Thermal-hydraulic analysis of an intermediate LOCA test at the ROSA facility including uncertainty evaluation

The goal of uncertainty analyses is to provide best-estimate simulations with uncertainty bands, which take into account uncertainties in the code modeling capabilities as well as uncertainties in the initial and boundary conditions of a given transient scenario. In the present paper, the experience acquired at the Paul Scherrer Institut (PSI) through participations to previous international programs (among these the OECD/NEA BEMUSE program) is used to evaluate a blind calculation of the ROSA-2 Test 1. The OECD/NEA ROSA-2 project aims at addressing thermal-hydraulic safety issues relevant for light water reactors by building up an experimental database at the ROSA Large Scale Test Facility (LSTF). The ROSA facility simulates a PWR Westinghouse design with a four-loop configuration and a power of 3423 MWth. Test 1 of the ROSA-2 program consists of an intermediate break located in one of the hot legs, in particular it represents the rupture of the pressurizer surge line. Using a TRACE model, previously validated against experiments of the ROSA-1 programme, namely small break LOCA Tests 6-1 and 6-2, a blind case calculation was performed for Test 1/ROSA-2. An uncertainty analysis was carried out together with the simulation of Test 1, in order to provide uncertainty bands for each time trend and finally determine whether the TRACE simulation is able to capture the experimental results within the uncertainty bands. Since the uncertainty bands did not envelop the experimental data, a post-test analysis was carried out. The post-test analysis was helpful in determining which relevant physical phenomena had not been included in the pre-test analysis. The new uncertainty bands derived after including all relevant phenomena enveloped the experimental data.

► Pre and post test calculations of an IBLOCA test at the ROSA facility are presented.
► Both analyses included an evaluation of the uncertainties.
► The differences between the two uncertainty evaluations are analysed.
► The article highlights the impact of the base case on the final uncertainty band.