CFD study of an innovative safety system for pressurized water reactors

After the FUKUSHIMA accident, passive systems become an important issue for the new projects of Pressurized Water Reactor (PWR). In response to the demands of his partners, the CEA has proposed suitable systems to deal with different accidental sequences. Then, it has dimensioned their components and modelled their integration in the reactor using the CATHARE2 system code. In particular, the CEA has worked on the ADS system (Automatic Depressurization System) devised to tackle with the well-known LOCA accident (Loss-of-coolant accident), resulting from a breach in the pipes of the primary circuit. Concretely, the ADS system is made up of a tank called IRWST (In-Containment Refueling Water Storage Tank), placed high in the containment, which has the role of achieving a safety passive injection to replace the conventional one using low-pressure pumps, during a LOCA accident. The ADS is in fact already present on the AP1000 reactor, the US PWR designed to be totally passive. The CEA has proposed to improve it by adding a LPWT tank (Low-Pressure Water Tank) pressurisable by steam issued from the pressurizer on the last line of the ADS system. It should allow a better core cooling by the additional water injection in the reactor core, anticipated in time compared to that of the IRWST. The study presented here exclusively focuses on the LPWT water pressurization phase prior to its injection into the reactor core. Three approaches are exposed:-An analytical approach, in which one considers the pressurization of the volume of gas present in the tank without taking into account the liquid phase;-A system approach with the CATHARE2 thermal-hydraulic code which allows to describe at the same time the volume of gas and the volume of liquid contained in the tank;-A CFD approach using the NEPTUNE_CFD code to determine the local evolution of the quantities that characterize the system. This approach will notably make it possible to describe the movement of the interface and the velocity profile of the water vapor. It will also help to bring a better understanding of the physical phenomena involved during the LPWT water pressurization phase. At the end, these different approaches show that the increase in pressure by the steam injection is an interesting option to pressurize the water tank. On the other hand, the results of the pressure evolution obtained by the system approach are very close to those obtained analytically, in the absence of liquid phase. However, they are overestimated compared to the CFD approach, which, thanks to a better modeling of the condensation of the injected vapor, leads to a more realistic evaluation