Uncertainty quantification method for RELAP5-3D© using RAVEN and application on NACIE experiments

Abstract

The best estimate plus uncertainty (BEPU) method plays a key role in the development of the innovative Generation IV nuclear reactors, for the improvement of knowledge and the good evaluation of the safety margins for new phenomena. The aim of this paper is to validate an uncertainty quantification (UQ) approach using RAVEN code. RAVEN, developed at the Idaho National Laboratory, is a multipurpose probabilistic and uncertainty quantification framework, capable to communicate with any system code, implemented with an integrated validation methodology involving several different metrics. In this activity, a coupled calculation RELAP5-3D/RAVEN has been performed to assess the validity of the embedded UQ approach. The simulations have concerned two tests conducted on NACIE (NAtural CIrculation Experiment) facility, a non-nuclear loop-type system using Lead-Bismuth Eutectic (LBE) as coolant, realized at the ENEA Brasimone Research Centre (Italy) to support the development of the GEN-IV reactors. The experimental tests are aimed to investigate the phenomena related to the natural and gas enhanced circulation flow regimes of heavy liquid metals (HLM), and to test and validate the main components in a LBE environment. A numerical 1D model of NACIE facility has been realized and the post-test analysis has been carried out using RELAP5-3D© ver. 4.3.4. Based on these results, a statistical analysis has been performed using RAVEN computer code, investigating the capability of three probabilistic comparison metrics fully integrated in the code.