Sensitivity and uncertainty analyses of the tritium production in the HCPB breeder blanket mock-up experiment

Abstract

A neutronics experiment performed at the Frascati Neutron Generator (FNG) on a mock-up of the Helium-Cooled Pebble Bed (HCPB) breeder test blanket module (TBM) has been analysed on the basis of neutron transport, sensitivity and uncertainty calculations using both deterministic and probabilistic computational methods. The calculations revealed a slight but systematic underestimation of the measured tritium activities by 5 to 10% on average, independent of the nuclear data used that were taken from the EFF-3 and FENDL-2.0, -2.1 data files. The uncertainty assessments showed data related uncertainties of the calculated tritium production rates at a level of 4% (2 σ confidence level). The beryllium cross-section data were shown to be the main source of these uncertainties. The total uncertainties of the tritium production prediction including the data uncertainties, the statistical uncertainties of the Monte Carlo calculation and the experimental uncertainties are in the order of 8 to 10% (2 σ). The observed underestimation of the measured tritium production is at the lower bound of the assessed uncertainty margin. The obtained results indicate that design calculations for the tritium breeding ratio (TBR) of fusion power reactors employing a HCPB type breeder blanket are conservative. Thus, an additional TBR margin is provided which allows compensation for potential other uncertainties.