On Using Code Emulators and Monte Carlo Estimation to Predict Assembly Attributes of Spent Fuel Assemblies for Safeguards Applications

Abstract

The quantification of the plutonium mass in spent nuclear fuel assemblies is an important measurement for nuclear safeguards practitioners. A program is well underway to develop nondestructive assay instruments that, when combined, will be able to quantify the plutonium content of a spent nuclear fuel assembly. Each instrument will quantify a specific attribute of the spent fuel assembly, e.g., the fissile content. In this paper, we present a Monte Carlo-based method of estimating the mean and distribution of some assembly attributes. An MCNPX model of each instrument has been created, and the response of the instrument was simulated for a range of spent fuel assemblies with discrete parameters (e.g., burnup, initial enrichment, and cooling time). The Monte Carlo-based method interpolates between the modeled results for an instrument to emulate a response for parameters not explicitly modeled. We demonstrate the usefulness of this technique in applying the technique to six different instruments under investigation. The results show that this Monte Carlo-based method can be used to estimate the assembly attributes of a spent fuel assembly based upon the measured response from the instrument.