Skip to main content

MELCOR Capability Development for Simulation of Debris Bed Coolability

Time: Thu 2021-09-23 09.30

Location: Via Zoom, (English)

Subject area: Physics, Nuclear Engineering

Doctoral student: Yangli Chen , Kärnkraftssäkerhet

Opponent: Dr.-Ing. Michael Buck,

Supervisor: Universitetslektor Weimin Ma, Kärnkraftssäkerhet


The severe accident management (SAM) strategy for a Nordic boiling water reactor (BWR) employs cavity flooding prior to vessel failure, so that the core melt (corium) discharged from the vessel could fragment and form a particulate debris bed. The key to the success of this SAM strategy is the coolability of ex-vessel debris beds.

The safety analysis involves knowledge about the reactor response to severe accidents under this SAM strategy, which requires the integral simulation of a system code such as MELCOR. Since currently the MELCOR code lacks the modeling of ex-vessel particulate debris beds, the present study aims to develop the capability of MELCOR for the simulation of debris bed coolability through the coupling of MELCOR with other codes, which are dedicated to this phenomenon.

The study is started from the qualification of a MELCOR model for severe accident analysis of a reference Nordic BWR, with the aim to help identify a proper core nodalization. For this purpose, three different core meshes (coarse, medium, and fine) are employed to obtain their impacts on corium release conditions. It is found the coarse mesh is sufficient in the present study, since it is not only computationally efficient, but also predicting earlier vessel failure and faster corium release, providing a more conservative condition for debris bed coolability analysis.

Two couplings are then adopted: (i) coupling of MELCOR with the COCOMO code, which is a mechanistic code for simulation of thermal hydraulics in debris beds; and (ii) coupling of MELCOR with a surrogate model developed in the present study. The first method can simulate the transient behavior of a debris bed during quench process. The second method can efficiently predict the coolability limit (dryout power) required in safety analysis. The surrogate model is developed based on the COCOMO prediction of two-dimensional debris beds.

The developed simulation tools, including the coupled codes and the surrogate model, are applied to the safety analysis of the reference Nordic BWR. The coupled MELCOR/COCOMO simulation is used to investigate the debris bed properties. The effective particle diameter is found as approximately 10% larger than the surface mean diameter of a debris bed with distributed sizes, quantified by the quench rate. For the effect of debris bed shape, it shows a faster quench process with a lower bed slope angle. The quench front propagation as well as the responses of local temperature and containment pressure are obtained.

The coupled MELCOR/surrogate model simulation is performed to estimate the coolability of ex-vessel vessel debris beds. The results show that debris beds are coolable under prototypical conditions with probable bed properties. The surrogate model is used to generate coolability maps, which show the debris bed coolability with the variation of bed properties. The sensitivity analysis indicates that the porosity and the geometry are the most influential to coolability limit. An uncertainty analysis methodology is proposed to obtain the probability of non-coolable debris beds.