Measurement uncertainties and minimum detectable concentrations for the in situ NaI gamma spectroscopy systems used at the Fernald site.

Abstract

This report determines the uncertainties associated with measurements made by using the mobile gamma-ray spectrometers deployed at the U.S. Department of Energy's Fernald Closure Project to characterize soil contaminated with {sup 238}U, {sup 226}Ra, and {sup 232}Th. It also examines minimum detectable concentrations (MDCs) for these instruments. The spectrometers use sodium iodide (NaI) detectors and are mounted on a variety of platforms that allow access to all areas of the site, including deep excavations. They are utilized for surveying large areas to obtain distribution patterns for radionuclides in soil, determining whether activity concentrations exceed action levels for hot spots, and determining if the concentration of total uranium exceeds the allowable level for Fernald's on-site disposal facility. Soil cleanup levels at Fernald are 82 parts per million (ppm) for total uranium (27.3 pCi/g for {sup 238}U), 1.7 pCi/g for {sup 226}Ra, and 1.5 pCi/g for {sup 232}Th. The waste acceptance criterion (WAC) for total uranium for the disposal facility is 1030 ppm. Uncertainties associated with counting, efficiency calibration, the calibration pad and sources used, the vertical distribution of contaminants in soil, the use of moisture corrections, and the use of corrections to account for the loss of radon from soil are examined. (Loss of radon is an important process because measurement of {sup 226}Ra relies on emissions from progeny of {sup 226}Ra and because {sup 222}Rn is an intermediate, highly mobile decay product.) The importance of each source of uncertainty depends on the radionuclide of interest and level of contamination. The combined relative uncertainty (relative standard deviation) in measurements of dry-weight concentrations near three times the cleanup levels (the action levels for hot spots) is about 30% for 4-second measurements of {sup 238}U, 40% for {sup 226}Ra, and 20% for {sup 232}Th. (Measurement uncertainties for {sup 226}Ra are elevated because of the magnifying effect of the correction process used to account for the loss of {sup 222}Rn from soil.) For measurements of total uranium near the WAC level, the total relative uncertainty is about 20% for 4-second measurements. When only uncertainties due to counting errors are considered, a trigger level of 900 ppm can be used with 4-second measurements to determine, with a 95% level of confidence, if concentrations of total uranium in soil exceed the WAC level. The MDCs for 4-second measurements are well below three times the relevant cleanup levels for all three radionuclides considered.