Bayesian Approach for Multigamma Radionuclide Quantification Applied on Weakly Attenuating Nuclear Waste Drums

Gamma spectrometry is a passive nondestructive assay method used to quantify radionuclides present in nuclear objects. Basic methods using empirical calibration with standard the activity of materials by determining coefficient are ineffective on nonreproducible objects such as waste packages. Package specifications composition or geometry change from one package another and exhibit large variability The current quantification process uses numerical modeling measured scene few available data composition, particular density, material, screen, geometric shape, matrix matrix, source distribution. Some them strongly dependent knowledge operator backgrounds. French Atomic Energy Commission (CEA) developing methodology packages drums without adjustment internal configuration knowledge. This suggests combining stochastic approach which uses, among others, surrogate models simulate gamma attenuation behavior, Bayesian considering conditional probability densities prior information problem inputs, Markov Chain Monte Carlo (MCMC) algorithms solve inverse problems, ray emission radionuclide spectra, outside dimensions interest. has been tested actinide low bulk density weakly attenuating compositions, distribution terms masses composing drums. Activity uncertainties taken into account.