Validation of the RELAP5 code for the modeling of flashing-induced instabilities under natural circulation conditions using experimental data from the CIRCUS test facility

This paper reports on the use of the RELAP5 code for the simulation of flashing-induced instabilities in natural circulation systems. The RELAP 5 code is used for the simulation of transient processes in the Russian RUTA reactor concept operating at atmospheric pressure with forced convection of coolant. However, during transient processes, natural circulation with flashing-induced instabilities might occur. The RELAP5 code is validated against measurement data from natural circulation experiments performed within the framework of a European project (NACUSP) on the CIRCUS facility. The facility, built at the Delft University of Technology in The Netherlands, is a water/steam 1:1 height-scaled loop of a typical natural-circulation-cooled BWR. It was shown that the RELAP5 code is able to model all relevant phenomena related to flashing induced instabilities. The magnitude and frequency of the oscillations were reproduced in a good agreement with the measurement data. The close correspondence to the experiments was reached by detailed modeling of all components of the CIRCUS facility including the heat exchanger, the buffer vessel and the steam dome at the top of the facility.
In the second part of the presentation, the impact of the modeling of subcooled boiling on natural circulation instability is considered. A transient with failure of all primary circulation pumps (initial event) and the failure of reactor SCRAM for the RUTA facility using the coupled neutron kinetics/thermo-hydaulic code systems DYN3D/ATHLET and DYN3D/RELAP5 was simulated. The predictions for initial and final reactor states given by the codes are in good agreement. However, the process of transition between these two states shows a qualitative difference. The DYN3D/RELAP5 code predicts unstable transient behavior of the reactor, while in the DYN3D/ATHLET simulation a smooth change of reactor parameters is observed during the whole accident. It was found that the different stability behavior is due to differences between the subcooled boiling models of the ATHLET and RELAP5 codes.
Despite the different performance of RUTA in the DYN3D/ATHLET and DYN3D/RELAP5 simulations of the accident the obtained results confirm a high intrinsic safety level for this reactor concept. In both compared calculations the allowed safety margins have not been reached. However, further validation of the subcooled boiling models on experiments at low pressure is necessary.