Revisiting ISP-13 with RELAP/SCDAPSIM/MOD3.5 using core SCDAP components

Abstract

The recent accident in the Fukushima Daiichi nuclear power plant opened a discussion on severe accident management that includes the analysis of the accident by means of computational tools that can predict the core behavior in such extreme conditions. The RELAP/SCDAPSIM/MOD3.5 code is designed to predict the behavior of Light Water Reactor (LWR) coolant systems during normal and accident conditions including severe accidents up to the point of reactor vessel failure. The code consists of two parts: the RELAP5 models calculate the overall Reactor Coolant System (RCS) thermal-hydraulic response, control system behavior, reactor kinetics and the behavior of special reactor system components such as valves and pumps, to predict the plant behavior under operational transients, Design Basis Accidents (DBAs) and Beyond DBAs; the SCDAP models calculate the behavior of the core and vessel structures under normal and severe accident conditions. Both portions of the code have been proven, separately, to accurately reproduce the response under its designed purpose, which are steady state, DBAs and BDBAs for the RELAP portion, and steady state and severe accident conditions for the SCDAP portion. The analysis of potential scenarios does not define a priori the final state of the fuel rods, and thus the most adequate tool is a system code such as RELAP/SCDAPSIM/MOD3.5 capable of simulating accident scenarios where severe accident phenomena may or may not occur. The present paper revisits the ISP-13 exercise, a cold leg double-ended guillotine LOCA conducted in the LOFT experimental facility, using two RELAP/SCDAPSIM/MOD3.5 models: the first one is entirely modeled with RELAP components, the second model keeps the RELAP nodalization with the exception of the core region, which is modeled with SCDAP components. The LOFT L2.5 experiment is a rather unique experiment since it features nuclear (UO2) fuel rods in a facility designed to simulate the major responses of a commercial pressurized water reactor (PWR). In addition, the fuel cladding of this experiment reached relatively high temperatures of around 1100 K. Even though this cladding temperature is far from the oxidation onset with steam, the LOFT L2-5 experiment challenges system behavior simulations by bringing the conditions close to those of severe accidents. The final goal is to evaluate whether the use of SCDAP components in LOFT L2-5 experiment reproduces similar results to those obtained with a RELAP standalone model, and that both simulations are in good agreement with experimental data.