Corium Pool Research Articles

Towards the accurate numerical prediction of thermal hydraulic phenomena in corium pools

The knowledge of heat transfer in corium pools is one of the important issues for corium retention, as it defines the safety margin for vessel integrity. In this regard, in the 90’s the BALI experimental program was performed at the CEA, France. The principal idea was to create a database regarding the heat transfer distribution at corium pool boundaries for in-vessel and ex-vessel configurations at high internal Rayleigh number (1015 to 1017). One of the tasks within the ongoing IVMR project, part of the HORIZON 2020 program, is to assess the up-to-date CFD turbulence models over a wide range of Rayleigh number for the homogenous pool tests of the BALI experiments. In the present study, the assessment of three different turbulence models is performed for two BALI test cases. These turbulence models include a linear k-ε model, a non-linear Reynolds stress model and an advanced turbulent heat flux model known as the AHFM-NRG. After an extensive and careful assessment it has been found that none of the previous models is able to correctly predict the complex heat transfer phenomena appearing in natural convection flow regimes at such high Rayleigh numbers. Hence, a new model is proposed to deal with a wide range of Rayleigh number flow regimes. Accordingly, the results obtained using this new approach are found to be in a very good agreement with the available experimental data.

Experimental study of transient phenomena in the three-liquid oxidic-metallic corium pool

Non-steady physicochemical phenomena in the three-liquid molten pool of prototypic corium are studied in the context of in-vessel melt retention problem. Experiments are made on the Rasplav-3 test facility within the CORDEB program. Structure of the initial molten pool consists of the surface light melt of molten steel, the intermediate layer of oxidic melt separated from steel melt by the crust; and the bottom layer of heavy metallic melt. It is determined that the three-layer pool structure can stay stable for a certain period of time, but the partitioning of steel and oxidic melt components through the crust brings the possibility of transformation of the three-layer pool to a two-layer structure.

Ablation and thermal stress analysis of RPV vessel under heating by core melt

After a reactor core melts and collapses suddenly, the melting core accumulates on the base surface of the bottom head of a reactor pressure vessel (RPV), causing severe thermal ablation and thermal stress, endangering the safety of the RPV bottom head. In this study, a 1000-MW pressurized water reactor is considered an example to study the heat transfer ablation and thermal stress of an RPV lower head after a core collapses, by performing numerical simulations. A two-dimensional (2D) heat transfer model is used to analyze the coupling heat transfer between the wall surface of the RPV, two-layer melting corium pool, and outer water chamber. The transient 2D temperature and ablation of the lower head wall surface are calculated. The thermal stress of the RPV lower head and deformation are also investigated using ANSYS software. The results show that (1) The upper crust is the least thick, with a thickness of approximately 0.01 m, whereas the side crust is the thickest at approximately 0.12 m at the base of the lower head. (2) The minimum thickness of the bottom wall decreases linearly with time, starting from the collapse time of the core to 2500 s, when it becomes 0.04 m. It does not change thereafter, but the melting zone is further expanded. (3) The lower head wall starts to melt from 200 s after the core collapses. The melting mass first increases sharply, and subsequently, increases slightly with time. The total melted material is 3000 kg at 5000 s. (4) The heat flux at the inner and outer surfaces of the lower head immersed in the uranium melt layer increases with the azimuth angle, reaching a maximum value of 750 kw/m2 and 250 kw/m2, respectively, at the interface with the metal melt. The heat flux at the RPV inner wall covered by the metal melt is approximately constant at 400 kw/m2, whereas that at the outer surface decreases with the azimuth angle. (5) The stress at the lower head is concentrated at the inner surface, with a maximum value of 625.65 MPa. The radial deformation increases with time only until 2200 s, with a maximum deformation of 28.39 mm occurring in the lower part of the RPV bottom.

Features of heat and deformation behavior of a VVER-600 reactor pressure vessel under conditions of inverse stratification of corium pool and worsened external vessel cooling during the severe accident. Part 2. Creep deformation and failure of the reactor pressure vessel

In the present manuscript the features of deformation of medium-power VVER-600 reactor pressure vessel (RPV) during a hypothetical severe accident (SA) in the case of worsened regimes of external vessel cooling are considered and analyzed. The analysis of thermal and deformation behavior of the RPV was performed for thermal loads on the reactor vessel that were defined before. The thermal loads acting on the reactor vessel from the corium melt were defined for two types of corium melt structures. A two-layer stratified melt, when a less dense layer of molten steel is located over the oxide heat-generating phase of the corium, was considered as the first structure of the molten pool. The inverse three-layer corium melt was considered as the second type of the melt structure. The numerical simulation of heating and creep deformation processes of the reactor vessel was performed with allowance for creep effect and failure of the RPV while varying both the value of in-vessel overpressure (from 0.2 to 0.8 MPa) and the conditions of external cooling of the RPV in the SA. The external cooling of the RPV were simulated by giving corresponding values of heat transfer coefficients (HTC) on the external surface of the vessel wall. The values of HTCs on the external surfaces of both cylindrical part of the vessel and the vessel bottom varied from 350 to 900 W/(m2K). The performed analysis resulted in the fact that under the conditions of worsened external cooling of the RPV in the SA, significant deformations of the RPV structure are observed. Particularly, during the SA the vertical displacement of the RPV lower head (LH) until the moment of its failure may attain 500 mm and more. Such considerable creep deformations of the reactor vessel structure are observed in the case of forming the inverse structure of the corium pool. Here, the LH deformations make up the bulk of total RPV deformation. Such a model of the vessel deformation may lead to partial or complete blockage of cooling gaps used for external cooling of the reactor vessel in the SA. In turn, the disturbance of the regime of external reactor vessel cooling may cause its overheating and premature RPV failure. The dependences of RPV failure time and the value of vertical displacement of the RPV structure on the value of in-vessel overpressure were obtained for the considered group of SA scenarios.

Effect of temperature gradient on chemical element partitioning in corium pool during in-vessel retention

The paper presents some results of the ISTC (International Science and Technology Center)-financed project 'Investigation of Corium Melt Interaction with NPP Reactor Vessel Steel' (METCOR). In the METCOR experiments the metallic phase of a two-liquid system was produced by the interaction between hot suboxidized corium and cooled VVER vessel steel, with the steel being corroded.Models of corrosion mechanisms in the considered conditions are used to systematize data on the limiting temperature of corrosion/(dissolution) of the vessel steel. A considerable influence of thermal gradient conditions is shown, which has to be taken into account in the analysis of molten pool behaviour.

Features of heat and deformation behavior of a VVER-600 reactor pressure vessel under conditions of inverse stratification of corium pool and worsened external vessel cooling during the severe accident. Part 1. The effect of the inverse melt stratification and in-vessel top cooling of corium pool on the thermal loads acting on VVER-600’s reactor pressure vessel during a severe accident

The problems touched on in this work are closely associated with the realization of in-vessel melt retention strategy through the external reactor vessel cooling and cooling of the molten corium pool inside the medium- power reactor VVER-600 (thermal power is ∼1600 MW) in the course of the SA. The general objective of the research was to determine a thermal state in two- and inverse three-layer molten corium pools, which can be formed in the reactor vessel during the SA. The second task was to estimate the efficiency of the top water flooding of corium pool for its cooling in SA by comparing the new results with those obtained in the previous investigation of the authors. Compositions and mass of the corium pools for the two-layer and inverse three-layer pool structures are analyzed and presented in the paper. Simulation of heat transfer in the molten pool was performed for time values 10, 24 and 72 h from the initiating event (IE) in the SA. To estimate the influence of decay heat generation in the bottom metallic layer of the inverse molten pool on the thermal state of molten pool, a series of model SA scenarios was considered in the work. To simplify the simulations the computation domain was bounded by only the pool with taking corresponding boundary conditions. Simulations of thermal state of the molten pool were carried out by means of the NARAL/FEM computer code in which the turbulent convection at the high-Rayleigh numbers was used through the use of the effective heat conduction properties of the corium materials. The numerical results obtained for two-layer corium pool brought out a series of features: (a) the top water flooding of the melt pool resulted in temperature decrease by ∼150 K only in the upper melt steel layer and had no effect on an essential temperature change in the oxide phase of the corium; (b) top flooding of the corium results in an essential decrease (by more than 40%) of maximal values of heat flux acting on the reactor vessel in the region of contact of the vessel wall with steel melt layer. Thus, the top water flooding of the pool surface yields an essential drop of the heat flux peak acting on the vessel wall from 1.65 to ∼1.2 MW/m2 in case at 24 h after IE; (c) the heat flux peaks acting on the vessel decrease from ∼1.65 MW/m2 (at 24 h after IE) to ∼1.15 MW/m2 (at 72 h) in case when the top flooding of the corium pool is absent, and decrease from 1.2 (24 h) to ∼0.6 MW/m2 (at 72 h) when using the top flooding. In the case of the inverse corium pool, the top water flooding essentially decreases (by more than 50%) the maximal value of heat flux in upper layer of steel melt; (d) in the case of melt inversion and redistribution of total decay heat generation in the corium pool between oxide and bottom metallic layers of the pool (parameter KOxide = QOxiide/(QOxiide + QBot_Me), the dependence of maximal values of thermal load on the lateral surface of the pool depending on KOxide value is observed. Thus, the increase of power of heat generation in the bottom metallic layer of the melt from 0.2 to 0.45 (the decrease of. KOxide from 0.8 to 0.55) causes the increase of heat flux value in the bottom layer by ∼1.5 times. Taking into account the fact that in this region of RPV lower head the CHF has low values (∼0.3…0.45 MW/m2), the probability of superheat and premature failure of the vessel bottom in this field increases. The maximal values of heat flux in the oxide phase and bottom heavy metal layer of the pool are observed near the boundary separating these layers. In this region of the VVER lower head, the heat flux attains the values that may exceed the corresponding values of CHF. Because of this, there is a high probability of superheat and the reactor vessel premature failure due to worsened heat transfer and cooling conditions on the external surface of the vessel wall. This fact should be necessarily taken into account when acting on the RPV lower head the thermal loads of moderate intensity (∼0.5…0.7 MW/m2) that may cause superheat and premature failure of the reactor vessel in the case of inverse molten pool formation during the SA in VVERs.

Large eddy simulation on turbulent heat transfer in reactor vessel lower head corium pools

Due to the high-Rayleigh-number natural convection and strong turbulence in the volumetrically heated corium melt pools, it is difficult to capture the heat transfer characteristics in detail. The Large Eddy Simulation (LES) methods are more and more applicable to nuclear industry and have become powerful tools to analyze the complicated turbulent and multi-phase flows. In this paper, the Wall-Modeled LES (WMLES) method was employed for the simulation of three typical corium pool heat transfer experiments: BALI, LIVE and COPRA experiments. The melt pool temperature and heat flux from numerical simulations were in good comparison with experimental data. Then numerical simulation for the transient formation of two-layer corium pool was performed with prototypical corium to obtain the heat transfer data in reactor situation. There was a distinctive gap between the temperatures in two layers. The heat flux showed a decrease at the two-layer interface and a huge jump at the metal layer side. The thin metal layer on the top at the early stage of the two-layer formation will threaten the vessel integrity.

On the treatment of plane fusion front in lumped parameter thermal models with convection

Within the framework of lumped parameter models for integral codes, this paper focuses on the modeling of a two-phase Stefan fusion problem with natural convection in the liquid phase. In particular, this specific Stefan problem is of interest when studying corium pool behavior in the framework of light water reactor severe accident analysis. The objective of this research is to analyze the applicability of different approximations related to the modeling of the solid phase in terms of boundary heat flux closure relations. Three different approximations are considered: a quadratic profile based model, a model where a parameter controls the power partitioning at the interface and the steady state conduction assumption. These models are compared with an accurate front-tracking solution of this plane fusion front problem. This “reference” is obtained by combining the same integral conservation equations as the approximate models with a mesh-based solution of the 1D heat equation. Numerical results are discussed for a typical configuration of interest for corium pool analysis. Different fusion transients (constructed from nondimensionalization considerations in terms of Biot and Stefan numbers) are used in order to highlight the potential and limitations of the different approximations.

COMPUTATIONAL FLUID DYNAMICS STUDY OF A CORIUM POOL IN A CANDU CALANDRIA

A self-heating pool of corium may form at the bottom of a CANDU calandria during a postulated severe accident. Computational fluid dynamics has been used to model natural convection in such a pool. The split in heat flow to the top, end shield, and bottom surfaces of the pool and the angular distribution of heat flux toward the main shell were calculated. The results indicate that about half of the heat flow occurs through the pool’s top surface. The influence of corium thermophysical properties was calculated to be minor.

Modeling of Water Ingression Mechanism for Corium Cooling with MELCOR

The water ingression mechanism can enhance the coolability of a pool of molten corium in containment during a severe accident. A water ingression model was added to the MELCOR code in 2015. The purpose of this work was to test the new model. It was found that the water ingression model performed satisfactorily in core-concrete–interaction experiments in which gas bubbles were released to the melt from decomposing concrete. The new model had little effect in the Small-Scale Water Ingression and Crust Strength (SSWICS) experiments that were done without gas bubbling through the melt. When applied to the Fukushima Daiichi Unit 1 accident, the water ingression model slowed down concrete ablation by 19% but did not quench the melt. Because the water ingression model was added to MELCOR so recently, the default treatment is still to use multipliers for the boiling heat transfer coefficient and thermal conductivity instead of the proper water ingression model. These default parameters significantly overestimated melt coolability in all the experiments that were calculated.

A simplified geometrical model for transient corium propagation in core for LWR with heavy reflector

In the context of the simulation of the Severe Accidents (SA) in Light Water Reactors (LWR), we are interested on the in-core corium pool propagation transient in order to evaluate the corium relocation in the vessel lower head. The goal is to characterize the corium and debris flows from the core to accurately evaluate the corium pool propagation transient in the lower head and so the associated risk of vessel failure. In the case of LWR with heavy reflector, to evaluate the corium relocation into the lower head, we have to study the risk associated with focusing effect and the possibility to stabilize laterally the corium in core with a flooded down-comer. It is necessary to characterize the core degradation and the stratification of the corium pool that is formed in core. We assume that the core degradation until the corium pool formation and the corium pool propagation could be modeled separately. In this document, we present a simplified geometrical model (0D model) for the in-core corium propagation transient. A degraded core with a formed corium pool is used as an initial state. This state can be obtained from a simulation computed with an integral code. This model does not use a grid for the core as integral codes do. Geometrical shapes and 0D models are associated with the corium pool and the other components of the degraded core (debris, heavy reflector, core plate…). During the transient, these shapes evolve taking into account the thermal and stratification behavior of the corium pool and the melting of the core surrounding components. Some results corresponding to the corium pool propagation in core transients obtained with this model on a LWR with a heavy reflector are given and compared to grid approach of the integral codes MAAP4.

Simulation of the in-vessel retention device heat-removal capability of AP-1000 during a core meltdown accident

Westinghouse Electric Company’s design for the Advanced Pressurized Water Reactor (AP-1000) incorporates a special in-vessel retention (IVR) device to enhance the capability of removing heat through the outer wall of the reactor vessel during a core meltdown. External Reactor Vessel Cooling (ERVC) is established by flooding the reactor cavity during a severe accident. The amount of heat removed is limited by the critical heat flux (CHF) at the outer surface of the vessel wall. In this paper, the CHF correlations developed specifically for this purpose are reviewed. The RELAP5-3D thermal-hydraulic code is used to simulate the natural convection flow within the water channel of the IVR device. The results of RELAP5-3D simulations are substituted into ULPU, SULTAN, SBLB, and KAIST critical heat flux correlations to determine the safety margin of the AP-1000 IVR design. Among these four correlations, the CHFs predicted by the ULPU and SBLB correlations are insensitive to the thermal-hydraulic conditions of the coolant channel. The CHF predicted by SULTAN is the lowest among the four correlations. The results demonstrate that the predicted CHF increases with the angle. Nevertheless, the ratio of CHF to heat flux is lowest around the upper quarter of the vessel. The heat load to the lower head is based on two bounding melt configurations. Configuration I involves a stratified light metallic layer on top of a molten ceramic pool, and Configuration II represents the condition that an additional heavy metal layer forms below the ceramic pool. The impact of corium pool configuration on the safety performance of IVR is very significant. Under Configuration II, as predicted by the SULTAN and KAIST correlations, the performance of the current AP-1000 IVR design is marginal. The results of the RELAP5-3D IVR simulation show that the total mass flow rate in the coolant channel increases sharply as the widths of the annular channels vary from 0.0762 to 0.762m. The mass-flow rate decreases gradually when the gap width increases beyond 0.762m. The KAIST CHF correlation predicts that CHF is maximal when the gap width is around 0.4572m and decreases gradually when the gap width increases.

Two-Dimensional Axisymmetric Thermostructural Analysis of APR1400 Reactor Vessel Lower Head for Full-Core Meltdown Accident

In severe accident conditions, the molten core material forms an internally heated debris bed and eventually becomes a molten pool of corium, which will cause or induce thermal and mechanical loads to the reactor vessel lower head (RVLH) resulting in penetrations leading to failure. A good understanding of the mechanical behavior of the RVLH is essential for estimating structural integrity and improving accident mitigation strategies.Coupled thermomechanical analysis using ANSYS, a general-purpose finite element method analysis code, was used to evaluate the possibility and timescale of failure. A two-dimensional axisymmetric finite element model was adopted based on APR1400 design data with relevant material properties including creep data.From the study, it was found that the possibility of plastic and creep failure of the RVLH for the APR1400 was considerably low for a full-core meltdown of the reactor core under ex-vessel cooling conditions with an ambient temperature of 130°C and constant decay heat from the corium, but the lower head may fail unless the increased internal pressure can be reduced on time. Plastic failure can be a major cause of lower head failure of a reactor vessel in high internal pressure conditions and creep failure is not negligible, since failure mechanisms under long-lasting periods are considered. This study found that the APR1400 RVLH failure time is around 220 h using 15% creep strain failure criteria from the postulated accident condition.

FISSION-PRODUCT RELEASES FROM A PHWR TERMINAL DEBRIS BED

During an unmitigated severe accident in a pressurized heavy water reactor (PHWR) with horizontal fuel channels, the core may disassemble and relocate to the bottom of the calandria vessel. The resulting heterogeneous in-vessel terminal debris bed (TDB) would likely be quenched by any remaining moderator, and some of the decay heat would be conducted through the calandria vessel shell to the surrounding reactor vault or shield tank water. As the moderator boiled off, the solid debris bed would transform into a more homogeneous molten corium pool located between top and bottom crusts. Until recently, the severe accident code MAAP-CANDU assumed that unreleased volatile and semi-volatile fission products remained in the TDB until after calandria vessel failure, due to low diffusivity through the top crust and the lack of gases or steam to flush released fission products from the debris. However, national and international experimental results indicate this assumption is unlikely; instead, high- and medium-volatility fission products would be released from a molten debris pool, and their volatility and transport should be taken into account in TDB modelling. The resulting change in the distribution of fission products within the reactor and containment, and the associated decay heat, can have significant effects upon the progression of the accident and fission-product releases to the environment. This article describes a postulated PHWR severe accident progression to generate a TDB and the effects of fission-product releases from the terminal debris, using the simple release model in the MAAP-CANDU severe accident code. It also provides insights from various experimental programs related to fission-product releases from core debris, and their applicability to the MAAP-CANDU TDB model.

Experimental study on influence of crust formation on heat transfer characteristic in corium pool

Experimental study on influence of crust formation on heat transfer characteristic in corium pool

Modelling of liquid phase segregation in the Uranium–Oxygen binary system

In the uranium–oxygen (U–O) binary system, the existence of a miscibility gap induces the segregation between two immiscible liquid phases at thermodynamic equilibrium: an oxidic phase and a metallic one. Within the framework of severe accidents in a Light Water Reactor (LWR), the knowledge of in-vessel corium (U–O–Zr-steel system) behaviour and the associated vessel failure risk are of prime interest. The corium pool configuration plays a key role on the heat flux distribution and therefore on the vessel integrity. A coupled thermochemistry-thermalhydraulics model is needed for investigating the stratification kinetics of a corium pool. The work presented here is a first step in the development of such a model, and is focused on the modelling of the kinetics of liquid phase segregation in the U–O binary system. With this aim, we have developed a Cahn–Hilliard model, derived from a free energy functional, for investigating the liquid phase segregation. The present paper discusses the coupling of the transient Cahn–Hilliard solver with a U–O CALPHAD thermodynamic database and the parameterization of such a kinetic model.

The COPRA experiments on the in-vessel melt pool behavior in the RPV lower head

The COPRA (COrium Pool Research Apparatus) facility was designed to investigate the in-vessel molten corium pool behavior for the in-vessel corium retention during severe accidents in Chinese large-scale advanced PWRs. The main part of the COPRA facility is a two-dimensional 1/4 circular slice test section with an inner radius of 2.2m to simulate the lower plenum of the reactor vessel at 1:1 scale for the Chinese large-scale advanced PWR. Up to now, the COPRA experiments have been performed in different melt volumes, heat generation rates and multiple injection of the melt in the one test. At present, the well-known simulant material NaNO3–KNO3 in non-eutectic composition (20mol% NaNO3–80mol% KNO3) is used. This work presents the behavior of a large-scale homogenous melt pool regarding the melt pool temperature, heat flux distribution through the vessel wall and crust thickness in transient and steady state conditions. And the heat transfer towards the vessel wall of the COPRA test was compared with previous experiments and experimental correlations. It showed that the downward heat transfer from the COPRA experiments was lower than those from other experiments within the same range of Rayleigh numbers (1015–1017).

COPRA: A large scale experiment on natural convection heat transfer in corium pools with internal heating

During a severe accident in light water reactors, the core may melt and relocate into the lower plenum of the reactor pressure vessel. The decay heat will threaten the structural and thermal integrity of the reactor vessel if there is no effective cooling mechanism. Natural convection plays an important role in determining the thermal-hydraulic behaviors inside the debris pool, which is directly relevant to the problem of retention of molten corium inside the lower plenum. The COPRA (COrium Pool Research Apparatus) experiments were performed to study the natural convection heat transfer behavior in an internally heated melt pool with high Rayleigh numbers. The COPRA test facility is a two-dimensional 1/4 circular slice vessel with an inner radius of 2.2 m to simulate the lower plenum of reactor vessel for the Chinese advanced PWR in a full scale. The inner width of the slice is 20 cm and the curved vessel wall has a thickness of 30 mm. 20 electrical heating rods, each with a diameter of 16 mm but different lengths according to their locations, are uniformly distributed in the vessel to simulate homogenous internal decay heat. They can provide a maximum of 30 kW power to the melt pool. The outside of the curved wall is enclosed with a regulated external cooling path to keep the boundary temperature nearly isothermal. The top surface of the pool can be maintained insulated with an adiabatic lid. 79 thermocouples are installed in the melt pool to measure the melt pool temperature field and 48 in the curved wall to obtain local heat flux distribution along the curved wall. In the first series of experiments, water was employed as the simulant material. Due to the full scale geometry, the Rayleigh number within the pool could reach up to 1016, matching that in the prototypical situation for PWR. The heat transfer characteristic with volumetrically internal heat was investigated by using the full scale facility COPRA. Relations of pool temperature and heat flux distribution, as well as Nudn-Ra′ were developed. The results have been compared with the results and correlations from other experiments.

COPRA experiments on natural convection heat transfer in a volumetrically heated slice pool with high Rayleigh numbers

Large scale COPRA experiments were performed to study the natural convection heat transfer in corium pools inside the reactor pressure vessel lower plenum during severe accidents. The test facility is designed to simulate the lower plenum of reactor vessel at 1:1 scale for the Chinese advanced PWR, which is a two-dimensional 1/4 circular slice structure with an inner radius of 2.2m. A non-eutectic binary mixture of 20mol%NaNO3–80mol%KNO3 compositions was selected as the simulant material. Twenty custom-designed heating rods were located in the facility to provide homogenous internal heating. Due to the full scale geometry, The Rayleigh numbers within the corium pools could reach to 1.188×1015–1.784×1016, matching those in the prototypical situation during severe accident. The pool temperature field and heat flux distribution, as well as the crust thickness distribution from tests with different pool heights and heating powers were obtained and compared in the experiments. Influences of the relocation position and upper cooling on the heat transfer characteristics were also studied in the experiments. Relations of pool temperature and heat flux distribution, as well as Nudn–Ra′ relation were developed. Comparison with previous experiments showed that the downward heat transfer Nudn prediction from COPRA experiments were in good agreement with SIMECO and LIVE salt experimental results.

Transient stratification modelling of a corium pool in a LWR vessel lower head

In the context of light water reactor severe accidents analysis, this paper is focused on one key parameter of in-vessel corium phenomenology: the immiscible phases stratification and its impact on the heat flux distribution at the corium pool lateral boundary with the so-called focusing effect related to a “thin” top metal phase and the potential vessel failure at that point. More particularly, based on the limited knowledge of the stratification transient phenomenon derived from the MASCA-RCW experiment, a basic model is proposed that can be used for corium in lower head sensitivity analyses. It has been implemented in the PROCOR platform developed at CEA Cadarache. A short parametric study on a simple hypothetical transient is presented in order to highlight the different focusing effect “modes” that can be encountered based on this in-vessel corium pool model. An early mode may occur during the formation of the top metal layer while two other modes may appear later during the thinning of this top metal layer because of thermochemically induced mass transfers. Some associated relevant parameters (model or scenario-dependent) and modelling issues are mentioned and illustrated with some results of a Monte-Carlo based sensitivity calculation on the transient behaviour of the corium in the lower head of a 1650MWe GenIII PWR. Within the limiting modelling hypotheses, the thermal modelling of the steel layer for small (centimetre) heights and the mass diffusivity (limited in this case to the uranium diffusivity in the oxidic layer) are main sensitive parameters.