and length of the field activity could be reduced
Table 6. Conversion coefficients for soil vapor
(mg/L) to soil matrix (mg/kg) concentrations
by as much as an order of magnitude.
for trichloroethylene.
When the quantitative performance of a soil-
vapor survey technique is assessed, special atten-
Percent
tion should be given to the collection and hand-
organic
Soil type
kg/L
carbon
Study
ling of both soil-vapor and matrix samples. It is
imperative that the dead volume of soil-vapor
55 soils
1.16
0.14
Rong (1997)
samplers be small, perhaps on the order of a few
CRREL siltysand
0.386
0.88
Hewitt (1998b)
CRREL sandysilt
1.36
0.10
milliliters. Although they were not presented
Wisconsin sand
1.09
0.17
here, several experimental findings indicated
Siltysand
0.806
<0.5*
This study
that it is important to create a small void from
* Estimate of organic carbon content based on soils pre-
which to remove the subsurface soil vapors. In
viously taken at this site from a depth below 30 cm.
principle, this void serves as a chamber, which,
with time, becomes filled with a vapor phase
except the samples were obtained both closer to
the surface (20 10 cm) and from a much greater
VOC concentration that is in quasi-equilibrium
with the in-situ soil matrix. Moreover, when com-
depth (greater than 30 m). Because these soils
paring soil-vapor VOC concentration to soil
were obtained from three different depths in the
matrix concentrations, it is important to use an
vadose zone, they have different levels of organic
in-field sample preparation and perhaps preser-
carbon associated with them, which most likely
vation method that limits VOC losses. This does
accounts for the discrepancies in the CO values
not necessarily preclude the use of the En Core
(Hewitt 1998b).
sampler, since it could be attached to a long rod to
These studies all support the concept that soil-
obtain a sample of soil matrix at the bottom of a
vapor VOC measurements can not only establish
narrow channel. Furthermore, since, currently,
spatial and temporal distributions, but, if done
there is no rigorous evaluation of how well sub-
properly, can also accurately predict the concen-
surface sample retrieval systems (e.g., geoprobes,
tration present in the soil matrix. To further
hollow stem augers, cone penetrometer, Veihm-
develop this relationship, more rigorous studies
eyer tubes, etc.) retain representative VOC con-
involving other soils and VOCs are necessary.
centrations in bulk samples, it is recommended
that these comparisons be initially made in the
SUMMARY
near-surface region.
This study has established a soil-vapor collec-
tion, handling, and analysis protocol that is not
LITERATURE CITED
only very precise but, when compared to collo-
Conant, B.H., R.W. Gillham, and C.A. Mendoza
cated soil matrix concentrations for the measure-
(1996) Vapor transport of trichloroethylene in the
ment of TCE, resulted in a linear and highly sig-
unsaturated zone: Field and numerical modeling
nificant correlation. Furthermore, the conversion
investigations. Water Resources Research, 32: 922.
coefficient established for the measurement of
Griffith, T.J., G.A. Robbins, and T.M. Spittler
TCE in these two media was consistent with
(1988) A new method for field analysis of soils
recent theoretical and empirical models. Clearly,
contaminated with aromatic hydrocarbon com-
the next logical step is to determine if this type of
pounds. In FOCUS Conference on Eastern Regional
relationship exists for other VOCs in other types
Water Issues, September, Stamford, Connecticut.
of soils. However, because there are two indepen-
National Water Well Association, p. 223248.
dent sources of information, it is very likely that
Hewitt, A.D. (1994) Comparison of methods for
soil-vapor surveys, accompanied by a sufficient
sampling vadose zone soils for the determination
number (5 to 20%) of confirmatory discrete soil
of trichloroethylene. Journal of the Association of
matrix analyses, will become an alternative
Analytical Chemistry, 77: 735737.
method for future site investigations for VOC con-
Hewitt, A.D. (1997) A tool for the collection and
tamination in the vadose zone.
storage of soil samples for volatile organic com-
The handling and analysis protocols presented
pound analysis. American Environmental Laboratory,
show that soil-vapor samples could either be
9: 1416.
measured on-site, making data acquisition
Hewitt, A.D. (1998a) Comparison of sample prep-
approach real time, or off-site, within 1 or 2 weeks.
aration methods for the analysis of volatile or-
In either case, if the bulk of site was characterized
ganic compounds in soil samples: Solvent extrac-
using a proper soil-vapor survey method, the cost
8
Previous Page
