Collection Methods and Laboratory Processing of Samples from Donnelly Training Area Firing Points, Alaska, 2003

Abstract : At firing points for 105-mm howitzers, 2,4-DNT is detectable in the surface soils. 2,4-DNT is listed as a hazardous substance by the EPA and several states, including Alaska. Sample collection methods and laboratory subsampling procedures were developed to estimate the mean concentration of 2,4-DNT at a sparsely vegetated firing point. Collection of replicate 50-increment samples, where the <2-mm fraction was approximately 3 kg, was found to be adequate to estimate a statistically valid upper confidence limit of the mean concentration of 2,4-DNT from a 10,800-m2 area. The 95% upper confidence limit was 0.7 μg/g for multi-increment samples collected by five different samplers. In contrast, collection of replicate 50-increment samples from heavily vegetated firing points did not provide normally distributed estimates of 2,4-DNT concentrations, indicating that more increments and more mass are needed per sample. Sample corers that yield uniform sampling depths of vegetated surfaces may also improve precision of the field samples. Accurate estimation of 2,4-DNT in the multi-increment samples required that the entire sample be extracted with solvent or the entire sample be subjected to grinding on a ring mill. Size fractionation studies revealed that most of the 2,4-DNT in the firing range soils was in the 0.595- to 2-mm size range, although the bulk of the soil was less than 0.595 mm prior to grinding. The 2,4-DNT appears to be in particulate form, most likely within fibers of the nitrocellulosebased propellant. Grinding for five minutes was needed to pulverize the propellant fibers sufficiently so that analytical subsamples could be obtained in a reproducible manner. We have adopted the practice of grinding firing point soils for five one-minute intervals, with time for heat dissipation between grinds, prior to obtaining replicate 10-g subsamples.