A Reliable Absolute and Relative Bayesian Method for Nuclear Decommissioning: Low-Level Radioactivity Detection With Gamma-Ray Spectrometry

Abstract

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Measurement of Low-Activity Uranium Contamination using Bayesian Statistical Decision Theory

“…Detection of low levels of uranium contamination is to determine its presence by means of an HPGe detector. To do so, we used the main gamma-ray emission lines from the decay chains of U-235 and U-238 [1]: (i) from the U-235 chain, one main line at 185.7 keV, and an additional triplet at 143.8; 163.4; 205.3 keV, and (ii) from the U-238 chain, one main line at 1001.0 keV.…”

“…In this paper, we aim to measure a low-activity uranium contamination on concrete surfaces, with varying enrichment encountered levels within a basic nuclear facility. Our surface contamination limit was taken as S A = 500 Bq/0.1 m 2 [1]. In this context, we have developed an advanced Bayesian method based on a Kibble bivariate gamma distribution, referred to as the Bayesian Kibble Dirichlet (BKD) test.…”

“…It allows to adjust the confidence degree in the stationarity of the background signal. In this perspective, the BKD test combines the properties of an absolute and a relative hypothesis statistical test, that were both developed with the same application at hand [1], and respectively referred to as the Bayesian Absolute test by mixture of Multivariate Poisson laws (BAM) and the Bayesian Relative test by mixture of Multinomial laws (BRM). Note that unlike the BKD test, the modeling of the BAM test was based on the univariate gamma ACP2021, March 7-11, 2022 -Virtual Edition distribution.…”