A new framework for sampling-based uncertainty quantification of the six-group reactor kinetic parameters

摘要

Delayed neutrons, which are described by kinetic parameters, are significant for nuclear reactor operation as they make nuclear reactors controllable. Modern lattice physics codes such as TRITON and Polaris in SCALE code system, generate kinetic parameters for a given material composition. The calculation is performed by summation of the delayed neutron data for each precursor isotope and then weighting is performed by real and/or adjoint neutron spectrum. Quantifying uncertainties in the weighted kinetic parameters is important for assembly and core calculations. Understanding uncertainty in modeling scenarios involve kinetic parameters (e.g. transients), requires propagating uncertainty in the weighted kinetic parameters due to uncertainties in the fundamental nuclear data libraries, including delayed neutron data. In this work, uncertainty analysis of the homogenized (also called weighted or macroscopic) kinetic parameters has been performed using Sampler, a module in SCALE code system, to investigate the effect of fundamental nuclear and delayed neutron data uncertainties on the weighted kinetic parameters. Two major sources of uncertainties were considered: (1) fundamental nuclear data (i.e. cross-sections, fission yield, decay data) and (2) nuclide-dependent group-wise delayed neutron data based on reported experimental measurements. In this study, a new capability was developed through SCALE code system to allow propagation of delayed neutron data uncertainties. Preliminary analysis demonstrated 7% uncertainty in beta(eff) at Beginning of Life (BOL) and increased to 15% after fuel burnup. Decay constant groups showed lower uncertainty than delayed neutron fraction groups, both at BOL and End of Life (EOL). Delayed neutron fraction responses showed high correlation to each other which is expected to be due to the cross-section covariances reported in SCALE data libraries as well as the normalization condition of the nuclide-dependent DNF. Different sources of U-235 delayed neutron data were compared, and the results showed that the uncertainty calculated by Tuttle data was bounded by other sources. The current study can be extended to calculate kinetic parameters and their uncertainties for more advanced LWR applications. Published by Elsevier Ltd.