United States
Environmental Protection
Agency
Risk Reduction
Engineering Laboratory
Cincinnati, OH 45268
Research and Development
EPA/600/SR-94/142 September 1994
EPA Project Summary
Field Investigation of
Effectiveness of Soil Vapor
Extraction Technology
Michael H. Corbin, Nancy A. Metzer, and Michael F. Kress
A research project was undertaken
to study the effectiveness of soil vapor
extraction (SVE), an emerging technol-
ogy for remediation of soils contami-
nated with volatile organic compounds
(VOCs). As part of the project, two soil
vapor extraction systems, Site D and
Site G at the Twin Cities Army Ammu-
nition Plant, were selected for evalua-
tion.
Site information regarding residual
soil concentrations before and after
treatment were gathered to compare
residual levels of volatile organics be-
fore and after treatment. Operational
data are analyzed to present the per-
formance of the systems and the pro-
gression of treatment with time. Capital
as well as operating and maintenance
costs are presented.
Results of the evaluation indicate that
soil vapor extraction has been effec-
tive in reducing the residual concen-
trations, generally by several orders of
magnitude. In most cases, residual con-
centrations were non-detectable. The
variability of the concentrations, when
detectable, also decreased. Samples
taken in silty clays and waste materials
showed the highest residual concen-
trations. Operational data indicated that
mass removal rates decreased rapidly
during the first few days of treatment
and, within a few months, reached a
level one tenth of the initial rates.
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
The purpose of this project was to char-
acterize and assess the effectiveness of
SVE in reducing the concentration of VOC
soil contamination.
SVE, an emerging technology for
rennedatiing soils contaminated with VOCs,
is done by mechanically drawing air
through the contaminated soils in the va-
dosse (unsaturated) zone. An array of sub-
surface vents is installed in the
contaminated area. A vacuum pump is
then manifolded to the vents to induce air
flow. The VOC-laden air is drawn from the
soils to the vents, through the manifold
and pump, and is either discharged to the
atmosphere or treated before discharge,
depending on specific site considerations.
Initially, attempts at determining the ef-
fectiveness of the technology used a mass
balance approach. This involved a
preremediation site characterization to de-
termine the quantity (in pounds) of con-
taminants in the soils, measurement of
the total mass of contaminants removed
during remediation, and a post-remediation
site characterization to determine the quan-
tity of residual contaminants remaining in
the soils after treatment. Because of the
high cost of soil sampling and analysis
and the heterogeneous nature of soil con-
tamination, the mass balance cannot be
done with much precision unless exten-
sive resources are expended. This has
not proven to be a reliable, cost-effective
means of assessing the system's effec-
tiveness.
Printed on Recycled Paper
-------
Subsequent efforts have focused on
determining the residual concentrations of
contaminants remaining in the soil after
the effluent air contaminant concentrations
drop to a very low level when compared
with the initial concentrations. The premise
behind this approach is that if the soils
have low or nondetectable levels of con-
taminants, they can be considered clean.
Many Superfund site remediation plans
specify soil cleanup concentration levels
using risk-based analysis or regulatory
standards. To date, no full-scale SVE sys-
tems documented in the literature have
reached a final site cleanup stage based
on stipulated soil cleanup levels and post-
treatment sampling.
The following approach was identified
to meet the project objectives. A literature
search and a review of internal projects
identified sites employing SVE technol-
ogy. Site evaluations were based on the
duration of the SVE system operation, the
quality of site characterization, the amount
and availability of operational data, the
willingness of the site owner/operators to
participate in the project, and the site char-
acteristics, such as soil type, contaminant
type, and size. After site selection, site
information regarding soil VOC concen-
trations before treatment was obtained.
Operational data from the system owners/
operators were obtained and evaluated
with respect to system performance. A
soil sampling program to evaluate the re-
sidual contamination levels remaining in
the treated soils was performed. The ini-
tial and current contamination levels in
terms of magnitude and distribution of con-
tamination were compared. The two SVE
systems (Site D and Site G) at the Twin
Cities Army Ammunition Plant in New
Brighton, MN, were selected for this
project.
Site D is the location of former leach-
ing/burn pits where solvents, explosive
primer wastes, and other combustibles
were disposed of through open burning.
The surficial geology of the site consists
of the Arsenal sand, stained sediments,
and residues from burning activities. The
Arsenal sand consists of brown-gray, fine-
to-cparse sands and gravels. The stained
sediments and residues consist of dark
gray-to-black, fine-to-coarse sands and
silts. The Arsenal sand extends below the
site to a depth of approximately 120 ft.
The contamination observed at Site D con-
sisted primarily of VOCs, ranging from
nondetectable (ND) to 7,000 mg/kg.
Trichloroethylene (TCE) and 1,1,1-
trichloroethane (TCA) comprised 71% and
20%, respectively, of the total VOCs and,
as such, were the primary contaminants.
The full-scale SVE system was installed
at Site D in January 1986 following the
placement of a soil cover. The system
consists of 39 air extraction vents, placed
from 34 to 54 ft below ground surface
(bgs). The system layout is shown in Fig-
ure 1.
Site G was an active landfill from the
1940s to the 1970s. The landfill contains
heterogeneous materials, ranging from
sands to clays to waste materials. Under-
lying the landfill materials is a silty clay
(Twin Cities Formation till). There were
some indications that the silty clay is con-
tinuous throughout the site. Beneath the
silty clay are the fine-to-medium-grained
Arsenal and Hillside sands. These sands
were encountered to a depth of 135.5 ft
bgs. The contamination observed at Site
G consisted primarily of VOCs with total
VOC concentrations ranging from ND to
960 mg/kg. Most of the sample with high
total VOC concentrations were taken from
the waste material. TCE comprised 16%
to 88% of the total VOC concentrations. A
full-scale SVE system was installed at Site
G in February 1986 following the place-
ment of a soil cover. The system con-
sisted of 89 air extraction vents, installed
at depths ranging from 32 to 54 ft bgs. An
activated carbon vapor control system was
installed for emissions treatment. The sys-
tem layout is shown in Figure 2.
As part of this project, seven soil borings
were installed at Site D and seven more
at Site G in May 1989. Site D soil borings
were installed to depths ranging from 30
to 35 ft; samples were collected at ap-
proximately 10-ft intervals. Site G soil
borings were installed to depths ranging
from 15 to 60 ft; samples were collected
at 15-ft intervals. The samples were ana-
lyzed for TCE and TCA (EPA Method
5030/8010), full VOC analysis (EPA
Method 8240), and moisture content
(ASTM Method D2216).
Results
Soils
In general, the concentrations at Site D
have been reduced by four or five orders
of magnitude to levels that are
nondetectable or in the very low parts per
billion (ppb). Only 5 of 21 samples had
detectable VOC concentrations; the high-
est was 29 ppb TCE. TCA was detected
in only one sample at a concentration of
0.8 ppb. In comparison, TCE concentra-
tions before treatment were reported as
high as 7,000 parts per million (ppm).
The moisture content of the soils at Site
D ranged from 1.7% to 14% and aver-
aged approximately 6.1%. This may be
compared with pretreatment levels of 3.3%
and 4.6% in two samples collected and
analyzed during the initial site investiga-
tion. These results do not indicate a sig-
nificant change in the soil moisture content
over the treatment period. This is interest-
ing: considering the large volume of air
that passed through these soils, one would
anticipate a significant decrease in the
soil moisture. Because the site is capped,
however, the air was forced to flow through
a large volume of uncapped soil before
reaching the treatment area, and, there-
fore, the total volume of soil was too large
to be significantly dried. Finally, the mois-
ture content results are consistent with
the previous description of well-drained
soils.
Split-spoon samples were obtained from
each soil boring installed at Site G. The
number of samples collected from each
boring varied because of the subsurface
conditions encountered as the borings
were advanced, e.g., a tar-like layer was
encountered at approximately 25 to 30 ft
in two borings. When this layer was en-
countered, it was not possible to advance
the auger through the layer.
The VOC concentrations at Site G
showed a higher degree of variability than
those at Site D. Concentrations of post-
treatment total VOCs ranged from
nondetectable to 0.420 ppm. TCE and
TCA were detected at maximum concen-
trations of 0.420 ppm and 0.200 ppm,
respectively. In comparison, the maximum
concentrations of TCE and TCA in the
pretreatment samples were 400 and 100
ppm, respectively, roughly three orders of
magnitude greater than after treatment.
Of the 21 samples, TCE or TCA was
detected in 15 samples; however, only 6
of these 15 samples showed concentra-
tions above the detection limit (the other
results were estimated values). Of the six
samples showing concentrations above the
detection limit, all were composed of a
waste material or had components of silt
or clay. The other samples were generally
composed of sandy soils. This would be
expected because the volatile compounds
should adsorb more strongly to silts, clays,
and waste materials and would therefore
be more difficult to remediate. Also, air
flow through these materials would be
much less than through sandy soils. There-
fore, the VOCs would be removed more
easily from the sand.
Operations
The mass removal rate (Ib/day) for Site
D is plotted against the days of operation
in Figure 3. The removal rate at the be-
ginning of operations was approximately
-------
, I i
Building \
I2760CFM ' **-"—I
Total
_!!_!_ _i
I ~~
j i
r r~
Legend
• Air Extraction Vent
Scale in Feet
Figure 1. Full-scale SVE system layout, Site D, Twin Cities Amy Ammunition Plant.
3
-------
Extent of Clay Cap
Legend
• Air Extraction Vent
SO
Scale in Feet
100
Figure 2. Full-scale SVE system layout, Site G, Twin Cities Army Ammunition Plant.
-------
1,200 Ib/day. It dropped to several hun-
dred pounds per day within 1 wk and
continued to decline with time. As of June
1990, a cumulative total of 108,460 Ib of
VOCs had been removed from the soils at
Site D.
The operational data plotted in Figure 3
suggest a logarithmic decay in the re-
moval rate. Two curves that approximate
the decay are shown on the graph. The
first curve, y = 895.7 - 305.64* log (x), is a
logarithmic function. The second curve,
y = (1,000 + 3 x)/(1 + 0.09 x), is a hyper-
bolic function. Both curves simulate the
high initial removals, the rapid decrease,
and the tailing in later treatment. The hy-
perbola indicates a long period of later
treatment characterized by low removals,
while the logarithmic decay indicates "zero"
removal at approximately 900 days of op-
eration.
The estimated installation cost for Site
D is $257,000. This cost does not include
design or construction management costs.
The operations and maintenance costs of
$194,000 are for the total operations pe-
riod from February 1986 to May 1990.
The labor costs were estimated at 70% of
the Site G system costs, since specific
information for the Site D system labor
was not available. The estimated present
worth was calculated by using a 4% infla-
tion rate compounded annually and an
additional 15% design/construction man-
agement cost on the capital costs. The
estimated volume of soil treated, assum-
ing a 17 ft radius of influence for each
vent, is 35,000 yd3. By estimating a present
worth of $573,000 and the volume of soil
treated, the treatment cost per cubic yard
is $17.
The mass removal rate (Ib/day) for Site
G is plotted against time (days of opera-
tion) in Figure 4. The maximum daily mass
removal rate, 5,015 Ib/day, was encoun-
tered on the second day of operations.
Within 2 wk of operation, the mass re-
moval rate was below 1,000 Ib/day. It must
be noted that the mass removal rates
were intentionally reduced during the ini-
tial stages of operations. The total VOC
mass removal rate continued to decrease,
reached a removal rate of approximately
200 Ib/day after 76 days of operation, and
dropped below 50 Ib/day after 195 days of
operation. The system, as of 21 May 1990,
has removed 97,727 Ib of VOCs. The
mass removal rate for the first 5 mo of
1990 ranged from 1 to 10 Ib/day.
As with Site D, hyperbolic and logarith-
mic functions were used to approximate
the operational data. The equations of
these lines are y = (1,500 - 1.1 x)/(1 +
0.08 x) and y = 2357.4 - 922.53* log (x)
for the hyperbolic and logarithmic func-
tions, respectively. Although these func-
tions can represent the general trends
seen during the SVE system operation,
these functions skew from the actual data
because they are unable to simulate the
sharp decline of mass removal rates dur-
ing the initial operation while still simulat-
ing the asymptotic nature of the curve
during later stages of operations. Addi-
tionally, the operational data vary unpre-
dictably because of periods of inoperation
and initial manipulation of the operating
parameters to control the mass removal
rate.
After September 1986, the Site G sys-
tem was operated with a vapor treatment
system. The vapor treatment system con-
sisted of two beds of carbon, initially con-
taining 6,500 to 9,600 Ib of carbon. The
length of time that the system was shut
down for changeouts varied greatly due to
logistical factors. The carbon treatment
system was deactivated in April 1989 as
the mass removal rate dropped below a
level where VOC emissions would pose
any health threat. Through April 1989, a
total of 248,000 Ib of carbon was spent at
Site G.
The estimated installation cost for Site
G is $257,000. This cost does not include
the design or construction management
of the system. The capital cost of imple-
menting the vapor control system was
$213,000, bringing the total capital costs
to $467,000. Operation and maintenance
costs include labor, power, system moni-
toring, and carbon changeouts (removing
and regenerating the spent carbon). These
costs totalled $500,000 for 4 yr of opera-
tions (February 1986 through June 1990).
The present worth was calculated based
on 4% inflation rate, compounded annu-
ally, and 15% design and construction
costs for the capital costs. Therefore, the
total cost to apply the SVE technology to
remediate Site G in 1990 dollars would be
$1,121,000. The estimated volume of soil
treated, assuming a 17-ft radius of influ-
ence for each vent, is 91,000 yd3. Treat-
ment costs may be expressed as dollars
per cubic yard of soil treated to date. For
Site G, the treatment costs are $13/yd3 of
soil treated.
Conclusions
The following conclusions can be made
from the results of the technical evalua-
tion and the soil sampling and operational
information collected for the sites.
1. SVE treatment at both Site D and Site
G effected significant treatment.
Comparison of pretreatment and post-
treatment data shows that TCE and
TCA concentrations have decreased
by several orders of magnitude at
both sites. The residual con-
centrations of TCE and TCA,
however, varied at each site due to
varying soil conditions.
2. At Site D, which consisted of a
uniform sand, most of the soils have
nondetectable TCE and TCA
concentrations (16 of 21 samples),
with a highest detected con-
centration of 0.029 mg/kg TCE.
Before treatment, the highest
concentration of TCE was 7,000 mg/
kg. Similar results were noted for
TCA, where the highest post-
treatment con-centration was 0.0008
mg/kg (an estimated value), and the
highest pretreatment concentration
was 1,000 mg/kg.
3. At Site G, which consisted of a more
variable strata including sands,
clays, and waste material, the
residual concentrations of TCE and
TCA were more variable. TCE and
TCA concentrations were below the
detection limit in 15 of 21 samples,
with maximum concentrations of
0.420 mg/kg and 0.200 mg/kg,
respectively. Before treatment, the
highest detected concentration of
. TCE was 400 mg/kg, and the
highest concentration of TCA was
100 mg/kg. All of the samples
showing maximum concentrations
1 (both pretreatment and post-
treatment) were taken from waste
material. The higher residual TCE
and TCA concentrations found in
the waste material and clays at Site
G may indicate the following:
The less permeable materials
or materials with a higher
organic content tend to absorb
or retain the contaminants to a
greater degree than do the
sands.
The air flow through the sands
is greater; therefore, the
contaminants are removed from
them more readily.
Site G is still operating and
residual levels may decrease
further with time.
4. The mass removal rate for VOCs
varies significantly over time. Initially,
the mass removal rate is very high,
but within days it decreases rapidly
and shortly, within a few months,
reaches levels that are one-tenth
-------
I
o
i
1000 •+
eoo -
goo 4 y=895.7-305.64*log(x)
(logarithmic function)
y=(1000+3x)/(1+0.09x)
(hyperbolic function)
NOTE: Y axis truncated to pro-
vide better resolution.
Days of Operation
Figure 3. Mass removal rate versus time. Site D, Twin Cities Army Ammunition Plant.
I
cr
2SOO -f-
2000 -
1500 -
1000
GOO - -
y=(1500-1.1x)/(1+.08x)
(hyperbolic function)
y=2357.4-922.53*log(x)
(logarithmic function)
NOTE: Y axis truncated to pro-
vide better resolution.
EOO
Days of Operation
1000
Figure 4. Mass removal rate versus time, Site G, Twin Cities Army Ammunition Plant.
-------
5.
those of the initial rates. This has
important implications for the design
of air emissions treatment units for 6.
SVE systems. An emissions
treatment unit sized for the initial
mass removal rates would be
completely oversized for the majority
of the systems' operational lifetime,
whereas a unit sized for the later low
removal rates could not handle the
initial removals.
The Site D system removed a total 7.
of 108,460 Ib of solvents between
January 1986 and May 1990 at an
estimated present worth total cost of
$573,000, or $17/yd3 soil ($5.28/lb
VOC) treated, for costs incurred as
of May 1990. Air emission controls
were not required for the Site D
system.
The Site G system removed a total
of 98,727 Ib of solvents between
February 1986 and May 1990. The
Site G soils have been treated at an
estimated present worth total cost of
$1,121,000 or $13/yd3 soil ($11.35/lb
VOC) treated, for costs incurred as
of May 1990. Air emission controls
were implemented at Site G.
Treatment costs for other sites will
depend on the following:
• site size and area! extent;
regulatory requirements for
: approvals, design, permitting and
operations;
• air emission controls;
• site and chemical specific
conditions;
• site cleanup criteria.
Treatment costs for other sites will likely
b& higher because of stricter regulatory
requirements, more detailed design re-
quirements, and more emissions monitor-
ing requirements. Therefore, treatment
costs for other sites should be evaluated
on a site-by-site basis.
The full report was submitted in
fulfillment of Contract No 68-03-3450 by
Roy F. Weston Inc under the sponsorship
of the U.S. Environmental Protection
Agency.
-------
M.H. Corbin, N.A. Metzer, and M.F. Kress are with Roy F. Weston Inc,
Westchester, PA 19380-1499
Janet Houthoofd is the EPA Project Officer (see below).
The complete report, entitled "Field Investigation of Effectiveness of Soil Vapor
Extraction Technology," (OrderNo. PB94-205531; Cost: $27.00, 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
Official Business
Penalty for Private Use
$300
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
EPA/600/SR-94/142
------- |