Assessment of Sampling Error Associated
with Collection and Analysis of
Soil Samples at Explosives-Contaminated Sites
THOMAS F. JENKINS, CLARENCE L. GRANT, GURDARSHAN S. BRAR,
PHILIP G. THORNE, THOMAS A. RANNEY AND PATRICIA W. SCHUMACHER
in press). Inexpensive on-site analysis methods
INTRODUCTION
for the most common explosives in munitions-
contaminated soils have been developed and are
Background
Determining the distribution of contaminants
now in common use. These procedures appear to
at hazardous waste sites is a fundamental prob-
be sufficiently accurate and precise to enable their
lem facing site investigators. In general, distribu-
use in mapping locations of contamination and, if
tions are very site-specific, depending on a num-
a sufficient number of samples are analyzed, in
ber of variables, including how the site was
providing estimates of spatial contaminant het-
contaminated, the physical and chemical proper-
erogeneity. With these field methods, sequential
ties of the contaminants involved, soil type, and
modifications in sampling plans are feasible be-
the geology and hydrogeology of the site. Lack-
cause data become available while sampling is
ing distribution information, it is impossible to
in progress.
devise an optimal sampling strategy.
On-site analytical methods are sometimes criti-
Accurate chemical characterization of a haz-
cized as having inadequate precision, accuracy
ardous waste site requires a well-designed sam-
and specificity. With respect to specificity, we agree
pling plan. After defining the area of interest (tar-
that the QA/QC plan must include laboratory-
get populations), which might be an entire site or
based confirmatory measurements on selected
several defined areas within a site, workers col-
samples. Similarly, accuracy should be verified
lect samples according to one of several possible
against reference methods for an appropriate num-
schemes. In the absence of reliable historical in-
ber of samples. The precision issue, however, is a
formation, it is difficult to choose among judg-
different matter. Historically, the precision of meth-
mental, random, systematic, stratified, or some
ods used in hazardous waste characterization has
combination of these sampling plans. Many refer-
received an inordinate amount of attention com-
ences recommend a preliminary study before de-
pared to sampling error. Contaminated soils are
vising a sampling plan (Gilbert 1987, van Ee et al.
often extremely heterogeneous, which causes the
1990, Huesemann 1994, Keith et al. 1995, Will-
major error source to be sampling and sub-
iams 1996).
sampling. No amount of improvement in analyti-
Until recently, most studies of hazardous waste
cal precision can significantly reduce total mea-
sites have relied on shipping samples to off-site
surement error when the analytical error is a minor
laboratories for analysis. Besides the high cost
contributor to the total. Williams (1996) noted that
and potential for sample contamination or degra-
the newly released U.S. EPA DQO guidelines fo-
dation of labile analytes, this arrangement does
cus on the uncertainty of a specific decision rather
not lend itself to the timely decisions that are
than the individual parameters that contribute to
necessary in a step-wise plan. Recently, this prob-
the overall uncertainty. This is an encouraging
lem has been addressed with the development
change.
and promotion of field analytical methods (Triegel
A sampling plan can only be optimized after
1988, Jenkins and Walsh 1992, EPA 1993, Triegel et
the process of obtaining representative samples
al. 1994, Keith et al. 1995, Williams 1996, Barnard,
has been adequately addressed. Numerous varia-
Previous Page
