Nuclear data sensitivity and uncertainty analysis on the neutronics study of the China Experimental Fast Reactor

The aim of this work is to perform sensitivity and uncertainty analyses on the neutronics study of the China Experimental Fast Reactor (CEFR), based on the last work package of the IAEA-CRP project on CEFR start-up tests. The input parameters that were perturbed for both sensitivity and uncertainty studies corresponded only to neutron-induced microscopic reactions of selected nuclides. Such perturbations were made with the aid of the SANDY code where for the sensitivity analyses, microscopic cross-sections were modified uniformly along the spectra and up to a certain degree (e.g. +/-1% or +/-5%). For the uncertainty analyses, cross-sections were randomly sampled based on covariance information included in the nuclear data library. Starting from a critical configuration model using the OpenMC code, the most sensitive nuclear reaction was found to be fission in U235. Regarding the results related to the uncertainty analysis, these were obtained by following a statistical approach, where neutron induced reaction uncertainty was propagated by means of the Total Monte Carlo (TMC) method. A total of 300 input random libraries based on both ENDF/B-VII.1 and JEFF-3.3. were created with SANDY for the computation of the effective multiplication factor sample. The number of input samples was chosen so, based on Wilks’ distribution free method for non-parametric tolerance limits, at least 95% of the population can be inferred with a 95% of confidence. Based on this theory, it can then be said that the inferred 95th percentile of the effective multiplication factor based on both the ENDF/BVII.1 and JEFF3.3 nuclear data libraries corresponded to be 1.05947 and1.03361, respectively.