Performance of Nondestructive Radioassay Systems for Nuclear Waste

A statistical analysis of the measurement performance of five mobile nondestructive assay systems currently in use for radioassay of nuclear waste was performed. Actual versus reported isotopic mass composition, reported versus actual measurement uncertainty, detection capability, and low level versus transuranic waste determination were assessed based on measurements of 11 surrogate waste drums. In terms of percent recovery the systems, for the most part, exhibited reasonable capabilities for estimating isotopic quantities in the surrogate waste drums. (By reasonable we mean, for example, that recovery values were in a range of 40% to 160% for most of the configurations tested.) While difference in performance on individual drums was quite large in some cases, nonparametric analysis of variance results for total plutonium and ${}^{241}\mathrm{Am}$ showed no statistically significant differences in the overall average percent recovery values for the systems evaluated. A similar analysis of the replicate standard deviations indicated significant differences between systems with regard to precision of measurements of total plutonium but not ${}^{241}\mathrm{Am}\text{.}$ The weakest aspect of system performance was the reporting of measurement errors. Reported measurement error values routinely underestimated actual values for certain surrogate waste configurations.