United States
Environmental Protection
Agency
Risk Reduction
Engineering Laboratory
Cincinnati OH 45268
Research and Development
EPA/600/S2-90/049 Jan. 1991
&EPA Project Summary
Radon Removal by POE
GAC Systems: Design,
Performance, and Cost
Jerry D. Lowry, Sylvia B. Lowry, and Julie K. Cline
Data were collected from 121
commercial point-of-entry (POE)
granular activated carbon (GAC)
units over an 8-yr period. These data
have been summarized to indicate
the effectiveness of these units to
remove radon (Rn) from drinking
water supplies. Although the long-
term data are limited to but a few of
the 121 units, it is clear that GAC is
effective for removing radon from
drinking water.
This Project Summary was
developed by EPA's Risk Reduction
Engineering 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 the report summarized here, field
data has been collected from sites that
use GAC to remove Rn from groundwater
supplies. The data which have been
collected from 1981 to 1989 are focused
primarily on POE applications and, to a
lesser extent, on small water supplies
that serve schools or small housing
developments. The full report includes
information about design, installation,
operation, monitoring, performance,
gamma exposure rates, shielding, and
costs.
There were 121 POE GAC units
(manufactured by Lowry Engineering, Inc.
[LEI])* located in 12 states that were sold,
installed, and monitored to varying
degrees over a 7-yr period. Most of the
POE GAC units were sold, installed, and
monitored as part of a private research
effort by Lowry Engineering, Inc., initiated
in October 1984. Most of the POE GAC
units are in Maine and New Hampshire
(Table 1), and the units were used with
waters having varying quality
characteristics.
Table 1 Number of POE GAC
Units by State
State
Maine
New Hampshire
New Jersey
Kentucky
Pennsylvania
Massachusetts
Colorado
Rhode Island
Connecticut
New York
North Carolina
Vermont
Number
61
20
12
1
6
5
3
3
6
1
1
1
*Mention ot trade names or commercial products
does not constitute endorsement or
recommendation for use
The U.S. Environmental Protection
Agency (EPA) under the Safe Drinking
Water Act is considering a maximum
contaminant level for Rn of between 100
and 2,000 pCi/L. The work and results
outlined in this report must be taken into
perspective with the current concern
about Rn levels in drinking water
supplies. All of the decisions associated
with the POE GAC units described in this
report were based on the respective
states safe guideline advisory levels in
the range of 10,000 to 20,000 pCi/L.
Many of these GAC units will not produce
a treated water Rn level below the new
proposed levels of 200 to 2,000 pCi/L.
Printed on Recycled Paper
-------
This does not mean that POE GAG units
will not be effective for future installations
after the new MCL is embraced; however,
different design and application decisions
would have to be considered to ensure
satisfactory performance.
Design, Installation, and
Monitoring of POE GAG Units
The POE GAG units were single
vessels housing 1.0 to 3.0 ft3 granular
activated carbon, depending upon the
model purchased. The majority of the
units contained 1.7 ft3 of carbon. A
typical POE GAG unit is illustrated in
Figure 1. These units were normally
installed downstream of an existing
pressure tank and were operated under
the existing household water pressure.
The GAG units were designed to operate
in the down flow to minimize
backwashing. At some locations, shields
(made of water, lead, or bricks) were
placed around the GAG unit to monitor
the reduction in gamma emissions from
the units.
Once the GAG unit was properly
installed and commissioned, it was
essentially maintenance free. The
sediment filter, when used, was typically
in the 30n to 50n range and required
replacement or washing approximately
two times a year. Occasionally, with a
water supply having abundant sediment,
the filter would need cleaning or
replacement as often as once a month.
Backwashing of GAG units was
recommended only if the hydraulic
capacity of the unit became noticeably
diminished, as indicated by a significant
drop in water pressure at the tap. Regular
once a week backwashing reportedly
caused a lower overall removal efficiency
and was not needed if a sediment filter
was in place. Field experience bears this
out for very few cases of loss of hydraulic
capacity have been reported.
The monitoring program since October
1984 consisted of:
1. an initial sampling and analysis
after 3 wk to confirm the success of
the installation,
2. sampling and analysis performance
check once every 6 mo for a period
of 2 yr.
In addition to these samplings, 11 of
the units were selected for a more
detailed analysis.
The sampling procedure used most
extensively involved the homeowner
collecting samples using 40-mL glass
septum, capped vials of the type used in
Volatile organic carbon (VOC) analysis.
Some of the first installed units were
sampled by the direct syringe method
with the use of prepared scintillation vials.
In both cases, the samples were normally
mailed directly to the Radon Research
Laboratory at the University of Maine
where they were analyzed by liquid
scintillation.
Performance of POE GAG Units
The performance of a GAG unit can be
accurately predicted if the following
information is known:
1. average water usage,
2. average influent radon
concentration,
3. GAG adsorption/decay constant
(Kss), and
4. volume of carbon used, with the
use of the equation:
Radon (out)/Radon (in) = exp(-
Kss*t)
where t is the empty bed contact
timo.
The theoretical and actual removal
performance for the POE GAG units are
given in Table 2. An average flow of 157
gal/d (21 ft3) and a Kss of 3.0/hr (Barneby
Cheney 299 GAG) were used to estimate
the expected performance. In actual field
operation, some units exceeded this
predicted performance; however, when all
field units are considered overall, removal
was something less than the theoretical
performance.
The actual removal percentages in
Table 2 are the average removal
percentages taken without regard to
specific knowledge of problem units.
Possibilities for errors in sampling point,
possible partial by-pass due to
equipment problems or possible
improper plumbing, etc., were not
investigated in every case where removal
was less than expected. Higher-than-
estimated water use could also have
affected the averages. We believe,
however that the effect of these
elements on the performance numbers is
very small. Unknown water quality factors
at specific sites are believed to be
responsible for the lower performance at
some sites.
Analysis of Selected GAG Units
Eleven individual sites were selected
for a more detailed analysis and
discussion. These sites have POE GAG
units with one or more of the following
characteristics: in service for a relatively
long period, significant number of
performance checks, a particular water
quality problem other than Rn, or a
progressive premature failure. Two of the
eleven GAG units are discussed in this
summary
Site No. 64
The POE GAG unit at site 64 was
routinely monitored during its 42-mo life,
which was intentionally terminated. (It
was replaced by an aeration system).
During its life, 10 samples were taken to
check performance. The average removal
was 96.3%, and the raw and treated
water Rn levels averaged 154,000 pCi/L
and 5,750 pCi/L, respectively. The last
two data sets for this unit indicated a
possible error in sampling or analysis.
For the treated water Rn, one set was
significantly higher and one significantly
lower. No investigation was made to
further define the possible problem.
Site No. 43
This site has had a GAG unit
operational for a period of 31 mo and has
been monitored 8 times. The site is
unique in that the amount of iron in the
raw water is significant and the iron
precipitates in the bed. On a site
inspection, the top of the GAG was found
to be beneath a 2.0-in. layer of bright
orange iron precipitate. Although a large
amount of precipitate was present and
the unit had never been backwashed, the
performance was documented to be
greater than 99.4%. In fact, this unit
consistently performed at a high level
and had not been hampered by the iron
level in the raw water. The Rn levels in
the raw and treated waters were 112,900
pCi/L and 638 pCi/L, respectively.
At the present time there are not
enough data to predict the long term life
of the Rn adsorption/decay steady state.
The report data indicate that each GAG
unit must be considered on a case-by-
case basis; few systems fail prematurely,
whereas others continue to show
theoretical removal efficiencies for
extended periods without any signs of
deterioration of the adsorption/decay
steady state. Other factors, such as Pb-
210 and its progeny buildup in the bed,
may dictate the service life of a GAG bed
in a state that regulates Pb-210.
Conclusions and
Recommendations
A large body of performance data for
over 100 GAG POE and small water
supply GAG systems was summarized to
document the effectiveness of these
systems to remove Rn in the field. Based
on the analyses of raw and treated water,
the following conclusions and
recommendations can be made:
Approximately 84% of the 121 POE
GAG systems in the field are still
achieving Rn removals of greater than
-------
Backwash
Fitting
Raw Water
10" Freeboard
Manual Control Valve
Treated Water
10" 0 Fiberglass Vessel
6" Support Gravel (0.25" 0)
Figure 1. POE GAC unit.
Table 2. Expected Versus Actual Removal Performance for POE GAC Units
Performance %
GAC
Model
GAC 10
GAC 17
GAC 30
Number of
Actual
Units
12
59
12
Flow
gpd
157
157
157
Empty Bed
Contact
Time, hr
1.14
1.94
3.43
Expected
96.7
99.7
99.99 +
Actual
90.7
92.5
98.6
95% (Oct. 1989). Approximately 113
units (94% of all 121 units) achieved
greater than 90% Rn reduction. Seven
POE GAC systems experienced a
premature failure believed to be related
to water quality. The problem of
premature failure is clearly associated
with particular regions within specific
states.
Some systems achieved removals
close to the theoretical steady state level
calculated with the use of a first-order
relation for the ratio of treated Rn to raw
water Rn. In general, however, the Kss
value in the field were slightly less than
that measured in the laboratory with a
different batch of GAC.
The ultimate bed life for a POE GAC
unit cannot be predicted based upon the
limited, long-term data collected to date.
For the 11 systems that were monitored
for 2 to 6 yr, there are no clear
indications of loss of efficiency over time.
The long-term data are limited to a few
units; however, it is clear that one unit
. S. GOVERNMENT PRINTING OFFICE: 1991/548-028/20156
-------
has performed at a high efficiency for
over 6 yr without signs of needing
replacement. Based on the 11 units, a
typical POE GAG unit may last a decade
giving removals of greater than 90%.
The levels of Pb-210 and its progeny in
the treated water produced by POE GAC
units should be documented for units that
have been operated for extended
periods.
The gamma exposure rate associated
with POE GAC units should be
documented more extensively. Actual
dosimetry measurements should be
made on occupants in households that
have these units.
The full report was submitted in
fulfillment of Contract No. 8C6155TTST
by Lowry Engineering, Inc., under
sponsorship of the U.S. Environmental
Protection Agency.
Jerry D. Lowry, Sylvia B. Lowry, and Julie K. Cline are with Lowry Engineering
Inc., Unity, ME 04988.
Kim R. Fox is the EPA Project Officer (see below).
The complete report, entitled "Radon Removal by POE GAC Systems: Design
Performance, and Cost," (Order No. PB91-125 6331AS- Cost-$1700
subject to change) 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:
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/600/S2-90/049
------- |