Application of the GRS method for estimation of uncertainties of LMFBR type reactor physics parameters with taking into account macroscopic experiments

A number of recent studies have been devoted to estimation of errors of fast reactor nuclear-physics parameters by the GRS (Generation Random Sampled) method. The method is based on direct sampling input data resulting in formation of random sets of input parameters which are used for multiple neutronics calculations. Once these calculations are performed, a statistical processing of calculation results is carried out to determine the mean value and the variance of each calculation parameter of interest. In this study the GRS full scale method was used for estimation of errors of calculation of nuclear-physics parameters (k-eff, power density, dose rate) for a perspective LMFBR type sodium-cooled fast reactor with taking into account different sets of integral and macroscopic experiments. The multiple neutronics calculations are performed by the nodal diffusion three-dimensional code TRIGEX and CONSYST/ABBN nuclear data system. The sampling of input data, nuclide nuclear cross-sections, is based on the ABBN-93 covariance matrices. (author)