Information driven safeguards approach for remote monitoring system of dry cask storage

The lack of an operational deep geological repository in the U.S. has left the nuclear power plant operators with the option of storing the more than five years water cooled spent nuclear fuel (SNF) assemblies in dry casks as an interim solution. A single dry cask that typically stores 32 pressurized water reactor (PWR) SNF assemblies can contain up to 20 significant quantities (SQ) of plutonium. A disadvantage of using dry casks from a nuclear safeguards perspective is that there is currently no effective way of re-verifying its contents after it has been sealed. Current safeguards measures rely on containment and surveillance (C&S) techniques primarily by using seals and optical surveillance systems. Nonetheless, if the seal is broken, there is no method to verify the contents of a cask without opening it. This endeavor of opening the cask would expose the personnel involved to high levels of radiation in addition to being expensive and undesirable. Hence, a remote monitoring system (RMS) is designed, analyzed and the results presented here. The RMS combines nondestructive analysis (NDA) and C&S to collect and transmit data continuously to the authorities so that the international atomic energy agency (IAEA) can fulfill its requirement of maintaining continuity of knowledge (CoK) and thus verifying the contents of the dry casks without having to ever open them. The study performed uses radiation transport simulations and analysis of the results, specifically neutron transport simulations for the radiation source term present inside the SNF assemblies. Simulations for this study includes removal of a selected number of SNF assemblies from a multipurpose canister (MPC) and substitution with dummy assemblies. Eight SNF assembly diversion scenarios are analyzed through neutron radiation transport simulations. A neutron transport simulation with all the 32 SNF assemblies present inside the MPC and the corresponding neutron signal on the top of the MPC is used as the reference to compare to the reduction in neutron signal for each of the diversion scenarios. A false alarm probability, (alpha) of 5% is set for the reference case to determine the threshold for detecting the SNF assembly diversion. Simulation results and the analysis showed that the non-detection probability (beta) for each SNF assembly diversion case is less than 20% and thus serve as a proof of concept that even diversion of a single SNF assembly is detectable. In addition, 32 Cf-252 point sources with source strengths mimicking that of SNF assemblies were simulated in a similar configuration within the MPC, which also provided similar results. This study is a proof of concept in order for experiments to be conducted to validate the simulation results. All the simulations are performed using the multi-purpose Monte Carlo Radiation Transport Code, MCNP 6.