ABSTRACT

During a D&D or ER process containers of radioactive waste are normally generated. The activity can commonly be determined by gamma spectroscopy, but frequently the measurement conditions are not conducive to precise sample-detector geometries, and usually the radioactive material is not in a homogeneous distribution. What is the best method to accurately assay these containers – sampling followed by laboratory analysis, or In-situ spectroscopy? What is the uncertainty of the final result? To help answer these questions, the Canberra tool ISOCS Uncertainty Estimator [IUE] was used to mathematically simulate and evaluate several different measurement scenarios and to estimate the uncertainty of the measurement and the sampling process.

Several representative containers and source distributions were mathematically defined and evaluated to determine the In-situ measurement uncertainty due to the sample non-uniformity. In the First example a typical field situation requiring the measurement of 200-liter drums was evaluated. A sensitivity analysis was done to show which parameters contributed the most to the uncertainty. Then an efficiency uncertainty calculation was performed. In the Second example, a group of 200-liter drums with various types of non-homogeneous distributions was created, and them measurements were simulated with different detector arrangements to see how the uncertainty varied. In the Third example, a truck filled with non-uniform soil was first measured with multiple In-situ detectors to determine the measurement uncertainty. Then composite samples were extracted and the sampling uncertainty computed for comparison to the field measurement uncertainty.