Severe Accident Code Research Articles

Degradation mechanism of stainless steel by U-Zr-O molten mixture during core degradation of BWR severe accident

ABSTRACT Degradation mechanism involving the interaction between the U-Zr-O phase from molten fuel cladding and steel was investigated at the lower structure of core region of the boiling water reactor, which could take place at low temperatures compared to the failure conditions typically assumed in the current severe accident codes. The non-isothermal interaction was simulated by the heat and mass transport equations with possible chemical reactions estimated by thermodynamic database and the following potential mechanisms were identified: (1) the U-Zr-O melt dissolves steel leading to serious damage, (2) the reaction results in partial solidification of U-Zr-O melt near the reaction surface forming (U, Zr)O2-x phase and U-Fe-rich metallic phase and (3) unsolidified U-Zr-O melt could drain to the lower plenum through the lower structure of core support plate. These mechanisms will propose new hypothesis in which the timing of core support plate degradation could get earlier due to liquefaction of U-Fe-rich metallic phase or the timing of lower head failure of the reactor pressure vessel could be affected by drained U-Zr-O melt.

Improvement of integrative severe accident analysis code, SPECTRA

Improvement of integrative severe accident analysis code, SPECTRA

Development of the sodium fire module in integrated severe accident analysis code, SPECTRA

Development of the sodium fire module in integrated severe accident analysis code, SPECTRA

Application of a severe accident code to the sensitivity and uncertainty analysis of hydrogen production in the WWER1000/V446

An uncertainty analysis based on the Latin Hypercube Sampling (LHS) method and MELCOR code for hydrogen production during a SA in WWER1000 nuclear reactor is presented. Firstly, the total amount of the produced hydrogen during the SBO with Large Break Loss of Coolant Accident (SBO-LBLOCA), SBO with Small Break Loss of Coolant Accident (SBO-SBLOCA) and Station Black Out (SBO) is estimated by the MELCOR code and benchmarked with the RELAP-SCDAP. In addition, the uncertainty of the effective parameters on the hydrogen generation is analyzed during both in-vessel and ex-vessel phases of the SA. The results showed that the SBO produces more hydrogen compared to other accident scenarios. Also, the range of the produced hydrogen varies between 255 and 525 kg and 812–1820 kg during the in-vessel and ex-vessel phases of the SBO, respectively. The falling debris heat transfer coefficient and Zr melt release temperature have the greatest effect on the hydrogen generation.

Modelling of fission products release in VERDON-1 experiment with cGEMS: Coupling of severe accident code MELCOR with GEMS thermodynamic modelling package

The treatment of chemistry and thermodynamics in the integral severe accident codes is typically limited. A more accurate treatment of the chemistry during the severe accident modelling is, therefore, of great interest. For this purpose, the work is focused on the development of coupling of the MELCOR code with chemical thermodynamic calculations using GEMS codes and HERACLES database. Developed coupling between the two codes, called cGEMS, allowed for the improved thermodynamic description of the fission product release from the nuclear fuel under severe accident conditions.VERDON-1 test was selected as an experimental reference for the simulations. Experimental release behaviour of Mo, Cs and Ba observed in VERDON-1 experiment was reproduced by the developed coupled code. Performed simulations provided detailed information about the fission product speciation at different redox conditions.The obtained information and the developed code provides a more accurate description of the fission product behaviour and release during severe accidents.

Preliminary validation of ASTEC V2.2.b with the QUENCH-20 BWR bundle experiment

ASTEC is an integral severe accident code which aims to simulate entire severe accident transients of the nuclear reactors. New BWR-related models are implemented in the new versions of European ASTEC integral code by IRSN for BWR-typical core structures such as absorber blades, canister, etc. during the In-vessel core degradation after a severe accident in a BWR. The goal is to improve the simulation of the whole severe accident sequence in a BWR, such as the Fukushima accidents. An important step is the validation of the new models (oxidation, molten material formation and melt relocation, etc.) before the code is used to simulate severe accident scenarios in BWRs. The QUENCH-20 BWR test was specifically performed to investigate the oxidation, hydrogen generation, melt formation and oxidation during the reflooding of the test section representing a quarter of a SVEA-96 Optima-2 fuel assembly. The data measured during the QUENCH-20 BWR test e.g. temperature of the fuel rod simulators, of the blade and of the canister as well as the hydrogen production, quench front movement is used for the validation of the code’s prediction capability. In this work, the code ASTEC V.2.2.b is used to simulate the behavior of the QUENCH-20 test section. For this purpose, a detailed model of the QUENCH facility and test section was developed. This model was used to analyze the severe accident phenomena taking into account the same boundary conditions i.e. electrical power, mass flow rates, pressure, etc. as in the test. It was found out that the evolution of the temperature distribution in the bundle is reasonably evaluated by ASTEC for the entire duration of the transient. Besides, oxidation processes are also well reproduced. ASTEC predicts a total hydrogen generation of about 53 g while the measured value is around 57 g. In addition, ASTEC rather well predicted that about 15% of this hydrogen amount comes from the B4C-oxidation. Such good overall agreement between ASTEC and experimental measurements leads to further validation possibilities on BWR plant applications.

Overview and outcomes of the OECD/NEA benchmark study of the accident at the Fukushima Daiichi NPS (BSAF), phase 2 – Results of severe accident analyses for unit 2

This is the second paper in a series of 3 in which results of severe accident analyses for Unit 2 of Fukushima Daiichi are presented, gained in phase 2 of the OECD/NEA project “Benchmark Study of the Accident at the Fukushima Daiichi Nuclear Power Plant (BSAF)”. Nine organizations of six countries (CNL Canada; CRIEPI, IAE, JAEA and NRA Japan; IRSN France; KAERI Korea; NRC/DOE/SNL U.S. and VTT Finland) submitted results of their calculated severe accident scenarios for Unit 2 of Fukushima Daiichi using different severe accident codes (ASTEC, MAAP, MELCOR, SAMPSON, THALES-KICHE).The present paper describes the findings of the comparison of the participants’ results for Unit 2 against each other and against plant data, the evaluation of the accident progression and the final status inside the reactors. Special focus is on reactor pressure vessel status, melt release and fission product behavior and release. Unit 2 specific aspects will be highlighted and points of consensus as well as remaining uncertainties and data needs will be summarized.

Overview and outcomes of the OECD/NEA benchmark study of the accident at the Fukushima Daiichi NPS (BSAF) Phase 2 – Results of severe accident analyses for Unit 1

Phase 2 of the OECD/NEA Project “Benchmark Study of the Accident at the Fukushima Daiichi Nuclear Power Plant (BSAF)” was established in mid-2015. The objectives have been similar to Phase 1 of the project but with an extended analysis period of 3 weeks from the occurrence of the earthquake, a major focus on fission product (FP) behaviour and releases to the environment and the comparison to various data including radiological data and results of backwards calculations of the source term. Nine organizations of six countries (Ciemat Spain; IAE, JAEA and NRA Japan; CEA, IRSN France; IBRAE Russia; KAERI Korea; NRC/DOE/SNL U.S.A.; VTT Finland) submitted results of their calculated severe accident scenarios for Unit 1 at the Fukushima Daiichi site using different severe accident codes (ASTEC, MAAP, MELCOR, SAMPSON, SOCRAT, THALES-KICHE). This paper describes the findings of the comparison of the participants results for Unit 1 against each other and against plant data, the evaluation of the accident progression and the final status inside the reactors. Special focus is on RPV status, melt release and FP behaviour and release. Unit specific aspects will be highlighted and points of consensus as well as remaining uncertainties and data needs will be summarised. The results for Units 2 and 3 are presented next in separated papers.

Finite Element Model of a Melting Terminal Debris Bed in a CANDU Calandria

Abstract Melting of the terminal debris bed at the bottom of a Canada deuterium uranium (CANDU) 6 calandria vessel during a severe accident is studied using a three-dimensional transient finite element analysis. Settling of the debris is modeled, as are the effects of changing temperature and porosity on debris material properties. The time and magnitude of maximum heat rejection from the debris bed, and the maximum mass of molten debris, are calculated as functions of the accident timing. The results are compared with those from the modular accident analysis program (MAAP)-CANDU severe accident code.

Numerical simulation of CS28-1 experiment by using CANDU severe accident analysis code, CAISER

Due to the demands for an accurate and realistic severe accident code in the PHWR safety analysis, a new severe accident code called CAISER is being developed at Korea Atomic Energy Research Institute (KAERI). The CAISER code simulates core degradation phenomena occurring in a calandria tank, and it consists of two main modules: A fuel rod degradation module and a fuel channel degradation module. The fuel rod degradation module simulates the severe accident phenomena in a fuel channel, which includes a core uncovery, fuel rods heatup, hydrogen generation due to steam-Zr oxidation, fuel rods slumping, fuel rods melting and relocation, and thermal interaction of relocated molten mass with a pressure tube or a calandria tube. The fuel channel degradation module simulates the overall severe accident phenomena in a calandria tank, including the sagging of a fuel channel, debris bed formation caused by a fuel channel failure, the molten pool formation, and the calandria tank failure. Each module is tracking mass and energy changes for the main components, which are fuel, cladding, pressure tube, calandria tube in a fuel rod degradation module, and debris bed, metallic and oxidic corium pools in a fuel channel degradation module. These two modules are closely interconnected to simulate phenomena both in a fuel channel and in a calandria tank at the same time. In this paper, the numerical modelling in the CAISER code is described, including the node system and the governing equations for each main component. For the purpose of CAISER code validation, numerical simulations were conducted for the CS28-1 experiment. The simulation results show that the temperature distributions calculated by CAISER are in line with the experimental data.

Experimental study on local fuel–coolant interaction in molten pool with different melts

In a Core Disruptive Accident (CDA) of Sodium-cooled Fast Reactor (SFR), by supposing rather pessimistic condition, a large molten fuel pool would be possibly formed that contains enough fuel with the potential to exceed prompt criticality due to fuel compactive motion. Local fuel-coolant interaction (FCI) in the molten pool is identified as one of the typical inducing factors that may result in such compactive motion. To understand the characteristics of this interaction, owing to the PMCI (Pressurization characteristics in Melt-Coolant Interaction) facility recently established at the Sun Yat-sen University, in our earlier publications two series of simulated experiments, i.e. Coolant-Injection (CI) experiments and Melt-Injection (MI) experiments, have been conducted, in which water and low-melting-point Bi-Sn-In alloy (60% Bi, 20% In and 20% Sn) were used respectively as the simulant materials for sodium and molten fuel. In this study, aimed at checking the generality of the experimental findings as well as achieving further enhanced understanding on this interaction, a large number of experiments are newly performed through delivering water into a simulated molten fuel pool comprised of Lead-Bismuth Eutectic (LBE) alloy (another low-melting-point alloy). Through detailed comparisons (with the CI-mode experiments using Bi-Sn-In alloy), it is found that a similar trend in the transient pressure and temperate history as well as the overall effect of experimental parameters (such as water quantity, melt and water temperature, water-lump shape and melt depth) on the pressurization characteristics can be reproduced for the new LBE experiments. Despite a much lower thermal diffusivity, for the LBE experiments it is confirmed that the most likely reason resulting in the limited pressurization as water volume increases should be also primarily owing to the isolation effect of vapor bubbles generated at the melt-water interface. Possibly due to the enhanced melt density which facilitates the penetration of melt through vapor bubbles, under the same parametric condition the pressurization in LBE experiments is generally much higher than that in Bi-Sn-In experiments. Compared to melt static pressure, the pressurization under different melt depths is supposed to be more influenced by the varied easiness of melt penetration. Current work provides a large amount of favorable experimental database and insight for promoted understanding on actual CDAs and improved verifications of SFR severe accident codes in China.

Three weeks analysis of the Fukushima Daiichi Unit 3 NPP by the SAMPSON code: Contribution to the BSAF-2 project

As a participant in the Benchmark Study of the Accident at Fukushima Daiichi phase 2 project, which aims to improve our understanding of the Fukushima Daiichi (1F) nuclear power plant conditions, the Institute of Applied Energy has performed a comprehensive evaluation and analyses of the 1F units using the SAMPSON code. The objective of this study is to clarify the accident progression and the distribution of debris/fission products in the plant as well as evaluate points of uncertainty in severe accident codes. The paper presents the results obtained for 1F Unit 3. Firstly, a likely accident scenario is evaluated taking into account all of the available measurements and recent information obtained from plant investigations. Subsequently, the SAMPSON code is applied to confirm the validity of the estimated phenomena and explain points of uncertainty in the the accident. The calculation is performed over the course of three weeks following the scram, which is representative of the time it takes to achieve stable plant conditions according to theories of thermal-hydraulics and fission product settling. The analysis confirms that a large portion of the core melted and relocated in the lower plenum of Unit 3. Thereafter, due to lack of cooling, the reactor pressure vessel failed, which caused the debris to be gradually transferred to the pedestal, where it interacted with concrete, generating flammable gases. It is estimated that a small amount of debris may have remained in the lower plenum in the form of a solidified crust, while a considerable portion relocated to the pedestal with limited spread to the dry well due to prolonged top-flooded conditions.

Main outcomes from the IVR code benchmark performed in the European IVMR project

In-Vessel Retention (IVR) of corium is one of the possible strategies for Severe Accident (SA) mitigation. Its main advantage lies in the fact that, by maintaining the corium within the vessel, it preserves the last containment barrier from corium aggression. One of the issues for the demonstration of the success of this strategy is the evaluation of the behaviour of the corium relocated in the lower head and how it stabilizes and affects the integrity of the vessel wall. The first modelling was developed in the nineties and assessed the heat transfers in a stratified corium pool with a top metal layer made of steel and Zirconium only. About 10 years later, the results of the MASCA program highlighted the possibility of having more complex stratified configurations, including a dense metal layer. Addressing thermochemical effects in the stratification makes the modelling of the corium pool in the lower head more difficult and, in addition, knowledge of the associated kinetics is still limited. As a consequence, available SA codes, either integral or dedicated to the lower head, can differ significantly in their models, which leads to discrepancies in the results when evaluating the IVR strategy.In order to identify the main modelling issues and to assess the capabilities of the codes, a benchmark exercise for code validation was made in the scope of the European H2020 project IVMR (In-Vessel Melt Retention). It is based on the definition of different IVR configurations at reactor scale with an increase in the complexity of the phenomena involved: starting from a steady-state stratified pool with metal on top, up to consideration of corium phase separation at thermochemical equilibrium and progressive ablation of metallic structures and vessel wall. As a last step, the more challenging transient configuration, with formation of a metal layer heavier than the oxide followed later on by stratification inversion, was also studied. It should be noted that mechanical resistance of the ablated vessel wall and cooling conditions out of the RPV in the cavity are not evaluated and considered out of the scope of this benchmark focused on the thermal load transferred from the corium pool through the vessel wall.Six organisations took part in this benchmark (CEA, EDF, GRS, IBRAE, IRSN, NRC-KI) and 6 different codes were used (ASTEC, ATHLET-CD, MAAP_EDF, PROCOR, HEFEST_URAN and HEFEST – stand-alone version of the corresponding module of the SOCRAT code). The main results and outcomes obtained are presented and discussed in this paper. Sensitivity studies have also been performed and allow obtaining a more consolidated range of results.Thanks to this benchmark exercise and to the approach followed with a progressive increase of complexity, the capabilities of codes to evaluate the heat flux profile applied to the vessel wall in steady-state are demonstrated. Then, for transient configurations, it is shown that different modelling approaches give rather consistent results since dispersion remains limited between code predictions, even in the most challenging configuration with stratification inversion: ±40% for the minimum vessel thickness (2.5 cm ± 1 cm). In addition, all codes predict that transient effects lead to more vessel ablation than in the final steady-state. Hence, the importance of the consideration of the progressive molten steel incorporation in the pool and of the consideration of thermochemical equilibrium to calculate the oxide and metal phases composition was highlighted in the benchmark.Regarding the dispersion of the results obtained, analysis shows that uncertainties on metal layer properties have a significant impact. In addition, differences in modelling assumptions were identified and discussed. The main issues are related to the following physical phenomena: (i) The interaction of the oxide crust with the molten steel; (ii) The kinetics of stratification inversion; (iii) The heat transfer in thin metal layer.

Modelling, verifications and safety feedback assessment of annular fast reactor fuel pins with severe accident code MITRA

Fuel of fast reactors is designed in the form of cylindrical pellets of solid or annular geometry. The annular configuration provides advantages in terms of higher achievable energy extraction and lower fuel-clad mechanical interaction. Another advantage of this configuration is the hydrodynamic flow of molten fuel inside the annular pellets during accidental meltdown, known as in-pin fuel motion. This motion can provide safety benefits during an unprotected transient overpower accident (UTOPA), if the molten fuel is dispersed significantly away from the core mid-plane. The physics behind the flow is complex. Intricate theoretical modelling and detailed experimental validation are pre-requisites for the reliable estimation of safety benefits. To meet this requirement, a Multi-phase In-pin Thermal hydraulic Relocation Algorithm (MITRA) is developed under the purview of the pre-disassembly analysis code, PREDIS. This paper presents a theoretical foundation of the algorithm followed by experimental verifications, burnup and top blanket design sensitivity analyses and whole core UTOPA simulations. Results show that regardless of the fuel burnup level, the melt tends to agglomerate into a column, slightly below the core mid-plane. Upon further melting, this column grows slowly, devoid of the rapid dispersion associated with fuel squirting. Minor deviations in the melt position arise due to variations in thermal parameters with burnup. UTOPA analysis shows a reduction in the safety feedback and increased melting. Modifying the conventionally solid top axial blanket to annular geometry enhances the safety feedback only after 34.2% melting.

Main Findings, Remaining Uncertainties and Lessons Learned from the OECD/NEA BSAF Project

The Organisation for Economic Co-operation and Development (OECD)/Nuclear Energy Agency (NEA) Benchmark Study of the Accident at the Fukushima Daiichi Nuclear Power Station (BSAF), which started in 2012 and continued until 2018, was one of the earliest responses to the accident at Fukushima Daiichi. The project, divided into two phases, addressed the investigation of the accident at Units 1, 2, and 3 by severe accident (SA) codes until 500 h, focusing on thermal hydraulics, core relocation, molten corium concrete interaction (MCCI), and fission product release and transport. The objectives of the BSAF were to make up plausible scenarios based primarily on SA forensic analysis, support the decommissioning, and inform SA code modeling. The analysis and comparison among the institutes have brought up vital insights regarding the accident progression, identifying periods of core meltdown and relocation and reactor pressure vessel (RPV) and primary containment vessel (PCV) leakage/failure through the comparison of pressure, water level, and containment atmosphere monitoring system (CAMS) signatures. The combination of code results and inspections (muon radiography, PCV inspection) has provided a picture of the current status of the debris distribution and plant status. All units present a large relocation of core materials and all of them present ex-vessel debris with Unit 1 and Unit 3 showing evidence of undergoing MCCI. Uncertainties have been identified, in particular on the time and magnitude of events such as corium relocation in the RPV and into the cavity floor and RPV and PCV rupture events. Main uncertainties resulting from the project are the large and continuous MCCI progression predicted by basically all the SA codes and the leak pathways from the RPV to the PCV and the PCV to the reactor building and environment. The BSAF project represents a pioneering exercise that has set the basis and provided lessons learned not only for code improvement but also for the development of new related projects to investigate in detail further aspects of the Fukushima Daiichi accident.

Evaporation of materials from the molten salt reactor fuel under elevated temperatures

The release of fission product and salt compounds from a molten salt reactor fuel under accident conditions was investigated with coupled computer simulations. The thermodynamic modeling of the salt and fission product mixture was performed in The Gibbs Energy Minimization Software GEMS and the obtained compound vapor pressures were exchanged with the severe accident code MELCOR, where the evaporation from a salt surface located at the bottom of a confinement building was simulated. The fuel salt considered in the simulations was LiF-ThF4-UF4 with fission products Cs and I. The composition of the fuel salt material was obtained from an equilibrium fuel cycle simulation of the salt using the EQL0D routine coupled to the Serpent 2 code. The results were compared to simulations using pure compound vapor pressures in the evaporation simulations. It was observed that by modeling the salt mixing the release of fission products and salt materials was reduced when compared to the pure compound simulations. The mixing effects in the salt, when compared to the pure compound simulation also affected evaporation temperatures and therefore the timing of the release of compounds. In an additional simulation in which the depressurization of the confinement was considered, the total evaporated mass of compounds increased due to increased mass transfer at the salt surface. The simulation process described in this paper can be used for a more comprehensive accident analysis of molten salt reactors once the detailed description of the reactor confinement and accident sequences are available and more fission product elements have been added to the analysis.

The explosions at Fukushima Daiichi Unit 3 and Unit 4 and implications on the evaluation of 1F3 accident

The explosions at Fukushima Daiichi accidents are investigated and discussed to provide a complete picture of the explosion events during the accidents. A ballpark estimate was performed to estimate the total amount of combustible gases in term of hydrogen equivalent that must be generated during the Fukushima Daiichi Unit 3 accident up to the time of the 1F3 explosion. The estimate is based on the available explosion and containment venting information together with reasonable, but highly uncertain, assumptions. About 1450 kg (after roundoff) of hydrogen equivalent would be needed to explain the observed 1F3 fireball explosion, and about 2100 ~ 2400 kg of hydrogen equivalent (from in-vessel oxidation of core components and ex-vessel generation during MCCI) must be generated in 1F3 in order to explain both explosions at 1F3 and 1F4 which involved the migration of combustible gases during venting from 1F3 to 1F4. The estimated amount of required generation of hydrogen equivalent is about 71.3~81.5% of the potential sources. The method for an estimate of combustible gases burned in the 1F3 fireball explosion, which is the largest contributor and the least well understood, is based on extensive hydrocarbon-based fireball explosion database and an empirical correlation that accurately predicts the fireball size of hydrocarbon fuel explosion. This estimate should be used as if it is supplementary plant data (called the “explosion data”) for 1F3 to help guide in the evaluation of severe accident analysis of 1F3. Recent results of the BSAF phase 2 study are compared with this estimate. Interestingly, one of the BASF phase 2 analyses comes very close to this estimate while most analyses significantly underpredict this estimate. Three time windows during the 1F3 accident that allowed a transfer of combustible gases from the drywell to the 5th floor of the reactor building through drywell upper head leakages have been identified to include (1) a 5-hour period during the 2nd water injection when the drywell pressure was higher than 4 bars, (2) a 3-hour period during the 3rd unsuccessful venting, and (3) a 4-hour period during the 4th unsuccessful venting. A discussion of the missing models in severe accident codes is also provided.

Experimental study on eutectic reaction between fuel debris and reactor structure using simulant materials

It is important to ascertain the mechanisms underlying the eutectic reaction between different reactor materials that might be encountered during a Core Disruptive Accident (CDA) of Sodium-cooled Fast Reactors (SFR) since such reactions will affect the accurate progression of a severe accident. In this study, motivated to understand the characteristics of the probable eutectic reaction between fuel debris and the lower head of reactor vessel, a series of simulated experiments has been conducted at the Sun Yat-sen University using a couple of rather lower-eutectic-point (456 K) materials (namely Sn particles and Pb pellet). The experiments were carried out in a self-designed experimental system, which mainly consists of a sample holder and a visible resistance furnace. To acquire a relatively comprehensive understanding, a variety of experimental parameters such as the reaction temperature (463–483 K), contact pressure (0.4–1.2 MPa), Pb pellet diameter (10–25 mm) along with the diameter (0.3–3 mm) and geometry (spherical, cylindrical and droplet-shaped) of the Sn particles have been taken. Through detailed analyses, it is found that the reaction temperature and contact pressure can have noticeable positive impact on the reaction rate. As for the size of Pb pellet and Sn particles, with increasing the diameter ratio of Sn particles to Pb pellet, a non-monotonous effect is observed due to the competing role between the contact area and contact pressure. An evident influence of Sn particle geometry on reaction rate has been verified in accordance with the variation of particle-bed porosity. The analyses in this work also suggest that the reaction rate in previous experiments using block-block samples is generally larger than present experiments using particle-pellet ones, especially at a higher temperature. Knowledge and evidence obtained from this work will be utilized for the design of future high-temperature experiments using actual reactor materials as well as for the improved validations of eutectic-reaction-related models incorporated in fast reactor severe accident codes.

Development of a reduced order model for severe accident analysis codes by singular value decomposition aiming probabilistic safety margin analysis

ABSTRACT Severe accident codes (e.g. MAAP, RELAP, and MELCORE) model various physical phenomena during severe accidents. Many analyses using these codes for safety margin evaluation are impractical due to large computational costs. Surrogate models have an advantage of quickly reproducing multiple results with a low computational cost. In this study, we apply the singular value decomposition to the time-series results of a severe accident code to develop a reduced order modeling (ROM). Using the ROM, the probabilistic safety margin analysis for the station blackout with a total loss of feedwater capabilities at a boiling water reactor is carried out. The dominant parameters to the accident progression are assumed to be the down-time and the recovery-time of the reactor core isolation cooling system, and decay heat. To reduce the number of RELAP5/SCDAPSIM analyses while maintaining the prediction accuracy of ROM, we develop a data sampling method based on adaptive sampling, which selects the new sampling data based on the dissimilarity with the existing training data for ROM. Our ROM can rapidly reproduce the time-series results and can estimate core conditions. By reproducing multiple results by ROM, a time-dependent core damage probability distribution is calculated instead of the direct use of RELAP5/SCDAPSIM.

Thermodynamic approach for determination of fuel relocation condition in severe accident progression

In a severe accident, a rapid temperature increase leads to thinning of the ZrO2 layer and the both of the UO2 dissolution by molten Zr, and relocation could take place within very short period. This paper considered such situation and proposed the relocation condition of “the quick dissolution scenario” with the melt composition at relocating temperature under a quasi-equilibrium assumption. The calculated values were compared with the UO2 solubility observed in the UO2–Zr dissolution tests representing “the slow dissolution scenario” in which the melt is held for a while by the ZrO2 layer. Two phases were recognized in the experiments with latter scenario: ceramic (U,Zr)O2−x phase and metallic liquid phase. The melt composition in each scenario was also converted from the U–Zr–O system to the UO2–ZrO2–Zr system as an example of user-defined criteria of dissolution which can be applied to the severe accident code analysis. The difference between two scenarios (“quick” and “slow”) can result in differing times of lower head failure through the high liquidus temperature of oxide solid phase or chemical activity of metallic liquid phase. Providing the conditions for different mechanisms in the consistent manner for the thermodynamic analysis enabled evaluation of the fuel dissolution behavior in the context of integral accident progression.