oxidized form in surface water.
Furthermore, the more naturally occurring substances
are generally found in ground water, and those
occurring from industrial pollution are usually found
in surface waters. This type of information is also
useful and can serve as a guide in selecting the proper
treatment technique. The form, however, may change
because of the oxidation-reduction conditions of the
water; therefore, it is advisable to identify the form
of the contaminant and routinely check it during
treatment.
Arsenic (As)
MCL: 0.05 mg/l
Common Valence Forms:
+3 (arsenite)
+5 (arsenate)
Most Likely Occurrence and Source of Contaminant:
+3 Ground water—natural occurrence
+5 Ground water—natural occurrence
Surface water—natural occurrence or indus-
trial pollutant
The standard analytical procedures used to deter-
mine the amount of arsenic in water measure only
total arsenic and do not distinguish between the two
valence forms. Because of significant differences in
removals of each form by conventional coagulation
and lime softening treatment methods, the arsenic
form should be determined before selecting a treat-
ment method or modifying an existing facility.
Furthermore, the treatment modification or design
should take into account the potential valence change
of the arsenic before treatment caused by the
oxidation-reduction characteristics of the raw water.
It is not as important to identify the valence of
arsenic if the treatment technique selected is either
ion exchange or reverse osmosis. The literature
indicates, however, that arsenic should be found in
the anion form in aqueous solutions as either AsO2 1
or AsO4~3 and therefore, the selection of the type of
ion exchange resin is significant (1).
As+5 REMOVAL
Laboratory experiments and pilot plant studies on
specific forms of arsenic have shown that As"1"5 can be
removed from water very effectively by conventional
alum or iron coagulation and by lime softening
treatment processes (2-5). These studies, however,
demonstrated that arsenic removals depend on the pH
of the treated water, the coagulant dose, and the
initial arsenic concentration, with pH being the most
important factor (figs. 4 through 7).
Alum and ferric sulfate coagulation (20-30 mg/l)
achieved over 90 percent removal of As*5 (0.3 mg/l)
between pH 5.0 and 7.5. Above pH 7.5, As"1"5
removals decreased, particularly with alum coagula-
tion. When the initial concentration was increased
above 1.0 mg/l, arsenic removals decreased as the
concentration increased, particularly with alum coag-
ulation. Larger doses of coagulant, however, pro-
duced higher removals and might be necessary to
achieve the MCL.
Lime softening was also found very effective for
As"1"5 removal. At pH 10.8 and above, 95 percent
removals were achieved with raw water concentration
of 0.1 to 10.0 mg/l. Below pH 10.8, removals
decreased as the pH decreased, to about 30 percent at
pH 8.5 (3).
As*3 REMOVAL
Laboratory and pilot plant experiments have
shown that As+3 is not removed as effectively from
water as As"1"5 either by iron or alum coagulation or
by lime softening treatment processes (figs. 4 through
7). In the pH range of 5.5-9.0, alum coagulation
(30 mg/l) removed less than 20 percent and ferric
sulfate (30 mg/l) 60 percent or less of 0.3 mg/l of
As+3. Furthermore, As+3 removals decreased with
increasing concentrations. Lime softening was shown
to be only slightly more effective, removing about 70
percent of 0.3 mg/l of As"1"3 at pH 10.8 and above.
Below this pH, removals decreased to less than 20
percent (3).
As+3 can be removed from water by conventional
coagulation and lime softening by oxidizing it to
As+5 before treatment. Laboratory studies have
demonstrated that the conventional chlorination dis-
infection process before treatment will result in As+3
removals similar to those achieved on As+5 by the
-------
100
80
60
40
20
*
Ferric sulfate 30 mg/l
As+3 0.3 mg/l
%: Chlorinated
O Not chlorinated
• As+5 0.05 mg/l
Pilot plant tests
XAs+5
•^ As+3 not chlorinated
7 8
pH OF TREATED WATER
10
FIGURE 4 ARSENIC REMOVAL BY IRON COAGULATION {2,3}
100
80
60
40
20
V
V
MCL for
0.3 mg/l
Alum 30 mg/l
As+3 0.3 mg/l
• Chlorinated
A Not chlorinated
D As+5 0.05 mg/l
Pilot plant tests
X As+5
•^ As*3 not chlorinated
X
1
FIGURE 5
789
pH OF TREATED WATER
ARSENIC REMOVAL BY ALUM COAGULATION (2,3)
10
-------
100
80
RCENT REMOVED
•J> cn
O 0
LU
a.
20
0
— MCL for
0.4 mg/l
.A^.-C*
Q^g^-m
As 0.4 mg/l j£ '
^B A A i r ^^r " Jifc
0 As+3 chlorinated ^iT^ f^ ^
3JC As+3 not chlorinated ^r^ •
— Pilot plant tests JF •
•^6 As+3 not chlorinated ^^ ^
A^^f /
-
—
"
8
^* •
^ * /
.*.-*—*-''>
I.I.I. I
9 10 11 12
pH OF TREATED WATER
FIGURE 6 ARSENIC REMOVAL BY LIME SOFTENING (3)
100
80
REMOVED
01
o
1-
z
UJ
2 40
UJ
a.
20
0
0
»....M.yQw4
Lime softening
pH 10.9-1 1.1
- ^ As+5
0
Ferric sulfate 30
OAs*3
Alum 30 mg/1
_ DAs+5
AAs+3 A
10
^"***» "*****•
LJ** ***
\\
mg/1 ^~°^ ^Q •
1 .11
0.5 1.0 5 10 20
ORIGINAL CONCENTRATION, mg/l
FIGURE 7 ARSENIC REMOVAL BY COAGULATION AND LIME SOFTENING
(3)
-------
same treatment processes (3). And studies have
shown that potassium permanganate should also be
an effective oxidant (6).
Recent investigations have found that the reaction
of chlorine with certain organic materials produces
chloroform and other related organic byproducts.
Consequently, the use of chlorine as an oxidant for
As"1"3 removal may not be advisable.
ARSENIC REMOVAL (GENERAL)
Laboratory and pilot plant studies and full-scale
treatment for arsenic removal have been conducted in
Taiwan on ground water and synthetic waters (6).
The valence form of arsenic was not identified in the
ground water, but the removal results by the various
treatment techniques studied suggest As+3. Labora-
tory and pilot plant study results showed that the
best removals, nearly 100 percent, were achieved with
iron coagulation when the raw water was oxidized
before treatment. Both chlorine and potassium per-
manganate were used as oxidants with about equal
success. Aeration was not effective.
Based upon the laboratory and field experiments,
a full-scale iron coagulation water plant was built in
1969 to serve 1,500 people in Taiwan. Arsenic
removal data over a 4-month period showed the raw
water to contain 0.36-0.56 mg/l of arsenic and the
treated water to be free of arsenic. The pH of the
finished water ranged from 7.7 to 8.3.
A few laboratory studies have been conducted on
the removal of arsenic by ion .exchange (7,8). Cation
exchangers, both of the H and Na form, produced no
removal. Several different anion exchange resins have
been tested and found to remove from 55 to 100
percent of the arsenic. This work confirms that
arsenic is found as an anion in water and that an
anion exchange resin is required to remove arsenic.
Activated alumina has also been found to remove
arsenic from water (9). Experiments on synthetic
water and a ground water containing arsenic showed
activated alumina to lower the arsenic content from
0.05-0.1 mg/l to 0.01 mg/l or less. No pilot plant or
full-scale treatment data are available.
REFERENCES
1. O'Connor, J. T. Removal of Trace Inorganic
Constituents by Conventional Water Treatment
Processes. In: Proceedings of the 16th Water
Quality Conference—Trace Metals in Water Sup-
plies: Occurrence, Significance, and Control. Uni-
versity of Illinois Bulletin No. 71 (108):99-110,
1974.
2. Gulledge, J. H., and J. T. O'Connor. Removal of
As(V) from Water by Adsorption on Aluminum
and Ferric Hydroxide. j.Am. Water Works Assoc.,
65(8) :548-554,1973.
3. Logsdon, G. S., T. J. Sorg, and J. M. Symons.
Removal of Heavy Metals by Conventional Treat-
ment. In: Proceedings of 16th Water Quality
Conference—Trace Metals in Water Supplies:
Occurrence, Significance, and Control. University
of Illinois Bulletin No. 71,1974. Pp. 111-133.
4. Logsdon, G. S., and J. M. Symons. Removal of
Heavy Metals by Conventional Treatment. In:
Proceedings of a Symposium on Trace Metals in
Water Removal Processes and Monitoring. U.S.
Environmental Protection Agency, New York,
N.Y., 1973. Pp. 225-256.
5. Logsdon, G. S., and J. M. Symons. Removal of
Trace Inorganics by Drinking Water Treatment
Unit Processes. AICE Symp. Ser., 70(136):367-
377,1974.
6. Shen, Y. S. Study of Arsenic Removal from
Drinking Water. /. Am. Water Works Assoc.,
65(8):543-548, 1973.
7. Calmon, C. Comments. J. Am. Water Works
Assoc., 65(8):586-589,1973.
8. Calmon, C. Removal Processes by Ion Exchange.
I n: Proceedings of Symposium of Traces of Heavy
Metals in Water Removal Processes and Monitor-
ing. U.S. Environmental Protection Agency, New
York, N.Y., 1973, Pp 7-42.
9. Bellack, E. Arsenic Removal from Potable Water.
j. Am. Water Works Assoc., 63(7):454-458, 1971.
10
-------
Barium (Ba)
MCL: 1.0mg/l
Common Valence Form: +2
Most Likely Occurrence and Source of Contaminant:
Ground water—natural occurrence
REMOVAL
Laboratory studies have been conducted on the
removal of barium from ground water by iron and
alum coagulation and lime softening (1-3). Lime
softening achieved greater than 90 percent removal in
the 10-11 pH range on well water containing 7-8.5
mg/1 of naturally occurring barium (fig. 8). Removals
decreased below and above this range.
Pilot plant studies and full-scale treatment infor-
mation on similar types of ground water verified the
laboratory data (4). Pilot plant test runs on water
containing 10-12 mg/l of barium at pH 9.2, 10.5, and
11.6 resulted in removals of 84, 93, and 82 percent,
respectively. Grab samples from two full-scale lime
softening plants showed removals of 88 and 95
percent. These plants operated at pH 10.5 and 10.3,
and the raw water barium concentrations were
measured at 7.5 and 17.4 mg/l, respectively (4).
Alum and ferric sulfate coagulation were not
effective for barium removal. Laboratory tests
showed that alum coagulation could achieve only
about 20 percent removal even when using 120 mg/l
of alum (figs. 9 and 10). Ferric sulfate coagulation
was only slightly better than alum, achieving 35
percent removal with 120 mg/l of coagulant (figs. 9
and 10). Conventional coagulation, therefore, is not
considered a good method for barium removal, unless
the barium concentration is only slightly above the
MCL
An alternative treatment method to lime softening
for barium removal is ion exchange. Field data from
two midwestern full-scale ion exchange softening
plants showed that barium removal was comparable
to hardness removal on well water containing 11-19
mg/l of barium and 225-230 mg/l of hardness as
CaCO3 (4). When these softening units were per-
forming efficiently and removing all of the hardness
from the water, they also removed all of the barium.
Furthermore, barium breakthrough occurred at about
the same time as hardness breakthrough. As a result
of the similarity in behavior of hardness and barium
in ion exchange treatment, the hardness test can be
used as a practical method to monitor barium during
treatment.
Although ion exchange softening treatment may
be very effective for barium removal, this technique
may not always be practical on waters containing
very high barium concentrations. In normal softening
treatment, raw water is blended with the treated
water to lower treatment costs and to provide a
finished water with a reasonable amount of hardness.
If the barium concentration is high, blending may not
be possible to maintain the barium concentration in
the finished water below the MCL. Furthermore,
hardness would also have to be added to the very soft
treated water, which would result in higher than
normal treatment cost.
REFERENCES
1. Logsdon, G. S., T. J. Sorg, and J. M. Symons.
Removal of Heavy Metals by Conventional Treat-
ment. In: Proceedings of T6th Water Quality
Conference-Trace Metals in Water Supplies: Oc-
currence, Significance, and Control, University of
Illinois Bulletin No. 71, 1974. Pp. 111-113.
2. Logsdon, G. S., and J. M. Symons. Removal of
Heavy Metals by Conventional Treatment. In:
Proceedings of a Symposium on Trace Metals in
Water Removal Processes and Monitoring. U.S.
Environmental Protection Agency, New York,
NY., 1973. Pp. 225-256.
3. Logsdon, G. S., and J. M. Symons. Removal of
Trace Inorganics by Drinking Water Treatment
Unit Processes. AICE Symp. Ser., 70(136):367-
377,1974.
4. Unpublished data, U.S. Environmental Protection
Agency, Office of Research and Development,
Municipal Environmental Research Laboratory,
Water Supply Research Division, Cincinnati, Ohio.
11
-------
100
80
60
40
20
MCLfor
8mg/l
• Ba 7-8 mg/l
n Hardness
__ X Pilot plant tests
***»,
°V
V
FIGURE 8
9 10
pH OF TREATED WATER
BARIUM REMOVAL BY LIME SOFTENING (1)
11
12
100
80
60
40
20
MCLfor
8 mg/l
Ba 7-8 mg/I
s£ Ferric sulfate 20-30 mg/I
O Alum 20-30 mg/l
FIGURE 9
78
pH OF TREATED WATER
BARIUM REMOVAL BY ALUM
COAGULATION (1)
10
AND FERRIC SULFATE
-------
TOO
80
MCLfor
Smg/1
60
Ba 7-8 mg/1
pH 7.5-8.0
% Ferric sulfate
O Alum
0 20
FIGURE 10
60 80
COAGULANT DOSE, mg/l
TOO
120
140
BARIUM REMOVAL
COAGULATION (1)
BY ALUM AND FERRIC SULFATE
Cadmium (Cd)
MCL: 0.010 mg/l
Common Valence Form: +2
Most Likely Occurrence and Source of Contaminant:
Surface water—industrial pollutant
REMOVAL
Laboratory experiments and pilot plant studies on
the removal of cadmium from water showed cad-
mium to be readily removed by lime softening, and to
a lesser extent by ferric sulfate and alum coagulation
treatment (1). Lime softening achieved removals of
greater than 98 percent in the 8.5-11.3 pH range on
well water containing 0.3 mg/l of cadmium (fig. 11).
Removals equally as good were obtained at pH
11.2-11.3 when the initial cadmium concentration
was increased up to 10 mg/l.
Cadmium removals by ferric sulfate and alum
coagulation were lower than those of lime softening
and were shown to depend on pH (fig. 12). Cadmium
hydroxide and carbonate are reported to be ex-
tremely soluble below pH 7, and the coagulation
studies confirmed the reports. Cadmium removals
increased with increasing pH. Ferric sulfate coagula-
tion studies on river water containing 0.3 mg/I of
cadmium showed removals to increase from 20
percent at pH 7.2 to above 90 percent at pH 8 and
above. Alum coagulation results on river water also
increased with pH, but the results were not reproduc-
ible above pH 8. The data indicate that above pH 8,
removals may depend on the turbidity of the raw
water. In some tests with low turbidity water (1-10
Jtu), removals decreased as the pH increased. Poor
formation of alum floe above pH 8 is probably
another reason for the lower removal results {1).
13
-------
The pH effect was also observed with ferric sulfate
and alum coagulation when the low turbidity well
water was used as the test water. In these laboratory
studies, ferric sulfate achieved higher removals than
alum, and removals increased as the pH increased
above pH 7.
Studies on the effect of varying the initial cad-
mium concentration showed that removals decreased
only slightly as the concentration increased when all
other conditions remained constant. Laboratory
studies also showed that cadmium removals by ferric
sulfate and alum coagulation can be increased by
increasing the amount of coagulant (fig. 13).
REFERENCE
1. Unpublished data, U.S. Environmental Protection
Agency, Office of Research and Development,
Municipal Environmental Research Laboratory,
Water Supply Research Division, Cincinnati, Ohio.
100
80
60
40
20
111 ....... H
X X
MCL for
0.03 mg/l
Cd 0.03 mg/l
• Well water
X Pilot plant tests
1
9 10
pH OF TREATED WATER
11
12
FIGURE 11 CADMIUM REMOVAL BY LIME SOFTENING (!)
-------
100
80
2 60
40
20
Cd 0.03 mg/l
River water
% Ferric sulfate
O Alum
Pilot plant tests
X Ferric sulfate
pH OF TREATED WATER
FIGURE 12 CADMIUM REMOVAL BY ALUM
COAGULATION (1)
10
AND FERRIC SULFATE
TOO
80
60
40
20
MCLfor
0.03 mg/1
Cd 0.03 mg/1
Well water
• Ferric sulfate pH 8.4
D Alum pH 7.7-8.0
kO1
0 20 40 60 80
COAGULANT DOSE, mg/l
FIGURE 13 CADMIUM REMOVAL BY ALUM
COAGULATION (1)
100
120
140
AND FERRIC SULFATE
-------
Chromium (Cr)
MCL: 0.05 mg/l
Common Valence Forms: +3
+6
Most Likely Occurrence and Source of Contaminant:
+3 Ground water—natural occurrence
+6 Surface water—industrial pollutant
There are two common valence forms of chromi-
um, Cr+3 and Cr+6. In aqueous solutions, Cr+3 will
exist as a cation and Cr+6 in an anion form as either
chromate (Cr04~2) or dichromate (Cr2O7~4) (1). The
standard analytical procedures used to determine the
amount of chromium in water measure only total
chromium and do not distinguish between the two
valence forms. Although the hexavalent form of
chromium is the most toxic, the MCL was established
for total chromium to minimize the analytical work
load and because the hexavalent form is that most
likely to be found in water. From a treatment
standpoint, however, the form of the contaminant is
significant because the hexavalent form is more
difficult to remove from water by conventional
coagulation treatment than the trivalent form. If
treatment is required, the form of the chromium
should be identified to select the proper type of
treatment system or modification.
Cr*3 REMOVAL
Laboratory studies have shown that alum coagula-
tion, iron coagulation, and lime softening are all very
effective methods for removing Cr+3 from water (2).
These studies have also shown that removals by lime
softening depend on pH, whereas pH has only a very
slight effect on removals by alum and iron coagu-
lation (figs. 14 and 15). For example, lime softening
achieved above 98 percent removal of Cr+3 (0.15
mg/l) in well water in the pH range of 10.6-11.3.
Below pH 10.6, removals decreased as the pH
decreased, to a low of 70 percent at pH 9.21.
Ferric sulfate achieved excellent Cr+3 removals-
greater than 98 percent throughout the 6.5-9.3 pH
range. Alum coagulation was not quite so effective as
100
80
a
LL>
| 60
iu
QL
40
20
MCL for
0.15 mg/l
Well water
D Cr+60.15mg/l
# Cr+30.15mg/l
Pilot Plant Tests
a
9 10
pH OF TREATED WATER
12
FIGURE 14 CHROMIUM REMOVAL BY LIME SOFTENING (2)
-------
ferric sulfate, but did obtain above 90 percent
removal in the 6.7-8.5 pH range. Above pH 8,5,
removals began to decrease; at pH 9.2 removal
dropped to 78 percent. When the Cr+3 concentration
was increased up to 10 mg/l, ferric sulfate and alum
coagulation achieved removals greater than 98 per-
cent in the optimum pH range (fig. 16). The same
excellent results were also obtained with lime soften-
ing when the Cr+3 concentration was increased up to
10 mg/l.
Because of the potential problem of oxidation of
Cr+3 to Cr"1"6 by chlorination, several experiments
were conducted to determine the effect of this factor.
Experiments on well water containing 0.15 mg/l of
Cr*3 showed that low chlorine doses of 2 mg/l for up
to 6 hours contact time lowered Cr*3 removals by
only about 10 percent with alum and ferric sulfate
coagulation (fig. 17). When the contact time was
extended to about 20 hours, however, alum removals
dropped to less than 10 percent total removal. This
work indicates that chlorination before treatment can
oxidize Cr+3 to the Cr+6 form, which is difficult to
remove, and that the extent of oxidation depends on
contact time and chlorine dose. If chlorination before
treatment is absolutely necessary, then ferrous sulfate
is recommended as the coagulant because it has been
found to be effective on Cr+6.
Cr+6 REMOVAL
Laboratory studies on the removal of Cr*6 from
river water showed that neither coagulation by alum
or ferric sulfate nor lime softening was very effective
(2). Of the three methods, ferric sulfate achieved the
best results, removing 35 percent at the low pH of 5.5
on river water containing 0.15 mg/l of Cr+6 (fig. 18).
Alum coagulation and lime softening could do no
better than 10 percent removal throughout their
entire pH range (figs. 14 and 18). These methods
would not be recommended, therefore, unless the
chromium concentration were only very slightly
above the MCL.
Ferrous sulfate coagulation was studied because of
its reducing characteristics as the ferrous iron oxidizes
to ferric iron in the formation of the ferric hydroxide
MCL for
0.15 mg
Cr+30.15mg
Well water
Ferric sulfate 30 mg/l
O Alum 30 mg
Pilot plant tests
River water
X Ferric sulfate
pH OF TREATED WATER
Cr+3 REMOVAL BY IRON AND ALUM COAGULATION (2)
FIGURE 15
-------
floe. Studies conducted on river water containing
0.15 mg/l of Cr+6 showed that ferrous sulfate
coagulation was capable of achieving above 98 per-
cent removals in the 6.5-9.3 pH range (fig. 18) (1).
With higher Cr+6 concentrations, however, it was
determined that removals depend on the time of pH
adjustment (fig. 19). For example, when the pH was
adjusted before coagulation, Cr+6 removals decreased
with increasing concentrations of Cr+6 in the raw
water. Further studies showed, however, that if the
pH of the water is adjusted several minutes after
coagulation, removals greater than 99 percent can be
achieved with Cr+6 concentrations of 10 mg/l. This
procedure of adjusting pH after coagulation is neces-
sary to provide time to reduce Cr+6 to Cr+3 before
floe formation.
REFERENCES
1. O'Connor, J. T. Removal of Trace Inorganic
Constituents by Conventional Water Treatment
Processes. In: Proceedings of the 16th Water
Quality Conference—Trace Metals in Water Sup-
plies: Occurrence, Significance, and Control. Uni-
versity of Illinois Bulletin No. 71 (108):99-110,
1974.
2. Unpublished data, U.S. Environmental Protection
Agency, Office of Research and Development,
Municipal Environmental Research Laboratory,
Water Supply Research Division, Cincinnati, Ohio.
100
80
60
40
20
Cr+3 pH 7.3-8.2
Well water
:Jc Ferric sulfate 30 mg/1
O Alum 30 mg/l
0.10
FIGURE 16
0.5 1.0
ORIGINAL CONCENTRATION, mg/l
10
20
Cr+3 REMOVAL BY ALUM AND FERRIC SULFATE COAGU-
LATION (2)
18
-------
100
80
60
40
20
0
Cr+3pH 7.3-7.6
Well water
Chlorine 2.2 mg/1
sfc 6 h contact time
O 20 h contact time
0.10
FIGURE 17
0.5 1.0
ORIGINAL CONCENTRATION, mg/l
20
EFFECT OF PRECHLORINATION ON Cr+3 REMOVAL BY ALUM
COAGULATION
100
80
Q
UJ
>
O
5
LU
LU
u
60
40
20
'smt>,
$*1
MCL for 0.15 mg/l
Cr+60.15mg/l
River water
• Ferric sulfate 30 mg/l
DAIum30/mg/l
• Ferrous sulfate 30 mg/l
Pilot Plant Tests
+ Ferric Sulfate
XAIum
-^-Ferrous Sulfate
I
10
pH OF TREATED WATER
FIGURE 18 Cr*6 REMOVAL BY ALUM AND IRON COAGULATION (2)
-------
100
80
60
40
20
0
\
Cr+6
River water
Ferrous sulfate 30 mg/1
pH adjustment
• Before coagulation
Q] After coagulation
0.10 0.5 1.0 5 10 20
ORIGINAL CONCENTRATION, mg/l
FIGURE 19 Cr+6 REMOVAL BY FERROUS SULFATE COAGULATION (2)
Fluoride (F)
MCL: 1.4-2.4 mg/l, depending on the annual average
air temperature
Common Valence Form: -1
Most Likely Occurrence and Source of Contaminant:
Ground water—natural occurrence
REMOVAL
Fluoride is added to many water supplies for the
prevention of dental caries, but some communities
have the problem of excessive amounts of natural
fluoride in their raw water. Fluoride has been shown
to be removed from water as a side reaction to excess
20
lime softening of high magnesium water. The removal
mechanism is coprecipitation with magnesium hy-
droxide; this process was demonstrated in Ohio in the
early 1930's (1). Although excess lime softening has
been shown to remove fluoride, the most common
method is ion exchange (or sorption) using either
bone char or activated alumina as the exchange resin.
Bone char is ground animal bones charred to remove
all organic matter. Activated alumina is calcined
granules of hydrated alumina. Both materials are
readily available.
Laboratory studies verified by actual plant prac-
tice have also shown that efficiency of removal of
fluoride with bone char and activated alumina is pH
dependent. The lower the pH, the more effective the
-------
removal of fluoride. Because both materials are
somewhat soluble in acid, however, and for reasons of
distribution and consumption, a pH slightly above 7 is
recommended. The capacities for fluoride removal
for both materials are somewhat similar, and both are
amenable to regeneration procedures. The media
should be selected based on laboratory tests on the
water to be treated to determine which material is
most effective.
Both activated alumina and bone char have been
used in full-scale treatment plants to remove fluoride
from water. The first activated alumina plant was
constructed in 1952 in Barlett, Texas (2,3). This plant
was designed to lower the fluoride concentration
from about 8 mg/l to less than 1 mg/l. The plant
operated sucessfully until early in 1977 when it was
closed down. Two other full-scale alumina plants built
in the 1970's continue to operate at Desert Center,
California and at the X-9 Ranch near Tuscon, Arizona.
These two plants have reported higher fluoride
removal capacity than the Barlett plant and their
costs estimated between 10-20 cents per 1,000 gallons
of treated water (excluding amortization costs). The
literature also indicates that there are several other
successful full-scale treatment plants using bone char
and activated alumina in California (4).
Recent laboratory studies have found that arsenic
can interfere with fluoride removal with bone char (5).
Arsenic has been shown to be removed readily from
water by both bone char and activated alumina. The
investigations showed that arsenic sorption on bone
char results in an irreversible change in the structure
of the char and ultimately renders it useless for fluoride
removal. On the other hand, activated alumina is
readily regenerated when both fluoride and arsenic
are removed. If, therefore, the raw water contains
arsenic as well as fluoride, activated alumina would be
the recommended media to use for fluoride removal.
Bone char could be selected, but it would have to be
thrown away when no longer effective for fluoride
removal.
REFERENCES
1. Scott, R. D., A. E. Kimberley, H. L. Van Horn, F.
F. Ey, and F. W. Waring, Jr. Fluorides in Ohio
Water Supplies, j. Am. Water Works Assoc.,
29(9):9-25,1937.
2. Maier, F. J. Defluoridation of Municipal Water
Supplies. J. Am. Water Works Assoc., 45(8):
879-888,1953.
3. Maier, F. J. Partial Defluoridation of Water. Public
Works, 9J :9Q-92, 196Q.
4. Harmon, J. A., and S. B. Balichman. Defluorida-
tion of Drinking Water in Southern California./
Am. Water Works Assoc., 57(2):245-254,1965.
5. Bellack, E. Arsenic Removal from Potable Water.
J. Am. Water Works Assoc., 63(7):454-458,1971.
Lead (Pb)
MCL: 0.05 mg/l
Common Valence Form: +2
Most Likely Occurrence and Source of Contaminant:
Surface water—industrial pollutant
REMOVAL
Literature on the solubility of lead indicates that
the carbonate and hydroxide forms are very in-
soluble, and, therefore, lead should be removed easily
from water by conventional treatment methods (1).
Laboratory studies on the removal of lead by
conventional treatment confirmed this finding (2,3).
Ferric sulfate and alum coagulation achieved greater
than 97 percent removals on river water containing
0,15 mg/l of lead in the pH range of 6-10 (figs. 20
and 21}. Experiments on well water under similar test
conditions showed ferric sulfate to achieve the same
high removals, greater than 97 percent, while alum
obtained slightly lower removals, 80-90 percent.
When the lead concentration was increased up to 10
mg/l, ferric sulfate continued to achieve excellent
removals of greater than 95 percent, whereas alum
achieved only about 80 percent (fig. 22).
Lime softening was also studied. Experiments on
well water with 0.15 mg/l of lead showed that this
treatment method could achieve greater than 98
percent removals in the 8.5-11.3 pH range (fig. 23).
Because of the low solubility of the hydroxide and
carbonate of lead and the ease of removal by
conventional treatment, several experiments were
conducted on lead removal by settling alone without
a coagulant. These tests showed lead removals of
85-90 percent by 1 hour of settling of river water
having a turbidity in the range of 9-40 Jtu (3). The
results indicate that surface waters should normally
contain very low amounts of lead because of the
natural settling process of streams and impound-
ments.
REFERENCES
1. Hem, J. D,, and W. H. Durum. Solubility and
Occurrence of Lead in Surface Water. /. Am.
Water Works Assoc., 65(8):562-568,1973.
2. Naylor, L. M., and R. R. Dague. Simulation of
Lead Removal by Chemical Treatment. /. Am.
Water Works Assoc., 67(lO):560-565,1975.
3. Unpublished data, U.S. Environmental Protection
Agency, Office of Research and Development,
Municipal Environmental Research Laboratory,
Water Supply Research Division, Cincinnati, Ohio.
21
-------
100
80
MCLfor
0.15mg/l
60
40
20
Pb 0.15 mg/l
Ferric sulfate 30 mg/l
D River water
HC Well water
Alum 30 mg/l
• River water
O Well water
Pilot plant tests
River water
X Ferric sulfate
FIGURE 20
789
pH OF TREATED WATER
LEAD REMOVAL BY ALUM AND IRON COAGULATION (3)
10
20
40 60 80
COAGULANT DOSE, mg/l
100
Pb 0.15 mg/1
Alum pH 7.5-7.95
• River water
O Well water
120
140
FIGURE 21 LEAD REMOVAL BY ALUM COAGULATION (3)
-------
100 r
80
60
2 40
20
O
Ferric sulfate 30 mg/1
Q River water pH 7.3
# Well water pH 7.5
Alum 30 mg/1
• River water pH 7.6
O Well water pH 7.6
• Lime softening pH 9.5
1
0.10
0.5 1.0
ORIGINAL CONCENTRATION, mg/l
10
20
FIGURE 22 LEAD REMOVAL BY COAGULATION AND LIME SOFTENING (3)
100
80
60
40
20
MCLfor
0.15 mg/1
Pb 0.15 mg/1
* Well water
i*—3&
I
9 10
pH OF TREATED WATER
11
12
FIGURE 23 LEAD REMOVAL BY LIME SOFTENING (3)
-------
Mercury (Hg)
MCL: 0.002 mg/l
Common Valence Forms: +2 mercury may be found
in the organic or inorganic form
Most Likely Occurrence and Source of Contaminant:
Surface water—industrial pollutant
Mercury may occur in either the inorganic form or
organic form. The organic form is the most important
because it is the more toxic of the two and is the
basis for establishing the iimit in drinking water.
Furthermore, organic mercury is the form most
likely to be found in water, and the more difficult
form to remove by conventional treatment. Conse-
quently, the form of the mercury contaminant should
be determined to select the proper treatment method.
INORGANIC MERCURY REMOVAL
Laboratory experiments and pilot plant studies on
the removal of mercury from drinking water have
been conducted by several investigators (1,2). These
studies showed inorganic mercury removals to depend
on pH of the treated water and turbidity. They also
showed that removals had little dependence on the
mercury concentration in the 0.003-0.116-mg/l range
(2). The best pH range reported for alum and iron
coagulation was 7-8 (1). Ferric sulfate coagulation,
17.8 mg/l, achieved 66 percent removal at pH 7 and
97 percent removal at pH 8 on water containing 0.05
mg/l of inorganic mercury. Alum coagulation was
shown to be much less effective; 47 percent of the
mercury was removed at pH 7 and 38 percent at
pH8.
Turbidity of the raw river water was shown to be
important only with alum coagulation. With 20-30
mg/l of alum, mercury removals increased from about
10 percent on 2-jtu water to 60 percent on 100-Jtu
water (fig. 24). Similar studies with ferric sulfate
showed turbidity to be less important, with removals
ranging from 30 to 60 percent (fig. 25).
Lime softening was moderately effective for in-
organic mercury removal and was pH dependent (2).
Removals increased as the pH increased from 8.5 to
11. In the 10.7-11.4-pH range, removals were 60-80
percent, whereas only about 30 percent removal was
achieved at pH 9.4 (fig. 26).
Powdered and granular activated carbon were
studied for inorganic mercury removal (2). Powdered
activated carbon was shown to increase removals
above those obtained with coagulation afone (fig. 27).
Much higher doses would be required to produce
significant increases, however, than those normally
used for taste and odor control. Granular activated
carbon was found to be fairly effective, although
removals depend on contact time and amount of
water treated. Removals of 80 percent of 20-29 jug/l
of mercury were achieved with 3.5 minutes' contact
time for up to 15,000 bed volumes of treatment (fig.
28).
Several preliminary ion exchange experiments have
been carried out for inorganic mercury removal (2).
These studies showed that as much as 98 percent of
inorganic mercury added to distilled water could be
removed by cation and anion exchange resins oper-
ated in series. Although these experiments were very
preliminary, the results indicated that ion exchange
should be an effective method for inorganic mercury
removal.
ORGANIC MERCURY REMOVAL
Laboratory experiments and pilot plant studies
have shown organic mercury to be more difficult to
remove from drinking water by conventional treat-
ment methods than inorganic mercury (1,2). Alum
and iron coagulation achieved lower organic mercury
removals than inorganic mercury under the same
initial test conditions. Studies on the effect of
turbidity on removal showed alum coagulation re-
movals to increase from 0 to about 40 percent when
the turbidity increased from 2 Jtu to 100 Jtu (fig.
24). Removals with ferric sulfate coagulation were
almost identical to the alum coagulation results. Lime
softening was studied and found • ineffective for
organic mercury removal; less than 5 percent was
removed in the pH range of 9.3-11.3 (fig. 25).
Preliminary studies were also carried out on ion
exchange for organic mercury removal (2}. Results of
these studies were similar to those on inorganic
mercury, with above 98 percent removals achieved by
passing distilled water containing organic mercury
through cation and anion exchange resins. These
results also indicate that ion exchange should be
effective for organic mercury removal.
24
-------
Powdered and granular activated carbon were
investigated for organic mercury removal and both
were found to be effective (2). Studies showed that
about 1 mg/l of powdered activated carbon is needed
to remove each 0.1 re/I of mercury from water to
reach a residual level of 2 Mg/l. Studies on the use of
granular activated carbon showed that removals de-
pend on contact time and the amount of water
treated similar to the finding for inorganic mercury.
Mercury removals of 80 percent or above were
achieved for 25,000 bed volumes of water wjtn 3.5
minutes contact time on water containing 20-29 ,ig/l
of organic mercury (fig. 29).
REFERENCES
1 Ebersole, G., and J.T. O'Connor.The Removal of
Mercury from Water by Conventional Water Treat-
ment Processes. Presented at 92nd Annual Con-
ference, American Water Works Association,
Chicago, HI., June 1972.
2. Logsdon, G. S, and ). M. Symons. Mercury
Removal by Conventional Water Treatment; Tech-
niques. 7. Am. Water Works Assoc., <55(8):554-
562,1973,
Alum 20-30 mg/l
Inorganic Hg
O Methyl Hg
3.0 3°
TURBIDITY OF UNTREATED WATER, tu
REMOVAL
EFFECT OF TURBIDITY ON MERCURY
WITH ALUM COAGULATION (2)
FIGURE 24
25
-------
Ferric sulfate 20-30 mg/l
Inorganic Hg
Methyl Hg
0
1.0
FIGURE 25
10 30
TURBIDITY OF UNTREATED WATER
tu
700
Inorganic Hg
O Methyl Hg
Pilot plant tests
X Inorganic Hg
FIGURE 26
9 10
OF TREATED WATER
MERCURY REMOVAL BY LIME SOFTENING (2)
-------
100 —
40
20
Alum 30 mg/l
% Initial inorganic Hg concentration = 9.3
I
J_
0
10
60
70
20 30 40 50
POWDERED ACTIVATED CARBON ADDED, mg/l
FIGURE 27 MERCURY REMOVAL BY POWDERED
ACTIVATED CARBON (2)
100
MCL for
40
20
_
0
O 2.4
• 1.7
DLO
• 0.5
A 0.3 ,
10,000
^^W ^
"^^^
"^^
A
1 I
20,000 30,000
BED VOLUMES TREATED IN COLUMNS
FIGURE 28
ORGANIC MERCURY REMOVAL BY GRANULAR
ACTIVATED CARBON COLUMNS (2)
-------
MCLfor
20wg/l
FIGURE 29
10,000 20,000 30,000
BED VOLUMES TREATED IN COLUMNS
INORGANIC MERCURY REMOVAL BY GRANU-
LAR ACTIVATED CARBON COLUMNS {2}
Nitrate (NO3)
MCL: 45mg/l (10 mg/I as N)
Common Valence Form: -1
Most Likely Occurrence and Source of Contaminant:
Ground water—agricultural pollutant
Surface water—agricultural pollutant
REMOVAL
Ion exchange is currently the only method in use
to remove nitrate from water. Conventional coagula-
tion and lime softening are not effective treatment
methods for the removal of this contaminant.
Laboratory experiments and pilot plant studies
have shown that some strong base and weak base ion
exchange resins are nitrate selective and can reduce
the nitrate concentration from as high as 50 mg/I (as
N) to 0.5 mg/I (1-4). One full-scale ion exchange
plant has been operating successfully in Long Island,
28
New York, since 1974 (2). This plant lowers the
nitrate level of 20-30 mg/1 in the raw water to 0.5
mg/I. The finished water is a blend of treated and raw
water, and contains about 5 mg/1 of nitrate (as N).
The plant was designed to treat 1,200 gal/min and
incorporates a continuously regenerated ion exchange
process, operated with the resin moving in a closed
loop and using a strong anion exchange resin. The
construction cost for the plant was $405,000 in
1974, and the estimated operating cost is 7 cents per
1,000 gallons of finished water.
Studies by EPA have shown that the nitrate
selectivity of strong base resins changes with sub-
stantial changes in the ion concentrations in the water
(5). Therefore, to evaluate the nitrate removal effi-
ciency for a specific water, tests should be conducted
with the actual water to be treated.
A research grant was funded by EPA to study the
capability of the strong-acid/weak-base ion exchange
system to remove nitrate. Preliminary results indicate
-------
that the system will remove nitrate effectively;
however, the operating cost may be about twice as
much as the strong base system. The potential
advantage of this system is a waste product that may
have fertilizer value for agricultural uses.
REFERENCES
1. Gauntlett, R. B. Nitrate Removal from Water by
Ion Exchange. Water Treat. Exam., 24(3):
172-190,1975.
2. Gregg, J. C. Nitrate Removal at Water Treatment
Plant. Civ. Eng., 43(4):45-47, 1973.
3. Holzmacher, R. G. Nitrate Removal from a
Ground Water Supply. Water Sewage Works,
7/S(7):210-213, 1971.
4. Korngold, E. Removal of Nitrates from Potable
Water by Ion Exchange. Water, Air, Soil Pollut.,
2:15-22,1973.
5. Beulow, R. W., K. l_. Kropp, J. Withered, and J.
M. Symons. Nitrate Removal by Anion-Exchange
Resins. J. Am. Water Works Assoc.,
67(9): 528-534, 1975.
Selenium (Se)
MCL: 0.01 mg/l
Common Valence Forms:
+4 (selenite)
+6 (selenate}
Most Likely Occurrence and Source of Contaminant:
+4 Ground water-natural occurrence
+6 Ground water—natural occurrence
Surface water—natural occurrence
The standard analytical procedures used to deter-
mine the amount of selenium in water measure only
total selenium and do not distinguish between the
two forms. Because of differences in removals of each
form by conventional coagulation and lime softening
methods, it is important that the form be determined
before selecting a treatment system or modifications
to an existing system. The literature indicates that the
two forms are fairly stable and act independently of
one another in the same solution (1). Because of this
stability factor, the oxidation-reduction character-
istics of the raw water should have little effect on
changing the form.
The literature also indicates that selenium should
be found as an anion in aqueous solutions as either
SeO3-2 (selenite) or SeO4-2 (selenate) (2). This fact is
important if ion exchange is being considered to
insure that the proper ion exchange media is selected.
Se+4 REMOVAL
Laboratory experiments and pilot plant studies
have shown that alum and ferric sulfate coagulation
and lime softening are only moderately effective on
the removal of Se+4 from water (3-6). Furthermore,
these studies have shown that removals depend on pH
and coagulant dose (figs. 30, 31, and 32). Ferric
sulfate coagulation (30 mg/l) achieved the best
results, with 85 percent removal at the low pH of 5.5
on river water containing 0.03 mg/l. Removals de-
creased to about 15 percent removal as the pH
increased to pH 9.2. Slightly lower removals were
achieved with low turbidity well water and the same
pH trend of decreasing removals with increasing pH
was observed.
Alum coagulation was much less effective than
ferric sulfate (fig. 30). Less than 20 percent removal
was achieved with 30 mg/l of alum on test waters
containing 0.03 mg/l of Se+4 throughout the pH
range. Even when the alum dose was increased to 100
mg/l at pH 6.9, only 32 percent of Se+4 was removed
(fig. 32). Studies on the effect of the initial selenium
concentration up to 10 mg/l showed that it was not a
factor for removal by either coagulant.
Lime softening was also studied. Results of these
tests showed that removals increased with increasing
pH, but that, at best, only about 45 percent could be
removed from well water containing 0.03 mg/l of
Se+4(fig.31).
Very limited laboratory studies have been con-
ducted to determine Se+4 removal from water by ion
exchange and reverse osmosis. Both methods achieved
excellent removals of greater than 97 percent on tap
or distilled water containing about 0.1 mg/l of
selenium (6). Although these studies were very brief
and were conducted under laboratory conditions, the
results indicate that both methods are capable of
achieving high removals of Se+4.
Se+6 REMOVAL
Laboratory tests and pilot plant studies have
shown that alum, ferric sulfate, and ferrous sulfate
coagulation and lime softening are ineffective for
selenate removal from water (3-6). Studies on water
containing 0.03-10 mg/l showed that none of these
conventional treatment methods could achieve more
than 10 percent removal. If, therefore, selenate has to
be removed from water, other methods must be used.
29
-------
Procedures for laboratory studies on the use of ion
exchange and reverse osmosis to remove selenate
from water were similar to those used for the selenite
studies (6). The results of the very limited tests
showed that both ion exchange and reverse osmosis
could achieve greater than 97 percent removals of
Se+6 from either tap or distilled water containing 0.1
mg/l of selenate. Although the studies were pre-
liminary, they indicate that selenate can be removed
effectively from water by these methods.
REFERENCES
1. Olson, E. E., and C. W. Jensen. The Adsorption of
Selenate and Selenite Selenium by Colloidal Ferric
Hydroxide. In: Proc. S. D. Acad. Sci., 20:115-121,
1940.
2. O'Connor, J. T. Removal of Trace Inorganic
Constituents by Conventional Water Treatment
Processes. In: Proceedings of the 76th Water
Quality Conference—Trace Metals in Water Sup-
plies: Occurrence, Significance, and Control. Uni-
versity of Illinois Bulletin No. 71 (108):99-110,
1974.
3. Logsdon, G. S., T. J. Sorg, and J. M. Symons.
Removal of Heavy Metals by Conventional Treat-
ment. In: Proceedings of J6th Water Quality
Conference—Trace Metals in Water Supplies:
Occurrence, Significance, and Control. University
of Illinois Bulletin No. 71, 1974. Pp. 111-133.
4. Logsdon, G. S., and J. M. Symons. Removal of
Heavy Metals by Conventional Treatment. In:
Proceedings of a Symposium on Trace Metals in
Water Removal Processes and Monitoring. U.S.
Environmental Protection Agency, New York,
N.Y., 1973. Pp 225-256.
5. Logsdon, G. S., and J. M. Symons. Removal of
Trace Inorganics by Drinking Water Treatment
Unit Processes. AICE Symp. Ser., 70(136):
367-377,1974.
6. Sorg, T. J., and G. S. Logsdon. Removal of
Selenium from Water-State of the Art. Presented
at 1976 Industrial Health Foundation, Inc., Sym-
posium on Selenium-Tellurium, University of
Notre Dame, Notre Dame, Ind., May 11-13,1976.
Se+4 0.03 mg/l
Ferric sulfate 25 mg/l
River water
Well water
Alum 25 mg/l
River water
O Well water
MCLfor
0.03 mg/l
Pilot plant tests
Gravel pit water
X Ferric sulfate
Alum
River Water
-
T Ferric Sulfate
Alum
67«9
pH OF TREATED WATER
FIGURE 30 Se+4 REMOVAL BY ALUM AND IRON COAGULATION (6)
-------
100
80
MCL for 0.1 mg/1
60
40
20
MCL for 0.3 mg/1
Se+4 Well water
• 0.1 mg/I
A 0.03 mg/I
X Pilot plant tests
Se+4 0.03 mg/I
FIGURE 31
9 10
pH OF TREATED WATER
REMOVAL BY LIME SOFTENING (6)
11
12
100
80
> 60
s
UJ
D;
40
20
MCL for
0.03 mg/I
0.03 mg/1
O Alum pH 6.9-7.4
* Ferric sulfatepH 6.9-7.2
0 20
FIGURE 32
40 60 80
COAGULANT DOSE, mg/I
100
120
140
Se+4 REMOVAL BY ALUM AND IRON COAGULATION (6)
-------
Silver (Ag)
MCL: 0.05 mg/l
Common Valence Form: +2
Most Likely Occurrence and Source of Contaminant:
Surface water—industrial pollutant
REMOVAL
Laboratory tests have been conducted on the
removal of silver from water (1). The tests showed
that silver should be easily removed from water by
conventional coagulation and lime softening treat-
ment methods. Alum and ferric sulfate coagulation
achieved greater than 70 percent removal in the 6-8
pH range on river water containing 0.15 mg/l of silver
{fig. 33). Above pH 8, alum removals decreased with
increasing pH. This decrease is attributed to the poor
alum floe formation above pH 8. Experiments with
both coagulants at pH 7.9-8.0 showed removals to
increase with increasing concentration from 0.1 S to
10 mg/l (fig. 34).
Lime softening was also studied. Tests conducted
on well water with 0.15 mg/l of silver found removals
to increase with pH from about 70 percent at pH 9 to
near 90 percent at pH 11.5 (fig. 35).
Because of the good removal results by chemical
coagulation, the effect of settling alone without a
coagulant was studied. An experiment with river
water of 39-Jtu turbidity and 0.15 mg/l of silver
showed that about 50 percent of the silver could be
removed by settling alone. This finding indicates that
silver probably should not be a serious problem in
surface waters because of the natural settling process
of the sediment in streams and impoundments.
REFERENCE
1. Unpublished data, US. Environmental Protection
Agency, Office of Research and Development,
Municipal Environmental Research Laboratory,
Water Supply Research Division, Cincinnati, Ohio.
100 -
80
60
40
20
O
Ag 0.15 mg/l
River water
O Ferric sulfate 30 mg/l
5jc Alum 30 mg/l
Pilot Plant Tests
+ Ferric Sulfate
7F" Alum
FIGURE 33
67 8 9 10
pH OF TREATED WATER
SILVER REMOVAL BY ALUM AND IRON COAGULATION (1)
-------
100
80
60
40
20
0
Ag 0.15 mg/1
River water pH 7.9-8.0
HS Alum 30 mg/1
O Ferric sulfate 30 mg/1
1
0.10 0.5 1.0 5 10 20
ORIGINAL CONCENTRATION, mg/1
FIGURE 34 SILVER REMOVAL BY ALUM AND IRON COAGULATION (1)
100
80
h-
z
UJ
(J
fit
LU
a.
MCL for
0.15mg/[
60
40
20
Ag 0.15 mg/l
• Well water
FIGURE 35
9 10 11
pH OF TREATED WATER
SILVER REMOVAL BY LIME SOFTENING (1)
12
-------
Treatment Costs
for the Removal
of Inorganic
Contaminants
The cost of reducing a contaminant or several
contaminants below the MCL's will depend on the
required treatment technique and on whether a
treatment facility exists. Modifications to, or changes
in operation of an existing plant may result in only a
slight operating cost increase, or possibly no addi-
tional cost at all. On the other hand, if there is no
treatment facility, the cost will be the expense of
constructing and operating a new treatment plant.
As indicated in the Introduction, the information
contained herein is general and should be used only as
a guide. Therefore, the reader is cautioned once again
that these data are intended for planning and not for
design purposes. Material and labor costs vary from
location to location and change almost daily. The
exact cost can only be determined by costing out a
specific plant according to the economic factors of
the location. In the final analysis, the actual cost
could be considerably different from that presented
in the following sections.
COST OF
MODIFICATIONS OR
OPERATIONAL CHANGES
AT EXISTING
TREATMENT PLANTS
The foregoing information on the removal of
specific contaminants by conventional treatment
techniques showed that removals depend on various
operational parameters such as pH, coagulant, coagu-
lant dose, and valence of the contaminant. A change
in any one of these operational variables to achieve
optimum removal of the contaminant should result in
only a very slight increase in operating cost of no
more than a cent or two per 1,000 gallons of water
treated.
If an existing plant must be modified to provide an
additional treatment step or two, or if new equip-
ment must be purchased, the cost will be higher than
that resulting from only an operational change.
Although some capital costs will occur, the increase
in operating cost should not be more than a cent or
two. It is not practical to list capital costs for all
potential modifications, and this document will omit
them,
COST OF NEW TREATMENT
FACILITIES
Although no one treatment method is effective for
the removal of all inorganic contaminants, some
grouping can be made. For example, lime softening is
a good technique for the removal of lead, cadmium,
Cr+3, As+s, and silver. Treatment costs for the
removal of these contaminants by lime softening,
therefore, should be approximately the same. For the
sake of simplicity and to avoid repetition, treatment
cost information presented is based on treatment
technique rather than on individual contaminant.
David Volkert and Associates prepared a report
entitled "Monograph of the Effectiveness and Cost of
Water Treatment Processes for the Removal of
Specific Contaminants" for EPA (1). This report
provides cost information on specific water treatment
processes based on economic indexes of July 1973.
Water and Air Research, Inc., updated some of this
information to October 1975 for use in a report for
EPA entitled "Cost Calculations Procedures for
Determination of Costs of Radium Removal from
Potable Water Supplies" (2). The following cost
information on new treatment facilities has been
developed primarily from these two reports. The
reader is again reminded that these cost data are
intended as a guide for general planning estimates and
should not be used for design purposes.
Capital and operating cost information on reverse
osmosis, ion exchange, and lime softening are shown
in figures 36 and 37. A band is shown for ion exchange
and lime softening. This band represents the cost of
softening about 80 percent of the total water flow
having a total dissolved solids (IDS) range of 2,000-
4,000 mg/l and hardness range of 150-750 mg/l (as
CaCOs). A single line is shown for reverse osmosis
because the capacity of the unit does not vary with
the TDS or hardness of the water. The reverse osmosis
34
-------
10,000
§ 1,000
o
u
_J
<
EC 100
o
10
j]7| Lime-soda
|J[[ Ion exchange
— Reverse osmosis
0.01
0.1
10
1.0
PLANT CAPACITY, mgd
FIGURE 36 CAPITAL COSTS OF WATER TREATMENT PLANTS
100
u
^
Q
O
or
10
1.0
0.1
0.01
Lime-soda
Ion exchange
Reverse osmosis
' ' ' ""II
0.01
FIGURE 37
0.1
10
100
1.0
PLANT CAPACITY, mgd
ANNUAL PRODUCTION COSTS: OPERATION, MAINTE-
NANCE, AMORTIZATION
-------
costs are also based on the treatment of about 80
percent of the total flow. The cost elements included
in the capital and operating costs figures are shown
in table 4.
The costs for chemical coagulation treatment are
about the same as those for lime softening because
the unit processes are almost identical. Lime soften-
ing costs can therefore be used for chemical coagula-
tion treatment cost estimates. The operating cost will
be somewhat conservative because lime softening
generally uses more chemicals and requires additional
pH adjustment equipment.
REFERENCES
1. Monograph of the Effectiveness and Cost of Water
Treatment Processes for the Removal of Specific
Contaminants. Vol. 1, Technical Manual. Contract
No. 68-01-1833. U.S. Environmental Protection
Agency, Washington, D.C., 1974. 324 pp.
2. Cost Calculations Procedures for Determination of
Costs of Radium Removal from Potable Water
Supplies. Contract No. 803854-01. U.S. Environ-
mental Protection Agency, Cincinnati, Ohio, 1976.
TABLE 4. Elements Included in Capital and
Operating Costs
Capital cost elements Construction for site prepara-
tion
Plant construction
Land costs, assumed at
$1,850 per acre
Interest during construction,
8 percent
Startup cost
Owners general expense, 12
percent of construction
Operating cost
Chemicals
Labor
Operation and maintenance
Amortization at 7-percent
compound interest for
depreciating capital
Useful life:
Lime-soda plants 40 years
Ion exchange 20 years
Reverse osmosis 20 years
36
-------
Treatment
Techniques
for the Removal
of Turbidity
from
Drinking Water
A maximum contaminant limit (MCL) has been
established for turbidity because certain types of
turbidity-causing solids, such as organic matter, can
interfere with disinfection or microbiological deter-
minations, or can prevent maintenance of an effective
disinfectant agent throughout the distribution system
(1). Suspended solids that cause turbidity can be
removed from water by coagulation, sedimentation,
and filtration. In addition to preparing the raw water
for disinfection, sedimentation and filtration offer
some fringe benefits. Very clear waters are more
esthetically appealing to consumers. Suspended
matter removed by filtration cannot settle in dead
ends in the distribution system and cause problems
with chlorine demand, microbiological growths, or
taste and odor. Also, removal of suspended matter
can result in removal of contaminants—for example,
heavy metals, pesticides, and asbestos—adsorbed or
attached to the suspended matter. Sedimentation and
filtration can also remove precursor substances that
could form trihalomethanes upon free residual
chlori nation.
The MCL's for turbidity apply to both community
and noncommunity water systems using surface water
sources in whole or in part. The MCL's for turbidity
in drinking water, measured at representative entry
points to the distribution system, are:
(a) One turbidity unit (TU), as determined by
a monthly average pursuant to § 141.22,
except that five or fewer turbidity units
may be allowed if the supplier of water can
demonstrate to the State that the higher tur-
bidity does not do any of the following:.
(}} Interfere with disinfection;
(2) Prevent maintenance of an effective
disinfectant agent throughout the dis-
tribution system; or
(3) Interfere with microbiological deter-
minations.
(b) Five turbidity units based on an average for
two consecutive days pursuant to § 141.22.
by GARY S. LOGSDON
Water Supply Research Division, MERL
U.S. Environmental Protection Agency
Cincinnati, Ohio
The MCL for turbidity is 1 tu, but under certain
circumstances it can be 5 tu. It is assumed for this
document that filtration plants will have an operating
goal of producing water meeting the 1-tu limit, or
better if possible.
The MCL for turbidity applies to systems treating
surface sources in whole or in part. Therefore, it is
not the purpose of this document to discuss
clarification of ground waters to remove iron or
manganese.
Filtration as a water treatment process has been
studied and applied on a municipal scale for many
years in the United States. Granular media filtration
with sand filters was thoroughly researched by Fuller
at Louisville at the turn of the century and has since
been investigated and used at many locations in the
United States. Mixed media filtration was pioneered
by Conley and Pitman at Hanford after World War II.
The use of dual media has been generally recognized
as superior to single media filtration (2). Direct
filtration, or filtration of water after chemical condi-
tioning but without settling, was attempted before
the turn of the century in the early days of filtration.
This process has been developed rationally and
applied effectively in a gradual fashion since World
War 11.
During World War II filter-aid filtration (diatoma-
ceous earth (DE) filtration) was developed for
potable water use by the U.S. Army Corps of
Engineers. Diatomaceous earth filtration has been
used for municipal water treatment since the 1950's
and is an alternative to granular media filtration.
Treatment Techniques
Both granular media and DE filtration can be used
to filter water and produce an effluent that meets the
MCL for turbidity. The nature and amount of
37
-------
treatment before filtration and the choice of granular
media or DE filters should be made case by case,
considering factors such as raw water quality, quality
fluctuations, plant size, and area available for plant
construction.
Because of the many papers, articles, symposia,
and books on filtration, a complete bibliography
would be virtually impossible. Two works that should
be mentioned, however, are Water Quality and
Treatment (2) and Water Treatment Plant Design (3),
both published by the American Water Works Asso-
ciation (AWWA). In addition, papers on filtration are
often found in the monthly issues of the Journal of
the American Water Works Association.
GRANULAR MEDIA FILTRATION
Most granular media filters are operated as gravity
filters, open to the atmosphere. Some are operated
inside closed vessels as pressure filters. The quality
and kind of pretreatment before filtration are deter-
mined by factors such as raw water turbidity,
filtration rate, and media size rather than by whether
the pressure or gravity mode is used. Pressure filters
would seldom have trouble with air binding, but
other granular media filtration problems, such as
mud balls or media upset, could occur. One dis-
advantage of pressure filters is that the operator
cannot see the media and recognize symptoms of
filter operating or backwashing problems. Pressure
filters are often used when raw water is supplied
under pressure, and is filtered and delivered to the
distribution system without re pumping—which is not
often done with surface waters.
Two key factors in successful operation of gran-
ular media filters are proper conditioning of the water
before filtration and thorough backwashing of media
at the end of each run. Variables affecting floe
strength and completeness of flocculation are rapid
mix time and energy (velocity gradient (G) in feet per
second per foot), flocculation time and energy, and
inorganic chemicals or polymers and doses used.
Conditioning chemicals are nearly always needed for
effective turbidity removal by granular media
filtration.
The complex interrelationships between these vari-
ables are best understood through pilot plant tests or
careful observation of treatment plant performance
during varying conditions of raw water quality and
treatment techniques. With proper pretreatment, in-
cluding rapid mix and/or flocculation, the operator
should be able to condition floe to be strong enough
to be retained in the media but not so strong as to
accumulate at the media-water interface in dual or
mixed media filters in the direct filtration process. At
plants having settling, floe should be properly con-
ditioned so that it will settle well but not penetrate
the filters if it does carry over from the settling basin.
The purpose of settling is to prepare a turbid water
for effective, cost-efficient filtration.
Continuous monitoring of filtered water turbidity
has been recommended by Gulp (4). While settling
and filtration processes can be monitored by using
laboratory turbidimeters to measure discrete samples,
continuous turbidimeters with recorders {which cost
under $1,000) provide the plant operator with a
positive, continuous record of the filtered water
turbidity.
For continued, effective filter operation, adequate
backwashing is a must. Poor backwash techniques
may not impair filter performance immediately.
Eventually, however, such problems as mud balls,
sand boils, and pulling away from sidewalls can occur.
Filter media must be cleaned thoroughly. Often
surface wash is used to augment the backwash action
and break up tougher floe in the media. Floe
containing polymer seems more difficult to wash out
of filters than floe from inorganic coagulants. In some
instances air-assisted wash has been used to remove
polymer floe.
D1ATOMACEOUS EARTH
FILTRATION
Because water is seldom, if ever, preconditioned at
DE filter plants, the principal operating variables
relate to the filter precoat and body feed during the
filter run. Both the DE particle size and amount
(pounds per square foot for precoat and milligrams
per liter for body feed) of DE used can be varied
according to filter design and raw water quality. The
operator's goal is to use a grade of DE fine enough to
yield acceptable filtered turbidity, while applying
enough body feed for a long filter run and, thus, for
an economical use of precoat. Insufficient body feed
causes filter blinding and short, uneconomical runs
that waste precoat filter aid.
Backwashing of diatomite filters also should be
thorough. All of the used filter cake should be
removed, and the septum should be cleaned thor-
oughly so that the new precoat will readily form on
it. Backwashing techniques for diatomite filters vary,
depending on the type of equipment and the manu-
facturer.
38
-------
Disposal of
Filter Plant
Sludge
The complexity of water plant Waste treatment
and disposal is so great that it cannot be dealt with in
this document; however, water utility operators
should be aware that, in the future, water plant
wastes will be considered pollutants, and will require
environmentally acceptable disposal methods.
Because sludge characteristics vary from plant to
plant and from process to process, the characteristics
of the waste material should be determined before
treatment and disposal methods are formulated fora
given plant. In general, because most water filtration
plant wastes have a high water content, they are
usually dewatered before ultimate disposal is under-
taken. Dewatering techniques that could be used for
sludge treatment include sand drying beds, lagoons,
thickeners, centrifuges, vacuum filters, pressure
filters, and natural freezing. Under certain circum-
stances, municipal water plant sludge can be dis-
charged to the sanitary sewer and treated at the
sewage treatment plant. Generally, the most accept-
able site for ultimate disposal of dewatered sludge is
on land.
Treatment Costs
for Turbidity
Removal
At existing filtration plants, costs for meeting the
MCL for turbidity might include increased operator
diligence, more frequent (perhaps continuous) turbid-
ity monitoring, slightly higher doses of chemicals, or
a different grade of DE body feed. For example,
existing granular media plants might begin to use
small amounts of polymer in addition to inorganic
coagulants. Or diatomite plants might switch to finer
sizes of precoat and body feed filter aid, and perhaps
use a slightly higher body feed dose, to attain the
desired clarity in the filter effluent. Costs for these
operational changes should be less than 1 cent per
1,000 gal Ions.
At some plants it might be necessary to convert
rapid sand filters to dual media or mixed media
filters. In addition to filter media, costs that might be
associated with filter conversion would be costs for
improved backwash capability. Additional infor-
mation on this subject can be found in AWWA's
Proceedings of the 1974 Seminar on Upgrading
Existing Water Treatment Plants (5).
Costs for plant upgrading are not easily predicted,
because local circumstances can be so varied.
Hudson's paper, "Plant Up-Rating Case Studies," is
informative on this topic (6). Hudson states that
some plants are designed for uprating, some are not,
and still others are difficult or impossible to uprate.
Hudson's work deals with increasing plant production
by modifying the plant. This approach is probably
more expensive than making modifications to im-
prove effluent quality. Nevertheless, the in-
dividualized approach to each plant is the type of
technique that would be needed in modifying a plant
to improve water quality. Plant modifications made
for the dual purposes of improved quality and higher
production rate should be expected to be more
costly.
For plants having raw water of such good quality
that filtration is not practiced now but will be
required in the future because the turbidity MCL will
be a primary, or mandatory, standard, direct filtra-
tion with granular media or DE should almost
certainly be sufficient. Costs for treatment plant
construction and for operation and maintenance have
been reviewed and estimates can be made. Capital
cost curves are included for four kinds of plants-
granular media plants built by conventional con-
struction (fig. 38), direct filtration plants (fig. 39),
granular media package plants (fig. 40), and DE filter
plants (fig. 41).
39
-------
Costs for conventionally built granular media and
DE filter plants were obtained and updated to
December 1975, using EPA's Sewage Treatment Plant
Construction Cost Index. Cost data are identified as
construction costs or estimated costs. For granular
media, data are also identified as for direct filtration
plants, plants with 1-hour contact basins, and plants
with settling basins. Costs along the upper line on
cost curves should be more typical of totally new
plants requiring intakes and raw water pumping, or
filtering at 2-3 gal/min/ft2. Costs on the lower curve
would be more typical of adding a filter plant to an
existing water system, or filtering at 5-6 gal/min/ft2.
Costs for package plants are based on tripling
package plant equipment cost to account for installed
cost and adding the cost of a tank to serve as a
clearwell. The concrete tank size was equal to 4
hours' filter plant production. Concrete tank costs
were estimated using "Cost Aspects of Water
Supply" (7).
Costs for operation and maintenance of filter
plants are not easily obtained, and available data
show a broad range of costs, probably because of the
various methods of cost allocation and accounting
used by different utilities. Costs suggested for in-
clusion in treatment operation and maintenance in
this report are those for raw water acquisition,
chemicals, labor, electricity and utilities at the treat-
ment plant, and maintenance. Costs of high service
pumping or meter repair shop operation were con-
sidered distribution costs, even if incurred at the
filtration plant.
Operation and maintenance costs for granular
media filtration range from 4 cents to 8 cents per
1,000 gallons of water treated. Operation and mainte-
nance costs for DE filtration range from 4 cents to 15
cents per 1,000 gallons treated. Because of the small
number of plants providing data for operation and
maintenance costs, no assurance exists that these
costs are representative of most plants or that the
cost range is all inclusive. Annual costs, including
capital plus operating and maintenance data, are
shown in figure 42.
These cost data are presented to give examples of
the nature of costs associated with water filtration.
For a cost estimate to have value and to be usable at a
specific site, it must be prepared for the locality
considered. Costs of capital, chemicals, material,
supplies, energy, and labor vary greatly and can
influence the choice of treatment selected.
References
1. Symons, J. M., and J. C. Hoff. Rationale for
Turbidity Maximum Contaminant Level. In: Pro-
ceedings of the Third Water Quality Technology
Conference, American Water Works Association,
Atlanta, Ga., Dec. 8-10, 1975.
2. American Water Works Association. Water Quality
and Treatment. 3rd ed. McGraw-Hill, New York,
N.Y., 1971.
3. American Water Works Association, American
Society of Civil Engineers, and Conference of
State Sanitary Engineers. Water Treatment Plant
Design. American Water Works Association,
Denver, Colo., 1969.
4. Gulp, R. L. Direct Filtration. Presented at
California Section, American Water Works Asso-
ciation. Water Treatment Forums V and VI,
Oakland and Los Angeles, Calif., Sept. 14, 15,
1976.
5. Proceedings of the 7974 Seminar on Upgrading
Existing Water Treatment Plants, American Water
Works Association, Denver, Colo., 1974.
6. Hudson, H. E., Jr. Plant Up-Rating Case Studies.
In: Up-Grading Existing Water Treatment Plants.
American Water Works Association, Denver, Colo.,
1974.
7. Cost Aspects of Water Supply. In: Proceedings of
the Eighth Annual Sanitary Engineering Con-
ference. University of Illinois, 1966.
40
-------
100
10
a- .E
<
L)
0.1
jjc Actual cost
D Estimated cost
-- ±50% average capital cost
i i i 1111 il i i i 111111 i i i 111 ill
0.1 1 10 100 1,000
PLANT CAPACITY, mgd
FIGURE 38 CAPITAL COST OF GRANULAR MEDIA PLANTS WITH
SETTLING
100
§R 10
= CTl
52
t-s
u
0.1
0.1
X Actual cost with 1-h contact basin
|c Actual cost
Estimated cost
- ±50% average capital cost
i i in
IMl
..I
1
1,000
I I I I I III
10 100
PLANT CAPACITY, mgd
FIGURE 39 CAPITAL COST OF GRANULAR MEDIA, DIRECT FIL-
TRATION PLANTS
-------
0.1
o.oi
i i i t i
III
Actual Cost
D Estimated Cost
— ±50% Average Capital Cost
A
Oil I ' ' '""I 1—I I III
0.01 0.1 1.0 10
PLANT CAPACITY, mgd
FIGURE 40 CAPITAL COST OF GRANULAR MEDIA
PACKAGE PLANTS, INSTALLED
10 r-
= <^
'E ":
32
"
-------
_ c
c J2
o £
U 13
o o
C C ft
U (J a-
<1> O
00 CX CN
O CL
c re
.2 a.
a o
O
H
U
O
1-
cn
in
oi
LU
O
CJ
Q
O
CL
LU
H
O
I-
u
o
CN
LU
u.
-------
Treatment
Techniques
for the Removal
of Coliform
Organisms from
Drinking Water
by GARY S. LOGSDON
Water Supply Research Division, MERL
U.S. Environmental Protection Agency
Cincinnati, Ohio
Modern water treatment generally includes a
process of disinfection designed to kill micro-
biological pathogens. Although a number of disease-
causing micro-organisms exist, their detection and
identification is difficult and tedious. Therefore, the
efficacy of the disinfection process is generally not
measured by tests for the absence of pathogens but
by measuring a group of indicator micro-organisms,
the coliform group of bacteria. Because coiiforms
originate primarily from the intestinal tract of warm-
blooded animals, including humans, they are indi-
cators of possible fecal contamination.
Certain pathogenic micro-organisms, particularly
cysts, are more resistant to disinfection than are
coliform bacteria. If a water is known or likely to be
contaminated by cysts of Endamoeba hystolytica and
Giardia lamblia, the treatment process should include
disinfection and either diatomaceous earth filtration
or coagulation and granular media filtration.
Maximum
Contaminant
Levels (MCL's)
The MCL's for coliform bacteria, applicable to
community and noncommunity water systems, are as
follows(l):
(a) When the membrane filter technique pur-
suant to § 141.21 (a) is used, the number
of coliform bacteria shall not exceed any
of the following:
(1) One per 100 milliliters as the arith-
metic mean of all samples examined
per month pursuant to § 141.2l{b) or
(c);
(2) Four per 100 milliliters in more than
one sample when less than 20 are
examined per month; or
(3) Four per 100 milliliters in more than
five percent of the samples when 20 or
more are examined per month.
(b) (1) When the fermentation tube method
and 10 miililiter standard portions pursu-
ant to § 141.21 (a) are used, coliform
bacteria shall not be present in any of the
following:
(i) more than 10 percent of the portions
in any month pursuant to
§ 141.21 (b) or (c);
(ii) three or more portions in more than
one sample when less than 20 samples
are examined per month; or
(iii) three or more portions in more than
five percent of the samples when 20
or more samples are examined per
month.
(2) When the fermentation tube method
and 100 miililiter standard portions pursu-
ant to §141.21(a) are used, coliform
bacteria shall not be present in any of the
following:
(i) more than 60 percent of the portions
in any month pursuant to
§ 141.21 (b) or (c);
(ii) five portions in more than one sample
when less than five samples are exam-
ined per month; or
(iii) five portions in more than 20 percent
of the samples when five or more
samples are examined per month.
(c) For community or non-community sys-
tems that are.required to sample at a rate
of less than 4 per month, compliance with
paragraphs (a), (b)(1), or (b)(2) of this
section shall be based upon sampling dur-
ing a 3-month period, except that, at the
discretion of the State, compliance may be
based upon sampling during a one-month
period.
44
-------
Disinfection
of Water
The methods by which water can be disinfected
include the use of chlorine, ozone, and chlorine
dioxide. Other methods not as practical or not
generally used for the treatment of drinking water,
and therefore not discussed in this document, include
gamma radiation, heat, silver, and ultraviolet light.
TURBIDITY
Disinfection efficacy is related to the clarity of the
water being treated. Disinfection of very turbid
waters can be difficult or impractical; therefore, in
the production of potable waters from turbid surface
sources, coagulation, flocculation, sedimentation, and
filtration are used to present a barrier to the passage
of micro-organisms into the water distribution system
and to produce water that is more easily disinfected.
It has been shown that sedimentation and filtration
of properly conditioned waters can remove a signifi-
cant part of the micro-organisms present. These
processes are not totally effective, however, so they
must be followed by disinfection. Disinfection is
more effective when the water is of high quality.
Therefore, the turbidity limit in the Drinking Water
Regulations (1) is 1 tu under most circumstances.
Treatment methods for producing adequately
clarified drinking water are discussed earlier in the
section on turbidity.
DISINFECTION BYPRODUCTS
In recent years, methods for detecting organic
compounds in water have been substantially im-
proved, and small quantities of organic compounds
that were previously undetectable can now be
measured. This improvement in analytical capabilities
has resulted in the discovery of organic byproducts
arising from the process of chlorine disinfection,
some of which may be hazardous to human
health (2-8).
Research is still being conducted to further the
understanding of the problem of chlorination by-
products. In general, however, the potential for the
formation of chlorination byproducts should be
recognized and minimized where possible. One
method of reducing chlorination byproducts is to add
the chlorine to water with the highest possible
quality. Some treatment before chlorination generally
reduces the amount of chlorination byproducts. The
microbiological quality of the water is of prime
importance, however, and changes in treatment prac-
tice must not result in a deterioration of the
microbiological quality of the finished drinking
water.
CHLORINATION
Much of the material in this section was drawn
from a chapter in Water Quality and Treatment by E.
J. Laubusch (9). For a more thorough review of
disinfection refer to this work, which provides a
bibliography of 207 entries.
The first continuous application of chlorination to
a municipal water supply was at the Boonton
Reservoir of the Jersey City Water Works in 1908.
Since that time, chlorination has become widely
accepted, and currently most water utilities use
chlorine for disinfection.
Chlorination may be accomplished using gaseous
chlorine or hypochlorite. Liquid chlorine is the least
expensive form of chlorine and is especially suitable
for larger water utilities. For small utilities, hypo-
chlorite can be added to the water by means of a
solution feed pump. For a ground water source, a
convenient means of hypochlorite feed is to have the
solution feed pump operating simultaneously with
the well pump.
The ability of chlorine to kill micro-organisms is
related directly to the chlorine concentration and
contact time, other factors being equal. Low chlorine
concentrations require longer contact time to achieve
equivalent kill. Other factors are its form (combined
(with ammonia) versus free chlorine), temperature,
and pH of the water. Free chlorine is a much more
effective disinfectant than combined chlorine (see
fig. 43) (10). Combined chlorine (chloramines) can
persist longer in some distribution systems because
they are less reactive. Some water utilities have had
success controlling bacterial aftergrowth in distribu-
tion systems using combined chlorine residuals.
The disinfecting efficiency of free available
chlorine residual decreases significantly as pH rises
(10). The chemistry of chlorination of pure water is
briefly summarized, as follows:
CI2 + H2O -» H+1 + CM + HOG
Chlorination of pure water causes the formation of
hypochlorous acid (HOCI), which dissociates to form
the hypochlorite ion (OCr1). Hypochlorite ion is a
45
-------
relatively poor disinfectant. The distribution of HOCI
and OCl~1 at different pH values is shown in figure
44(11). The concentration of titratable free
available chlorine needed to obtain 99-percent E. coli
kill in 30 minutes is shown as a function of pH in
figure 45 (10). Chlorine is a less effective disinfectant
at high pH than at low pH. It has been demonstrated
that the excess lime softening process has disinfecting
capabilities (12). Riehl listed three factors important
in lime sterilization:
• Quantity of excess lime used
• Time of reaction
• Amount of particulates, organics, and micro-
organisms in water
Fair, Geyer, and Okun (11) state that pathogens
do not survive long at pH values above 11. The
reduced efficiency of chlorine at high pH may be
compensated for by the bactericidal effect of high pH
in the excess lime softening process. At plants using
this process it would be prudent to evaluate the
disinfection effect of excess lime softening when any
changes in the disinfection process are considered.
In many water utilities that chlorinate to produce
a free residual, observation of plots of chlorine
residual versus applied chlorine shows that the
chlorine residual increases, then decreases, and finally
increases again. Such a curve is shown in figure 46
(13) and is termed a breakpoint curve. Research
results from a number of laboratories have shown
that the breakpoint phenomenon is related to the
chemistry of chlorine and nitrogen compounds.
Ammonia, when chlorinated, forms chloramines that
are subsequently destroyed in the breakpoint chiori-
nation process.
Viewing breakpoint chlorination as the relation-
ship between water, chlorine, and nitrogen com-
pounds has provided a useful theory for laboratory
work, but it is probably too simplistic when we
consider what actually happens in water treatment
plants. White reported that Griffin found tastes and
odors that occurred at chlorine doses below the
breakpoint to disappear above the breakpoint (10).
This finding suggests the possible involvement of
substances other than chlorine and ammonia
nitrogen.
The concept of "destroying" compounds in break-
point chlorination was mentioned by Cox (13), and
White referred to "disappearance" of tastes and
odors. It is vital to remember that matter is neither
created nor destroyed; it merely changes form. White
shows that an end product of ammonia chlorination
is nitrogen gas. Nitrogen gas may escape from the
water, but nitrogen atoms have not been destroyed.
This matter is of practical importance—not just
theoretical concern—because of the possibility that
organic compounds could react with chlorine before
the breakpoint is reached. The disappearance of tastes
and odors hints at this phenomenon. There is no
assurance at present that breakpoint chlorination will
destroy undesirable organics, or that chlorinating less
than to the breakpoint will assure a water free of
hazardous chlorinated organic compounds.
OZONE
Ozone has been used as a disinfectant since about
1900, primarily in Europe and Canada. Because few
water utilities in the United States now use ozone,
information on ozonation practices is based mostly
on foreign experience.
Ozone is reported to be a more effective dis-
infectant than chlorine, and it is effective over a
wider range of pH and temperature. Ozone can be
used in lower doses than chlorine to achieve equiva-
lent disinfectant kill, and it is effective in reducing or
eliminating tastes and odors. Ozonation does not
cause the formation of trihalomethanes (THM's)
during disinfection. Water that has been disinfected
with ozone, however, may form THM's if chlorine is
applied to provide a free chlorine residual in the
distribution system. Disinfection-level ozone doses do
little to remove THM precursor (8).
Use of ozone as a disinfectant has had a number of
disadvantages in the past, but a number of these are
being overcome. Laubusch (9) indicated that ozona-
tion energy requirements and operating costs were
higher, "about 10 to 75 times higher than chlorine."
Laubusch also indicated that analytical methods were
not sufficiently specific or sensitive for effective
process control. Both of these observations are no
longer true. Ozonation costs are compared with
chlorination costs in the cost section of this docu-
ment, and they are similar. Closed-loop instru-
mentation for ozonation control in wastewater treat-
ment has now been developed and marketed.
Ozone must be generated on site because it is
highly reactive and thus cannot be shipped as chlorine
can. Ozone also decomposes very rapidly after genera-
tion. Its half-life in water is approximately 20
minutes or less. In plants big enough to produce
ozone from pure oxygen economically, ozone pro-
duction is not difficult. Small plants usually produce
ozone from air that has been dried by refrigeration.
46
-------
10 h-
1.0
0.1
0.01
% Hypochlorous acid
V
0.001
J—I I I I 11
J I I I I I I frj
10 100
99% DESTRUCTION OF E COLI AT 2-6° C, min
1,000
FIGURE 43
COMPARISON OF GERM1C1DAL EFFICIEN-
CY OF HYPOCHLOROUS ACID, HYPO-
CHLORITE ION, AND MONOCHLORAMINE
(10)
-------
II11 I I I I
INN I
LL-66
LJ _i a.
LU
LJ
oo x H Z ce
00 a < LU ^O
riypx
LLJ
<*Q in -3-
odd
CN
d
q
d
o
d
O
o
o
CN
O
CO
i-
o
O
r-
o
CO
o
Oi
o
o
o x Q
LU
LU
LJ
IDOH
-------
c*
UJ
I-
LJJ
Qi
O
_J
X
u
U.
l/l
O
h-
U
<
UJ
^
vo
LQ
Qi
O
O
-------
Because ozone is a gas and is not highly soluble in
water, ozonation facilities differ from chlorination
facilities. The use of tall columns and special diffusers
to maximize ozone-water contact is common. Ozone
contact is often done in various proprietary devices.
Multiple-contact stages may also be used. Ozone
application facilities and technology probably have
more in common with pure oxygen activated sludge
technology than with ordinary water treatment facil-
ities. Ozone treatment still has some disadvantages,
including the lack of ozonation experience in the
American water utility industry and an associated
tendency to favor the use of the old, familiar
disinfectant, chlorine.
There are other disadvantages, including the lack
of ozone residual in the distribution system, diffi-
culty in operating ozonators at varying rates to match
water production and ozone demand, complicated
equipment needed for drying air at plants too small
to use pure oxygen, and the lack of data on the
organic byproducts of ozonated raw waters. The
organic oxidation products problem is not limited to
ozonation of raw water, but may occur if ozone is
applied to settled or finished water. Regrowth is
another disadvantage of using ozone without a
residual in the distribution system.
Some problems associated with ozonation may be
solved more readily than others. For example, in-
stallation of modular ozonators might permit variable
ozone production by starting or stopping individual
units. One of the more serious considerations in
treating water with ozone is the question of oxidation
product formation. It has been shown that ozone
appled at disinfection-level doses does not eliminate
the total organic carbon content of water. Thus, even
though chlorine or chlorine dioxide could be used to
maintain a distribution system residual, THM produc-
tion could occur with the use of free chlorine or with
the use of excess chlorine in chlorine dioxide pro-
duction. Further research is needed to resolve this
question.
Based on current knowledge about ozone as a
water disinfectant, its use could be considered seri-
ously by water utilities. Some of the problems
formerly cited as disadvantages have been resolved.
One that remains is that lack of knowledge of
ozonation byproducts.
CHLORINE DIOXIDE
Chlorine dioxide, like ozone, is not widely used as a
disinfectant in the United States, although it is effec-
tive for that purpose. Because chlorine dioxide is a
powerful oxidant, it has been used to control phenolic
tastes and odors. It also has been used to control tastes
and odors resulting from algal bloom and decaying
vegetation in open reservoirs.
Disinfectant comparisons of chlorine dioxide and
chlorine at doses less than I.Omg/l reveal some of the
beneficial characteristics of chlorine dioxide. Chlorine
dioxide is slightly less effective as a disinfectant than
chlorine at pH 6.5. As pH increases above 7, however,
chlorine dioxide maintains its disinfecting capability
while hypochlorous acid (HOC1) dissociates to form
hypochlorite ion (OCl"1) which is a much less effec-
tive disinfectant. At pH values above 10 chlorine
dioxide reportedly dissociates to chlorite (ClC^"1)
and chlorate (CIC^"1), neither of which is an effective
disinfectant.
Chlorine dioxide cannot be transported because of
its potential explosjveness, so it must be generated
at its point of use. Aqueous sodium chlorite
(Na2ClO2) reacts with aqueous chlorine to form
chlorine dioxide. Depending on generator control,
untreated chlorine or chlorite may be found in the
generator's effluent. Chlorate is also reported to be
formed. Chlorine dioxide can also be formed from
sodium chlorite, sodium hypochlorite, and sulfuric
acid, or from sodium chlorate, sodium chloride, and
sulfuric acid. These methods are not commonly
employed.
The problems associated with the use of chlorine
dioxide are related to its byproducts and end
products. The fate of oxidized organic compounds is
not completely understood. Work is now underway
to define the reaction byproducts that occur when
humic acids, synthetic organic chemicals, or natural
waters are treated with chlorine dioxide. It has been
reported that chlorine dioxide minimizes the forma-
tion of chloroform and other THM's if properly
controlled, but it is not known what other possibly
toxic or carcinogenic substances are formed as by-
products in reaction with natural waters.
50
-------
Another problem is caused by formation of
chlorite, which reportedly has detrimental health
effects related to the blood. Work is under way to
better define these health effects and to relate them
to chlorite concentrations. Chlorite is reported to be
an end product of the reaction between chlorine
dioxide and natural waters.
Cost of Water
Disinfection
Cost of water disinfection was estimated by
Symons and colleagues in Interim Treatment Guide
for the Control of Chloroform and Other Tri-
halomethanes (8). Table 5 was taken from that publi-
cation. Disinfection costs may vary by a factor of 10
or more, depending on plant size, disinfectant used,
and dose required to attain the desired residual. In
general, chlorination is cheaper than using ozone or
chlorine dioxide.
Summary
The goal of disinfection has been and still is to
produce water that is safe to drink. In the past, this
goal was attained by killing pathogens in the water.
This remains the purpose of disinfection. In addition,
the production of potentially hazardous chemicals
during disinfection makes it necessary that all aspects
of the local situation be carefully considered before
adopting or changing disinfection processes.
TABLE 5. Estimated Cost of Disinfection
Costs,a cents per 1,000 gal
Item
Chlorination 2 mg/l, 30-minute contact time:
Chlorine at 10 cents per pound
Chlorine at 20 cents per pound
Ozone 1 mg/l, 20-minute contact time:
Ozone generated by air
Ozone generated by oxygen
Chlorine dioxide 1 mg/l, 30-minute contact time:
Chlorine at 10 cents per pound
Chlorine at 20 cents per pound
1-mgd design
capacity
1.8
1.9
4
5
3
3
10-mgd design
capacity
0.6
0.8
1.2
1.3
1.5
1.5
1 00-mgd design
capacity
0.4
0.5
0.7
0.7
1.2
1.2
aThese costs will vary at different locations, so should be considered approximate.
Note.—Sodium chlorite cost = 70 cents per pound.
51
-------
References
1. National Interim Primary Drinking Water Regula-
tions, Fed. Reg., Dec. 24, 1975.
2. Rook, J. J. Formation of Haloforms During
Chlorination of Natural Waters. Water Treat.
Exam., 23(2):234,1974.
3. Bellar, T. A., J. J. Lichtenberg, and R. C. Kroner.
The Occurrence of Organohalides in Chlorinated
Drinking Water. /. Am. Water Works Assoc.,
66:703,1974.
4. Symons, J. M., T. A. Bellar, J. K. Carswell, J.
DeMarco, K. L Kropp, G. G. Robeck, D. R.
Seeger, C. J. Slocum, B. L. Smith, and A. A.
Stevens. National Organics Reconnaissance
Survey for Halogenated Organics in Drinking
Water. /. Am. Water Works Assoc., 67(11):
634-647,1975.
5. Love, O. T., jr., J. K. Carswell, A. A. Stevens, T.
J. Sorg, G. S. Logsdon, and J. M. Symons.
Preliminary Assessment of Suspected Carcino-
gens in Drinking Water—Interim Report to
Congress. App. VI. U.S. Environmental Pro-
tection Agency, Washington, D.C., June 1975.
6. Love, O. T., Jr., J. K. Carswell, A. A. Stevens,
and J. M. Symons. Treatment of Drinking Water
for Prevention and Removal of Halogenated
Organic Compounds {An EPA Progress Report).
Presented at the 95th Annual Conference of the
American Water Works Association, Minneapolis,
Minn., June 8-12,1975.
7. Love, O. T., Jr., J. K. Carswell, A. A. Stevens,
and J. M. Symons. Pilot Plant Studies and
Measurement of Organics. Presented at 1975
Water Quality Technology Conference, American
Water Works Association, Atlanta, Ga., Dec.
8-10,1975.
8. Symons, J. M. Interim Treatment Guide for the
Control of Chloroform and Other Tri-
halomethanes. U.S. Environmental Protection
Agency, Municipal Environmental Research
Laboratory, Water Supply Research Division,
Cincinnati, Ohio, 1976.
9. Laubusch, E. J. Chlorination and Other Dis-
infection Processes. In: Water Quality and Treat-
ment. 3rd ed. McGraw-Hill, New York, N.Y.,
1971.CH. 5.
10. White, G. C. Handbook of Chlorination, Van
Nostrand Reinhold, 1972.
11. Fair, G. M., J. C. Geyer, and D. A. Okun. Water
and Wastewater Engineering. Vol. 2. John Wiley,
New York, N.Y., 1968.
12. Riehl, M. L. Water Supply and Treatment. 9th
ed. Bulletin No. 211, National Lime Association,
Washington, D.C., 1962.
13. Cox, C. R. Operation and Control of Water
Treatment Processes, World Health Organization
Monograph Series No. 49,1964.
For a more comprehensive and thorough review of disinfection, the reader is referred to "Ozone,
Chlorine Dioxide and Chloramines as Alternatives to Chlorine for Disinfection of Drinking Water:
State-of-the-Art." This document was prepared in November 1977. It is available from Water Supply
Research, Office of Research and Development, U.S. Environmental Protection Agency, Cincinnati,
Ohio 45268.
52
-------
Treatment
Techniques
for the Removal
of Organic
Contaminants
from
Drinking Water
by O.THOMAS LOVE, JR.
Water Supply Research Division, MERL
U.S. Environmental Protection Agency
Cincinnati, Ohio
The National Interim Primary Drinking Water
Regulations established maximum contaminant levels
(MCL's) for six organic chemicals: endrin, lindane,
methoxychlor, and Toxaphene—which are chlorinated
hydrocarbons-and two chlorophenoxys, 2,4-D and
2,4,5-TP (Silvex). These six specific organic contami-
nants can be grouped under the general term "pesti-
cides" and this section summarizes pertinent results
from investigators who have examined pesticide
removal (i.e., reduction in concentration) by various
water treatment techniques.
Occurrence of
Pesticides in
Water Supplies
Because of the vast and diversified use of these six
pesticides in the United States, there is certainly a
potential or opportunity for contaminating water
supplies with these materials. These organic pesticides
are not naturally occurring. They may, for example,
enter a drinking water from direct application for
control of nuisance vegetation, fish, and aquatic
insects; from nonpoint sources such as runoff from
agricultural, urban, and suburban areas; from acci-
dental spills; or, of course, from direct wastewater
discharge from a point source. A bibliography on the
occurrence and fate of pesticides in soils, aquifers,
impoundments, lakes, and water courses appears at
the end of this section, and pesticide removal by
natural processes will not be discussed further.
Because pesticide contamination is likely to be
intermittent, it presents a potentially troublesome
problem to the water plant operator. Distribution
system contamination from pesticides is the result
either of an inadvertent cross-connection or of
sabotage, and control of these situations is not
discussed in this section.
Endrin
1,2,3,4,10,10-Hexachloro-6,7-epoxy-
l,4,4a,5,6,7,8,8a-octanydro-'I,4-endo,
endo-5,8-dime thane-naphthalene
MCL: 0.0002 mg/l
Molecular Weight: 381
Threshold Odor Concentration: 0.009-0.018 mg/l
(1,2)
Odor Type: Musty and chlorinous (2)
Other Names (3,4): Mendrin, experimental insecticide
269, nendrin
Endrin, a chlorinated hydrocarbon, is a potent
organic insecticide introduced in the United States in
1951. It received a U.S. patent in 1959. Currently
there are several registered uses of endrin. This
pesticide is used primarily on field crops because it is
nonsystemic and persistent.
Lauer et al. (5) investigated a situation where
endrin, applied in sugar cane farming, contaminated a
drinking water source. The incident showed that
conventional treatment (coagulation, sedimentation,
and sand filtration) was ineffective in reducing the
contaminant. Similar reports containing field data on
treatment plant removal of endrin are scarce because
contamination is random and monitoring is not
continuous. In the community water supply
study (6), performed in 1969, 80 of the 160 samples
collected nationwide for pesticide analysis showed
53
-------
detectable but nonquantifiable traces of endrin.
These samples, however, were all collected on
finished water and the treatment processes could not
be evaluated for pesticide removal.
Most of the information on reducing various
concentrations of endrin has been gathered through
laboratory studies and pilot-scale water treatment
plant experiments. The foremost work on this topic
was conducted in the early and mid-1960's by
Robeck, Dostal, Cohen,, and Kreissl (1). This work
will be referred to frequently throughout this or-
ganics section because two of the six pesticides
studied in detail (endrin and lindane) now have
established MCL's. Table 6 summarizes the expected
removals of endrin by chlorine oxidation, conven-
tional treatment, and conventional treatment supple-
mented with powdered activated carbon (PAC) or
granular activated carbon (GAC).
TABLE 6. Endrin3 Removal at 0.010-mg/l
Load (1)
Unit process
Endrin removal,
percent
Chlorination, 5 mg/l
Coagulation and filtration
Powdered activated carbon;
5 mg/l
10 mg/l
20 mg/l
Granular activated carbon,
0.5 gal/min/ft3
35
85
92
94
>99
aMCL = 0.0002 mg/l.
Lindane
a ci
gamma (-y)isomerof 1, 2,3,4,5,6— hexa-
chlorocyclohexane
MCL: 0.004 mg/l
Molecular Weight: 291
Threshold Odor Concentration: 0.33-12 mg/l (1,2,7)
Odor Type: Chlorinous medicinal (2)
Other Names (3,4): Gammaexane, Gammopaz,
Gexane, Kwell, Lindex, Lindust, Lintox, among
many others
Lindane, a chlorinated hydrocarbon, is the most
toxic isomer of benzene hexachloride. It was dis-
covered in 1942 and is widely used as an insecticide
to control cotton insects and grasshoppers. Nicholson
et al. (8,9) conducted a study on a watershed in the
southeastern United States, where lindane and other
pesticides were used on cotton fields. Varying con-
centrations of lindane were detected throughout the
study in both the river and the effluent from a water
treatment plant using the river as a source of drinking
water. The water treatment processes consisting of
coagulation, sedimentation, filtration, and chlorina-
tion were ineffective in reducing the insecticide levels.
With field data lacking, removal of lindane has
been the target for several laboratory and pilot plant
experiments. Buescher et al. (10) took distilled, de-
ionized, and carbon-filtered water, spiked it with
several mg/l lindane and subjected it to chlorine (40
mg/l), hydrogen peroxide (40 mg/l), sodium peroxide
(40 mg/l), potassium permanganate (40 mg/l), ozone,
and aeration. It is highly unlikely (barring a lindane
spill) that the concentration of this pesticide would
ever be found in the milligram-per-liter range in a
drinking water source; however, it is of interest that
only ozone (in concentrations far in excess of
disinfection doses) had any appreciable effect on
reducing the lindane concentration. Robeck et al. (1 ),
on the other hand, used much lower pesticide
concentrations (10-20 Mg/0, yet conventional treat-
ment was ineffective, and 5 mg/l chlorine and up to
40 mg/l potassium permanganate did not oxidize or
destroy this insecticide. Ozone, however, reduced 20
-------
pressure differential, however, applied to the feed
solution was 100 atmospheres (1,470 psi). Unless the
untreated ' water is very low in turbidity some
pretreatment. for participate removal is necessary
before reverse osmosis can be effective.
TABLE 7. Lindane3 Removal by Activated
Carbon (3)
Unit process
Lindane removal,
percent
Powdered activated carbon:
5mg/l
lOmg/l
20 mg/l
Granular activated carbon,
0.5gal/min/ft3
30
55
80
>99
initial lindane level = 0.010 mg/l. MCL = 0.004
mg/l.
Toxaphene
chlorinated carnphene, 67-69 per cent
chlorine. Where average n = 8
MCL: 0.005 mg/l
Molecular Weight: 412
Threshold Odor Concentration: 0.005-0.14 mg/l (2)
Odor Type: Musty to moldy (2)
Other Names (3,4): Alltox, Estonox, Chem-phene,
Geni-phene, Gy-phene, Phenacide, Phenatox, Tox-
adust
Toxaphene, a chlorinated hydrocarbon, was intro-
duced in the United States in 1948, patented in 1951,
and registered uses are common. A major use is in
cotton farming to combat such insects as boll weevils,
bollworms, aphid, and leafworm. Contamination of a
public drinking water supply by agricultural runoff
was reported by Nicholson et al. (9), who found that
conventional water treatment practice (coagulation,
settling, filtration, and chlorination) was ineffective
in reducing Toxaphene concentrations that were
variable, but never found to exceed 0.41 /Ltg/L
Cohen et al. (15,16) experienced similar dis-
couraging removal results in the early 1960's in
laboratory studies using much larger concentrations of
this insecticide and dosing up to 100 mg/l alum for
coagulation. Moreover, neither chlorine nor chlorine
dioxide had any effect on removing Toxaphene.
Adsorption with activated carbon was more en-
couraging. An initial concentration of 0.1 mg/l Toxa-
phene was reduced to 0.007 mg/l by 5 mg/l PAC, so
the authors concluded that "no common treatment
other than that with activated carbon will remove
Toxaphene."
2,4-D
Cl—< >— O— Cft,—COOH
2,4-dichlorophenoxyacetic acid
MCL: 0.1 mg/l
Molecular Weight: 221
Threshold Odor Concentration: 3.13 mg/l (2)
Odor Type: Chlorophenol to musty (2)
2,4-D is a systemic herbicide discovered in 1944. It
was patented in the United States in 1949. Registered
products are in use (3,4). 2,4-D was one of the first
organic compounds used for weed control, and it
remains popular for that use.
Aly and Faust (17) made an excellent contribution
to the literature by evaluating the common water
treatment processes of coagulation, oxidation, and
adsorption for the removal of 2,4-D derivatives. Five
materials (the sodium salt of 2,4-D, 2,4-
Dichlorophenol, and isopropyl, butyl, and isooctyl
esters) were selected. It was found that at a 1.0-mg/l
load none of these compounds was significantly
removed by either aluminum or ferric sulfate
(100 mg/l dose) in laboratory coagulation and settling
studies. Further, chlorination up to 100 mg/l and the
addition of potassium permanganate up to 10 mg/l
were ineffective in 2,4-D removal. Powdered activated
carbon was effective, and doses required to reduce
various levels of the 2,4-D derivatives to MCL are
shown in table 8.
Whitehouse (18) conducted laboratory studies to
determine the effect of pH and types of PAC in
removing 2,4-D from solution; however, the concen-
trations of adsorbantsand adsorbates (100 mg/l level)
are thought to be too atypical in water treatment to
warrant further details on the results.
Reverse osmosis needs additional study before it
can be suggested as an effective technique for
removing 2,4-D from drinking water. Lonsdale et
al. (19) reported a 92.8-percent rejection of 2,4-D
from a 1-percent NaCI solution having an initial
herbicide concentration of 35 mg/l. Edwards and
Schubert (20) found that 2,4-D rejections from a
50-mg/l solution never exceeded 65 percent initially,
and near the end of each run rejection efficiencies
ranged from 1 to 51 percent. These studies used small
55
-------
reverse osmosis cells at pressures ranging from 80 to
1,500 psj. In addition to pretreatment considerations
with reverse osmosis, the reject water constitutes a
separate disposal problem that must be included in
the overall evaluation and cost estimation of the
method.
TABLE 8. Carbon Doses Required to Reduce
the Concentration of 2,4-D Compounds to
Maximum Contaminant Limit3 (MCL) {17}
Initial
concen-
Powdered activated carbon dose, mg/l
tration,b Sodium Isopropyl Butyl Isooctyl
mg/l salt ester ester ester
10
5
3
1
306
153
92
31
150
74
44
14
165
82
49
15
179
89
53
16
aMCL = 0.1 mg/l.
b Expressed as the acid equivalent.
2,4,5-TP (Silvex)
|—CH—COOH
CH,
2,4,5 -t ri ch I o rophen oxy p rop io n ic ac i d
MCL: 0.01 mg/l
Molecular Weight: 269
Threshold Odor Concentration: 0.78 mg/l (2)
Odor Type: Idodform (2)
Other Names (3,4): Kuron, Kurosol, Aqua Vex,
O-X-D
Silvex is a postemergent herbicide used for con-
trolling brush, aquatic vegetation, woody plants, and
certain weeds not susceptible to 2,4-D. It was
introduced in the United States in 1952. Currently
there are several registered uses of this herbicide.
Treatment data for the removal of this compound
from drinking water are not available in 1977.
Robeck et al. (1) examined the butoxy ethanol ester
of 2,4,5-T; if it is assumed (until more definitive
information becomes available) that the two com-
pounds would behave similarly in a dilute aqueous
solution, table 9 may be useful as a summary of
expected removals.
TABLE 9. 2,4,5-T
mg/l Load (3)
Ester Removal at 0.01-
Unit process
2,4,5-T ester
removed,
percent
Chlorination, 5 mg/l
Coagulation and filtration
Powdered activated carbon:
5 mg/l
10 mg/l
20 mg/l
Granular activated carbon
65
80
80
95
>99
Methoxychlor
ca,
1,1,1-trichloro-2,2-bis(p-methoxy-
phenylje thane
MCL: 0.1 mg/l
Molecular Weight: 346
Threshold Odor Concentration: 4.7 mg/i (2)
Odor Type: Musty to chlorinous {2}
Other Names (3,4): DMDT, dimethoxy-DT, dianisyl
trichloroethane, Marlate
Methoxychlor is a chlorinated insecticide used to
control external parasites on animals. Treatment
information is not available on this insecticide in
1977. It is very likely, however, that adsorption with
GAG would remove this contaminant effectively from
drinking water. (See author's note, p. 61)
Summary of
Treatment
Techniques
The organic compounds for which MCL's have
been established are the pesticides endrin, lindane,
Toxaphene, 2,4-D, 2,4,5-TP, and methoxychlor.
There is a varying amount of information on the
removal of the first four materials from drinking
water, but the author was unable to locate pertinent
studies referenced on the removal of 2,4,5-TP and
methoxychlor. Data gaps in table 10 emphasize the
need for additional research on a number of aspects
of the removal of organics from drinking water.
In spite of limited specific information, it is
apparent that adsorption is more effective than
56
-------
>s
rt
E
E
3
vT
-O
T3
O
E
OJ
'-^
t/1
_(J
'c
00
6
c
s
k.
u
Q_
O
LU
_J
m
(5
p-
^
"""*
c
.2
*d
o
D
T3
0
^
•a
Q
,j-
r-T
<5
'S
u
u
o
O
—
^
U
V
">•
3
CO
u
w
">.
Q.
O
a.
O
J2
M
«
E
3
O
t/1
o
w c
S ,2 ^o
^ o T~~
O £ ^"
.0 C
WO
JS U
oi o
D.
V
•a
•S
w
Z
*!
v
o
"O
o
Z
u £
o
Q.
TD
w
w
2
• u
U *j
^ o
Q.
U
T3
W
•a
w
w
Z
t U
w «
« 5
O.
8
u
u
o
ol
o
Q**
o
V
o
CTi
O
V
o
Oi
o
V
o
01
o
V
t^
^-^
o
V
o
\o
o
V
00
Oi
1A
^r
.2
ra
w
C
"
.2
_TS
3>
6
C 13
^ fS
o .-
^fl •£ M
" 'I -o
iC c £
- Q «
•= & o
.2 o a.
S "°
3 ™
O 1"-
p p-
O r-
OP-
O P- OO
o r- ofl
o o o
i^~ o o
o o o
V V V
i—
V
o o
V V
cc oo o oo
oeosooci
oo oo
c
Us?
E
(J
U
.2 -
S *
o .r
o
_ £
o 5
HI
0 0 0 0 .0
ll II II II o
57
-------
TABLE 11. Activated Carbon, mg/l, Required to Reduce the Pesticide Level in Distilled Water
and in Little Miami River Water (23)
10-jug/l initial level
1.0-jug/l initial level
Pesticide
2,4,5-T ester
Endrin
Lindane
Method
Jar test3
Plant treatment6
Jartesta
Plant treatment6
Jar test3
Plant treatment6
1 -0 Mg/l
after
treatment
2.5
14
1.8
11
2
29
0.1 MS/'
after
treatment
17
44
14
126
12
70
0.1 Mg/l
after
treatment
1.5
3
1.3
11
1,1
6
0.05 Mg/l
after
treatment
3
5
2.5
23
2
9
aPesticide is removed from distilled water by activated carbon alone, with a contact time of an hour.
bPesticide is removed from river water by conventional treatment and activated carbon.
conventional treatment or oxidation for pesticide
removal. The effectiveness of adsorption is influenced
by the temperature and pH of the water, but to a
greater degree adsorption depends on:
• Concentrations of adsorbant and adsorbate
• Contact or residence time
• Competition for available adsorption sites
Competitive adsorption is important to recognize
because it dramatically affects the amount of ad-
sorbant available for the contaminant. Table 11 illus-
trates this point. In one instance the pesticides are
removed from distilled water by PAC alone, and in
the other, a similar pesticide concentration is re-
moved from river water by conventional treatment
supplemented with PAC. Where pesticide was added
to the river water (i.e., where competition for the
adsorption sites existed), the PAC doses were 4 to 14
times higher than the companion noncompetitive
situation. Studies are underway to further examine
competitive adsorption (21,22). Until the phenom-
enon is better understood, it is impossible to predict
with much certainty a PAC dose effective for
variations in water quality.
Robeck (23) commented that, because pesticides
and their carrier solvents have odors, water plant
operators treating for odor removal will provide
some incidental protection by reducing certain pesti-
cides. When the comment was made, in 1972, the
MCL's in the EPA Guidelines for Pesticides in
Water (24) were quite similar to the threshold odor
concentrations (ThOC's). In 1977, however, the MCL
is more rigid and differs from the odor detection level
for some pesticides by several orders of magnitude
(see table 12). Note that the T^OC's given for these
pesticides were acquired under laboratory conditions
using "pure" compounds and odor-free water. The
odor of these pesticides (table 12) or of their reaction
products after chlorination is not known, however,
and could vary significantly (1,16,25). Relying on
odors to signal pesticide contamination, or relying on
intermittent odor control by PAC to insure a safe
pesticide level is risky and considered poor practice.
TABLE 12. Threshold Odor Concentration
(TnOC) and Maximum Contaminant Level
(MCL) of Pesticides in Water
Pesticide
Endrin
Lindane
Toxaphene
Methoxychlor
2,4-D
2,4,5-TP
ThOC (3,4,17,15),
mg/1
0.009-0.01 8
0.33-1 2
0.005-0.14
4.7
3.13
0.78
MCL, mg/l
0.0002
0.004
0.005
0.1
0.1
0.01
A few milligrams per liter of PAC may be adequate
for odor control, but, according to available pesticide
treatment information, several milligrams per liter are
required to effect organic removals, and the sludge
created is sometimes troublesome and difficult to
manage. A plant operator using PAC should consider
multiple points of injection for maximum efficiency
58
-------
of the adsorbant and possibly for better
removals (1,26).
Because of the uncertainties involved in pesticide
occurrence, GAC beds that are continuously on line
offer the best barrier against pesticide contamination
(23,26). Organic pesticides have been demonstrated
to be very strongly adsorbed both on virgin GAC and
on GAC considered exhausted for odor control (1).
The life of a GAC bed for pesticide removal is not
indefinite, but it has been suggested that replacing or
reactivating a GAC bed because of odor penetration is
an adequate guideline for controlling pesticides (23).
Granular activated carbon can be used directly (no
pretreatment) with low-turbidity ground waters or as
a filter/adsorber in a conventional or direct filtration
plant.
Estimating Cost
for Reducing
Trace Organics
(Pesticides)
Below MCL
ADSORPTION WITH PAC
The following assumptions are made in estimating
the cost of adding PAC for controlling pesticides in
drinking water:
• A filtration plant already exists.
• PAC is already being fed for odor control.
• Sludge-handling facilities and disposal are adequate
for the additional imposed loadings.
• Analytical costs for pesticide analysis are
excluded.
• PAC costs 30 cents per pound.
• A PAC dose of 5-80 mg/l would allow for a
contamination level from 2.5 to 50 times the
MCL
On these assumptions the estimated cost would be
1.2-20 cents per 1,000 gallons.
ADSORPTION WITH GAC
One of the most rigorous and complete documents
for estimating GAC costs was developed by Clark and
coworkers (27). Although not addressed to pesticide
removal, most of the parameters and indices used are
applicable to any fixed-bed adsorption process. If it is
assumed that penetration of odor compounds pre-
cedes pesticide breakthrough, costs for odor removal
by GAC should be an estimated cost for pesticide
removal. For example, an existing 1-mgd filtration
plant operating at 70-percent capacity replaces 30
inches of sand with GAC, which effectively removes
odors for 12 months. The adsorption {pesticide
removal) process would cost an estimated 3.3 cents
per 1,000 gallons. It is assumed that:
• GAC costs 38 cents per pound and the amount
needed varies directly with the plant size.
• Interest and labor costs are negligible.
• Attrition loss is small and GAC is replaced when
exhausted (no on-site reactivation).
Table 13 then summarizes the expenditures that
might be expected for pesticide removal by GAC for,
time periods appropriate for the expected life of GAC
for odor removal. Hansen(1,28) has studied very
closely the actual costs for GAC adsorption at his
water utility in Mount Clements, Michigan. His costs,
shown in table 13, are lower than the estimate
because of differences in the unit costs and amounts
of GAC used in each situation.
TABLE 13. Costs of Granular Activated Car-
bon (GAC) Adsorption for Pesticide
Removal
GAC replacement frequency,
Costs, cents per months
12
Estimated 3.3
Actual (12,28)
24 36
1.6 1.1
0.5
48
0.8
0.4
References
1. Robeck, Gordon G., Kenneth Dostal, Jesse
Cohen, and James Kreissl. Effectiveness of Water
Treatment Processes in Pesticide Removal. J.
Am. Water Works Assoc., 57:181-199, 1965.
2. Sigworth, E. A. Identification and Removal of
Pesticides and Herbicides, j. Am. Water Works
Assoc., 57:1016-1022,1965.
3. Cleaning Our Environment—The Chemical Basis
for Action. Report, American Chemical Society,
Committee on Chemistry and Public Affairs,
Subcommittee on Environmental Improvement,
Washington, D.C., 1969. Pp. 193-197.
59
-------
4. Packer, K., ed. Nanogen Index—A Dictionary of
Pesticides and Chemical Pollutants. Nanogen
International, Freedom, Calif., 1975. Pp. 36-90.
5. Lauer, E. J., H. P. Nicholson, W. S. Cox, and J. I.
Teasley. Pesticide Contamination of Surface
Waters by Sugar Cane Farming in Louisiana.
Trans. Am. Fish, Soc., 95:310,1966.
6. McCabe, L. J., j. M. Symons, R. D. Lee, and G.
G. Robeck. Survey of Community Water Supply
Systems, J. Am. Water Works Assoc., 62:670,
1970.
7. Faust, S. D., and O. M. Aly. Water Pollution by
Organic Pesticides, j. Am. Water Works Assoc.,
56:267-274,1965.
8. Nicholson, H. P. Pesticide Pollution Studies in
the Southeastern U.S. Robert A. Taft Engi-
neering Center, Cincinnati, Ohio, 1962.
9. Nicholson, H. P., A. R. Grzenda, and J. I.
Teasley. Water Pollution by Insecticides, A Six
and One-half Year Study of a Water Shed. In:
Proceedings of the Symposium on Agricultural
Waste Water, U.S. Environmental Protection
Agency, Athens, Ga., 1966. P 132.
10. Buescher, C. A., J. H. Dougherty, and R. T.
Skrinde. Chemical Oxidation of Selected Organic
Pesticides. /. Water Pollut. Control Fed., 36:
1005-1112,1964.
11. Moergeli, B. Removal of Pesticides from Drinking
Water, Sulzer Tech. Rev., 54(2):91-6, 1972.
12. Hansen, R. E., Organics Removal of Mount
Clemens, Michigan. Presented to Michigan Sec-
tion of the American Water Works Association,
Ann Arbor, Mich., Jan. 1977.
13. Smola, D. J. Removal of Toxic Pesticides by
Reverse Osmosis Water Treatment. Master's
thesis, Massachusetts University, Amherst De-
partment of Civil Engineering, Dec, 1968. P. 104.
14. Hindin, E., Baul J. Bennett, and S. S. Narayanan.
Organic Compounds Removed by Reverse Osmo-
sis. Water Sewage Works, 776:466^71,1969.
15. Cohen, J. M., L. J. Kamphake, A. E. Lemke, C.
Henderson, and R. L. Woodward. Effects of Fish
Poison on Water Supplies. Part 1. Removal of
Toxic Materials. /. Am. Water Works Assoc.,
52:1551-1566,1960.
16. Cohen, J. M., G. A. Rourke, and R. L. Wood-
ward. Effect of Fish Poisons on Water Supplies.
Part 2. Odor Problems. /. Am. Water Works
Assoc., 53:49-57,1961.
17. Aly, O. M., and S. D. Faust. Removal of 2,4-D
Derivatives from Natural Waters. J. Am. Water
Works Assoc., 57:221-230,1965.
18. Whitehouse, J. D. A Study of the Removal of
Pesticides from Water. University of Kentucky,
Water Resources Institute, Research Report No.
8, Lexington, Ky. 1967.
19. Lonsdale, H. K., C. E. Milstead, B. P. Cross, and
F. M. Graber. Study of Rejection of Various
Solutes by Reverse Osmosis Membranes. Rand D
Report No. 447. Office of Saline Water, Washing-
ton, D.C., July 1969. 72 pp.
20. Edwards, V. H., and Paul F. Schubert. Removal
of 2,4-D and Other Persistent Organic Molecules
from Water Supplies by Reverse Osmosis. /. Am.
Water Works Assoc., 66:610-616,1974.
21. EPA Grant No. R8034730. Activated Carbon
Adsorption of Trace Organic Compounds. Re-
search in progress. Principal investigator: V.
Snoeyink, University of Illinois. Project officer:
A. Stevens, MERL, WSRD, Cincinnati, Ohio
45268. Completion date: summer 1977.
22. EPA Grant No. R804639. Effectiveness of Acti-
vated Carbon for Removal of Toxic and/or
Carcinogenic Components from Water Supplies.
Research in progress. Principal investigator: W.
Weber, University of Michigan. Project Officer:
A. Stevens, MERL, WSRD, Cincinnati, Ohio
45268. Completion date: 1979.
23. Robeck, Gordon G. Purification of Drinking
Water to Remove Pesticides and Other Poisonous
Chemicals: The American Practice. In: Proceed-
ings of the 9th Congress of the Internationa/
Water Supply Association, London, U.K.,
September 11-14,1972.
24. Unpublished working manuscript, U.S. Environ-
mental Protection Agency, Water Supply Pro-
grams, Washington, D.C., Oct. 1971.
25. Woodward, R. L. Significance of Pesticides in
Water Supplies, j. Am. Water Works Assoc.,
52:1367-1372,1960.
26. Love,O. T., Jr., J. K. Carswell, A. A. Stevens, and
J. M. Symons, Evaluation of Activated Carbon as
a Drin king Water Treatment Unit Process.
Mimeo, U.S. Environmental Protection Agency,
Cincinnati, Ohio, Mar. 3,1975. 17 pp.
27. Clark, R, M., D. L. Guttman, J. L. Crawford, and
J. A. Machifko. The Cost of Removing Chloro-
form and Other Trihalomethanes From Drinking
Water Supplies. In: James M. Symons, Interim
Treatment Guide for the Control of Chloroform
and Other Trihalomethanes. U.S. Environmental
Protection Agency, Cincinnati, Ohio, June 1976.
App. 1.
60
-------
28. Hansen, Robert E. Problems Solved During 92
Months of Operation of Activated Granular
Carbon Filters. In: Proceedings of the 3rd
Annual AWWA Water Quality Technology Con-
ference. Atlanta, Ga., Dec. 1975.
Bibliography—
Occurrence and
Fate of Pesticides
in the Environment
Aly, O. M., and S. D. Faust. Studies on the Fate of
2,4-D and Ester Derivatives in Natural Surface
Waters. J. Agric. Food Chem., 72:541-544,1964.
Boucher, Francis R., and G. Fred Lee. Adsorption of
Lindane and Dieidrin Pesticides on Unconsoli-
dated Aquifer Sands. Environ. Sci. Techno!.,
6.6;53S-543.
Brown, E., and Y. A. Nishioka. Pesticides in Western
Streams, A Contribution to the National Pro-
grams. Pest. Monit. J., 7:38, 1967.
Crosby, D. G., and H. O. Tutass. Photodecomposition
of 2,4-D, J. Agric. Food Chem., 74:596-599,
1966.
DeMarco, J., J. M. Symons, and G. G. Robeck.
Synthetic Organics in Stratified Impoundments,
J. Am. Water Works Assoc., 59, 1967.
Edwards, C. A. Persistent Pesticides in the Environ-
ment. Chemical Rubber Company, Cleveland,
Ohio, 112-113,1973.
Huang, Ju-chang. Effect of Selected Factors on
Pesticide Sorption and Desorption in the Aquatic
System. /. Water Pollut. Control Fed., 43(8):
1739-1748,1971.
Huang, Ju-Chang, and Cheng-Sun Liao. Adsorption of
Pesticides by Clay Minerals. ASCE Sanit. Eng.
£>/c., 5/15:1003-1077, 1970.
King, Paul H., H. H. Yeh, Pierre S. Warren, and
Clifford Randall. Distribution of Pesticides in
Surface Waters, /. Am. Water Works Assoc.,
67:483-486,1969.
Leigh, G. M. Degradation of Selected Chlorination
Hydrocarbon Insecticides. J. Water Pollut. Con-
fro/Fed., 47(11 ):R450,1969.
Rosen, A. A., and F. M. Middleton, Chlorinated
Insecticides in Surface Waters. Anal. Chem.,
37:1729-1733,1959.
Weaver, L. G., G. G. Gunnerson, A. W. Breidenbach,
and J. L. Lichtenberg. Chlorinated Hydrocarbon
Pesticides in U.S. River Basins. U.S. Public
Health Report No. 80, 1965. Pp. 481-493.
Author's note: Re: Methoxychlor
The effectiveness of granular activated carbon to remove methoxychlor from drinking water has been demon-
strated in West Germany* at the ng/1 level and recently in the United States** at the ug/1 level.
*Schmidt, K., Effectiveness of Standard Drinking Water Preparation for the Eliminating Pesticides and Other
Pollutants. Gas. Wasser-facn Wasser-Abwasser, 115, No. 2:72-76,1974.
**Steiner, John IV and J.E. Singley. Methoxychlor Removal from Potable Water. Report submitted for publica-
tion. Department of Environmental Engineering Sciences, University of Florida, Gainsville, FL, 1977.
61
-------
Treatment
Techniques
for the Removal
of Radioactive
Contaminants
from
Drinking Water
by GARY S. LOGSDON
Water Supply Research Laboratory, MERL
U.S. Environmental Protection Agency
Cincinnati, Ohio
The National Interim Primary Drinking Water
Regulations (1) established maximum contaminant
levels (MCL's) for two categories of radioactive
contaminants—alpha emitters and beta and photon
emitters. The beta and photon emitters are generally
manmade radioisotopes rather than naturally occur-
ring ones. Because terms and definitions associated
with radioactivity are not in everyday use by most
water utility people, a glossary is included at the end
of this section.
Different types of ionizing radiation may cause
different levels of biological damage, even though
the amount of energy involved is the same for each
type of radiation. For this effect, the term "quality
factor" (QF) is used. The QF is a property of the
nature and energy of the absorbed radiation. A high
QF indicates that the type of radiation in question has
a greater potential for causing biological damage,
whereas a low QF indicates that the radiation in
question would be less biologically damaging if the
absorbed energy were equal in both cases. Table 14
gives some of the relationships among the various
radiation units.
In general, for beta and gamma radiation the
absorbed dose and biological effect or damage are
related 1 to 1. Alpha particles, however, are assumed
to be 10 times more damaging, as compared to beta
and gamma radiation, rad for rad, because one rad of
alpha radiation would be absorbed in a very small
volume of biological material, so that each cell would
be exposed to more ionizing radiation.
TABLE 14. Relationships Among Radiation
Units
Quality factor
Type of radiation
R rads (QF) (2) rem
X-rays and gamma rays 1
Beta particles —
Thermal neutrons —
Fast neutrons —
Alpha particles —
1
1
1
1
1
1
1a
5
10
10
1
1a
5
10
10
aln 1966, the International Commission on Radia-
tion Protection (ICRP) recommended that the QF for
low energy beta radiation less than 0.03 MeV be
assigned a value of 1.7. In 1969 the ICRP amended its
earlier recommendation, suggesting the use of 1 as a
QF for all beta radiation. The National Council on
Radiation Protection and Measurements recommends
aQFofl.
Alpha Emitters
MAXIMUM CONTAMINANT LEVELS
The MCL's for alpha emitters are 5 pCi/l for
radium-226 and radium-228, and 15pCi/l for gross
alpha activity including radium-226, but excluding
radon and uranium.
62
-------
RADIUM IN WATER SUPPLIES
Radium is present in water as a naturally occurring
element, primarily in ground waters and to a lesser
extent in surface waters. Studies by Hursh (3) on the
water supply sources of 41 cities in the United States
showed the average radium-226 concentration of
those municipal water supplies that used surface
sources to be less than 0.3 pCi/l, ranging from 0.002
to 3.7 pCi/l. Numerous studies on ground water
supplies in areas of radium-bearing deposits have
demonstrated radium-226 concentrations ranging
from about 0.5 pCi/l to more than 50 pCi/I. The
EPA (4) has estimated that as many as 500 public
water supplies may exceed the 5-pCi/l radium
concentration.
The major radiation problems in ground waters are
thought to be caused by leaching of radium from
radium-bearing rock strata into the deep sandstone
aquifers in Iowa and Illinois, and by leaching of
radium from phosphate rock deposits into the Florida
aquifer. Elevated radium levels have also been asso-
ciated with surface run-off water in the vicinity of
uranium-rich deposits in Colorado and New Mexico.
REMOVAL OF RADIUM
FROM WATER
The water treatment plant is the point of control
between radium dissolved or suspended in raw water
supplies and the consumer. To meet the interim
standard of 5 pCi/l in a cost-effective manner, it is
important to understand how the treatment process
affects the level of radium in potable water.
The major problem of concern in public water
supplies is soluble radium in ground water. Soluble
radium exists in water as a divalent ion, similar in
chemical behavior to calcium and magnesium. Soften-
ing treatment methods have been shown to be
effective in removing 70-99 percent of dissolved
radium, as have membrane desalting methods. Coagu-
lation without softening may remove up to 25
percent of radium; however, the results are variable
and difficult to control.
The three methods selected for analysis in this
document are:
• Lime or lime-soda softening (precipitative soft-
ening)
• Ion exchange softening
• Reverse osmosis
Radium removal efficiencies and associated oper-
ating data from water treatment plants in Illinois (5),
Iowa (6), and Florida (7)'have been compiled and
analyzed herein. Removal efficiencies for each of
three treatment methods and associated costs of
treatment and waste disposal are reported for each
type of plant in the following sections.
Lime or Lime-Soda Softening. Radium removal by
lime softening can be related to hardness removal
(fig. 47) and pH of treatment (fig. 48). Lime soften-
ing can remove 80-90 percent of the radium; there-
fore, it is suitable for raw waters containing up to
25 pCi/l. To achieve these removals, the process pH
would have to be above 10.0.
Ion Exchange Softening. Radium removal by ion
exchange is related to hardness removal. Well-
operated ion exchange plants can remove 95 percent
or more of the radium in raw water (fig. 49). Because
radium removal still takes place for a period of time
after the resin ceases to remove hardness (Ca+2 and
M§+2)> regeneration to achieve good hardness re-
moval will assure good radium removal. Naturally,
blending of raw and softened water recontarninates
the treated water with radium. This practice is
common for ordinary municipal zeolite softening
plants, but it could result in production of a water
exceeding the radium MCL in some instances. Blend-
ing must be given careful study before it is used at a
radium removal plant.
Reverse Osmosis. Osmosis is the spontaneous
passage of liquid from a dilute to a more concen-
trated solution across an ideal semipermeable mem-
brane that allows passage of the liquid but not of
dissolved solids. Reverse osmosis is a process in which
the natural osmotic flow is reversed by the applica-
tion of pressure to the concentrated solution suffi-
cient to overcome the natural osmotic pressure of the
less concentrated (dilute) solution. When the amount
of water passing in either direction is equal, the
applied pressure can be defined as the osmotic
pressure of the dilute solution having that particular
concentration of solutes.
In practical applications, pumps are used to supply
the pressure to overcome osmotic pressure. The water
flow rate through the membrane depends primarily
on the net driving pressure. The solute flow rate
through the membrane depends almost solely on the
solute concentration of the feed water.
The pumping pressure required to provide the
driving force in the reverse osmosis process is a direct
function of the concentration of dissolved solids in
the feed. Reverse osmosis applications have been
63
-------
1.00
0.80
0.60
0.40
O Elgin, III., water at EPA pilot plant (8)
• West Des Moines, Iowa
^c Webster City, Iowa, without soda ash
3|c Webster City, Iowa, with soda ash
X Peru, III., three dates A
A Englewood, Fla.
A Venice, Fla.
0 0.20
FIGURE 47
*
1 .00
1 .20
0.40 0.60 0.80
RADIUM REMOVAL FRACTION
LIME-SODA PROCESS, TOTAL HARDNESS RE-
MOVAL FRACTION VERSUS RADIUM REMOVAL
FRACTION (7)
1.00
o 0.80
H
Ll_
_I
>0.60
5
LU
&.
s
3
9 0.40
<
a:
0.20
C^^
0 *&^^^
_j; ***^*\
"A O Elgin, III., water at EPA pilot plant
(8)
• West Des Moines, Iowa
-%. Webster City, Iowa, without soda ash
3JC Webster City, Iowa, with soda ash
X Peru, III.
^ Elgin, III.
A Englewood, Fla.
A Venice, Fla
1 1 1 1 1 L
FIGURE 48
9 10 11
pH OF TREATMENT
RADIUM REMOVAL FRACTION VERSUS pH
OF TREATMENT, LIME-SODA PROCESS (7)
-------
Ill III II
0.60 0.80
TOTAL HARDNESS REMOVAL FRACTION
RADIUM REMOVAL FRACTION VERSUS TOTAL HARD-
NESS REMOVAL FRACTION IN ION EXCHANGE PLANTS,
BEFORE BLENDING (7)
FIGURE 49
primarily for feed water with total dissolved solids
(TDS) above a minimum of 2,000 mg/l, and usually
in the range of 4,000-35,000 (sea water) mg/l TDS.
A characteristic of semipermeable membranes used
for reverse osmosis is that their rejection is greater of
multivalent ions, such as Ca*2, Mg"1"2, Ra+2, and
SC>4~2, than of monovalent ions Ha*1, C!"1, and so
forth. The primary advantages of reverse osmosis are
its high rate of rejection of dissolved solids in the raw
water and its suitability for use in small systems.
There are some disadvantages to reverse osmosis,
including:
• High initial and operating costs
• Need for pretreatment of raw water with turbidity
removal; treatment with acid and other chemicals
to prevent fouling of the membranes by slimes,
suspended solids, iron, and manganese; and precip-
itation of calcium carbonate and magnesium hy-
droxide
• Need to stabilize finished water with lime or other
chemicals to prevent corrosion in distribution
system
Table 15 presents radium removal data from two
reverse osmosis plants. The Greenfield plant removed
93 percent of the TDS and 96 percent of the
radium-226. The difference between TDS removal
and radium removal results from the amount of
monovalent ions that passed through the membrane.
It will be assumed, for purposes of this report, that
a well-operated reverse osmosis unit can remove 95
percent of the influent radium activity.
DISPOSAL OF
TREATMENT WASTE
Each of the treatment processes for removing
radium from potable water generates a waste stream
of some sort. These wastes must be disposed of in an
environmentally acceptable manner. This section is
based on a more comprehensive report prepared by
Singley etal. (7).
Methods for Lime Sludge Disposal. Alternatives
for disposal of lime sludges are numerous and varied.
There follow several of the more important..
65
-------
TABLE 15. Radium Removal in Reverse Osmosis Plants
Radium
IDS
Plant
Greenfield, Iowa
Sarasota Bay MHP, Fla. (9)
In, pCi/l
14.0
22.0
Out, pCi/l
0.6
0.8
Percent
removal
96
96
In, mg/l
2,160
Out, mg/l
164
Percent
removal
92
• Discharge
— To sanitary sewers
— To local receiving water
— By wet pumping or trucking to local sanitary
landfill
• Storage
— Permanent lagooning
— Sanitary landfill
a. with prior temporary lagooning
b. with prior mechanical dewatering: vacuum
filtration, centrifugation, others
— Other natural or manmade depressions (all with
some dewatering before transportation)
a. strip mine areas
b. borrow pits and quarries
c. others
• Use
— Direct without drying: farmland and pasture-
lands
— With prior dewatering
a. farmland and pastureland
b. road stabilization
c. calcination and recycle
• Disposal
— Direct, recharge to aquifers
— With prior dewatering: salt mines, coal mines,
and so forth
— As a nuclear waste
Methods for Lime Softening Backwash Disposal.
Alternatives for disposal of filter backwash are fewer
than for the lime sludges. Some methods will depend
on location, plant capacity, and operational factors.
Several of the more important alternatives follow.
• Discharge
— To sanitary sewer
— To local receiving water
• Storage
— Tanks or lagoons
a. for settling and decanting into receiving
water
b. for settling and pumping supernatant back
to plant
• Disposal as a nuclear waste
Ion Exchange Brine Disposal. One of the problems
created by sodium cycle ion exchange softening is the
disposal of spent brine from the regeneration cycle.
In view of the increasing water pollution control
requirements, these high salinity waters may face
severe limits on discharge. The problem becomes even
more sensitive when the waste contains elevated levels
of radium.
The waste products from the brine and rinse cycle
are composed primarily of the chlorides of calcium
and magnesium and the excess salt necessary for
regeneration. The total solids in a composite sample
may vary from an average concentration of 50,000-
100,000 mg/l to a maximum of 70,000-200,000 mg/l.
Disposal techniques may be limited by considera-
tions of salinity rather than radium concentration. A
list of potential alternatives for handling the waste-
water streams follows:
• Discharge
— To sanitary sewer
— To local receiving water
a. streams
b. oceans
• Storage
— Evaporation lagoons
— Land spreading
• Use-recovery
• Disposal
— In deep aquifers
— In oil well fields
— As nuclear wastes
Disposal of Reverse Osmosis Waste. Dissolved
solids rejected by the membrane in a reverse osmosis
unit flow from the unit in a more concentrated waste
stream in a continuous flow. The Greenfield plant
was reported to convert 67 percent of the flow to
potable water, wasting 33 percent of the raw water
flow as brine (10). Because the waste is produced
continuously in large volumes, waste strength (3-4
times the raw water concentration) is lower than ion
exchange brine strength. Disposal to a sewer may be
feasible for reverse osmosis waste.
66
-------
TREATMENT AND
DISPOSAL COSTS
TO REMOVE
ALPHA EMITTERS
Costs of water treatment for radium removal were
calculated by Singley et al. (7). Their computations
were based in part on work by Volkert (11). Singley
et al. based cost calculations on the need to treat
waters with low, medium, and high TDS concentra-
tions and low, medium, and high radium concen-
trations {see table 16).
The costs of treatment for radium removal are
indicated in figures 50 through 55, based on treating
to the radium standard of 5 pCi/l raw waters with
radium concentrations of 7.5, 20, and 50 pCi/l,
respectively.
Unit costs of treatment for radium removal de-
crease as plant capacity increases, but increase with
higher TDS values and higher raw water radium
concentrations. The estimated costs in figures 50-55
do not include waste treatment costs. Reverse osmo-
sis is the most expensive process, although it is the
TABLE 16. Raw Water Quality Concentra-
tions Assumed for Calculations of Radium
Removal Costs
Item
High level solids:
TDS
TH
Ca+2
Mg+2
Alk
HCO3-1
Medium level solids:
TDS
TH
Ca+2
Mg+2
Alk
HCO3-1
Low level solids:
TDS
TH
Ca+2
Mg+2
Alk
Radium concentrations:
Low level
Medium level
High level
mg/l
as CaCO3
-
750
500
250
300
-
-
300
200
100
200
-
-
150
100
50
100
mg/' pCi/l
as ion
2,000
-
200
60
360
-
1,000
-
80
24
244
-
400
-
40
12
122
7.5
20
50
most suitable for automated plant operation and use
in small plants. Ion exchange, which is generally used
in a batch process, is estimated to be the least costly.
Lime softening has most frequently been the process
of choice for large treatment plants.
Manmade
Radionuclides,
or Beta and Photon
Emitters
MAXIMUM CONTAMINANT LEVELS
The average annual concentration of beta particle
and photon radioactivity from manmade radio-
nuclides in drinking water shall not produce an
annual dose equivalent to the total body or any
internal organ greater than 4 mrem/yr.
Except for the radionuclides listed in table 17, the
concentration of manmade radionuclides causing
4 mrem total body or organ dose equivalents shall be
calculated on the basis of a 2-1/d drinking water
intake using the 168-hour data listed in Maximum
Permissible Body Burdens and Maximum Permissible
Concentration of Radionuclides in Air or Water for
Occupational Exposure (12), as amended August
1963. If two or more radionuclides are present, the
sum of their annual dose equivalent to the total body
or to any organ shall not exceed 4 mrem/yr.
TABLE 17. Average Annual Concentrations
Assumed to Produce a Total Body or Organ
Dose of 4 mrem/yr
RadionucMde
Critical organ
pCi/l
Tritium
Strontium-90
Total body
Bone marrow
20,000
8
If other beta or gamma emitters are present in the
water, the nuclides should be identified and quanti-
fied, so that a health physics expert can calculate the
estimated dose in millirems per year resulting from
drinking 2 l/d of the water. Article 141.16 of the
Interim Drinking Water Regulations states that this
shall be done if the gross beta particle activity exceeds
50 pCi/l.
It is not anticipated that the limits for beta and pho-
ton radiation will be exceeded. Rather, the standard
has been promulgated to prevent the degra-
67
-------
10,000
£ 1,000
100
^ Reverse osmosis
ITTl Lime softening
OJ Ion exchange
J L
J 1—L
0.01
FIGURE 50
0.1
100
500
1 10
PLANT CAPACITY, mgd
CAPITAL COST FOR WATER TREATMENT PLANTS FOR
7.5pCi/l RADIUM-226IN RAW WATER
10
0.1
-" Reverse osmosis
QD Lime softening
HE Ion exchange
0.01
FIGURE 51
0.1
100
500
1 10
PLANT CAPACITY, mgd
ANNUAL PRODUCTION COSTS (OPERATION, MAINTE-
NANCE, AMORTIZATION) FOR 7.5 pCi/l RADIUM-226 IN
RAW WATER
-------
10,000
t 1,000
100
— Reverse osmosis
[HI Lime softening
[[[] Ion exchange
I , , , . , ,,.!
0.01
FIGURE 52
0.1
100
500
1 10
PLANT CAPACITY, mgd
CAPITAL COST FOR WATER TREATMENT PLANTS FOR
20pCi/l RADIUM-226 IN RAW WATER
•^ Reverse osmosis
QE Lime softening
OH Ion exchange
0.1
0.01
FIGURE 53
0.1
100
500
1 10
PLANT CAPACITY, mgd
ANNUAL PRODUCTION COSTS (OPERATION, MAINTE-
NANCE, AMORTIZATION) FOR 20 pCi/l RADIUM-226 IN
RAW WATER
-------
10,000
1,000
100
0.01
FIGURE 54
0.1
Reverse osmosis
Lime softening
ITQ Ion exchange
100
1 10
PLANT CAPACITY, mgd
CAPITAL COST FOR WATER TREATMENT PLANTS
50pCi/l RADIUM-226 IN RAW WATER
500
FOR
10
0.1
— Reverse osmosis
01 Lime softening
Hfl Ion exchange
11 nl
0.01
FIGURE 55
0.1
100
500
1 10
PLANT CAPACITY, mgd
ANNUAL PRODUCTION COSTS (OPERATION, MAINTE-
NANCE, AMORTIZATION) FOR 50 pCi/l RADIUM-226 IN
RAW WATER
-------
dation of existing waters rather than to bring about a
massive cleanup program. The Interim Drinking Water
Regulations (13) stated:
The 4 millirem per year standard for man-made
radioactivity was chosen on the basis of avoid-
ing undesirable future contamination of public
water supplies as a result of controllable human
activities. Given current levels of fallout radio-
activity in public water supply systems and
their expected future declines and the degree of
control on effluents from the nuclear industry
that will be exercised by regulatory authorities,
it is not anticipated that the maximum con-
taminant levels for man-made radioactivity will
be exceeded except in extraordinary circum-
stances.
MONITORING FOR BETA
AND PHOTON EMITTERS
Monitoring requirements for manmade radiation
are stated in the Drinking Water Regulations, in part,
as:
(1) Within two years of the effective date of
this part, systems using surface water sources
and serving more than 100,000 persons and
such other community water systems as are
designated by the State shall be monitored for
compliance ...
Required monitoring is limited to systems serving
more than 100,000 persons and using surface water,
because EPA felt that these systems would be more
likely to have manmade radioactive contamination.
Small supplies are not required to be monitored
because of analytical costs and the number of
laboratories available to do the radiochemical
analyses.
REMOVAL OF MANMADE
RADIOACTIVITY
Because the beta and photon radiation limit is not
expected to be exceeded, detailed treatment guide-
lines are not necessary. However, certain funda-
mentals should be set forth. It is the purpose of this
guideline to discuss only the basic concepts of
radioisotope removal.
If a drinking water has gross beta activity ex-
ceeding 50pCi/l, Interim Drinking Water Regulations
require that an analysis be performed to determine
the major radioactive constituents present. The regu-
lations also require that organ and body doses be
calculated. If the 4-mrem/yr dose limit is exceeded,
treatment for removal of a portion of the radio-
activity would be required.
Although it would be necessary to remove a
radioisotope because of its radioactive properties, the
actual removal technique should be related to the
chemical properties of the radionuclide. In water
treatment processes, radioactive substances, such as
strontiu m-90, behave the same as the nonradioactive,
or stable, element. Thus, treatment techniques cannot
be given for beta emitters or gamma emitters as a
class, but each radioisotope must be considered
separately when it is found in water.
Data on removal of a number of specific radio-
nuclides and on fission product mixtures have been
summarized by Straub (14). The most effective con-
ventional water treatment techniques were lime soft-
ening and ion exchange softening. Excess lime
softening followed by filtration achieved strontium
removals of 87-96 percent. Ion exchange with green-
sand gave 75 percent removal for yttrium-91 and
96+ percent removal of scandium-46, strontium-89,
and a barium-140-lanthanum-140 mixture.
Other data indicate that reverse osmosis should be
effective for removing radioactivity when the contam-
inants are in a dissolved ionic form. Hauck and
Sourirajan reported 96.5 percent removal for stron-
tium chloride fed at 485 mg/l (15). In a discussion of
heavy metals removal tests with 500 mg/l of solutes,
the authors stated that "the same degree of separa-
tion can be expected at lower concentrations..."
The ability of reverse osmosis to remove very
dilute ions from solution was demonstrated in the
data on radium removal given earlier. The quantities
of radium in the raw or feed water were in the
picogram range, yet reverse osmosis gave radium
removals of greater than 90 percent.
Investigations have shown that heavy metals (Cd,
Cr, Cu, Zn), barium, and cesium can be removed
by reverse osmosis. Mixon (16) demonstrated re-
movals of 90 percent and more for Ba, Cd, Cr, Cu,
and Zn. Lonsdale et al. (17) reported removals of
98 percent or more for cesium chloride and strontium
chloride.
Studies by Russian investigators (18) showed
reverse osmosis to be effective for treating low level
wastes. A model wastewater containing NaNO3 in
amounts from 0.5-32 g/l was spiked with beta radio-
activity (0.1-100^Ci/l) by adding iodine-131,
cerium-144, cesium-137, zirconium-95, strontium-83,
or cobalt-60. The authors reported 90 percent re-
movals. Unpublished data show 95 percent or better
removal of cesium-134, cesium-137, and cobalt-60 by
reverse osmosis when these nuclides were present in
71
-------
the feed water in the microcurie-per-liter concentra-
tion range. Treatment of a low-level radioactive
laundry waste by reverse osmosis resulted in removal
of more than 99 percent of the radioactivity (19).
Reverse osmosis appears to be a very promising
way to remove most of the multitude of possible beta
and photon emitters. Good engineering practice,
however, would be to perform pilot-scale tests with a
small reverse osmosis unit before installing a full-scale
treatment plant.
Costs for removal of beta and photon emitters
were not given by the authors. For a preliminary
estimate of costs, the radium removal costs given for
lime softening, ion exchange, and reverse osmosis
should suffice.
Glossary
Dose equivalent. The product of the absorbed dose
from ionizing radiation and such factors as ac-
count for differences in biological effectiveness
caused by the type of radiation and its distribution
in the body as specified by the International
Commission on Radiological Units and Measure-
ments.
Gross alpha particle activity. The total radioactivity
resulting from alpha particle emission as inferred
from measurements on a dry sample.
Gross beta particle activity. The total radioactivity
resulting from beta particle emission as inferred
from measurements on a dry sample.
Manmade beta particle and photon emitters. All
radionuclides emitting beta particles and/or
photons listed in NBS Handbook 69 (12), except
the daughter products of thorium-232, uranium-
235, and uranium-238.
Picocurie (pCi). That quantity of radioactive material
producing 2.22 nuclear transformations per
minute.
Quality factor (QF). Related to the potential a type
of radiation has for causing biological damage. QF
is related to the energy deposited by radiation per
unit distance in absorbing tissue. It is more
harmful biologically to deposit a unit of energy in
a very short distance than to distribute the energy
deposit over a long distance.
Rad. The energy released when ionizing radiation
absorbed is measured in rads, the radiation ab-
sorbed dose. The rad is defined as the dose of any
ionizing radiation that is accompanied by the
liberation of TOO ergs of energy per gram of
absorbing material.
Rem. The unit of dose equivalent from ionizing
radiation to the total body or any internal organ
or organ system. A millirem (mrem) is 1/1000 of a
rem. Rem was also defined and explained by
Glasstone (20) as, "1 rem is taken to be the
quantity of radiation which produces the same
biological damage in man as that resulting from
the absorption of 1 rad of X-rays or gamma rays."
References
1. National Primary Drinking Water Regulations.
Fed. Reg., July 9,1976.
2. Basic Radiation Protection Criteria. NCRP Re-
port No. 39. National Council on Radiation
Protection and Measurements, Bethesda, Md.,
1971.
3. Hursh, John B. Radium Content of Public Water
Supplies. J. Am. Water Works Assoc., 63(Jan.),
1954.
4. Statement of Basis and Purpose for the Proposed
National Interim Primary Drinking Water Regula-
tions, Radioactivity. U.S. Environmental Protec-
tion Agency, Washington, D.C., Aug. 15, 1975.
Pp. 64-67.
5. Determination of Radium Removal Efficiencies
in Illinois Water Supply Treatment Processes.
Technical Note ORP/TAD-76-2, U.S. Environ-
mental Protection Agency, Washington, D.C.,
May 1976.
6. Determination of Radium Removal Efficiencies
in Iowa Water Supply Treatment Processes.
Technical Note ORP/TAD-76-1, U.S. Environ-
mental Protection Agency, Washington, D.C.,
June 1976.
7. Singley, J. E., et al. Costs of Radium Removal
from Potable Water Supplies. EPA-600/2-77-073.
U.S. Environmental Protection Agency,
Cincinnati, Ohio, 1977. In press.
8. Unpublished data, U.S. Environmental Pro-
tection Agency, Water Supply Research Division,
MERL, Cincinnati, Ohio.
9. Sarasota County Health Department news re-
lease, Oct. 26, 1975.
10. Moore, D. H. Greenfield, Iowa, Reverse Osmosis
Plant, J. Am. Water Works Assoc., 64(Nov.):781,
1972.
72
-------
11. Monograph of the Effectiveness and Cost of
Water Treatment Processes for the Removal of
Specific Contaminants. Vol. 1. Technical Manual.
Prepared for the EPA Office of Air and Water
Programs by David Volkert and Associates, Aug.
1974.
12. Maximum Permissible Body Burdens and Maxi-
mum Permissible Concentration of Radionuc/ides
in Air or Water for Occupational Exposure. NBS
Handbook No. 69, U.S. Department of Com-
merce, Washington, D.C., Aug. 1963.
13. National Primary Interim Drinking Water Regu-
lations. Fed. Reg., Aug. 14,1975.
14. Straub, C. P. Removal of Radioactivity by
Water-Treatment Processes. In: Low Level Radio-
active Wastes—Treatment, Handling, Disposal.
U.S. Atomic Energy Commission, Washington,
D.C., 1964.Ch.8-
15. Hauck, A. R., and S. Sourirajan. Performance of
Porous Cellulose Acetate Membranes for the
Reverse Osmosis Treatment of Hard and Waste
Waters. Environ. Sci. Techno!., 5(Dec.):
1272-1274,1969.
16. Mixon, F. O. The Removal of Toxic Metals from
Water by Reverse Osmosis. Office of Saline Water
Report 73-889, Sept. 1973. Pp. 3-5.
17. Lonsdale, H. K., et al. Research on Improved
Reverse Osmosis Membranes. Office of Saline
Water Report No. 577, Oct. 1970. P. 110.
18. Dytnerskii, Y. F., et al. Purification and Concen-
tration of Liquid Wastes with a Low Level of
Radioactivity by Reverse Osmosis. At. Energy,
35{6):405408, 1973. Cited in: Chem. Abstr.,
57:130,1974.
19. Koenst, J. W., et al. The Treatment of PWR
Nuclear Process Wastes Using Membrane Sys-
tems. In: Proceedings of the 28th Industrial
Waste Conference. Purdue University, Lafayette,
Indiana, May 1-3,1973. P. 765.
20. Glasstone, S. Sourcebook on Atomic Energy, D.
Van Nostrand and Company, Inc., New York,
N.Y., 1967. P. 741.
73
-------
-------
-------
------- |