Conference Agenda

An ex-vessel core catcher for advanced light water reactors is being developed to stabilize molten corium outside the reactor vessel and prevent molten corium concrete interaction (MCCI) during severe accidents. The core catcher includes a sacrificial concrete (SC) layer, carbon steel body, protective material, and an external cooling channel. Its key function is to collect molten corium and remove heat, while preventing re-criticality and excessive hydrogen generation. During severe accidents, molten corium discharge and some parts of a reactor pressure vessel (RPV) are expected to impact directly on the core catcher body. Given their high momentum and large internal energy, it might degrade the integrity of the structural bodies, affecting safety performance of the core catcher. To investigate these scenarios, we developed a smoothed particle hydrodynamics (SPH) framework to analyze the interaction among the molten corium, reactor pressure vessel, and core catcher. As a fully Lagrangian particle-based method, SPH is suitable for handling free surface of corium flows and structural topology changes in the RPV and core catcher body. The explicit incompressible SPH (EISPH) model simulates corium behavior with an implicit viscosity solver for computational efficiency. A finite multiplicative model based on Total Lagrangian SPH (TLSPH) handles large-strain elastoplasticity in the structures. The coupled EISPH-TLSPH model is used to investigate mainly corium relocation behaviors and structural integrity under several potential accident scenarios. We expect that the proposed model will be useful for accurate safety analyses of the accident mitigation strategy using the ex-vessel core catcher.

This paper presents the analysis of the effect of Molten Corium Stratification (MCS) during Molten Corium Concrete Interaction (MCCI) using developed the PUMBAA (Prediction modUle of MCCIs with Basemat Attack and Ablation) code under the severe accident analysis platform called SAFARI (Safety Analysis Code For Severe Accident Risk Identification) currently being developed in Korea. PUMBAA built upon CORQUECH developed by Argonne National Laboratory in US, as a reference code focuses on the MCCI phenomena, aiming to extend its ability to analyze MCCIs phenomena interlaced with Molten Corium-Water Interactions (MCWIs) and Containment Thermal Hydraulics (CTHs) along with Severe Accident Management (SAM) actions during severe accidents. In this study, the development and performance of PUMBAA’s corium stratification analysis capability are introduced and validated against the MOCKA 3.1 experiment. Part of the MOCKA series, this experiment simulates a dry 2D cylindrical siliceous concrete cavity, where continuous corium pouring induces various physical phenomena. The paper first describes the modeling of corium stratification in PUMBAA and key physical phenomena in the MOCKA 3.1 experiment. It then presents the validation process and result analysis. The results show that the PUMBAA code, through its corium stratification analysis, effectively accounted for the physical phenomena of the stratified oxide and metal layers observed in the MOCKA 3.1 experiment, accurately predicting the trends in the experimental data.

The melt jet breakup is an important phenomenon for assessment of the ex-vessel phase severe accident, which is highly related to the debris bed coolability. The violent two-phase boiling is accompanied by the melt jet breakup phenomenon due to the high temperature of the melt, and it can affect the jet breakup behavior. The effect of the vapor generation intensity, which represents the two-phase mixing zone behavior, was investigated by controlling both the melt and water temperatures. The melt jet breakup length was observed by visualization using high speed cameras. Based on the experimental observations, the effect of the vapor generation intensity was confirmed. As the vapor generation intensity increases, the jet breakup length became longer. Therefore, the parameter for the vapor generation intensity was suggested to develop the new jet breakup length correlation including the vapor generation intensity parameter so that the existing correlations (Saito correlation and Epstein & Fauske correlation) could be integrated into single correlation.

Hualong One is the first independently developed million-kilowatt-class pressurized water reactor nuclear power plant in China, which meets the design standards of the third generation nuclear power technology. The Cavity Injection and Cooling System (CIS) for the reactor vessel is one of the measures to mitigate severe accidents in Hualong One, and its unique combination of active and passive technologies can effectively prevent the rupture of the reactor pressure vessel and achieve the retention of molten debris within the reactor. Based on the layout features of the CIS system in Hualong One and the design experience of the first unit, this paper proposes design optimization solutions and improvement measures from the perspective of layout design for subsequent projects, improving the compact arrangement of pumps and valves in the plant, and providing valuable design experience for future PWR nuclear power plant design in China.

In a postulated station blackout accident, the fuel channels inside of a Canada Deuterium Uranium (CANDU) reactor would dry out, heat up, and collapse to the calandria vessel bottom. Due to decay heat generation, the debris bed would continue to heat, compact, and melt, forming a molten corium pool. Here, the transient heating and melting of a compacted debris bed is simulated using unsteady Reyolds-averaged Navier-Stokes based computational fluid dynamics.

A time-varying decay heat is used with the starting conditions representing post moderator dry-out. A source-based enthalpy-porosity phase change model is employed to capture the non-eutectic melting process, accounting for a 500K difference between the solidus and liquidus temperatures of the corium. The developing molten region, characterized by a maximum modified Rayleigh number around 10¹², is modelled with the k-ω turbulence model. Turbulence is allowed to develop from an imposed very low level in the melting region consisting of a growing liquid pocket and a partially molten layer, as the temperature locally exceeds the solidus temperature.

Heat flow through the vessel wall to the surrounding water is modelled with a conjugate boundary, and a convection-radiation boundary is applied to the corium top surface. Verification and validation cases are done based on previous studies using molten-salt corium simulants.

The evolution of the molten corium velocity and temperature fields, the unmolten crust thickness, as well as their impact on the exiting heat flux are presented and analyzed. These findings assist the in-vessel retention studies of CANDU reactors and inform future modelling efforts.

Hydrogen distribution in the reactor building of Fukushima Daiichi NPP Unit 3 during the accident was analyzed using the CFD code BAROC. The main feature of BAROC is that it solves the pressure Poisson equation based on energy conservation. This makes the calculation stable and fast, even under sudden changes of fluid conditions. The total number of spatial meshes with 50 cm cubic was approximately 750,000. Transient from the hydrogen inflow to 18 hours later were analyzed in 14 parallel using an Intel Gold 5218 2.3GHz CPU, and the computational time was almost real time. Initial conditions were assumed to be that the building was filled with air at room temperature and atmospheric pressure. Hydrogen was assumed to have entered the building from a shield plug on the 5^th top floor of the building. The analysis assumed that 75 tons of steam and 650 kg of hydrogen would both enter the building. Although the blowout panel in the 5^th floor was not opened at the accident, it was assumed that there was a certain amount of leakage because the building was not a leak-tight structure. Analyses were performed for 3 cases with 2%, 3.3%, and 6.6% leakage of the blowout area. The analysis showed that the hydrogen concentration in the 5^th floor was within the flammable range regardless of the leak area. It was also found that when the amount of inflow hydrogen was increased to 1,300 kg, the hydrogen concentration was within the detonation range in the 5^th floor.