Uncertainty propagation calculation from model parameters in thermal scattering of H-1 bound in H ₂ O to neutronics parameters with deterministic sampling method and sandwich formula

In order to reduce the nuclear data-induced uncertainties in neutronics parameters of nuclear reactors, the data assimilation using integral data is promising. Generally, integral data used in the data assimilation are obtained at room temperature whereas concerned neutronics parameters are at high temperature, and thus this inconsistency in temperature can bring an issue. In the present paper, this is addressed by focusing on the thermal scattering cross sections of H-1 bound in H2O. Uncertainties of neutronics parameters due to the model parameter uncertainties in the TSL data evaluation is quantified with the deterministic sampling method and the sandwich formula using covariance matrices and sensitivity coefficients. It is demonstrated that the deterministic sampling method is effective and that uses of covariance matrices and sensitivities not for scattering cross sections but for scattering matrices are mandatory in the sandwich formula. Standard deviations of k∞ at high temperature of the LWR fuel pincells and correlations between them and keff of the critical experiments at room temperature are calculated. Results of the uncertainty propagation calculations are significantly dependent on the assumed uncertainties in the model parameters of thermal scattering.