agents can be stored at room temperature. The
Recommended quality assurance/
EnSys RISc colorimetric kits have shelf lives of at
quality control
The recommended quality assurance/quality
least 2 months and up to 1 or 2 years. Before or-
control (QA/QC) procedures vary considerably
dering test kits, it is important to ensure that they
with the screening procedure. Some test methods
will be used before the expiration date.
do not specify QA/QC procedures and leave to
the investigator the determination of the numbers
Skill level
of blanks, duplicates, replicates, and standards that
The skill level necessary or required to run these
are run. During field application of these meth-
tests varies from low to moderate, requiring a few
ods, it is common to send at least 10 to 20% of the
hours to a day of training. The manufacturers of
positive samples to an off-site laboratory for analy-
the kits generally provide on-site training. A free
sis by EPA Method 8330, and a smaller fraction of
training videotape on the CRREL TNT and RDX
the nondetect samples also may be verified. In
procedures (also useful for the EnSys RISc colori-
some cases, field methods are used to identify
metric kits) is available by submitting a written
samples containing explosives residues. Samples
request to Commander, U.S. Army Environmen-
containing explosives are sent for on-site analy-
tal Center, Attn: SFIM-AEC-ETT/Martin H. Stutz,
sis. In any case, the QC samples recommended by
Aberdeen Proving Ground, MD 21010. Training
the method developer should be used.
videotapes are also available from some kit
While it is essential to ensure that field meth-
suppliers.
ods perform as intended, laboratory-type QC
requirements may be inappropriate for on-site
Cost
analytical methods. Because site characterization
As shown in Table 3, routine sample costs vary
efforts may be cost constrained, excess QC samples
by method. The per-sample cost is affected by con-
reduce the number of field samples that can be
sumable items and instrument costs to run the
analyzed. Because sampling error (variability) is
method. In figuring costs per sample, it is impor-
typically much greater than analytical error
tant to include the costs of reruns for out-of-range
(Jenkins et al. 1996a, b), especially for explosives
analyses. With the EnSys RISc colorimetric TNT
residues, overall error is more effectively reduced
kit, the color-developed extract may be simply
by increasing the number of fields as opposed to
diluted and reread with the spectrometer. With all
the number of QC samples. Good sample prepa-
other methods, the original soil extract needs to
ration procedures and correlation of the field meth-
be reanalyzed, which in the case of immunoassay
ods with the laboratory HPLC method over the
procedures requires the use of another kit. Colori-
concentration range of interest should be the pri-
metric methods typically have sufficient extra re-
mary performance criteria. Documentation of pro-
agents to rerun samples with no increase in cost.
cedures and results must be emphasized.
It should be noted that the per-sample costs do
During the initial evaluation of on-site and off-
not include labor hours.
site analytical methods, it may be desirable to ana-
lyze a variety of QC samples to determine sources
Comparisons to laboratory method,
of error. The methods can then be modified to
SW-846 Method 8330
minimize error as efficiently as practical. This may
The objectives of the study or investigation, the
involve collection and analysis of composite ver-
site-specific contaminants of concern, the concen-
sus grab samples, duplicates, replicates, splits of
tration ranges encountered or expected, and their
samples, splits of extracts, etc. For more complete
relative concentration ratios affect the selection of
information on the types and uses of various QC
a particular on-site method. The accuracy of an
samples, see A Rationale for the Assessment of Er-
on-site method is another selection criteria, but
rors in the Sampling of Soils (EPA 1990).
care must be used in interpreting accuracy results
from comparisons between reference analytical
methods and on-site methods.
Storage conditions and shelf life
Storage conditions and shelf life of immunoas-
Colorimetric methods actually measure groups
say kits are more critical than colorimetric meth-
of compounds (i.e., nitroaromatics or nitramines),
ods. The reagents for some immunoassay kits
and immunoassay methods are more compound
should be refrigerated but not frozen or exposed
specific. Therefore the reported accuracy of a
to high temperatures. Their shelf life can vary from
method may depend on the mix of explosives in
3 months to more than 1 year. Colorimetric re-
the soil and the reference method data used for
20
Previous Page
