Semi-Empirical Modeling and Implementation of Hybrid K-Edge Densitometry for Pyroprocessing Applications

The United States is currently looking at options for handling of spent nuclear fuel. Currently, there are ≈ 70, 000 metric tons of spent fuel in storage in the US alone. Pyroprocessing is a possible method for spent nuclear fuel reprocessing which was proven to work at Argonne National Laboratory. This masters thesis showcases a method for empirically modeling hybrid k-edge densitometry, one of the numerous possible safeguards needed for a reprocessing facility. This is accomplished by using MCNP to perform 54 sets of 2-stage simulations for KED and XRF, respectively. The end results are empirical linear functions for the magnitude of the k-edge drop of uranium and plutonium, as well as empirical functions for the XRF peaks for uranium and plutonium. These two empirical functions are functions of uranium concentration and plutonium to uranium mass fraction ratio, respectively. The semi-empirical functions are then implemented into the Sandia National Laboratory Separation and Safeguards Performance Model EChem (SSPM Echem) Simulink model. The SSPM is a Simulink model which can use modular safeguards functions for methods such as HKED, passive gamma, and passive neutron detection. These types of safeguards modules help to evaluate how much the standard error of inventory difference (SEID) is affected by additional safeguards, providing a quantifiable value. The empirical functions as well as multiple representative figures and tables are presented, showcasing the ability of the Simulink module to correctly predict the KED drops as well as the XRF peaks. The ultimate goal is to combine the KED and XRF into HKED measurements to get a value for the mass fraction of plutonium in salt. Plots showcasing the differences between the HKED module output and SSPM\u27s internal mass tracking are shown