Safeguards modeling for studies of pyroprocessing input accountancy and fuel impacts on ID and SEID

Input accountancy is especially challenging for pyroprocessing because destructive assay (DA) accountancy methods for safeguarding special nuclear material (SNM) in aqueous reprocessing are not compatible with pyroprocessing. A team of researchers at Idaho State University (ISU) on the Resonance Absorption Densitometry for Materials Assay Security Safeguards (RADMASS) project awarded to General Electric Global Research (GE) and funded through ARPA-E ONWARDS is modeling a new approach of input accountancy in pyroprocessing in hopes of improving upon existing nondestructive assay (NDA) methods.This paper provides a description of a novel nondestructive assay method (NDA) called dual isotope notch observer (DINO) detection using a laser Compton scattering (LCS) source. The source is tuned to the energy of nuclear resonance fluorescence (NRF) for SNM isotopes and promises a decrease in accountancy error over existing NDA technologies. This detection method has been modeled in MCNP6.2 for study and optimization.The Fast Reactor Pyroprocessing Safeguards Performance Model (FRPSPM) has been developed in SimEvents to explore the effects of changing beam parameters for an LCS source and DINO detection system on inventory difference (ID) and standard error of inventory difference (SEID). This paper describes the modeling of process and monitoring systems of pyroprocessing for purposes of calculating ID and SEID with model outputs.The FRPSPM will also be used to discover which contributions from the process itself are most impactful on SEID and investigate the boundaries of input fuel enrichments, burnups, and compositions capable of meeting SEID requirements with sufficiently long inventory cleanout periods to render pyroprocessing commercially practical. The results to date of the parametric studies will be presented.This study will be able to advise future regulatory frameworks for pyroprocessing, especially with respect to handling and management of advanced reactor fuels, as well as contribute towards the development of a figure of merit (FOM) indicating viability of a commercial pyroprocessing facility. Additionally, GE will use the data produced to train a machine learning (ML) model and develop a digital twin for a pyroprocessing facility. It is hoped that these combined efforts will inspire a reconsideration of pyroprocessing as a viable commercial practice to minimize nuclear waste through utilization of recovered actinides in fast reactor fuel.