Nuclear data uncertainty propagation on a sodium fast reactor

Uncertainties on criticality parameters and reactivity effects in a SFR (Sodium Fast Reactor) were calculated and characterized. An SFR model loaded with ≈ 10% MAs (Minor Actinides) was chosen as test case of a burner reactor, its core geometry was implemented in MCNP-5 according to benchmarking specifications. Monte Carlo random sampling on the ENDF/B-VII.1 neutron-induced nuclear data was performed using the SANDY code developed at the Belgian nuclear research center SCK•CEN. SANDY was proved to be a successful tool for uncertainty quantification in the model case of a MA burner SFR. The results show that despite the high content in MAs, the contribution of U-238 data is still dominant in most instances, especially the inelastic scattering cross section data. Minor actinides together with Pu make up to 33% of the total calculated keff uncertainty, which amounts to ± 1345 pcm of reactivity. Relative uncertainty greater than ± 3.47% was estimated for βeff with significant contribution from the prompt multiplicity data of U-238 and Pu-239. Relative uncertainty on the Doppler reactivity worth was found to be at least ± 12.6%. Coolant voiding in different core zones yields reactivity worth with relative uncertainties whose lower bounds range from ± 6.37% to ± 22.47% depending on the voiding conditions, with important contributions due to the Na-23 nuclear data uncertainty.