Development of an Inductive Method using Ordinary Least Squares Regression Technique in Nuclear Forensics

Nuclear forensics involves examining nuclear materials and other radioactive substances, seized out of regulatory control, to determine their origin.In the case of spent nuclear fuel, the goal is to identify its initial composition (enrichment, plutonium content), reactor type, and the conditions of its irradiation (burnup, cooling time), based on the isotopic composition of the analyzed sample.In order to solve the inverse problem, several methods can be employed.CEA has developed a range of methods for that purpose, one of them being the "inductive method".Using a numerical model solving the problem of depletion under irradiation and during cooling, this method consists in determining the expected target ( 235 U enrichment, plutonium content, reactor type, burnup and cooling time) based on marker isotopes.We have identified relevant marker isotopes and established differentiation criteria able to differentiate between light water-moderated reactors.Our method allows predicting the expected targets with an ordinary least squares regression technique in the case of pressurized water reactors and boiling water reactors.