Design and Description of the NGSI Spent Fuel Library with Emphasis on the Passive Gamma Signal

One goal of the Next Generation Safeguards Initiative SpentFuel Project is to estimate the amount of plutonium in an assemblyusing nondestructive assay (NDA). For the purpose ofquantifying how a wide range of NDA techniques are expectedto perform as a function of various reactor conditions (initial enrichment,burnup, and cooling time), two sets of virtual spentfuel assemblies in the form of MCNP (Monte Carlo N particletransport code) input files were developed to represent pressurizedwater reactor (PWR) assemblies under those conditions. Thefirst library was created using infinitely reflected boundary conditionsfor four different values of initial enrichment, burnup, andcooling time (sixty-four total combinations). The second libraryconsisted of more realistic irradiation conditions with a 1/8 coregeometry model of a PWR with shuffling of fuel assemblies andmore realistic combinations of input parameters. Several differentassembly-shuffling sequences were examined, and impacts ofshuffling assemblies (moving the assemblies to different locationswithin the core) on the resulting passive gamma count rate areexplored in this paper. The main goal of these spent fuel librariesis to predict isotopic concentrations within each pin of a fuelassembly under expected operational conditions. These concentrationsare then used to assist in the design and assessment ofthe various candidate NDA instrument techniques, primarilyevaluated through further MCNP simulations. The passive gammageometry used to simulate the detector response involved atwo-step process using MCNP. In the first step, we computed theenergy-dependent photon spectrum crossing a collimator face inthe direction toward the detector. This spectrum was then used asthe source for a detector response calculation. This same processwas performed for each of the three sides of the assembly in thesecond library, thus generating three passive gamma signals foreach assembly using various values of initial enrichment, burnup,and cooling time, since each side of the assembly will emit uniquesignals dependent on the shuffling sequence employed. This approachfor passive gamma signal estimation was performed foreach side of each assembly in the second library so that a large databasefor cross-checking passive gamma signals could be created.