New method for estimating uranium and plutonium masses using differential die-away signals

Abstract

This study presents a novel method for estimating the masses of uranium-235 and total plutonium in spent nuclear fuel assemblies using only Differential Die-Away (DDA) analysis results and cooling time without relying on passive or delayed neutron measurements. A database of DDA signals was generated through MCNP6 simulations based on spent nuclear fuel compositions derived from ORIGAMI depletion calculations under a wide range of initial enrichments, burnups, and cooling times. The correlations were developed between DDA results (i.e., die-away time and total neutron counts) and the isotopic masses. For235U, a linear function of the signal ratio was used with the coefficients depending on cooling time. For Pu, a separate correlation was introduced using the die-away time difference between spent and fresh fuels, which required an estimation of initial enrichment. A Gaussian Process Regression (GPR) model was trained for this purpose, achieving high accuracy. Validation with 50 independent SNFs showed high prediction performance with maximum relative errors of 7.52 % and 2.41 % for235U and Pu, respectively. This method provides a non-destructive and efficient technique for characterizing spent nuclear fuel assemblies, offering potential applications in nuclear safeguards, spent fuel management, and nuclear material accountancy.