Safety analyses show that core degradation during a severe accident would not be uniform. This was confirmed by TMI2 examinations. In fact, a central region of the core may overheat, melt and flow down to the lower plenum of the reactor while peripheral regions of the core would remain almost intact.
In case of rupture of the vessel by molten debris, air may be drawn from the containment by natural convection into the reactor vessel and react with the intact rods.
Studying air ingress into the reactor vessel is of interest because the interaction of air with Zircaloy cladding can strongly affect the evolution of severe accident scenarios. The main consequences that may be expected are an additional heat generation, an increased cladding degradation, an additional fission product release and nitriding. In case of air/steam recirculation in the vessel, significant nitriding of cladding can occur. The resulting ZrN phase is characterized by its brittleness and instability under oxidizing conditions, oxidation of pre-existing ZrN phase layers has been observed to result in violent oxidation and heat release.
Therefore, the first consequence for safety is a risk of strong deflagration in the vessel if a large number of rods on which a substantial layer of ZrN has grown are suddenly in contact with oxygen or steam.
The second consequence is a late melting of core materials due to the very exothermic oxidation reaction, leading to a late release of materials out of the reactor pressure vessel (RPV).
In the paper we present an ICARE/CATHARE V2.0 calculation simulating air ingress into the vessel and with a focus on the nitriding effect due to the oxygen consumption in the reactor vessel. The basic modelling and the necessary extensions of both ICARE and CATHARE codes are explained. The two-dimensional flow is calculated to predict the regions of oxygen starvation where nitriding takes place.