(DWRD0003)
UTILIZATION OF VARIOUS TREATMENT UNIT PROCESSES AND
TREATMENT MODIFICATION FOR TRIHALOMETHANE CONTROL
by
James M. S virions
From
Proceedings - Control of Organic Chemical Contaminants in Drinking Water
A Series of Seminars Sponsored by
The Office of Drinking Water
U. S. Environmental Protection Agency
Physical & Chemical Contaminant Removal Branch
Drinking Water Research Division
Municipal Environmental Research Laboratory
U. S. Environmental Protection Agency
Cincinnati, Ohio 45268
January 1980
-------
A r UTILIZATION OF VARIOUS TREATMENT UNIT PROCESSES
AND TREATMENT MODIFICATION FOR TRIHALOMETHANE CONTROL
by
James M. Symons
INTRODUCTION
As noted in another paper^ in these Proceedings,
trihalomethanes arise from the reaction of free chlorine
with organic precursors, thought to be largely aquatic
humlc and fulvic acids, to form trihalomethanes and
other chlorinated organic compounds. Because of the
carcinogenicity of one of the trihalomethanes, chloroform,
the U.S. Environmental Protection Agency has proposed a
Maximum Contaminant Level (MCL) for the arithmetic sum
of the trihalomethanes, called Total Trihalomethanes
(TTHM)2.
If and when a final MCL is promulgated, some water
utilities will be required to alter their treatment
practices in order to comply with the Regulation. The
choice of the treatment technique or treatment modification
to be used in order to achieve compliance is up to the
utility, their consulting engineer, if any, and the
appropriate State Regulatory Agency, in the States that
have primacy. The purpose of this paper is to review the
various choices that can he considered when attempting to
reduce the total trihalomethane concentration.
Because the basic equation of TTHM formation involves
one product, TTHM, to be controlled, and two reactants,
organic precursor(s) and free chlorine, three approaches
are possible. These are: 1) treatment for the removal of
TTHM, 2) treatment for the removal of organic precursors,
-------
2
and 3) the use of disinfectants other than free chlorine.
Each of these techniques will be discussed.
Ae ra t i on
Because the purging technique for the concentration step
in the purge-and-trap technique for measuring trihalomethanes
is effective for removing trihalomethanes from a water sample,
one possible method for removing trihalomethanes would be
to purge or strip them from solution. Figure 1 shows that
chloroform could be removed from Cincinati, Ohio tap water
using diffused-air aeration, the removal of chloroform
improving as the air to water ratio (volume to volume)
increased from 1:1 to 20:1. Figure 1 also shows, however,
the precursor material, as measured by the chloroform
formation potential, is not removed by aeration. This is
not surprising, because these materials are thought to be
largely non-purgeable. This means, however, that although
aeration might reduce the concentration of trihalomethanes,
the presence of precursor would allow the formation reaction
to continue in the distriubution system and TTHM might be
reformed in the distributed water if the reaction had not
gone to completion in the water treatment plant.
Oxidation
The possibility of removing trihalomethanes by oxidation,
using either ozone or chlorine dioxide as the oxidant, was
investigated. Table I shows that neither of these oxidants,
even at rather large doses was effective for removing the
trihalomethanes.
Table I
REMOVAL OF TRIHALOMETHANES BY OXIDATION
Cincinnati, Ohio Tap Water
Ozonation
Applied O3 Dose - 25 mg/L
Contact Tiume - 4-5 Minutes
REMOVAL OF TRIHALOMETHANES
Tap Water
Mixer Only
Mixer + O2
Mixer + O3
TTHM, ug/L
26
29
27
27
-------
3
/T-
40 IP
CHLORO-
FORM,
Jig/I
30
20
10
f-T-'l
f—
I
1
&
y
&
§
1
I
S
§&
i
1
a
|4-r
*4—
c
j
fi---
II
r
CHLOROFORM CONCENTRATION
AFTER AERATION
CHLOROFORM CONCENTRATION
AFTER AERATION, RECHLORINATION
AND TWO DAYS STORAGE @25°C
(UNREACTED CHLOROFORM
FORMATION POTENTIAL)
I I
I I
1:1
4:1
8:1
16:1
20:1
AIR TO WATER RATIOS
FIGURE 1. REMOVAL OF CHLOROFORM FROM CINCINNATI,
OHIO TAP WATER BY AERATION
-------
4
Table I (Contd.)
CHLORINE DIOXIDE TREATMENT
Cincinnati, Ohio Tap Water
Dos e
rag/L
C10 2
Contact Time
days
TTHM
ug/L
0
7
7
0
10
0
1
2
0
2
52
58
51
75
78
Adsorption
Synthetic Resins
The Rohm and Haas Company has developed a synthetic
resin reported to be specifically designed to effectively
remove trihalomethanes by adsorption. This material has the
trade name Ambersorb XE-340*. Figure 2 shows that this
material was effective for the removal of trihalomethanes
during the 40 weeks of the test on Cincinnati, Ohio tap
water.
Powdered Activated Carbon (PAC)
Several brands of powdered activated carbon were
investigated in jar tests for their ability to adsorb
trihalomethanes from Cincinnati, Ohio tap water. Table II,
the results from one experiment with a commercially available
brand of powdered activated carbon, shows that complete removal
was impossible even with uneconomically high doses, but the
first mg/L added caused a 20% reduction. This type of
experiment was repeated with 9 different types of powdered
activated carbon. Following these experiments, for
comparison of effectiveness, the geometric mean of the
trihalomethane loading on the PAC in ug/gm was calculated
for all doses. These loadings were then ranked, arbitrarily
taking the loading on the poorest powdered activated
carbon as unity. Figure 3 shows that the most effective
powdered activated carbon has a geometric mean trihalomethane
loading eight times higher than that of the least effective.
The data in Table II was taken from Brand "G". This has
been confirmed by others.
*Mention of commercial products does not imply endorsement
by the U.S. Environmental Protection Agency.
-------
5
150
TTHM, COLUMN INFLUENT
125--
100--
TTHM
CONC.75 -
ug/l
50--
i "Sf TTHM. COLUK
EFFLUENT
25-
40
35
15 20 25
TIME IN OPERATION, WEEKS
30
10
FIGURE 2. REMOVAL OF TRIHALOMETHANES
BY AMBERSORB® XE-340 (empty bed
contact time=10 min)
-------
ARBITRARY SCALE OF INCREASING EFFECTIVENESS
UNITY
8
h tt t t r
ype pac A B C D E
G
T
H
T
I
Figure 3. Comparison of Gm THM Loading on
Powdered Activated Carbon
-------
7
Upper Ohio Valley Water Treatment Plant
INST.
TOTAL
TRIHALO-
METHANE
CONC.
jug/i
TOTAL TRIHALOMETHANES
ADSORBER INFLUENT
SAND REPLACEMENT
ADSORBER
EFFLUENT
j—i i i i i
i„ i ,,i i I > t
5 10 15 19 24 29 33 38 43 47
FIGURE 4. TIME IN SERVICE, DAYS
52
REMOVAL OF TRIHALOMETHANES BY GRANULAR ACTIVATED CARBON BEDS-
6.5 MINUTE EMPTY BED CONTACT TIME - VIRGIN WVW 14x40 GRANULAR
ACTIVATED CARBON
-------
8
Table II
POWDERED ACTIVATED CARBON FOR THE REMOVAL OF TRIHALOMETHANES
Cincinnati, Ohio Tap Water
PAC
TTHM
Dose
Cone
mg/L
ug/L
0
75
1
60
2
61
4
59
8
60
16
57
32
52
64
40
100
33
Granular Activated Carbon (GAC)
Following the replacement of sand media with virgin
granular activated carbon at a water treatment plant, the
breakthrough curve for total trihalomethanes was determined.
Figure 4 shows the first breakthrough, effluent containing
approximately 10% of the influent, occurred after approximately
1 week of operation, while exhaustion, influent equals
effluent, occurred after approximately 7 weeks. This adsorption
bed with a 6.5 minute empty bed contact time (volume/flow ¦
EBCT) showed that granular activated carbon is effective
for trihalomethane removal, when fresh, but breakthrough
occurred soon after the bed was put into operation.
Summa ry
The major advantage of attempting to control trihalomethane
concentrations by removing the trihalomethanes themselves
is that no change in disinfection practice would be required.
The disadvantages are first, that trihalomethanes cannot be
easily and economically removed (at this time, Spring 1979,
the costs of XE-340 treatment have not been determined) by
any of the treatment processes discussed, and second, that
-------
9
REMOVAL OF PRECURSOR
The second method for reducing the concentration of
trihalomethanes is to reduce the concentration of one of
the reactants in the basic equation of formation; free
chlorine + precursor yields trihalomethanes, plus other
chlorinated by-products. This section of the paper will
discuss the effectiveness of three unit processes for
reducing precursor concentrations, precipitation, oxidation,
and adsorption.
Precipitation
In most water treatment plants, the precipitation step,
either using iron or aluminum salts as coagulants, or calcium
hydroxide if softening is also a goal, is used to remove
color and turbidity. Because the major trihalomethane
precursors are thought to be humic and fulvic acids, which
are also associated with natural color, studies were
conducted to determine whether or not precipitation for
color and turbidity removal would also reduce trihalomethane
precursors•
The first of these studies by USEPA was conducted using
a pilot plant treating Ohio River water. In this study,
chlorine was added at different stages of the treatment
process and the terminal chloroform concentration following
two days of simulated distribution residence time was
determined. In the first phases of the study, chlorine was
added to the raw water prior to the addition of coagulants
in the rapid mix. In the second phase of the study, chlorine
was added to the settled/water to determine whether or not
coagulation and precipitation in the settling basin would
have reduced the precursor concentration. In the final phase
of the study, filtered water was chlorinated to determine
whether or not the additional clarification during filtration
would further influence trihalomethane precursor concentration.
These three phases were conducted using ferric sulfate as'
a coagulant in one case and aluminum sulfate as the coagulant
in the other.
Figure 5 shows that if the terminal chloroform
concentration in the "raw water chlorination-aluminum
sulfate coagulant" study is taken as unity, the removal
of precursor during sedimentation and filtration is seen
by the proportionately lower concentration of chloroform
rresulting when the point of chlorine application was moved
from the raw water to the settled water and finally to
the filtered water. Also shown in Figure 5 is the improved
effectiveness, at least in this water, for ferric sulfate,
as opposed to aluminum sulfate as a coagulant.
-------
1.25
RELATIVE TERMINAL
CHLOROFORM
CONCENTRATION IN
FINISHED WATER
AFTER 2 DAYS AT
GIVEN POINT OF
CHLORINATION
COMPARED TO
CHLOROFORM
FORMATION POTENTIAL
IN RAW WATER
.5
.7 5 -
.2 5 -
RIVER
RAW WATER TERM
chci3 CONC
ALUMINUM
SULFATE
FERRIC
SULFATE
CO AG UL ANTICOAGULANT
_ DISTRIBUTION
SYSTEM
Q) COAGULATION, (T,
2 DAYS RAW FLOCCULATION, FILTRATION
WATER STORAGE SETTLING
CHLOROFORM IN FINISHED WATER RELATIVE
TO POINT OF CHLORINATION
IPILOT PLANT STUDIES!
FIGURE 5.
-------
11
The improvement in ferric sulfate as a coagulant is seen
in all three bars in Figure 5. The reason that differences
between the two sets of data is seen even when chlorine was
added to the raw water is because, although the chlorine
and all the precursor is initially intimately mixed in the
rapid mixer, in the settling basin the precursor begins
to settle out and is removed from intimate contact with the
free chlorine. Therefore, even when raw water is
chlorinated, a better coagulant will produce a lower terminal
chloroform concentration because the separation of chlorine
and precursor in the sedimentation basin is hastened. Recall,
however, that pH influences the rate of TTHM formation1,
therefore, treatment plants using lime softening may not
experience the same effect as noted in Figure 5.
In an attempt to demonstrate whether or not these
pilot plant results could be duplicated at full-scale, the
City of Cincinnati water utility agreed to move their point
of chlorination. Figure 6 is a schematic diagram of a water
treatment plant using Ohio River water as a source. Because
in the regular scheme of treatment, both coagulant and
chlorine were added at Point A, considerable clarification
occurs in the storage reservoirs that exist between Point A
and Point B. Therefore, if precipitation is an effective
method of precursor removal, less precursor should be
available at Point B for reaction with the free chlorine.
Figure 7 shows that when the point of chlorination was
moved in mid-1975, see arrow, the chloroform concentration
in the distributed water declined sharply. This generally
lower concentration of chloroform was maintained even during
the warmer months of 1976.
To determine whether or not changes in the raw water
had occurred during this time period that might account for
this decline in chloroform concentration, trihalomethane
formation potential determinations were made on the raw
water at various times during the test period. Although
these data are somewhat scattered, they do indicate the
decline in distributed water chloroform concentration was
not caused by a change in the precursor content of the
raw river water.
Note, also, that although a sharp decline in distributed
water chloroform concentration occurred, a similar decline
in the brominated trihalomethanes did not occur. This is
probably because these materials are formed faster than
chloroform and chloroform is formed from any precursor
that remains once the bromide has been consumed.
-------
OFF-STREAM STORAGE RESERVOIRS
POINT B
R. MIX \
FLOC, SETT.
WATER
TREATMENT
PLANT
CHLORINE, ALUM
POINT A
FILTERS
PUMPING
STATION'
POINT C
CLEAR WELL
INTAKE
DISTRIBUTION SYSTEM
OHIO RIVER
FIGURE 6. SCHEMATIC OF CINCINNATI WATERWORKS
-------
300
MOVE CHLORINATION
2600 FROM POINT "A"
TO POINT "B"
[-JULY 14, 1975
y
220
1
180
TRIHALO
METHANE
CONG. 140
Jug/L
100
60
20
NF
A
CHLOROFORM
A
O
KEY
RAW WATER CHLORINATED
AND STORED:
03 DAY TERMINAL CHLO
ROFORM CONCENTRATION
~ 4 DAY
A 6 DAY
©
~
~
-
E
(V
W
/_
BROMODICHLOROMETHANE
DIBROMOCHLOROMETHANE
A A
-\VA J* Vx ^ A/\
X/--'
JUL
IaugIsepIoctInovIdecI
V/
1975
NOVjDEC|JAN|FEB|MAR|APR|MAYIJUNIJUL
1976
U)
Figure 7.TRIHALOMETHANES IN CINCINNATI, OHIO TAP WATER
-------
14
The results in Table IV are from an attempt by this
same water utility to determine whether or not moving the
point of application of chlorine from Point B in Figure 6
to the effluent from the sedimentation basin in the treatment
plant would result in a further decline in distributed water
trihalomethane concentration because of the removal of
additional precursor through the settling process in the
actual treatment plant itself. Table IV compares the
distributed water trihalomethane concentrations during
September and October of the control year when the chlorine
was being added at the rapid mix, Point B in Figure 6,
to data collected during the same time period of 1977.
Assuming 1976 can be taken as a "control" on the experiment,
these data would tend to indicate generally lower trihalomethane
levels occurred in the distributed water when chlorine is
added to the effluent of the sedimentation basins, although the
decline is not as dramatic as the 1975 experiment described
earlier.
Another example of the use of precipitation for the
removal of precursor was at a softening plant in Daytona
Beach, Florida. Figure 8 is a schematic diagram of the
Daytona Beach, Florida water treatment plant and shows
the three points of application of chlorine during the
study period. In this system, sampling for instantaneous
trihalomethane concentrations was several hours after
chlorination, Point 3.
Figure 9 shows the decline in the weekly mean
instantaneous trihalomethane concentrations when
chlorination followed the reduction in precursor concentration
through settling and filtration. As in the previous study,
weekly mean terminal trihalomethane values of the raw
water showed that this decline in instantaneous trihalomethane
concentration was caused by moving the point of chlorine
application and was not caused by a change in the
characteristics of the raw water. Figure 9 also shows the
terminal trihalomethane values after some storage period.
In this case, filtration for improved clarification was
required to obtain a significant decline in terminal
trihalomethane concentrations.
In summary, these studies, and those conducted at many
other utilities, have shown that precipitation will remove
trihalomethane precursors, and that chlorination at some
point in the treatment process following a clarification
step will result in lower trihalomethane concentrations in
the distibuted water.
-------
SLAKED LIME
POLYMER NALCO 1174
5AMPLI
POINT
iNn.3,
SAMPLE'
. POINT ,
My/
RECARB
BASIN
SAMPLI
k POINT
SLUDGE
PUMP {
SUPPLY
WELLS
DUAL MEDIA
FILTERS
GROUND
STOKAGE
SAMPLE
, POINT
TANK
HIGH SERVICE PUMPS
© TKSFR
pump
DISTRIBUTION
SYSTEM
FILTERED WATER
BACKWASH PUMP
CLEARWELL
FLOW DIAGRAM
RALPH F. BRENNAN WATER PLANT
DAYTONA BEACH. FLORIDA
Figure 8
-------
450
400
350
300
250
200
150
100
75
50
25
0
16
DAYTONA BEACH, FLA.
Cl2
RAW
NO 002
C02
1
i
TERM
TTHM
IN RAW
WATER
CI2
SETTLED
|—|TTHM FORM, .
NO POTENTIAL
C02 ^~~TERM TTHM
C02 c,2
FILTERED
NO CO
C02
I
I
II
Figure 9.
-------
17
TABLE IV
INFLUENCE OF MOVING POINT OF APPLICATION OF CHLORINE AT
THE CINCINNATI WATER WORKS
CONTROL YEAR EXPERIMENTAL YEAR
TTHM in TTHM in
Point of Appl. Dist Point of Appl. Dist.
Date of Chlorine Water Date of Chlorine
ug/L
Water
ug/L
9/13/77 Rapid Mix 121
9/16/76 Rapid Mix 120
21 " 117 21
23 " 118
28 120 28 Settled Water
29 " 109 29
30 " 116 30
10/1
2
3
4
10/5 " 110 5
6
7 " 114 7
12 " 92 11
14 " 109
107
82
102
76
89
87
116
78
78
78
81
10 Rapid Mix 113
90
-------
18
Oxidation
Ozonation
In an effort to determine whether or not ozone would
have any influence on trihalomethane precursors, unchlorinated
Ohio River water was ozonated with various applied ozone
dosages. Following ozonation the water was chlorinated and
the resulting trihalomethane concentration measured after
6 days of storage. As a control, replicate Ohio River
samples were chlorinated without ozonation. The data in
Table V shows ozonation alone did not produce trihalomethanes,
but, except for the highest does, ozonation prior to
chlorination did not reduce the trihalomethane concentration
after six days of storage. This implied that, at least
in this USEPA experiment, ozone was not effective for removing
trihalomethane precursors.
TABLE V
EFFECT OF OZONATION OF TRIHALOMETHANE PRECURSORS
Ohio River Water
Contact Time - 5-6 minutes
Applied O3 Chlorine TTHM Cone.
Dose Dose After 6 days
mg/L mg/L ug/L
0.7 0 <0.2
0 8 20
0.7 8 23
18.6 0 <0.2
0 8 30
18.6 8 30
0 8 123
227 8 70
-------
19
A similar experiment was conducted on Ohio River water
by another investigator, (Glaze, et al., personal communi-
cation), however, and, as shown in Figure 10, the ozonation
did have an effect on the seven-day total trihalomethane
formation potential. This showed that under these conditions
ozone was effective for removing trihalomethane precursors.
Trussell-* reviewed the literature and determined, as shown
in Figure 11, that the percent reduction in trihalomethane
precursor was quite variable from experiment to experiment.
Percent removal ranged widely even at the same ozone
dosage. This implies that the conditions of ozonation;
contactor design, dosage, temperature, and so forth, as
well as the specific nature of the organic material
being ozonated might have an influence on the destruction
of precursor by ozonation.
Chlorine Dioxide
In an effort to determine whether or not chlorine dioxide
has an effect on trihalomethane precursor, Ohio River water
was chlorinated and a duplicate sample treated with chlorine
dioxide and chlorine. The data in figure 12 show that the
treatment with chlorine dioxide did have some effect on
chloroform precursor as indicated by the lower chloroform
concentration, even in the presence of excess chlorine
following chlorine dioxide treatment.
Potassium Permanganate
To investigate whether or not treatment by potassium
permanganate would remove trihalomethane precursors, Ohio
River water was dosed with potassium permanganate stored
and subsequently chlorinated. The resulting data, (see
Table VI) showed that in the Ohio River, potassium
permanganate had little influence on precursor. These
data also indicate that the effect that did occur was
reduced even further as the pH was raised.
-------
20
TABLE VI
TRIHALOMETHANE PRECURSOR REMOVAL BY KMNO4
Ohio River Water
KMnO^ Reaction
Cl2
Amount Reaction Reaction
Added Time Time %
(mg/L) (Hours) pH (Hours) Removal
0
1.5
7.1
2
5
1.5
7.1
2
15. A
0
1.5
9.3
2
5
1.5
9. 3
2
2.7
0
1.5
10. 2
2
5
1.5
10. 2
2
5.6
0
0.5
7.0
30
5
0. 5
7.0
30
19.0
Adsorpt ion
Powdered Activated Carbon
Studies were conducted to determine whether or not
powdered activated carbon could be used for adsorbing
trihalomethane precursors. In these tests, Ohio River
water was treated with powdered activated carbon,
coagulated, settled and subsequently chlorinated. The
resulting chloroform formation potential as related to
powdered activated carbon dose, Figure 13, shows that
very high doses of powdered activated carbon were required
to obtain substantial reductions in the chloroform formation
potential. Furthermore, even at these high doses, chloroform
formation potential was not eliminated, but was only
reduced two-thirds.
Table VII compares the performance of two brands of
powdered activated carbon for the reduction of trihalo-
methane (see Figure 3) and total trihalomethane formation
potential. Brand "I" showed no improvement in trihalo-
methane removal as the dose was increased from 2 to 30 mg/L,
while precursor removal increased from 3 to 29 percent.
In contrast, Brand "B" showed a large improvement in
performance for trihalomethane removal as dose was
increased, but the increasing dose had little effect on
precursor removal. This demonstrates the variability of
various brands of powdered activated carbon for the control
of trihalomethanes and trihalomethane precursors, and
demonstrates the need for on-site studies if this type
-------
21
280
240
^ 200
a.
LL
160
120
(0
"O
I
h-
¦
o
(0
o
80
40
OHIO RIVER WATER (2/77)
Oxygen, only
Ozone, only
Dose: 20mg 03/min. in 22 liter tank
0 20 40 60 80
Figure 10. Time, min.
100
120
-------
100
80 -
OZONE FOR PRECURSOR REMOVAL
(After Trussell) 3
60 -
T3
-------
50
40
* 30
nr
20
10
1
2
3
4
Contact Time, Days
Figure 12. Comparison Of Chlorine And Chlorine Dioxide
On The Formation Of Chloroform In Filtered Water
-------
2 DAY
25 ° C
CHLORO-
FORM
20
FORM-
POTENTIAL15
M9/'
10
2 MINUTES RAPID MIX
5 MINUTES SLOW MIX
30 MINUTES SETTLING
S^lQ
*OF
°A Y.
fOr
ma
25 <
r/OA/
p°tbn
__ ~'WiAl
10
80
90
20 30 40 50 60 70
POWDERED ACTIVATED CARBON DOSE, mg/l
FIGURE 13. INFLUENCE OF POWDERED ACTIVATED CARBON ON ON CHLOROFORM
FORMATION POTENTIAL
100
-------
25
of adsorption is contemplated
TABLE VII
PERCENT REDUCTION OF TRIHALOMETHANES AND TRIHALOMETHANE
PRECURSORS WITH POWDERED ACTIVATED CARBON
Brand I
Dose 2 mg/L 30 mg/L
Brand B
2 mg/L 30 mg/L Sample
TTHM -33%
-32%
-3%
-16% Cinti
Tap
Water
2-day
TTHMFP -3%
-29%
-21%
Ohio
-37% River
Water
Granular Activated Carbon
As an example of the ability of granular activated
carbon to remove trihalomethane precursors, three sets of
data at different empty bed contact times (6.5 minutes,
9 minutes and 24 minutes) are chosen as examples. These
data, Figures 14, 15, and 16 shows improved performance
for the removal of trihalomethane precursors by granular
activated carbon as EBCT^ is lengthened. For the longer
two empty bed contact times, Figures 15 and 16, a "steady-
state" condition, less than exhaustion, seems to have
developed towards the end of the run.* This is probably
caused by biological activity naturally occurring on
the granular activated carbon surfaces.^
Ozone + Granular Activated Carbon
Reports in the literature^ indicate that the biological
activity mentioned above can be Increased if ozone is added
to the water prior to granular activated carbon adsorption.
In an attempt to investigate this reported effect, a pilot
plant treating Ohio River water was used to compare the
behavior of preozonated settled water both passing through
parallel granular acitvated carbon adsorption beds.
Figure 17 shows that the effluent from the ozone + GAC
system always contained less trihalomethane precursor than
did the effluent from the utiozonated system. This combination
of oxidation and adsprption is being investigated by USEPA,
Office of Research and Development, Drinking Water Research
Division, Cincinnati, Ohio.
-------
26
0.260
0.240
0.220
0.200
WVW 14x40 GRANULAR
ACTIVATED CARBON
fsAND REPLACEMENT
I ADSORBER
INFLUENT
SAND REPLACEMENT
ADSORBER EFFLUENT
0.180
TOTAL
TRIHALO-160
METHANE
FORM. o.i4o
POTENT.
GONG. 0.120
mg/l
0.100
0.080
0.060
0.040
0.020
APPROX. EMPTY BED CONTACT TIME = 6.5 MINUTES
i i i i i i ¦ ¦ i i
11 15 20 25 27 291 3 5 8 101215171922 24 26 29 31
JULY AUGUST
FIGURE 14. REMOVAL OF TRIHALOMETHANE
PRECURSORS BY GRANULAR
ACTIVATED CARBON BEDS
-------
0.081-
9 MIN. EBCT
2 DAY, 00*.
25°C.
T THMFP
CONC. 0 041
mg/l
ADSORBER
INFLUENT
EFFLUENT-GAC
0.021-
4 5 6
TIME IN SERVICE, MONTHS
INFLUENCE OF ADSORPTION ON
TRIHALOMETHANE FORMATION POTENTIAL REMOVAL
FIGURE 15.
N>
-J
-------
o>
£
0.30
0.25
0.20
24 MlN. EBCT
INFLUENT
5
0.15
E
5 0.10
*
: *
0.050
Ef* — H£i
8
EFFLUENT
to
00
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180
Days of run
Fig. 16 Performance of terminal TTHM through Post Filter Adsorber
run#1 (Feb.-Oct. 1977)-Jefferson Parish,La.
-------
0.1
0.08
VALUES COMPUTED FROM MONTHLY AVERAGES OF
WEEKLY DETERMINATIONS 2 DAY, 25°C TOTAL
L TRIHALOMETHANE FORMATION POTENTIAL
2 DAY.
'0.06 k
25 C,
TTHMFP
CONC.
mg/l 004
0.02
ADSORBER
INFLUENT
EFFLUENT-GAC ONLY
to
vo
EFFLUENT-GAC+OZONE
1 1 I L
3 4 5 6 7 8 9
TIME IN SERVICE, MONTHS
17 INFLUENCE OF OZONATION PRIOR TO ADSORPTION ON
Fig.
TRIHALOMETHANE FORMATION POTENTIAL REMOVAL
-------
30
Summary
Advantages
If trihalomethane control is accomplished by removing
trihalomethane precursors, the following four advantages
will be realized:
1. Disinfectant demand is reduced when precursor is
removed, resulting in a cost saving.
2. Chlorine can still be used for disinfection.
3. Coagulation and sedimentation, a treatment process
practiced by many water utilities already, is effective for
reducing the precursor.
4. Several oxidative and adsorptive techniques are
effective to varying degrees for removing precursors,
giving engineers and utilities a wide choice of unit
processes to consider.
Disadvantages
On the other hand, this treatment approach has four
disadvantages:
1. If raw water disinfection is needed, trihalomethanes
will be formed by this step prior to any precursor removal
treatment process.
2. Complete removal of precursor, should very low
final trihalomethane concentrations be required, is difficult.
3. If bromide is present, bromine containing tri-
halomethanes will be affected last by the treatment as
they are formed faster than chloroform by the reaction of
chlorine and bromide with any remaining precursor.
4. Because chlorine would continue to be used as a
disinfectant, other organic halogens (OX) will be formed,
see Table III, although their concentration should be
reduced in a manner similar .to the TTHMs.
-------
31
ALTERNATE DISINFECTANTS
The third method of reducing the trihalomethane
concentrations is to change the other reactant, free
chlorine, to some other disinfectant. If an alternate
disinfectant were effective for controlling pathogenic
organisms, and did not react with precursors to form
trihalomethanes, effective control of trihalomethanes
would have been accomplished. For this discussion, four
alternate disinfectants will be considered: 1) chlorine
dioxide, 2) ozone, 3) chloramines and 4) bromine chloride.
Chlorine Dioxide
Chlorine dioxide has five advantages. First, it is
a good disinfectant, being active against both bacterial
pathogens and viruses. Furthermore, in contrast to free
chlorine, pH has little effect on the disinfecting power
of chlorine dioxide. Second, chlorine dioxide is easy
to feed. Feeding equipment usually consists of a chlorinator,
and one or two solution feeders, feeding sodium chlorite
alone or sodium chlorite and acid for pH control. The
mixture is then allowed to pass through a reaction column
and the chlorine dioxide then fed into the water to be
disinfected. Third, chlorine dioxide forms a persistent
residual so that some biocide can be maintained throughout
the distribution system. This has the advantage of helping
to keep the distribution system clean and guarding against
the effects of minor cross-connections. Fourth, if excess
chlorine is not present, no trihalomethanes will be formed
if chlorine dioxide is the disinfectant, see Figure 18.
Fifth, disinfectant demand is often reduced compared to
free chlorine, when chlorine dioxide is used, see Figure 19 .
This is true largely because chlorine dioxide does not
react with ammonia.
Disadvantages
On the other hand, chlorine dioxide has four disadvantages.
The first is that through disproportionation and as one end
product of the reaction of chlorine dioxide with organic
compounds, chlorite, CIO2"" is formed. At the present time,
Spring, 1979, intensive research is being conducted on the
toxic potential of chlorite."* Chlorine dioxide will not
achieve widespread usefulness until the question of the
possible toxicity of chlorite is settled. Second, because
chlorine dioxide is frequently formed using chlorine as
one of the reactants, if the reaction is not properly
controlled, excess chlorine will be present with the
chlorine dioxide. As shown in Figure 12, if chlorine is
present, some trihalomethanes will be formed, but less than
with chlorine alone. Third, the byproducts from the
reaction of chlorine dioxide and organic material have not
been thoroughly evaluated either by chemical analysis^
-------
50
—
40
30
—
AC.®
Chloro-
form
y
/
y
/
~
/
/»
/jg/i 20
/
X
— y
s
S
/
/
/
s
s
/
10
r
i
n
i
i
^"Chlorine Dioxide Alone
Contact Time, Days
FIGURE 18. COMPARISON OF CHLORINE AND CHLORINE DIOXIDE ON
THE FORMATION OF CHLOROFORM IN FILTERED WATER
-------
33
DISINF. «
NOV.17,1975
OHIO RIYER WATER
TURBIDITY = 23FTU
pH = 7.5
STO. PLANT COUNT = 10,000/ml
COLIFORH = 700/100ml
4 6 8 10 12
APPLIED DOSAGE, mji/l
FIGURE 19. COMPARISON OF DISINFECTANT DEMANDS FOR OHIO RIVER WATER
-------
34
or toxicological evaluation.^ Fourth, chlorine dioxide
will produce some organic halogen material. As shown
in Table VIII, much less organic halogen material is
produced when chlorine dioxide is reacted with humic
acid as compared to chlorine, but organic halogen was not
absent as an end product when chlorine dioxide was reacted
with humic material.
TABLE VIII
REACTION OF HUMIC ACID WITH CHLORINE AND CHLORINE DIOXIDE
CI2/C Cl2,C102 Reaction CHCI3 Organic
M/M mg/L Time, hr. ug/L Halogen, ug/L
1/3 3.8 1 39 198
5/3 19.4 1 32 278
C102/C
M/M
1/15 0.75 1 0.4 23
1/3 3.7 2 1.6 52.5
After Stevens, A.A., et al^
Oz one
Advantages
As an alternate disinfectant, ozone has four advantages.
First, ozone is a good disinfectant showing excellent
biocidal properties with respect to bacterial pathogens
and viruses. Furthermore, the disinfecting power is not
affected by pH. Second, although the possibility of ozone
oxidizing chloride to chlorine thereby producing trihalo-
methanes exists, studies have shown that this does not
occur, see Table IX. Further, at this time (Spring 1979),
the production of organic halogen from ozonation has not
been studied, but none is expected. Third, bromide is not
oxidized by ozone to bromine, so the brominated trihalo-
methanes are not formed during ozonation, see Table X.
Fourth, no disinfectant residual is created. This is listed
as an advantage because future investigations may reveal
that the presence of a biocide in distributed water is not
advantageous because of some health effect. If this ever
became the case, ozone would have the advantage of not
producing a disinfectant residual.
-------
35
TABLE IX
EFFECT OF OZONATION ON DUAL MEDIA EFFLUENT
Treated Ohio River Water
CONTACT TIME - 5-6 minutes
Applied O3 Chlorine TTHM Cone.
Dose Dose After 6 days
rcg /L mg/L ug/L
0.7 0 <0.2
0 8 20
18.6 0 <0.2
0 8 23
TABLE X
EFFECT OF OZONATION ON DUAL-MEDIA EFFLUENT
Treated Ohio River Water
Continuous Flow Studies
O3 Contact Time - 5-6 Minutes
Applied Bromo- Dibromo-
Ozone Chlorine Chloroform dichloro- chloro-
Dose Dose ug/L methane methane
mg/L mg/L "g/L ug/L
0.7 0 0.2 None Found None Found
0 8 6 14 4
18.6 0 0.2 None Found None Found
0 8 12 9 2
-------
36
Disadvantages
As an alternate disinfectant, ozone has five disad-
vantages. First, no disinfectant residual is created. This
is listed both as an advantage and a disadvantage because
conventional wisdom recommends the maintenance of a biocide
to the extremities of the distribution system. If futher
research should not alter this opinion, the lack of a
residual becomes a disadvantage for ozonation. Second,
biodegradation in the distribution system is enhanced
following ozonation. In some instances in the United
Kingdom (Packham, personal communication) this has lead to
problems of excessive bacterial growth in the distributrion
system caused by ozonation making the organic matter in
the water more biodegradable. Third, relatively elaborate
generating equipment is required. To generate ozone,
very dry air is required and quantities of electrical
energy are necessary to maintain the corona discharge
essential for ozone prodution. Furthermore, specialized
materials, frequently stainless steel, are required for the
generating equipment and gas piping. This will increase
costs somewhat. Fourth, ozone decomposes very rapidly making
long contact times between the water and the disinfectant
difficult to achieve. Fifth, the organic byproducts
from ozonation have not been thoroughly studied either
chemically or through toxicological testing.5
Chloramines
Advantages
As an alternate to free chlorine for disinfection
chloramines have four advantages. First, chloramines do
not form trihalomethanes as free chlorine residual does,
see Figure 20. As a practical example of this effect,
Figure 21 is a schematic diagram of a water treatment plant
which adds ammonia folllowing filtration. The data in
Table XI show that the chloroform concentration does
not increase from the plant tap to a storage tank 12 hours
away because of the presence of excess ammonia in the water.
This demonstrates on a plant scale that chloramines will
prevent the formation of trihalomethanes.^
-------
37
120
110
100
90
80 1-
CHLORO-
FORM 70
CONC.
/*g/l 60
50 ^
40
30 \-
20
¦f! &\
2.8 mg/l
FREE Cl2
RESIDUAL
OHIO RIVER WATER
10 U
5.5 mg/l Cl2 AND 5.2 mg/l NH 3 -N ADDED
COMBINED CHLORINE RESIDUAL I
2.4 mg/l COMBINED Cl2 RESIDUAL
' ' ' ' 1 ' 1
X
10 20 30 40 50 60
REACTION TIME, HOURS
70
FIGURE 20. CHLOROFORM PRODUCTION
WITH FREE AND COMBINED
CHLORINE RESIDUAL
-------
BLOCK DIAGRAM OF WATER TREATMENT PLANT
HEXA-
META-
Fe2(S04)3 PHOSPHATE
SOFT. SETT. FLOC. SETT. FILT.
LIME
PUMP
PUMP
pH 10.2
FREE CI9 5 to 1 mg/l
8 HOURS
8 HOURS
NH
U>
00
FIGURE 21. TREATMENT GIVEN RIVER WATER "B" 7
-------
39
TABLE XI
INFLUENCE OF AMMONIA ADDITION ON TRIHALOMETHANE FORMATION
WATER B
Plant Tap Storage Tank
12 Hours Away
Chloroform Cl£ NH3 Chloroform
ug/L mg/L mg/L ug/L
38 1.8 0.55 34
36 1.6 0.50 35
34 2.2 0.35 36
38 2.3 0.40 35
Note; Data Collected on Four Different Days
Second, chloramines do not form as much organic halogen
as occurs with chlorine. In Table XII, data of Sontheimer's®
show that both organic halogen as measured by dissolved
organic chlorine (D0C1) and trihalomethanes (TTHM) are
reduced when breakpoint chlorination is not practiced.
TABLE XII
INFLUENCE OF CHLORINATION PRACTICE*
Breakpoint Non-Breakpoint
D0C1 TTHM D0C1 TTHM
ug/L ug/L ug/L ug/L
River 50 0.2 23 0.1
Settled 640 53 72 5
GAC Eff. 340 18 18 1
*After Sontheimer^ _____
Third, chloramines are easy to feed, only requiring a
chlorine and ammonia gas feeder or a chlorine feeder and
a feeder of ammonium salts, such as ammonia sulfate. Fourth,
chloramines produce a persistent residual so that some
biocide can be carried throughout the distribution system.
-------
40
Disadvantages
Chloramines have two disadvantages, one, it is a much
weaker biocide than free chlorine, chlorine dioxide, or
ozone, and its biocldal activity is reduced as pH increases.
Second, the disinfection byproducts from chloramination have
not been thoroughly studied either chemically or toxicologi-
cally.5
Bromine Chloride
Bromine chloride is not an acceptable alternative to
free chlorine because of the formation of large quantities
of bromine containing trihalomethanes. The data in Table
XIII show that when free chlorine is used as a disinfectant
the primary trihalomethane is chloroform, but when bromine
chloride is used as a disinfectant almost all of the
trihalomethane appears as bromoform. Furthermore, the
data in Figure 22 show that more total trihalomethanes
are formed when bromine chloride is used as a disinfectant
instead of free chlorine. Both of these sets of data
indicate that the use of bromine chloride is not desirable.
TABLE XIII
THM FORMATION
Chlorine
CHBr2Cl
CHBr 3
CHCI3
Bromine
Chloride
CHBr 3
CHC13
CHBrCl2
CHBrCl2
CHBr2C1
ug/L
ug/L
6 hrs.
44
16
3.4
0. 2
0. 3
<0.1
1.7
149
24 hrs.
85
23
4.5
1.3
0.4
<0.1
2.0
177
2 days
106
28
5. 2
0.3
0.5
0.1
2. 7
194
3 days
116
30
5.8
0.2
0.6
0.2
3.2
209
4 days
118
31
5. 9
0. 3
0. 5
0.1
3.4
209
-------
41
* *
175-
150- f
BROMINE CHLORIDE
5 125
100
CHLORINE
*Chloroform=106 jug /1
* *Bromoform=194 jug/I
0 10 20 30 40 50
Figure 22.Contact Time, Hours
-------
42
CONCLUSIONS
This paper has attempted to summarize the alternate
treatment possibilities for the three methods of controlling
trihalomethane concentrations: one, removing the trihalo-
methanes themselves, two, reducing the concentration of
precursor materials, or three, using an alternate disinfectant.
Further details on these treatment techniques are contained
in the Interim Treatment Guide for the Control of Chloroform
and Other Trihalomethanes.^
Two major conclusions are important. One, prevent the
formation of trihalomethanes, rather than trying to remove
them. This should be done by one, reviewing disinfection
practices, applying disinfectant to the best quality water
possible, two, improving pretreatment where possible, and
three, possibly using a disinfectant other than chlorine.
The second major consideration is that no relaxation of
microbiological quality of drinking water should be permitted
because of trihalomethane control.
-------
43
REFERENCES
1. Stevens, A.A., "Formation of Trihalomethanes by the
Reaction of Chlorine with Precursors," Proceedings,
Control of Organic Chemical Contaminants in Drinking
Water, U. S. Environmental Protection Agency,
Washington, D. C., In Press.
2. Control of Organic Chemical Contaminants in Drinking
Water, Federal Register, 43, No. 28, 5756-5780,
(Feb. 9, 1978).
3. Trussel, R.R. and Umphres, M.D., "The Formation of
Trihalomethanes," Journal of the American Water
Works Association, 70, 11, 604-612 (Nov. 1978).
4. Symons, J.M., Carswell, J.K., DeMarco, J. and
Love, O.T., Jr., "Removal of Organic Contaminants
from Drinking Water Using Techniques Other than
Granular Activated Carbon Alone - A Progres
Report," Proceedings, Practical Application of
Adsorption Techniques in Drinking Water,
Washington, D.C., April 30 - May 1-2, 1979,
U.S. Environmental Protection Agency, Washington, D.C.,
In Press.
5. Bull, R.J., "Health Effects of Alternate Disinfectants
and their Reaction Products," Presented at 99th
Annual Conference of the American Water Works
Association, June 24-28, 1979, San Francisco, Calif.
6. Stevens, A.A., Personal Communication.
7. Tuepker, J.L., "Sampling and Analysis of Chloro-
organics in the Distribution System," Proceedings,
Water Quality Technology Conference, Paper 3A,
December 6-7, 1976, San Diego, California, American
Water Works Association, Denver, Colorado.
8. Sontheimer, H.A., "Effectiveness of Granular Activated
Carbon for Organics Removal," Presented at 98th
Annual conference, American Water Works Association,
June 25-29, 1978, Atlantic City, New Jersey.
9. Symons, J.M., et al., "Interim Treatment Guide for
the Control of Chloroform and Other Trihalomethanes,"
U.S. Environmental Protection Agency, Cincinnati, Ohio,
Unpublished, 48 pp., plus 4 Appendices (June 1976).
* U.S. GOVERNMENT PRINTING OFFICII: 1930-657-146/5586
------- |