was substantial and similar in magnitude to those
the rapid biotransformation of TNT to amino-
documented in an earlier study of TNT, DNT,
DNTs, and rapid leaching or chemi-sorption (bind-
and ammonium picrate (Jenkins et al. 1996).
ing) of the amino metabolites (Thorne and Leggett
Using the results from the on-site HMX method,
in press).
the standard deviation due to sampling error was
11 times that due to analysis when discrete core
samples were used for characterization. The use
CONCLUDING REMARKS
of on-site sample homogenization and composit-
ing of discrete soil samples and the use of 20-g
The ultimate goal of site characterization ac-
subsamples of soil for analysis were shown to
tivities is to provide sufficient information so that
minimize the problems caused by short-range
informed decisions can be made about the devel-
spatial heterogeneity.
opment of an optimal approach for remediation
when it is required. To accomplish this goal, po-
tentially contaminated sites are generally divided
Midscale spatial heterogeneity
The degree of midscale spatial heterogeneity
into small geographically defined units (grids),
of explosives contamination was assessed for the
and samples are collected and analyzed to char-
first time at CFB-Valcartier. The site was divided
acterize the concentrations of contaminants within
into sixteen 6- 6-m square grids, and each grid
these zones. The dimensions of these grids can
was further subdivided into four 3- 3-m sub-
range from tens of meters on a side to hundreds
grids. Analysis of replicate samples from nine
of meters on a side and often a single core sample
randomly selected subgrids indicated that we
is collected within the grid, divided into depth
were generally able to reproduce results within a
intervals, and the various depth related samples
factor of two, using a simple integrating method
analyzed at an off-site commercial laboratory.
that samples about 10% of the surface within the
Decisions regarding the need for cleanup are
subgrid.
made by comparing contaminant concentrations
The variability of the concentrations in the four
obtained from sample analysis to action levels
subgrids making up a geographically defined
determined by risk assessment.
grid area was also estimated and found to aver-
An unstated assumption of this approach is
age about 5 when concentrations were above 50
that the concentration of contaminants of interest
mg/kg. Even in the presence of this substantial
in the samples collected and analyzed adequately
spatial heterogeneity, we were able to produce
represents the average concentration of those con-
grid-composite samples from the four area inte-
taminants at the collected depths within grid
grated subgrid samples that were in good agree-
boundaries. Discussion with both the personnel
ment with the arithmetic mean of the four. The
conducting site characterization activities and
mean difference observed for the 10 grids where
those within the government that oversee these
this was evaluated was about 20% with values
activities indicate that some of the shortcomings
never differing by a factor greater than about 2.
of this approach are recognized, but that financial
considerations preclude the analysis of sufficient
numbers of samples to adequately address the
Accumulation of explosives residues
problem. The high cost of laboratory analyses, in
at an active antitank firing range
This study documented, apparently for the first
particular, is often quoted as an impediment to
time, the levels of explosives residues that accu-
the analysis of the required number of samples to
mulate in the soil at an active firing range. HMX
truly characterize the distribution of contaminants.
accumulated in the surface soils to a much greater
Explosives represent a fairly unusual set of en-
degree than did TNT, the other major component
vironmental contaminants. The most commonly
in the melt-cast explosives used in the 66-mm
encountered secondary explosives are solids at
M72 rocket fired at CFB-Valcartier. Concentra-
environmental temperatures and were generally
tions of HMX as high as 2000 mg/kg were found
released into the environment as particulates. Thus
in surface soil near to the targets. This concentra-
the bulk of contamination generally resides in
tion appears to be too high to be due to residues
near surface soils. These explosives have low
from detonations and may have resulted from
vapor pressures and hence no special precautions
rupture of unexploded rockets and ejection of
are needed during sample collection to mitigate
particles of the explosives. The much lower accu-
vapor losses. They are relatively polar, particu-
mulation of TNT in these soils is probably due to
larly for neutral organic compounds, and do not
25
Previous Page
