This research is based on the concept that the diversion of nuclear fuel pins from Light Water Reactor (LWR) spent fuel assemblies is feasible by a careful comparison of spontaneous fission neutron and gamma levels in the guide tube locations of the fuel assemblies. The goal is to be able to determine whether some of the assembly fuel pins are either missing or have been replaced with dummy or fresh fuel pins.
It is known that for typical commercial power spent fuel assemblies, the dominant spontaneous neutron emissions come from Cm-242 and Cm-244. Because of the shorter half-life of Cm-242 (0.45 yr) relative to that of Cm-244 (18.1 yr), Cm-244 is practically the only neutron source contributing to the neutron source term after the spent fuel assemblies are more than two years old.
Initially, this research focused upon developing MCNP5 models of PWR fuel assemblies, modeling their depletion using the MONTEBURNS code, and by carrying out a preliminary depletion of a ¼ model 17x17 assembly from the TAKAHAMA-3 PWR. Later, the depletion and more accurate isotopic distribution in the pins at discharge was modeled using the TRITON depletion module of the SCALE computer code. Benchmarking comparisons were performed with the MONTEBURNS and TRITON results.
Subsequently, the neutron flux in each of the guide tubes of the TAKAHAMA-3 PWR assembly at two years after discharge as calculated by the MCNP5 computer code was determined for various scenarios. Cases were considered for all spent fuel pins present and for replacement of a single pin at a position near the center of the assembly (10,9) and at the corner (17,1). Some scenarios were duplicated with a gamma flux calculation for high energies associated with Cm-244. For each case, the difference between the flux (neutron or gamma) for all spent fuel pins and with a pin removed or replaced is calculated for each guide tube. Different detection criteria were established. The first was whether the relative error of the difference was less than 1.00, allowing for the existence of the difference within the margin of error. The second was whether the difference between the two values was big enough to prevent their error bars from overlapping. Error analysis was performed both using a one second count and pseudo-Maxwell statistics for a projected 60 second count, giving four criteria for detection. The number of guide tubes meeting these criteria was compared and graphed for each case. Further analysis at extremes of high and low enrichment and long and short burnup times was done using data from assemblies at the Beaver Valley 1 and 2 PWR.
In all neutron flux cases, at least two guide tube locations meet all the criteria for detection of pin diversion. At least one location in almost all of the gamma flux cases does. These results show that placing detectors in the empty guide tubes of spent fuel bundles to identify possible pin diversion is feasible.