Uncertainty Analyses of Tritium Production and Gamma Heating Rates in the FNS Clean Benchmark Experiments

Abstract

The propagation of nuclear data uncertainties in fusion neutronics calculations is presented in this paper. The uncertainty propagation employs the random samples of neutron cross sections and secondary particle energy/angular distributions generated by the SANDY code as nuclear data in the transport simulation of the Monte Carlo (MC) code OpenMC. The random samples are obtained from stochastic sampling employing covariances in nuclear data libraries. In this work, uncertainties in nuclear data result in perturbed neutron flux distributions that are then propagated to the gamma heating and tritium production rates in the Fusion Neutron Source clean benchmark experiments on vanadium, beryllium, tungsten, iron, copper, and graphite assemblies, which were irradiated with a 14-MeV deuterium-tritium neutron source from the Shielding Integral Benchmark Archive and Database (SINBAD). The uncertainty analysis results show that for the beryllium assembly, the tritium production uncertainties are dominated by the 9Be cross sections, while the cross sections of 6Li and the impurities present have an insignificant effect on the tritium production. In addition, the gamma heating in the vanadium assembly has the largest uncertainty (up to 23%, with impurities contributing less) among the materials analyzed, followed by graphite (~ 20%), tungsten (17%), iron (14%), and copper (< 6%). These results are important for the application of best estimate plus uncertainty methods, verification and validation, and design of fusion reactors and power plants.