&EPA
United States
Environmental Protection
Agency
Research and Development
r,
.-*
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
EPA-600/S2-81-077,078:079 July 1981
Project Summary
Removing Trace Organics
From Drinking Water Using
Activated Carbon and
Polymeric Adsorbents
C. S. Oulman, V. L. Snoeyink, J. T. O'Connor, and M. J. Taras
"Bench-Scale Evaluation of Resins
and Activated Carbons for Water
Purification." by V.L. Snoeyink, W.A.
Chudyk, D.O. Beckmann, P.M. Boening,
and T.J. Temperly. In the first of a
three-volume study, adsorption iso-
therms and bench-scale column studies
were used to com pare the performance
of five types of commercially available
activated carbons and four types of
resins for removing humic acids, f ulvic
acids, 2-methylisoborneol (MIB), and
chloroform from water. For adsorbing
humic materials, some of the activated
carbons and the weak base phenol-
formaldehyde resins performed satis-
factorily. The same activated carbons
provided satisfactory removal of MIB,
although the capacity was reduced
somewhat in the presence of humic
acid. The carbonaceous resin and one
of the activated carbons has about the
same capacity for chloroform removal
at concentrations under 0.5 mg/L.
The presence of 10 mg/L of humic
acid had little effect on their capacity
for adsorbing chloroform.
"The Removal of Trace Organics
from Drinking Water Using Activated
Carbon and Polymeric Adsorbents,"
by J.T. O'Connor, D. Badorek, and L.
Thiem. In the second volume, a pilot
plant was operated at the Kansas City,
Missouri, Water Treatment Plant to
study adsorption as a means of remov-
ing trihalomethanes (TTHM) and total
organic carbon (TOC) from drinking
water. The pilot plant consisted of 15
columns, 15 cm (6 in.) in diameter,
and each containing about a 0.9-m (3-
ft) depth of a granular adsorbent.
Granular activated carbons and poly-
meric adsorbents were compared in
four extended tests conducted over
periods of 183, 111, 65, and 129
days. The pilot-plant studies demon-
strated the effects of regeneration,
variations in trace organic concentra-
tion, and depth of adsorbent on trace
organic removal including effluent
concentration and adsorption capacity.
"Trace Organics Removal Using
Activated Carbon and Polymeric Ad-
sorbents," by C.S. Oulman. In the
third volume, a survey was made to
determine the trace organic matter in
raw and treated water from 14 water
utilities across the United States.
Monthly analyses were made for TTHM
and TOC. Analyses were made on
carbon/resin adsorbable ether extracts
from each utility for a number of
indicator compounds and for bacterial
mutagenicity as measured by the
Ames test. The results of the water
quality survey indicated that most of
the water utilities are able to produce
an acceptable finished water with
conventional treatment methods. In
those places where additional treat-
ment is needed for trace organics
removal, activated carbon will probably
-------
be the more versatile adsorbent to use.
The results from this effort have been
published by Glatz, et al. in the Journal
American Water Works Association.
70(8):465-468, 1978.
EPA did not participate in this por-
tion of the overall project, but some of
the results have been included in the
full report. This third volume also
contains an executive summary of
results obtained from the first and
second volumes. Only the bench-scale
activities (first volume) and the pilot-
scale adsorption studies (second vol-
ume) are discussed in this Project
Summary.
This Project Summary was devel-
oped by EPA's Municipal Environ-
mental Research Laboratory, Cincin-
nati, OH, Jo announce key findings of
the research project that is fully docu-
mented in three separate reports fsee
Project Report ordering information at
back).
Introduction
In 1975, the American Water Works
Association (AWWA) listed 15 high-
priority problems needing study. One
problem was called "Reliable Screening
Tests and Techniques for Determining
an Evaluation of Organics in Drinking
Water," and another was "Method for
Removal of Organics in Drinking Water."
Likewise, the U.S. Environmental Pro-
tection Agency (EPA) was concerned
that many organic compounds were not
being removed by conventional water
treatment practice; and further, that
chlorine used for disinfection was
shown to produce halogenated by-
products. A cooperative research effort
was initiated to (1) examine the oc-
currence of trace organics in drinking
water; and (2) evaluate the efficiency of
their removal by a "broad spectrum"
adsorbent, granular activated carbon,
and more selective adsorbents—poly-
meric resins.
The project was divided into three
parts and reported separately. Bench-
scale studies (first volume) were con-
ducted at the University of Illinois to
select the adsorbents for pilot-scale
column use. In the second volume,
adsorption columns were designed,
fabricated, and operated by personnel
from the University of Missouri—Co-
lumbia and located at the Kansas City,
Missouri, water treatment plant The
third volume of the project was a survey
of trace organics in 14 different water
utilities across the United States. That
work was supported by the AWWA
Research Foundation and its partici-
pating members and conducted by the
Ames Laboratory at Iowa State Uni-
versity. The results and conclusions
from the first and second volumes will
be discussed individually in this Project
Summary.
A major objective of the overall study
was an evaluation of the removal of
taste-and-odor-producing compounds
as measured by threshold odor number
(TON). The influent TON values were,
however, consistently low, which made
clear-cut evaluations of the removal of
odor-producing compounds inconclusive.
Bench-Scale Studies
(first volume)
Bench-scale studies were made to
determine which of the various com-
mercially available adsorbents should
be used in the side-by-side comparison
pilot-scale study of activated carbon and
polymeric adsorbents. Adsorption iso-
therms, using water-containing chloro-
form, humic and f ulvic acids, and MIB as
solutes, were determined on five types
of activated carbons and four different
resins.
The macroporous, phenol-formalde-
hyde, weak-base resin had a high
capacity for humic substances and
could be regenerated with sodium
hydroxide but was not able to remove
the earthy-musty odor compound, MIB.
The styrene-divinyl benzene resin did
not adsorb humic substances, but it did
have some capacity for MIB. The acrylic
and carbonaceous resins did not adsorb
humic materials or MIB, butthecarbon-
aceous resin had an excellent capacity
for chloroform. The activated carbons
could remove the humic substances and
the MIB but had a relatively small
capacity for chloroform.
A number of adsorbents were recom-
mended for use in the pilot-plant tests,
based on the adsorption isotherms and
the mini-column studies. Initially, one
carbon, Westvaco Nuchar® WVG,* was
recommended for a side-by-side com-
parison with polymeric adsorbents
because of its good capacity for both
humic substances (27.6 mg/g**) and
MIB (112.6 mg/g). Later, other carbons
were selected for inclusion in the pilbt
plant tests.
No one polymeric adsorbent could be(
recommended as having the capacity to
remove organic matter in such a wide
range of molecular weights as did any of
the activated carbons. Therefore, two
materials were selected to be used in
tandem — an adsorbent for high molecu-
lar weight compounds such as humic
acids and an adsorbent for low molecu-
lar weight compounds such as MIB and
chloroform. Diamond Shamrock ES-
561 was recommended for the humic
acid removal application because of its
reasonably low swelling properties and
fairly high capacity for humic acids in
the neutral pH range. This was not one
of the polymeric adsorbents evaluated
in the bench-scale tests, but its proper-
ties are similar to one of the phenol-
formaldehyde resins tested. Rohm and
Haas Ambersorb® XE340 was the other
polymeric adsorbent tested; it has a high
capacity (18.2 mg/g) for chloroform
removal. In the later phases of the pilot-
plant experimentation, however, addi-
tional resins were selected for study.
Pilot-Scale Adsorption Studies
(second volume)
Results
Fifteen glass columns, each 1 5 cm in
diameter and containing approximately
0.9 m of adsorbent, were located at the
Kansas City, Missouri, water treatment
plant and operated in a post-filtration
mode. Table 1 indicates the way the
various columns were loaded during
each phase of the study. The average
concentrations of TTHM and TOC were
42 fjg/L and 2.5 mg/L, respectively.
Phase I
Over an initial period (Phase I) of 133
days, a 0.9-m bed of bituminous-base
granular activated carbon (Nuchar®
WVG) removed 70%* of the influent
TTHM. Approximately 2.7 m of Nuchar®
WVG were required to remove 99% of
the influent TTHM. A 0.9-m-deep bed of
a carbonaceous resin Ambersorb® XE-
340 provided 98% TTHM removal during
the same period. Steaming the Nuchar®
WVG increased TTHM removal to 87%,
whereas steaming the weak-base anion
exchange resin (Diamond Shamrock
ES-561) had no significant measurable
effect on an initially low removal ef-
ficiency. The periodic steaming of the
'Mention of commercial products does not con-
stitute approval or endorsement by EPA.
"Isotherm capacities reported for an equilibrium
concentration of 1,000 /jg/L.
*AII removal efficiencies are based on influent and
effluent concentrations averaged over the project
phase period
-------
Table 1. Identification of Adsorbents Utilized in Kansas City, Missouri Pilot
Plant Evaluation of Removal of Organic Substances from Drinking
Water
Column
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Phase 1
1(1 33 days)
Feb.-Aug. '77
WVGM
WVGM
WVGM
ES-561M
ES-561M
XE-340M**
—
—
—
—
—
—
—
XE-340M
EX-561M**
Phase II
11(111 days)
Aug. -Dec. '77
WVGM
WVGM
WVGM
ES-561 nm
ES-561, 1R)
XE-340nn)**
HD-1030M
HD-1030M
HD-1030M
LCKM
C-THMM
ROWO.SM
Sand
WVGM**
ES-561 am**
Phase III
III (65 days)
Mar-June '78
WVGnm
WVGnm
WVGnm
XE-340vm**
IRA-904M
WVGM
HD-1030nm
HD-1030nm
HD-1030nm
A-162M
WVGM
HD-1030M
Sand
LCKM
WVGM
Phase IV
IV (129 days)
July-Nov. '78
WVG(2m
WVGun
WVG(2m
XE-340l3m*
IRA-904nm*
WVGM*
HD- 1030 urn
HD-1030(2n
HD-1030(2m
WVGnm
WVGM
HD-1030M
HD-1030M
WVGM
WVGM**
Empty Bed Contact Time:
[11.2 min @ 2 gpm/sf
*2.2 min @ 10 gpm/sf
WVG Bituminous Base Carbon - Westvaco
HD-1030 Lignite Base Carbon - ICI
LCK Petroleum Base Carbon - Union Carbide
C- THM Bituminous Base Carbon developed by Calgon for enhanced removal
of trihalomethanes.
ROW 0.8 Extruded Peat Base Carbon - American Nor it
ES-561 Weak Base Anion Exchange Resin - Diamond Shamrock
XE-340 Carbonaceous Resin - Rohm and Haas
IRA-904 Strong Base Anion Exchange Resin - Rohm and Haas
A-162 Strong Base Anion Exchange Resin - Diamond Shamrock
[**Adsorbents subjected to steaming
(v) Virgin adsorbent
(1R) Once regenerated
(2R) Twice regenerated
(3R) Thrice regenerated
Nuchar® WVG reduced the number of
microorganisms recovered from the
adsorbent from 89,000 to 5,800 colo-
nies/gram.
Throughout Phase I, the influent con-
centration of TTHM steadily increased
from less than 10 /jg/L (February) to
over 80 fjg/L TTHM (June) as influent
water temperature increased. An im-
portant outgrowth of the present study
was the establishment of the seasonal
pattern of TTHM formation in the finished
water at Kansas City. This facilitated
subsequent decisions as to when virgin
and regenerated adsorbents should be
placed in service. TOC measurements
were included in the sampling protocol
near the end of Phase I.
Phase II
During Phase II, all 15 columns were in
operation; this permitted the perfor-
mance of the carbons made from bitu-
minous coal, lignite, peat, and petroleum
to be compared. TTHM levels had in-
creased to a peak of approximately 200
fjg/L in the late summer, establishing a
pattern that was to be repeated in the
following year. This maximum TTHM
level provided a more significant chal-
lenge for the adsorbent than was present
in Phase I. Over the 111 days of Phase II
operation, the TTHM removals were
comparable to those observed during
the first 111 days of Phase I with the
following removals: bituminous, 80%;
lignite, 83%; petroleum, 82%; bitumi-
nous base carbon enhanced for TTHM
removal, 85%; and extruded peat base
carbon, 68%. Steaming of the Nuchar®
WVG column again increased TTHM
removal. The steamed column removed
92% of the TTHM over a period of 111
days, exactly equaling the percent
removal observed over the first 111 -day
period of Phase I.
Once again, the Ambersorb® XE-340
was effective in removing 90% of the
influent TTHM. Because the Diamond
Shamrock ES-561 continued to be
erratic and generally ineffective, it was
eliminated from further pilot-plant
testing. The 0.9-m-deep beds of granu-
lar activated carbon removals of TOC
were bituminous, 51%; lignite, 37%;
petroleum, 30%; bituminous carbon
enhanced for TTHM removal, 19%; and
extruded peat base carbon, 45%. It
became evident that there were far
greater differences in carbon perform-
ances with TOC than with the removal
of the small amounts of TTHM in the
influent. Moreover, the Calgon carbon
(Filtrasorb® C) developed to enhance
TTHM removal did achieve superior
removal of TTHM but at the expense of
reduced TOC removal capability. Ap-
parently, the pore size distribution that
results in more effective TTHM removal
retards the removal of a range of other,
larger adsorbates. Steaming of the
Nuchar® WVG column appeared to in-
crease TOC removal modestly to 56%.
Perhaps most significant was that the
2.7-m depth of Nuchar® WVG was in-
capable of removing more than 75% of
the influent TOC at the 5 m/hr (2
gpm/ft2) application rate.
Neither Diamond Shamrock ES-561
nor Ambersorb® XE-340 showed any
significant TOC removal. At the end of
Phase II, the Nuchar® WVG and Hydro-
darco® 1030 were returned to the
respective manufacturers for thermal
reactivation to prepare for a study of the
effect of reactivation on adsorbent per-
formance (Phase 111).
-------
Phase III
Reactivating the carbons restored their
virgin adsorption capacity, as measured
by Iodine Number and Decolorizing Index.
During Phase II of the study, comparison
of TOC removal indicated similar
performance between 0.9-m-deep beds
of Nuchar® WVG, whether it was a
once-reactivated (60%), virgin (62%), or
regularly backwashed virgin (61 %)
product. Once-reactivated and virgin
lignite base carbon (Hydrodarco® 1030)
removed 43% and 51% of the TOC,
respectively, over the period of Phase III.
No steaming was done during Phase
III. Since the influent concentrations of
TTHM were generally less than 5 /jg/L,
little information on TTHM removal was
obtained. Thus, Phase III was terminated
after only 65 days so that Phase IV could
be initiated immediately before the time
when the maximum TTHM influent
concentration was expected, based on
the previous year's seasonal pattern.
The Ambersorb® XE-340 was again
ineffective in TOC removal during Phase
III, whereas a 0.9-m-deep bed of a
strong-base anion exchange resin (Rohm
and Haas Amberlite® IRA-904) appeared
to remove roughly one-third of the
influent TOC. Another bed of a different
strong-base anion exchange resin (Dia-
mond Shamrock A-162) removed 47%
of the influent TOC. On a weight basis.
Diamond Shamrock A-162 was only
one-half as effective as Nuchar® WVG
activated carbon.
The Ambersorb® XE-340 was placed
first in the series of three columns to
observe the effect of high molecular
weight components of TOC on the ad-
sorption of TTHM by the resin. Laboratory
studies had indicated that high molecu-
lar weight organic substances might be
irreversibly adsorbed to the carbonace-
ous resin leading to "fouling" and loss
of TTHM removal capability. Since this
resin is still under development and
evaluation, the manufacturer was un-
certain of the appropriate regeneration
procedure. After subsequent EPA stu-
dies, a far more vigorous steaming
procedure is now being recommended
for the Ambersorb® XE-340 than was
recommended at the time of the present
study.
Large numbers of microorganisms
were dislodged from the adsorbents at
the end of Phase III in June 1978. The
inert sand media and the resins harbored
few microorganisms whereas the acti-
vated carbon supported significant
growth, particularly in those columns
that had the most TOC removed. Back-
washing of a column containing virgin
Nuchar® WVG resulted in more than an
order of magnitude reduction in bacterial
count. For example, 850,000 colonies/
gram were found on the activated
carbon in the undisturbed column,
compared with 52,000 colonies/gram
on the backwashed activated carbon,
indicating the effectiveness of back-
wash in controlling accumulations of
organisms.
Phase IV
Nuchar® WVG and Hydrodarco® 1030
were both reactivated to levels beyond
their virgin capacities by the manu-
facturers in preparation for Phase IV.
This resulted in the following TTHM
removals by 0.9-m-deep beds of Nuchar®
WVG: twice-reactivated, 75%; once-
reactivated, 67%; virgin, 66%; and virgin
(replicated), 67%. The regularly steamed
column of virgin Nuchar® WVG removed
96% of the influent TTHM overthe 129-
day period. The twice-reactivated, once-
reactivated, and virgin Hydrodarco®
removed 77%, 72%, and 62% of the
influent TTHM. Again, the reactivation
of this activated carbon beyond its virgin
capacity resulted in slightly enhanced
TTHM removal.
TOC removals were marginal, as
before. The 0.9-m-deep beds of Nuchar®
WVG removals were twice-reactivated,
55%; once-reactivated, 50%; virgin,
56%; and virgin (replicated), 53%. Steam-
ing increased the TOC removal of the
virgin Nuchar® WVG to 63%. Approxi-
mately 2.7 m of the twice-reactivated
Nuchar® WVG removed a total of 78%
of the TOC from fairly constant influent
levels of 2 mg/L. The 0.9-m-deep beds
of Hydrodarco® 1030, twice-reactivated,
once-reactivated, and virgin removed
39%, 38%, and 36% of the influent TOC.
Approximately 2.7 m of the Hydrodarco®
1030 removed just under 58% of the
TOC.
The Ambersorb® XE-340 was regen-
erated after each of the first three
phases with 1 -1 /2-bed volumes of low
pressure (12 psig) steam. Despite this
procedure, the performance of the
Ambersorb® XE-340 declined. In Phase
IV, at the higher (20 m/hr) application
rate, Ambersorb® XE-340 removed only
54% of the influent TTHM and 5% of the
influent TOC. As observed a year earlier,
the cool (November) water temperatures
suppressed organism growth on all of
the adsorbents.
Overall, the influent adsorbate con-|
centrations and column removals were
consistent between replicates, in suc-
cessive phases of operation, and with
successive regenerations. Granular
activated carbon exhibited the potential
for prolonged removal of both TTHM and
TOC.
Conclusions
1. In studies using pilot plant adsorp-
tion columns to adsorb halogenated
organic substances from softened,
filtered Kansas City water, acti-
vated cabon and carbonaceous
resin were able to remove TTHM
for extended periods. Conversely,
strong and weak-base anion ex-
change resins were not able to
remove the TTHM. Onlythe granu-
lar activated carbons were effective
in removing significant amounts
of the TOC present. Periodic steam-
ing of the activated carbon columns
reduced bacterial growth and en-
hanced TTHM and TOC removal.
Regular backwashing of the acti-
vated carbon columns was also
effective in reducing the accumu-
lation of bacterial growth.
2- Little or no difference was observed
in removals obtained by twice-
reactivated, once-reactivated, and
virgin activated carbons, indicating
that calcium carbonate deposits
from lime-softened water did not
coat the adsorbent and impair its
adsorption capacity.
The full three-volume report was
submitted in fulfillment of Grant No. R-
804433 by the University of Illinois,
University of Missouri—Columbia, and
Iowa State University, under the spon-
sorship of the U.S. Environmental Pro-
tection Agency.
4
-------
C. S. Oulman, V. L Snoeyink, J, T. O'Connor, and M. J. Taras are with Iowa State
University, Ames, IA 50010; University of Illinois, Urbana, 1L 61801; Univer-
sity of Missouri, Columbia, MO 65211; and A WWA Research Foundation,
Denver, CO 80235, respectively.
Thomas Love, Jr. is the EPA Project Officer (see below).
The complete reports, entitled:
"The Removal of Trace Organics from Drinking Water Using Activated Carbon
and Polymeric Adsorbents," (Order No. PB 81-196 768; Cost:$11.00)
"Bench-Scale Evaluation nf Resins and Activated Carbons for Water Purifica-
tion," (Order No. PB 81-196 776; Cost: $8.00)
"Trace Organics Removal Using Activated Carbon and Polymeric Adsorbents,"
(Order No. PB 81-196 784; Cost: $8.00)
The above reports will be available only from: (prices are subject to change)
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
i US GOVERNMENT PRINTING OFFICE 1981-757-012/7ZOZ
-------
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 S300
RETURN POSTAGE GUARANTEED
------- |