Impact of different correlation structures in cross-section covariance matrices on the inventory and inventory-related parameters

At present, cross section correlation data are very poorly known. This work can be helpful to evaluate the effort that the international community should devote to the improvement of correlation data for inventory prediction in nuclear systems. Our purpose is to assess the impact of different correlation structures on the uncertainties in relevant fuel cycle and repository parameters, using fixed variances/diagonal values in the cross section covariance matrices. At this stage, only correlation in energy is taken into account, and no correlation among different types of nuclear reactions or different isotopes is considered. To accomplish this goal, the inventory code ACAB is used to estimate the uncertainties in the actinide concentrations of the irradiated fuel, related decay heat and dose (radiotoxicity) in a representative ADS irradiation scenario. It is shown that the nature of the introduced correlations is very relevant to estimate the overall uncertainty in those parameters, and then, the need of more scientifically based correlation data is clearly justified. (authors)