\r/
United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-83-032 June 1983
Project Summary
Organic Contaminant Removal in
Lower Mississippi River Drinking
Water by Granular Activated
Carbon Adsorption
Wayne E. Koffskey, Noel V. Brodtmann, and Ben W. Lykins, Jr.
Concern over the detection of known
and suspected carcinogens in the New
Orleans, Louisiana, area water supplies
resulted in a research effort to provide
data for removal of these compounds.
The primary objective was to examine
the efficiency of using granular acti-
vated carbon (GAC) for the removal of
organic contaminants in drinking water.
Two full-scale systems were compared
and evaluated-a post filtration adsorp-
tion GAC filter in series with a sand
filter and a combined filtration adsorp-
tion GAC filter. Also examined were
the ability of pilot GAC columns to
predict the organics removal efficiency
of full-scale GAC filters, the effects of
varying empty-bed contact times, and
variations in the organic contents of
drinking water resulting from different
types of water treatment processes.
Correlations between certain nonspe-
cific parameters (such as total organic
carbon) and specific individual param-
eters were examined in the hope that
one or more nonspecific parameters
could be used as a surrogate monitor-
ing parameter.
Both GAC systems effectively re-
moved organic contaminants with the
same relative adsorption efficiencies.
Efficiencies were not reduced because
of pore blockage by turbidity. A few
organic substances (phthalates, n-
alkanes, and substituted benzene de-
rivatives) had little or no adsorptive
affinity for GAC, however, at the ng/L
level.
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
The Mississippi River, in its 3,862-
kilometer (2,400-mile) course to the Gulf
of Mexico, ultimately drains nearly two-
thirds of the continental United States.
Water from the Mississippi and its major
tributaries support many cities and indus-
tries and serves as the receiver for vast
quantities of municipal and industrial
wastes on a continuous basis. Moreover,
some of the Nation's largest petrochem-
ical manufacturing complexes are located
along both banks of the river from Baton
Rouge to below New Orleans. These large
industrial complexes continuously cycle
vast quantities of cooling and process
water back to the river, in many cases
adding dissolved organic contaminants to
the river's total burden. It is this water that
is the source of drinking water for the New
Orleans metropolitan area, including Jeffer-
son Parish.
This project, which was one of the first
of its kind in the country on a full size plant
scale, is of major importance particularly in
light of the growing public and official
concern about contaminants in the Nation's
drinking water supplies. Furthermore,
new and impending Federal legislation
and regulations have required an expan-
sion of the collective knowledge and abili-
ties in the area of removing microcon-
taminants from drinking water during the
treatment process.
-------
Procedures
Study Design
To meet the study objectives, a four-
phase study was designed. Four non-
specific organic parameters, 1 8 volatile
organic species, 66 nonvolatile organic
species, and general chemical and micro-
biological water quality parameters were
monitored for each step in the full-scale
treatment process. This process included
the post-filtration adsorption and the com-
bined filtration adsorption GAC filters.
During the latter two phases, parallel ef-
fluent quality data were collected from two
7.6-cm. diameter (3-m.-diameter) glass
columns simulating both types of full-
scale GAC filtration systems. Similar data
were collected for four 10.2-cm.-diameter
(4-in.-diameter) glass columns in series,
each with an empty bed contact time
(EBCT) of 10 min. ThecumulativeEBCTof
each consecutive column in series was
thus 10, 20, 30, and 40 min.
Flow rates, hydraulic loadings, and EBCT
for all of the GAC systems, both full- and
pilot-scale, were evaluated and used in
comparative interpretations of the analyti-
cal data. GAC losses for the adsorber and
filter adsorber were also closely monitored.
Treatment.Schemes
During each of four sequential phases of
the project (I, MA, MB, and 111), two different
GAC filter configurations were evaluated
simultaneously to determine the effect of
influent turbidity on GAC adsorption and
the suitability of 1 2x40 mesh GAC as a
filtration media. The first filter (the filter
adsorber) received clarified chloraminated
water and served the dual purpose of
removing filterable turbidity and adsorb-
ing dissolved organic compounds. The
second GAC filter (the adsorber) received
sand filter effluent and functioned as a
post-adsorption treatment step.
During Phases I and IIA, the filter ad-
sorber consisted of a 15.2-cm. (6-m.)
layer of sand covered by about 61-cm.
(24-m.) of GAC (WestvacoWVG 12x40).*
For these two phases (I and IIA), the
adsorber consisted of 76.2-cm. (30-m.) of
the same type of carbon that was used in
the filter adsorber. During Phases MB and
III, the sand layer was removed and the
filter adsorber contained approximately
76.2-cm. (30-m.) of GAC (WestvacoWVG
1 2x40 for Phase MB, and Calgon Filtrasorb
400 for Phase III). The adsorber for these
two phases (MB and III) also contained
76.2-cm. (30-in.) of the same type of GAC
used in the filter adsorber.
All of the treatment variations along with
the dosages used during each phase of the
project are shown in Table 1.
To determine how well pilot columns
could predict the performance of a similar
full-scale system, two GAC pilot columns
were operated under the same conditions
and in parallel with the full-scale adsorber
and filter adsorber units during Phases MB
and III. Both of these Pyrex glass columns
were 7.6 cm. (3-in.) in diameter and 1.5-
m. (5-ft.) high, and contained 76 cm. (30-
m.) of GAC supported by 1 5.2-cm. (6-m.)
of gravel. These pilot columns were desig-
nated as the adsorber simulator and the
filter adsorber simulator. Each received
influent water from the same sources and
at the same relative hydraulic loadings as
their full-scale counterparts.
To observe the effect of EBCT upon
carbon adsorption, four other GAC pilot
columns (contactor Nos. 1, 2, 3, and 4)
were operated in series during Phases MB
and III at a flow rate of 741 miymin. to
simulate the effects of 10, 20, 30, and 40
min. of EBCT on a system operating as an
adsorber. These Pyrex glass columns
were 10.2 cm. (4-in.) in diameter and 1.8-
m. (6-ft.) high. Each column contained 91 -
cm. (36-in.) of GAC supported by 1 5.2-
cm. (6-in.) of gravel. The influent water for
these series contactors was the same as
that for the adsorber filter and the ad-
sorber simulator column.
Results and Conclusions
1. The conventional treatment process
effectively removed some organic contam-
inants, notably the surrogate parameters:
total organic carbon (TOC) (3% to 25%)
and trihaiomethane formation potential
(THMFP) (22% to40%) across the precip-
itator; and ultraviolet (UV) (50% to 55%),
rapid fluorescence measurement (RFM)
(14% to 28%), and emission fluorescence
scan(EMF) (10% to 31%) upon the addi-
tion of 0.5 to 1.0 mg/L potassium per-
manganate. These results indicate that
conventional water treatment is signifi-
cantly, but not totally, effective in remov-
ing general types of organic contaminants
from drinking water. No significant differ-
ence with respect to organics removal was
observed between lime softening and poly-
electrolyte polymer treatment. Because of
the nonreactive nature of monochloramine
with the THM precursor substances, the
formation of trihalomethanes (THM) was
not affected by the difference in pH of
these treatment processes.
2. The GAC systems were effective in
removing organic contaminants studied.
Table 1. Treatment Schemes of All Phases During the Research Project
Treatment Step
Phase I
2/77 to
8/77
Phase IIA
11/77 to
4/78
Phase IIB
7/78 to
1/79
Phase III
4/79 to
10/79
* Mention of trade names or commercial products does
not constitute endorsement or recommendation for
use
Potassium
permanganate (mg/L) 0.5-1.0 0.5-1.0 0.5-1.0
DADM*-type polymer (mg/L) 0.5-4 4-8 2-8
DMAt-type polymer (mg/L) - 0-1 0-3
Ferric chloride (mg/L)
Lime (mg/L) 50-115 7-10 8-10
Ferrous sulfate (mg/L) 3.5-34
Sodium hexameta-
phosphate (mg/L) 1.0 1.0 1.0
Monochloramine
residual (mg/L) 1.0-2.1 0.9-2.0 0.8-2.2
GAC type WVG 12x40 WVG 12x40 WVG 12x40
GAC-adsorber
(depth in cm^m3) 76.2:27 76.2:27 76.2:27
(depth in in.:ft3) 30:954 30:954 30:954
GAC-filter adsorber
(depth in cm.:™3) 61:21.6 61:21.6 76.2:27
(depth in in.:ft3) 24:763 24:763 30:954
0.5-1.0
1-8
0-3
0-25
0.6-2.4
Filtrasorb
400
76.2:27
30:954
76.2:27
30:954
* - diallyldimethylammonium chloride
t - dimethylamine
-------
Those GAC systems operating in the ad-
sorber mode exhibited the same relative
adsorption efficiencies as those operating
in the filter adsorber mode; this indicated
the applicability of either type of system
for organics removal. No observable re-
duction in the adsorption efficiency of the
filter adsorber existed because of pore
blockage by turbidity.
3. The GAC pilot column simulators
were generally quite effective in predicting
the adsorption efficiencies of their respec-
tive full-scale counterparts after taking
into account the variations caused by GAC
losses during backwashing in the full-
scale systems. Thus the pilot column sys-
tem can be an effective and economical
tool in the design of full-scale GAC facilities.
4. The breakthrough profiles of the sur-
rogate parameters through the various
GAC systems did not correlate reliability
with those of the individual parameters
under study when simple linear correla-
tions were used. Multiple, nonlinear-type
correlations were indicated; but none
were found that correlated well with the
data.
5. Evidence indicates that a critical GAC
bed depth (which varies for different sub-
stances) is required to remove a particular
organic contaminant. This critical bed
depth is generally deeper for the THMs
than it is for the higher-molecular-weight
substances such as the chlorinated hydro-
carbon insecticides because these THM
substances break through the GAC beds
at different rates. For example, the GAC
beds with a 20-min. EBCT were saturated
with THMs within 100 days, but contactor
No. 1 with a 10-min. EBCT continued to
remove more than 90% of the chlorinated
hydrocarbon insecticides at the end of
each phase.
The increase in EBCT across the series
contactors resulted in increased adsorp-
tion efficiency for most of the parameters
under study. But the increment of the
increase in efficiency across each con-
secutive contactor decreased as EBCT
increased.
6. A steady-state condition was reached
for the surrogate parameters of TOC, UV,
RFM, EMF, and THMFP after approxi-
mately 100 days where a relatively con-
stant removal of these constituents was
observed. The relative removal levels for
each constituent increased with EBCT up
to 30 min.; afterthat, no further significant
removal was observed.
7. The use of chloramination (specific-
ally monochloramine) as the sole means of
disinfection without the use of GAC filtra-
tion was effective in maintaining the THM
level below 10 ppb in the distribution
system. A mean coliform distribution of
zero and total bacteria level of less than 50
counts/mL were observed during this
time.
8. A concentration gradient effect was
observed for both types of GAC systems.
Even after saturation, large surges in in-
fluent concentration were effectively ad-
sorbed with relatively no apparent adverse
effects in the effluent of the GAC systems.
This phenomenon greatly improved the
effectiveness of GAC filtration in com-
bating chemical spills in the surface water
source.
9. The following organic substances
had little or no adsorptive affinity for GAC
at the ng/L level and did not exhibit the
typical s-shaped breakthrough curve:
phthalates, n-alkanes, and substituted
benzene derivatives. Some of these sub-
stances exhibited a low degree of constant
removal. Also, the surrogate parameters
(TOC, UV, RFM, EMF, and THMFP) showed
less than 100% removal after 40 min. of
EBCT during the series contactor studies.
10. Variations in the effectiveness of
the different types of GAC used during this
project were observed for the removal of
the lower molecular-weight volatile sub-
stances. The WVG 12x40 mesh carbon
used during the first three phases ap-
peared to have had a higher adsorptive
efficiency for these substances than did
the Filtrasorb 400 used during Phase III.
11. The 1 2x40 mesh GAC medium in
the filter adsorber appeared to remove
turbidity as well, if not better, than the
sand medium when equal depths of the
media were compared. Thus since the
adsorption efficiency of the two GAC sys-
tems (adsorber and filter adsorber) are
similar, conversion from sand filtration to
GAC filtration (sand replacement) would
appear to be advantageous for those seek-
ing organic removal and turbidity reduction.
Wayne E. Koffskey is with the Jefferson Parish Department of Public Utilities.
Jefferson. LA 70121; Noel V. Brodtmann is presently with Environmental
Professionals Ltd.. Metairie, LA 70001; the EPA author Ben W. Lykins.Jr. (also
the present EPA Project Officer see below) is with the Municipal Environmental
Research Laboratory, Cincinnati, OH 45268.
Jack DeMarco was the EPA Project Officer.
The complete report, entitled "Organic Contaminant Removal in Lower Mississippi
River Drinking Water by Granular A ctivated Carbon A dsorption," (Order No. PB
83-194 506; Cost: $47.0O. subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, MA 22161
Telephone: 703-487-4650
The present EPA Project Officer can be contacted at:
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
-------
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
CHICAGO IL 60604
------- |