Analytical investigations on a core catcher concept for ex-vessel melt retention by water injection through porous concrete from below

Abstract

The aim of this work is to investigate important aspects of ex-vessel molten corium cooling by bottom flooding through a porous concrete core catcher. These are the transport of cooling water through the porous concrete to the melt and the fragmenta-tion and porosity formation in the melt due to the interaction with cooling water/steam. Firstly, this work investigates the hydraulic parameters of the porous concrete core catcher. Cooling water flow simulations are performed by means of the code COCOMO3D for the passive distribution of the cooling water into the melt layer from below with sufficient flow rates over the large reactor cavity. These investigations show that the permeability values of two different layers of the porous concrete and their relation to one another have a significant effect on the rising superficial velocity of the cooling water and on the required pressure head for the coolant supply. A methodology is developed in this work in order to optimize the core catcher for hydraulic properties, which can be applied to various boundary conditions. Based on this methodology, for the inlet configuration that provides the cooling water around to the core catcher over the whole perimeter, various concrete pairings can be chosen to provide sufficient cooling water into the molten corium uniformly, with a feasible pressure head. Due to the restrictions on the design for back fitting, in these cases the water supply can be provided to the core catcher only from a very small inlet connection. Thus, very high velocities of cooling water are expected around the inlet region, and the linear friction laws are not adequate anymore. To gather the needed data for improving the modelling a dedicated experiment set-up is built within the framework of this work, and the relation between pressure and superficial velocity of water for porous concrete samples from CometPC core catcher, which are provided by KIT, is measured. The falling head method is used which enables the measurements for a wide range of pressure values. The non-linear friction law with the values for permeability and passability obtained from the measurements is then implemented into COCOMO3D. The simula-tions are performed for restricted water inlet case with the quadratic friction law for porous concretes. These simulations show that back fitting of the porous concrete core catcher device with limited water inlet configuration can raise many challenges. Increasing the area of the water inlet and providing the water uniformly from the perimeter of the porous core catcher device is the more feasible approach for the reactor application. Finally, the fragmentation and porosity formation phenomenon caused by bottom flooding is modelled in this work. The initial molten corium-steam interaction is assumed as the decisive phenomenon having a lasting effect on the fragmentation. Therefore, the fragmentation is modelled as a void fraction modeling of the two-phase flow of molten corium and steam in thermal equilibrium. In order to simulate this two-phase flow a new model for the interfacial friction force for molten corium-steam two-phase flow has been necessary. A new Bubbly-Channel two-phase flow interfacial friction force model is developed in this work in order to model the porosity formation during corium cooling by bottom flooding. With a stand-alone simulation program, this new model is validated against the COMET and CometPC experiments with varying boundary conditions and material properties. The results show very good agreement with the final porosity range achieved in the experiments. The new interfacial friction model and the material properties of liquid corium are implemented into COCOMO3D code. The COCOMO3D simulations of twophase flow of molten corium and steam are performed for CometPC plus experiment data. The flow pattern of steam and melt during these simulations shows good replication of the post-test morphology of solidified porous thermite from the experiments as well as the dispersion and ejection that happened during the experiment. Two further cases are simulated for the reactor case, which show that the partial ablation of the sacrificial concrete between core catcher and molten corium has an effect on the coolant distribution in the compact melt layer. These investigations show that being able to simulate molten corium as a moving liquid, due to the new model, provides more realistic modelling of coolant ingression into the compact melt layer via bottom flooding and the space that coolant actually occupies, hence the fragmentation phenomenon. The two-phase flow modelling of molten corium and steam presented in this work can be extended further in the future to three-phase flow of molten corium-steam-water with evaporation phenomenon in order to model the entire cooling and solidification process by bottom cooling.