Criticality and reactor physics benchmark experiments. Influence of nuclear data uncertainties

A number of LWR-type criticality and reactor physics experiments, mainly from the ICSBEP and IRPhEP Handbooks, many of which were already used in the past for international benchmarks in the framework of OECD/NEA working groups, are being evaluated with respect to uncertainties in the basic nuclear data. For this, the sampling based uncertainty and sensitivity analysis tool XSUSA along with the Monte Carlo code KE-NO-V.a as transport solver is employed. Particular emphasis is put on experiments where differential quantities, mainly reaction rate distributions, were measured; the uncertainties of such quantities are not directly accessible to tools based on first order perturbation theory. With respect to multiplication factors and reactivity differences, all results are compared with corresponding results obtained with TSUNAMI-3D from the SCALE 6.1 system; the agreement is very good for all assemblies under consideration. With respect to fission rate distributions, the uncertainty analyses yield only moderate uncertainties from nuclear data; therefore, in general the total uncertainty is dominated by measurement uncertainties, which include the uncertainties of technological parameters. The work is continuously being extended; in the future, also non-LWR specific assemblies, mainly relevant for GEN-IV reactors, will be investigated. (author)