Investigations on the flooding behaviour of a partially degraded reactor core

Abstract

This work investigates a mid-sized break loss of coolant accident (MBLOCA) with a 200 cm2 leak in the hot leg next to the pressurizer surge line in the most common reactor type which is the 1; 300MWel KONVOI class pressurized water reactor (PWR). The scope was to analyze different reflooding scenarios, where the reactor core has already started melting when the reflooding is initiated. For this scenario a basic simulation called „BaseCase“ is conducted to get an overview of the accident progression and all the phenomena that occur. It also serves as a basis for all the following studies. In the MBLOCA scenario all active emergency core cooling systems are supposed to work correctly until the flooding pool is empty. Regularly, a switch to sump circulation would take place at that time, but in the simulation it is assumed to fail. So from then on no more water injection into the primary system takes place, which will eventually lead to complete core destruction. In order to investigate whether stable and coolable core conditions can be reached by feeding water into the primary loops during core melting, a diverse set of reflooding scenarios were selected to be simulated based the findings of the BaseCase simulation The feeding can be realized by turning on the high or low pressure injection pumps (HPI / LPI) in sump circulation mode for example, and it can take place at different locations. This leads to different reflooding scenarios, which are analyzed for differences and cooling potential. From study A „Variation of the beginning of reflooding“, it could be conducted that at least up to 40 t of melt can be cooled in the core, if the reflooding system has a sufficient injection rate. The results of study B „Variation of the reflooding system“ show that the reflooding via the LPI creates a self-blocking effect, which compromises the injection, thus preventing the core quenching for a time. This effect does not occur while using the HPI, leading to the question whether a middle pressure injection (MPI) system might be the optimum for the reflooding of a partially degraded PWR core, since it combines a higher mass flow rate than the HPI with a higher pressure head than the LPI. The injection location varied in study C „Variation of the reflooding location“ does affect the simulations, but to a much lower extend than the used reflooding system. Nevertheless, a combined reflooding using both the hot and the cold legs seems to be the optimum. Study D „Variation of the number of available pumps“ revealed that at least two LPI pumps are necessary to performa successful quenching of the core area. The simulation with only one pump did end with a core status that seems to be stable as well, but even after a simulation time twice as high as usual, the core could not be flooded with water and has to be cooled by steam.