On gas-cooled fast reactor designs – Nuclear data processing with sensitivity, uncertainty and similarity analyses

Abstract

To license the gas-cooled fast reactor design, sufficient data should be known about its safety and operation capability. The first step in deployment of the commercial GFR 2400 is to prove the design and physical parameters by its demonstrator, referred as ALLEGRO. This study summarizes sensitivity, uncertainty and similarity analyses of GFR 2400 and ALLEGRO cores utilizing standard and in-house developed codes. In the part focused on the sensitivity analyses, the in-house developed calculation scheme is compared and validated with available commercial codes for both gas-cooled fast reactor designs. The following part is aimed to estimate the uncertainty propagated from nuclear data to keff for the ALLEGRO design. The highest contributor to keff uncertainty induced by nuclear data is 239Pu fission neutron yield (nubar), due to the high sensitivity profile and inelastic scattering reaction on 238U, due to the associated large cross-section uncertainty in the energy spectra of interest. The last part is focused on the similarity assessment of both gas-cooled fast reactor designs. The results show that the ALLEGRO MOX starting core is not sufficient representation of the commercial GFR 2400 accommodating carbide fuel, therefore carbide fuel type has to be used also in the next fuel loading patterns in ALLEGRO design to demonstrate the feasibility of GFR 2400 design.