Performance of a Tomographic-Gamma-Scanner on RFETS Wastes Containing macroscopicLumps of Plutonium

摘要

This document describes various sources of tomographic gamma scanning errors andquantifies the random error and bias error for Rocky Flats Environment Technology Site (lWETS) molten salt extraction (MSE) residues measured in cans on the skid-mounted tomographic gamma scanner (TGS). The random error was quantified using the scatter of replicate measurements about the corresponding mean (average) values. The bias error was obtained by comparing TGS/FRAM measurements of the total plutonium content of cans to the corresponding values obtained by calorimetric assay (Cal/gamma). FW( 1) is a gamma-ray spectrum analysis code that is used to analyze TGS emission spectra to determine the ratio between the total plutonium and the 239Pu within a given container, The bias error should not be confused with a global bias where the average measurements from many different samples are all a fixed percentage difference from the corresponding true values. If the global measurement bias is zero, then the average bias error for a large number of different samples will be zero. If the Cal/gamma errors are assumed to be zero, then the extracted TGS/FRAM bias error would exist solely because of the TGS/lWQvl bias errors. Although small, the Cal/gamma errors are not zero and will contribute to the extracted bias error. The extracted bias error can thus be viewed as a slight overestimate of the TGS/FRAM bias error.