Methodology for core analyses with nuclear data uncertainty quantification and application to Swiss PWR operated cycles

At the Paul Scherrer Institut, the development of the SHARK-X methodology to propagate nuclear data uncertainties in CASMO-5 2-D assembly depletion calculations started few years ago. Before that, the CMSYS platform had been established to serve as framework for the continuous development and validation of core models based on CASMO/SIMULATE for all the Swiss operating reactors and cycles. Recently, a first attempt to integrate SHARK-X within CMSYS was initiated in order to complement the core analyses with nuclear data uncertainty quantifications. A proof-of-principle is presented in this paper on the basis of real operated UO2 and MOX cycles of a Swiss PWR plant. In this framework, the nuclear data uncertainties related to cross-sections, neutron multiplicity and fission spectrum, provided via SCALE-6.2b4 based variance-covariance matrices, are propagated using the stochastic sampling method in all types of core calculations, including cycle depletion, validation against flux measurements and safety calculations typically required for reload licensing. For each calculation type and core design, the statistical convergence of relevant quantities of interest is studied in order to determine the required optimal number of samples. Thereafter, the uncertainties of selected quantities of interest are presented and compared between both core designs. For instance, the estimated uncertainties are shown for critical boron concentration as well as 3-D power/burnup distributions for cycle depletion analyses. Regarding the validation of the core models against in-core detector measurements, the uncertainties in biases between calculated and experimental 3-D reaction rates are also evaluated. Finally, for calculations of safety parameters related to core design and licensing, uncertainties in reactivity coefficients, control rod worths and kinetic parameters are presented.