that the composition of the fortifying solution is
transfer or expose soil subsamples to the atmo-
paramount. With respect to the potential for vol-
sphere and by desiccating the soil prior to spiking,
ume production of subsamples for distribution,
thereby minimizing biological activity.
these observations are reassuring.
The availability of reliable QA/QC soil sub-
In the absence of significant differences in batch
samples for the analysis of VOCs will allow us to
means, all values for a given analyte in a given soil
judge laboratory performance in a matrix that is
were combined to yield an overall mean, standard
consistent with vadose-zone samples. Moreover,
deviation, percent relative standard deviation and
of perhaps even greater importance would be the
95% confidence limits around the mean (Table 5).
use of such materials for comparing different
With the exception of TDCE, which is the most
methods of analysis. Currently site assessments
volatile of the analytes tested, the RSDs and 95%
are carried out by a variety of analytical sample
confidence limits are excellent, despite the residual
preparation and collection methods, including
systematic error known to remain. Although the
passive and active soil vapor, thermal desorption,
larger uncertainties associated with TDCE were
solvent extraction and water-based static and dy-
not unexpected, the results are still quite good for
namic extraction. Furthermore, several detection
the analysis of VOCs in a porous matrix.
systems are available for quantification. The op-
portunity to use performance evaluation soil sub-
samples for correlations between these analytical
DISCUSSION
methods would broaden our understanding of
This study shows that laboratories following the
their capabilities and applicability.
procedures outlined in Method 8240 (U.S. Environ-
mental Protection Agency 1986) are capable of reli-
CONCLUSION
ably determining VOC concentrations in either
spiked MeOH or soil materials that have been de-
The results from this round-robin study show
signed for performance evaluation. The compari-
that vapor fortification treatment and glass am-
son of the pooled RSDs for the analytes common to
poule confinement is a relatively simple, cost-ef-
both the supplied QA MeOH solvent-based stan-
fective means to produce reliable, precise and re-
dard and the soil subsamples (Ben, Tol and TCE)
producible soil subsamples spiked with VOCs
resulted in values of 7.8% and 10.4%, respectively.
with good concentration stability and ease of dis-
The differences between these values are minor,
tribution for use as QA and QC materials. These
considering that the QA samples were distributed
features also make vapor-fortified soils suitable
with specified analyte concentrations, whereas an-
for assessing the precision and accuracy of proce-
alysts were not informed of the concentrations
dures for sample preparation (extraction) and
present in the soil subsamples, and the latter re-
analysis (determination).
quired an extraction step prior to analysis (Hewitt
et al. 1994).
LITERATURE CITED
Unfortunately this degree of determinative un-
Dixon, W.J. (1953) Processing data for outliers. Bio-
certainty cannot often be extended to cover esti-
metrics, March, 7489.
mates of VOC concentrations in vadose-zone soil
subsamples. This is because vadose-zone soils are
Hewitt, A.D., P.H. Miyares, D.C. Leggett and T.F.
Jenkins (1992) Comparison of analytical methods
usually transferred from the field to a laboratory as
for determination of volatile organic compounds
bulk samples. This requires that the sample matrix
in soils. Environmental Science and Technology, 26:
be handled at least twice prior to analysis: once
19321938.
when the bulk sample is placed in a shipping or
Hewitt, A.D. (1993) Feasibility study of preparing
storage bottle and again when a subsample is re-
performance evaluation soils for analyzing vola-
moved for analysis. During these handling opera-
tile organic compounds. USA Cold Regions Re-
tions the VOC concentrations can become signifi-
search and Engineering Laboratory, Special Re-
cantly reduced due to volatilization losses (Urban
port 93-5.
et al. 1989, Siegrist and Jenssen 1990, Siegrist and
Hewitt, A.D. (1994a) Losses of trichloroethylene
van Ee 1993, Hewitt 1994a,c). Also, while being
held at 4C, analyte biodegradation has been
from soil during sample collection, storage and
laboratory handling. USA Cold Regions Research
shown to cause rapid reductions in the concentra-
and Engineering Laboratory, Special Report 94-8.
tion of some VOCs (Hewitt 1994b). These losses
Hewitt, A.D. (1994b) Concentration stability of
were avoided here by eliminating the need to
7
Previous Page
