United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-83-003 Mar. 1983
Project Summary
Trihalomethane Removal by
Coagulation Techniques in a
Softening Process
J. C. Thurrott, H. Zwart-Duryea, J. R. Sloane, and D. Williamson
This research program investigated
various potable water treatment proc-
esses in combination with lime soften-
ing to effect maximum removal of
trihalomethane precursor compounds.
A study of the literature was used to
guide the initial test work. Bench-scale
jar tests investigated various combina-
tions of coagulants that earlier studies
indicated would be promising. The test
work evaluated the relative effective-
ness of lime softening, alum coagula-
tion, ferric coagulation, and clay coagu-
lation with respect to their ability to
remove THM precursors by themselves
and in various combinations with lime
softening.
The bench-scale test work was fol-
lowed by a series of eight pilot-plant
test runs using the U.S. Environmental
Protection Agency's trailer-mounted
pilot facility at the 45.4-MLD(12-MGD)
Ralph F. Brennan Water Treatment Plant
in Daytona Beach, Florida. The raw
water studied is a moderately colored,
high-hardness groundwater emanating
from the Floridan aquifer.
Various treatment processes studied
included single-stage coagulation/lime
softening, two-stage coagulation/lime
softening, lime softening/coagulation,
bentonite clay with lime softening, and
polymeric coagulant/clay coagulation/
lime softening.
Extensive analytical data were col-
lected and summarized on raw water
for test samples, pilot-plant process
waters, Brennan Treatment Plant sam-
ples, and Daytona Beach distribution
system samples. These data were used
to evaluate the effectiveness of each
process studied, as well as to compare
pilot-plant performance with full-scale
plant results.
This Project Summary was developed
by EPA's Municipal Environmental Re-
search Laboratory. Cincinnati, OH, to
announce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
In 1979, the U.S. Environmental Pro-
tection Agency (EPA) promulgated an
amendment to the National Interim Pri-
mary Drinking Water Regulations estab-
lishing a maximum contaminant level
(MCL) of 0.10 mg/Lfor total trihalometh-
anes (TTHM's). Two previous investiga-
tions at the Ralph F. Brennan Municipal
Water Treatment Plant in Daytona Beach,
Florida, indicated that the level of trihalo-
methanes (THM's) in the finished water
exceeded the MCL by several fold.
These reports further concluded that
the level of THM production could be
reduced by improving coagulation and by
altering the point of chlorination in the
treatment process. These studies did not,
however, determine a treatment process
that would reduce the effluent THM
concentration below the legislated maxi-
mum.
Before radical alternatives such as air
stripping, activated carbon, or alternative
disinfectants were investigated, this pro-
ject was undertaken to study further THM
reductions by means of improved coagu-
lation/lime softening techniques to re-
move precursor organics.
Procedures
Literature Search
To guide and optimize jar test work and
-------
pilot-plant efforts, reports on previous
work by others (including those at the
Brennan Plant) were reviewed. Results of
various investigations generally agreed
that THM precursors were definitely
removed through coagulation with alum
and ferric salts, but the removal efficiency
varied significantly with the raw water
source. These studies generally indicate
that coagulation of THM precursors is a
function of pH and chemical dose. Wide-
spread disagreement exists, however, as
to the dosage levels required for maxi-
mum THM reduction.
Jar Test Program
The purpose of the jar test was to refine
and confirm the results of the two earlier
studies done at the Brennan Plant and to
investigate those parameters that the
literature indicated might be effective.
The raw water studied was a moderately
colored, high-hardness groundwater em-
anating from the Floridan aquifer (see
Table 1).
Eight areas of study for THM precursor
removal efficiency were identified:
1. Aluminum sulfate (alum) versus fer-
ric sulfate (ferric) as a coagulant;
2. Coagulant dose (alum and ferric);
3. Coagulant pH (alum and ferric);
4. Coagulation followed by lime soften-
ing;
5. Lime softening followed by coagula-
tion;
6. Softening at a high pH (approximately
11) with magnesium carbonate;
7. Several other coagulant aids includ-
ing sodium aluminate, a bentonite
clay, and CatFloc-T* and;
8. Bentonite clay with lime softening.
To pursue these eight areas effectively,
an initial series of jar tests investigated
the following parameters:
1. Optimization of lime dose for soften-
ing and THM precursor removal.
2. Effect of pH on color and THM
precursor removal.
3. Effect of pH and alum dosage on color
and THM precursor removal when
treating lime-softened water.
4. Effect on THM precursor removal
when softening with lime after coag-
ulating with alum.
5. Effect on THM precursor removal
when using clay and CatFloc-T in
combination with lime.
6. Effect of sodium aluminate on THM
precursor removal when treating
lime-softened water.
'Mention of trade names or commercial products
does not constitute endorsement or recommenda-
tion for use
Table 1. Daytona Beach Raw Water Analysis. 1976-81
Parameter 5/76 11/77 .8/78
5/79
10/80 11/81
pH
Color, PCU
TDS, mg/L
Turbidity. NTU
Total Hardness. mg/L as CaCOz
Calcium, mg/L as Ca*2
Magnesium, mg/L as Mg*z
Alkalinity, mg/L as CaCO3
Chloride, mg/L
Sulfate, mg/L as SO4"2
7.31
34
392
1.8
294
104
8.5
276
50
2.0
6.8
20
323
0.26
316
114
7.2
254
37
4.3
7.0
20
424
0.25
284
101
7.4
274
33
4.3
7.1
15
362
0.4
298
95
14.3
260
42
2.0
6.9
15
383
0.3
294
111
3.8
276
36
9.8
6.8
15
452
0.1
296
102
9.5
274
54
6.2
Once the results of this first series were
analyzed, additional jar tests studied the
following:
1. Effectiveness of ferric sulfate for
THM precursor and color removal at
various dosages and pH levels.
2. Effect on THM precursor removal by
adding magnesium sulfate and sodi-
um carbonate at high pH.
3. Effect on THM precursor removal by
using CA-35 clay in combination
with Rohm and Haas XE-392 poly-
meric coagulant.
Pilot-Plant Program
The results of the jar tests and continu-
ous re-evaluations during the pilot-plant
phase of the study ultimately led to eight
pilot-plant test runs:
1. Simulation of full-scale Brennan
Plant conditions for development of
baseline data.
2. Clay and Calgon CatFloc-T in combi-
nation with lime softening.
3. Alum at 20 mg/L with American
Cyanamid Magnifloc 985N in combi-
nation with lime softening.
4. Alum at 40 mg/L with Magnifloc
985N in combination with lime soft-
ening.
5. Two-stage coagulation/softening in-
volving 40 mg/L alum with sulfuric
acid pH adjustment followed by lime
softening.
6. Two-stage coagulation/softening in-
volving sulfuric acid pH adjustment
with 80 mg/L alum followed by lime
softening.
7. Two-stage softening/coagulation in-
volving Nalco 8184 with lime fol-
lowed by 50 mg/L alum with final
NaOH pH adjustment.
8. Use of Calgon CA-35 clay in conjunc-
tion with Rohm and Haas XE 391 and
XE 392 coagulants.
Results and Conclusions
Jar Test Results
The initial jar tests showed that maxi-
mum hardness reduction could be effec-
ted at a lime dose rate of 250 mg/L. For
subsequent jar tests, 220 mg/L lime feed
was used to duplicate actual Brennan
Treatment Plant practice. This work con-
firmed earlier results, which showed that
lime coagulated with a polyelectrolyte
was inefficient as a precursor removal
process on the well water at the Daytona
Beach plant.
Subsequent jar tests compared ferric
sulfate in combination with lime soften-
ing as opposed to alum with lime. Dosage
of the coagulants was varied, as was
order of treatment. Alum at higher doses
removed color more effectively than ferric
sulfate, though both performed similarly
with respect to precursor removal. Ani-
onic and cationic polymers were tested in
combination with alum to improve floe
settling characteristics of the light alum
floe. Results were poor, but further jar
tests indicated that little benefit would be
realized in removing this floe even through
an intermediate filtration step.
Low pH coagulation of previously soft-
ened high pH water was also attempted
and resulted in effluent THM values in the
200 to 300 //g/L range.
Additional jar tests were conducted
using clay with CatFloc-T and with sodi-
um aluminate, both in combination with
lime softening. Neither process showed
results that were substantially better
than the conventional alum/lime combi-
nation.
Brief initial jar tests involving magne-
sium sulfate and sodium carbonate in
combination with lime softening showed
very promising results, with THM values
in the 200 /ug/L range. Subsequent
retesting failed to duplicate this removal
efficiency, however, so further testing
was discontinued.
-------
Extensive jar testing was accomplished
using the experimental Rohm and Haas
polymeric coagulant. This material was
tested at various doses as well as various
ratios of cationic to anionic polymer.
Further tests evaluated the use of clay in
conjunction with the polymers, as well as
the effect of chlorinating before filtration
as opposed to chlorinating after filtration.
The summary results of the jar tests
were used to establish the pilot-plant
operating program.
Pilot-Plant Test Results
The average THM concentration for the
eight pilot-plant runs are listed in Table 2.
The first pilot-plant run was intended to
simulate the full-scale operation of the
Brennan Treatment Plant to establish
baseline parameters. Pilot-plant perform-
ance generally paralleled treatment plant
performance, though higher 3-day THM's
were realized in the pilot-plant effluent
samples. This result is thought to be
caused by the lack of sludge recirculation
capability in the pilot plant, which results
in decreased efficiency.
The second pilot test showed a slight
improvement overthe baseline condition,
but the results were not promising enough
to warrant further investigation of the
clay-polymer combination.
The use of alum as studied in the third
and fourth runs showed additional im-
provement in THM precursor removal,
but the 300+ Aig/L levels obtained in
stored samples still fell far short of the
desired 100,ug/L MCL
The two-stage coagulation/softening
process attempted in the fifth and sixth
pilot runs showed continued improve-
ment, but at the expense of a significant
increase in process complexity and oper-
ational costs. Of the conventional soft-
ening/coagulation schemes studied, this
one shows the most promise; but it still
did not approach the 1 00 //g/L goal.
The seventh pilot run, in which soften-
ing was followed by coagulation, resulted
in 3-day THM readings comparable to
those in the single-stage process evalu-
ated in the third pilot run.
The final pilot series deviated signifi-
cantly from the rest of the study in that an
experimental polymeric coagulant was
used in combination with clay and lime.
This series was the most promising by far
(Figure 1), resulting in a 3-day THM
average of 158 yug/L. Because the pilot-
plant baseline tests indicated that full-
scale operation of the Brennan Plant
exceeded the performance of the pilot
unit, it is expected that this process might
Table 2. Average THM Concentration in Pilot Plant (ug/L)
Type of
Water
and THM Run #1 Run #2 Run #3 Run #4 Run #5 Run #6
RW:
CHCI3 1.14 .743 1.15 1.61 0.00 0.00
CHBrC/2 0.00 0.00 0.00 0.00 0.00 0.00
CHBrzCI 0.00 0.00 0.00 0.00 0.00 0.00
CHBr3 0.00 0.00 0.00 0.00 0.00 0.00
TTHM 1.14 .743 1.15 1.61 0.00 0.00
RWSC:
CHCI3 268 377 376 364 356 273
CHBrC/s 81.4 75.3 76.8 61.7 73.5 56.3
CHBr2CI 11.0 7.77 8.37 6.66 8.76 6.78
CHBr3 0.00 0.00 0.00 0.00 0.00 0.00
TTHM 360 460 461 432 438 336
p\A/irt.
rvVIU.
CHCI3 51.0 41.8 44.0 54.4 36.7 31.4
CHBrCh 15.7 9.90 11.1 11.2 7.35 9.6
CHBrtCI 3.63 4.28 3.12 3.10 2.23 2.49
CHBr3 0.00 0.00 0.00 0.00 0.00 0.00
TTHM 70.3 56.0 58.2 68.7 46.3 43.5
' FWSC:
CHCI3 249 262 248 238 219 155
CHBrCh 79.3 72.6 66.6 59.6 57.1 54.1
CHBrzCI 22.8 15.8 14.4 14.7 17.2 10.4
CHBr3 0.00 .43 0.28 0.00 1.23 .234
TTHM 351 351 329 312 295 220
ACSC: ACSC ACSC
CHCI3 205 195
CHBrCli 39.3 43.0
CHBrzCI 4.66 5.80
CHBr3 0.00 0.00
TTHM 243 244
*AbbrGvidtions'
0.00 - Below detection limits.
LCSC - Lime clarifier, super chlorinated, stored 72 hours.
RW - Raw water.
RWSC - Raw water, super chlorinated, stored 72 hours.
FWID - Finished water, immediately dechlorinated.
FWSC - Finished water, super chlorinated.
ACSC - Alum clarifier, super chlorinated, stored 72 hours.
Run #7 Run it 8
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
0.00 0.00
314 272
58.6 57.5
7.43 8.38
0.00 0.00
380 338
41.6 21.1
10.6 8.60
2.62 3.13
0.00 0.00
54.8 32.8
203 98.3
57.0 44.0
12.6 14.9
0.00 0.831
273 158
LCSC
339
71.2
13.4
0.00
424
approach the required 100 //g/L in a full- 01 by Russell &Axon, Engineers-Planners-
scale sludge recirculation operation. Architects, Inc., under the sponsorship of
the U.S. Environmental
Recommendations
Full-scale testing of the Rohm and
Haas polymeric coagulant with clay and
lime should be accomplished at the
Brennan Water Treatment Plant. To im-
prove the mechanism of THM removal,
precursor materials should be identified
through additional analytical work.
The full report was submitted in f ulf ill-
ment of Cooperative Agreement CR807426-
Protection Agency.
-------
60
I
to
•g
0)
I
50
40
30
20,
10
Condition
1, 220 mg/L lime + 0.06 mg/L polymer
2. 220 mg/L lime + 1 mg/L clay + 1 mg/L polymer
3. 22 mg/L alum + 220 mg/L lime + 0.3 mg/L polymer
4. 40 mg/L alum + 220 mg/L lime + 0.3 mg/L polymer
5. 40 mg/L alum @ pH-5 + 220 mg/L lime
6. 80 mg/L alum @ pH-5 + 220 mg/L lime
7. 220 mg/L lime+ 0.06 mg/L polymer
@ pH-8 + 50 mg/L alum
8. 5 mg/L clay + 20 mg/L XE-392 + 3.5 mg/L XE-391 +
220 mg/L lime
Average removal - THMFP,^ - THMFP,in,shea
THMFP,K
345
Condition No.
Figure 1. THM formation potential removal in the pilot plant
-------
J. C. Thurrott andH. Zwart-Duryea are with the City ofDaytona Beach, FL 32015;
J. R. Sloane is with Russell and Axon. Engineers-Planners-Architects, Inc.,
Daytona Beach, FL 32015; and D. Williamson is with Environmental Science
and Engineering, Inc., Gainesville, FL 32601.
O. Thomas Love, Jr. is the EPA Project Officer (see below).
The complete report, entitled "Trihalomethane Removal by Coagulation Tech-
niques in a Softening Process," (Order No. PB 83-151 845; Cost: $11.50,
subject to changej will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
•&U. S. GOVERNMENT PRINTING OFFICE: 1983/659-095/1910
-------
J
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
RETURN POSTAGE GUARANTEED
Third-Class
Bulk Rate
MERL0063240
LOU N T1LLEY
REGION V EPA
SI
CHICAGO It 60604
------- |