In front of the target in this range, 100 discrete minigrid samples and 4 composite
samples were collected from a 10 x 10-m grid located in B1. Analytical results for the
four replicate 30-increment composite samples are shown in Table IX. Results are
shown for the individual subsamples for each of the four composite samples along
with the mean concentrations for each replicate sample and the grand mean for the
10-m x 10-m grid. Six explosives-related compounds were detected in all of these
samples with TNT being present at a factor of 11 or greater than any of the other
compounds. The other compounds detected are either manufacturing impurities in
military grade TNT (2,4-DNT) or environmental transformation products of TNT
(1,3,5-TNB, 1,3-DNA, 2-ADNT, 4-ADNT).
The grand mean concentration of TNT was 10.7 mg/kg in these four replicate samples
and the relative standard deviation was 5.55%. The mean concentrations for the other
analytes detected in these composite samples (in the order of decreasing
concentrations) were 2-ADNT at 1.19 mg/kg, 4-ADNT at 0.776 mg/kg, 3,5DNA at
0.263mg/kg, 1,3,5-TNB at 0.107 mg/kg, and 2,4-DNT at 0.098 mg/kg. The relative
standard deviations for these compounds ranged from 5.83% to 9.20%. Thus very
repeatable sampling was achieved for this 10 x 10-m grid area using 30-increment
composite samples, even with respect to the minor components present at low
concentrations. Subsampling error for each of the four replicates was estimated by
pooling the relative standard deviations for the six compounds. These pooled %
RSDs ranged from 2.95 to 5.93% (Table IX) indicating that the method used to
process these samples was effective at minimizing subsampling error for these large
composite samples weighing about a kilogram.
Analytical results for the 100 discrete minigrid samples are presented in Table X.
The same six compounds detected in the 30-increment composite samples from this
area were also detected in almost all of these discrete samples, although the
concentrations varied by as much as three orders of magnitude. RDX and HMX were
also detected in 19 and 4 of these discrete samples, respectively, but were not detected
in any of the four 30-increment composite samples.
TNT concentrations varied from 0.381 to 289 mg/kg in these discrete samples, a
range of nearly three orders of magnitude, with a mean value of 16.2 mg/kg. Clearly,
use of a single discrete or even several discrete samples to estimate the mean
concentrations in this 10 x 10-m area would be prone to large sampling errors. The
highest RDX concentration was 35.4 mg/kg in minigrid # 42, but the TNT in this
sample was only 0.657 mg/kg indicating that the source of the RDX was probably not
the same source that led to widespread TNT concentrations within the overall 10 x
10-m grid. The source of the RDX is uncertain, but could been C4 used as a donor
charge to detonate duded bombs on this range. The fact that none of the other
minigrids surrounding minigrid # 42 had RDX concentrations in excess of 0.179
mg/kg indicates that a small piece of RDX-based explosive may have been present in
the discrete sample collected from minigrid #42.
The distribution of TNT values for these 100 minigrid samples is shown in Figure 19
as a histogram with a bin size of 5 mg/kg. Clearly, this distribution is non-Gaussian
as has been found elsewhere for energetic compounds at other types of training ranges
[30-31]. 68 of the TNT concentrations for discrete samples had concentrations that
were less than the mean of 16.2 mg/kg. The concentration of TNT versus position
18
DRDC Valcartier TR 2004-204
Previous Page
