Verification and validation of SEMRA code shock propagation module

The ex-vessel severe accident (SA) mitigation strategy in Nordic Boiling Water Reactors (BWRs) relies on drywell flooding. In the event of the reactor lower head failure corium is released into the water pool in the drywell. The corium jet is expected to fragment, quench, and form a coolable debris bed, ultimately preventing containment failure and release of radioactive products into the environment.

During corium fragmentation in water, a vapor film is formed around the melt preventing direct melt-water contact and limiting the heat transfer between the two liquids. In case of vapor film collapse an explosive conversion of thermal energy of the melt into the mechanical energy of the evaporating volatile coolant may be triggered. These phenomena are often called steam explosion (SE). The resulting pressure wave may propagate through the water-corium mixture, escalate and form a shock wave with the potential to challenge containment integrity. There are significant phenomenological and scenario uncertainties associated with steam explosion. The problem of the uncertainty quantification in the risk analysis is exacerbated further by (i) the chaotic nature of the steam explosion phenomena and (ii) the lack of steam explosion modelling codes based on the modern numerical methods with increased stability and accuracy.

The goal of this work is to develop a numerical code SEMRA (Steam Explosion Modelling and Risk Analysis) for modelling of melt-coolant interactions and assessment of the risk of containment failure due to steam explosion. In this paper, we focus on the development of the deterministic part of the code that utilizes improved numerical methods to assess the propagation of steam explosions. The objective is to verify the implemented numerical schemes for pressure propagation and to establish a reference solution for the next stage of code development which incorporates more comprehensive thermodynamic modelling and transport phenomena relevant to steam explosion.

Specifically, we address the phenomena shock wave triggering and propagation. We implement a numerically stable code using AUSM+ -up, Godunov and HLLC schemes to model multiphase flow. We evaluate the performance of SEMRA code against several known verification and validation problems. Then we use SEMRA code to simulate triggering tests carried out in KROTOS facility and compare the results against the experiment and TEXAS-V code calculations. We analyze the effect of the flux reconstruction method, the vanishing phase treatment and the spatial discretization on the results. We discuss the results and their contribution to the enhancement of triggering and propagation modelling in a SE code.