PRELIMINARY UNCERTAINTY ASSESSMENT OF THE CONTAINMENT BEHAVIOR

抄録

In recent years, an ever-growing attention has been paid to the assessment of the uncertainties linked to simulation results from computer codes employed in the safety analysis of Nuclear Power Plants (NPPs). These codes have been improved over the years, and they have reached a high level of maturity with regards to their ability to simulate complex systems and scenarios. This is particularly important when it comes to Severe Accidents (SAs). However, a thorough quantification of the uncertainties related to SA codes predictions is still needed. In this framework, the present paper reports a preliminary Uncertainty And Sensitivity Analysis (UASA) of the containment behavior for the PHEBUS FPT1 test. The MELCOR 2.2 code has been employed to model the containment vessel of the PHEBUS facility, whereas the uncertainty investigation has been carried out through the propagation of a limited set of uncertain input parameters. The Wilks formula has been then used to calculate the minimum number of code runs (minimum sample size) necessary to achieve a 95%/95% confidence and probability levels. As for the outcomes, considering that the correct estimation of a possible radiological source term to the environment is of high relevance to the safety assessment of NPPs, the focus of the entire analysis has been placed on the mass of aerosol suspended in the containment atmosphere. The uncertainty band has been derived for the selected Figure of Merit (FOM), and an initial search for the parameters with the highest influence on the response has been conducted. Results show that the uncertainty band varies in time along with the scenario unfolding, and it presents a wide broadening concurrently with the reaching of the peak of aerosol in the containment atmosphere.