Monte Carlo based sensitivity and uncertainty analysis of the HCPB Test Blanket Module in ITER

This paper presents the first real test application of the Monte Carlo based sensitivity and uncertainty computational approach to the Test Blanket Module (TBM) in ITER. To this end the standard three-dimensional MCNP model of ITER (20° torus sector) with integrated TBM of the Helium-cooled Pebble Bed (HCPB) type was employed. Sensitivity profiles and integrated sensitivities were calculated for the total tritium production in the HCPB TBM by using the track length estimator approach of MCSEN, a local extension to the Monte Carlo code MCNP. The associated uncertainties of the tritium production due to nuclear data uncertainties were assessed by making use of available co-variance data from different sources. The major contribution to the sensitivities was shown to come from the materials contained in the TBM, in particular the beryllium neutron multiplier and the breeder ceramics constituents 6Li and 16O. With the calculated sensitivity profiles and the available co-variance data, the cross-section induced uncertainties of the tritium production have been assessed at a level of 2.8%.