Since the structures inside the pedestal of Fukushima Daiichi Nuclear Power Station Unit 2 are relatively intact, the temperature of fuel debris relocated from the reactor pressure vessel (RPV) to the pedestal region is estimated to be rather low; oxide components remained solid and metallic components are molten. In order to predict the RPV failure, the molten pool behavior in the lower head is a key factor. Only a few experiments, however, addresses the transient heat transfer of solid-liquid molten pool. To enrich experimental database, melting and heat transfer behavior are investigated using the LIVE facility at Karlsruhe Institute of Technology. The results showed that convective heat transfer could take place in a solid-liquid mixture pool and the thermal loads on the vessel wall shifted from bottom upwards.

In the decommissioning of the Fukushima Daiichi Nuclear Power Plant (NPP), the quantitative analysis of residual boron and borides in the reactor core and the identification of boron compound states are one of the important issues to be investigated. In this presentation, we report on the neutron energy-resolved analysis of boron samples irradiated with neutrons at J-PARC/MLF. We also investigated the possibility of identifying the compounds using peak broadening of the prompt gamma rays for each boride. The prompt gamma-ray peak-widths of metallic and non-metallic borides were significantly different from each other, while those of zirconium boride and iron boride were slightly different. The differences between these metal borides were measured and evaluated in detail by gamma-ray energy spectrum analysis. Finally, we will briefly introduce the current experiments and analysis results of our efforts toward two- and three-dimensional quantitative measurements using energy-analyzed two-dimensional detectors.

The Moving Particle Semi-implicit (MPS) method is being developed for simulation of multi-component liquid/solid relocation with solid-liquid phase changes. Main model developments and validation of the developed code against the simulated spreading and ablation experiments are summarized in the current paper.

福島第一原子力発電所事故Fukushima Daiichi Nuclear Power Station Accident

In the decommissioning of the Fukushima-Daiichi (1F) Nuclear Power Plant, it is essential to understand characteristics of the melted core materials. The estimation of boride in the real debris is of great importance to develop safe debris removal plans. Hence, it is required to investigate the amount of boron in the melted core materials with nondestructive methods. Prompt gamma-ray activation analysis (PGAA) is one of the useful techniques to determine the amount of borides by means of the 478 keV prompt gamma-ray from neutron absorption reaction of boron. Moreover, it is well known that the width of the 478 keV gamma-ray peak is typically broadened due to the Doppler effect. The degree of the broadening is affected by coexisting materials, and can be recognized by the width of the prompt gamma-ray peak. As a feasibility study, the prompt gamma-ray from boride samples were measured using the NOBORU, and RADEN beamlines at the Materials and Life Science Experimental Facility (MLF) of Japan Proton Accelerator Complex (J-PARC).

The test bundle of the latest test CMMR-4, Core-Material Melting and Relocation experiment, consists of 48 fuel rods filled with ZrO$_{2}$ simulant pellets with Zircaloy claddings, a control blade with B$_{4}$C particles in SS tube and sheath, two Zircaloy channel box walls, and lower support structures. The height of the test bundle was 80 cm and the heating system of the test was the plasma heating, which enabled melting of the oxide simulant fuel pellets. The test confirmed that macroscopic gas permeability existed until the ceramic-fuel melted and that the hot fuel rods tended to remain as columns in the core region, which suggests the heating of the support structure in earlier phase is unlikely. This information is useful not only for 1F decommissioning but also for further understanding of a BWR severe accident progression. The test bundle was cut by using the abrasive waterjet (AWJ) technique that uses abrasive garnet of 150-300 micro m with feed rate of approximately 1.5 kg/min. In order to cut off about 30 mm of ZrB$_{2}$ spot contained in the relocated melts, 750 liters of water, 84 kg of garnet and one nozzle replacement were necessary. The EPMA and XRD analyses of the cross-section showed that the place where repelled the garnet-contained waterjet contained ZrB$_{2}$. Since the cutting by AWJ technique has the property of selectively abrading the soft spots of the material, it must be noted that, in case of utilizing the technique in 1F decommissioning, garnet might be repelled by a hard boride and abrades places which were not expected.

Based on the analysis of the samples collected in the PCV interior of Fukushima Daiichi NPP, potential issues regarding the fuel debris retrieval are summarized. The analytical targets, analytical method and obtainable knowledge that are necessary to evaluate each issue are discussed. Valuable information can be obtained even from small samples to understand the current situation of the plant and the accident progression. Although the amount of collectable samples is limited, we considered how to evaluate the whole fuel debris and how to utilize the analytical results for recriticality, burnup and safety issues.

Steam and hydrogen generation history and leakage scale from RPV of Fukushima Daiichi Nuclear Power Plant unit 3 to reproduce the measured pressure history of RPV and PCV were inversely analyzed using a thermal hydraulic code GOTHIC. The leakage area from RPV to reproduce the measured decrease behavior of RPV pressure is evaluated to be around 1 cm$^{2}$, regardless of the assumed leakage paths. Since some of SRVs are supposed to have been open at the time of the major slumping (~12:00 of March 13), provided that the leakage area was kept ~1cm$^{2}$, its effect on PCV pressure would have been negligible. In this case, the gas flow at the time of the main slumping would have been through S/C, where vapor condensation was effective, thus certain contribution of non-condensable gases like hydrogen seems necessary to explain the observed D/W and S/C pressure increase.

In Fukushima Daiichi Nuclear Power Station Units 1-3, some part of the core materials including fuel relocated through the reactor pressure vessel (RPV) into the pedestal area. As a result of comprehensive analysis of data such as pressure gauges and water level gauges,it was evaluated that the time duration of the core material relocation to the pedestal was, less than 0.5 hr for Unit 1, around 2.5 hours for Unit 2 and around 7 hours for Unit 3 respectively.

Steam and hydrogen generation history and gas leakage area are inversely evaluated by a thermal hydraulic analysis code GOTHIC. The analyzed period in the accident progression is from the arrival of reactor liquid level at the top of active fuel (TAF) until start of depressurization of reactor pressure vessel(RPV) by activation of automatic depressurization system(ADS). Based on the measured behaviors of the RPV and PCV pressures from 6:30 of March 13th until the ADS activation, some leakage from RPV to PCV is supposed during this period. The leakage path and area are inversely derived on plural possible accident scenarios. The leakage area are estimated to be no greater than 1 cm$^{2}$. This result suggests that the gas flow at the time of the main slumping would have been through S/C, where vapor condensation was effective, thus certain contribution of non-condensable gases like hydrogen seems necessary to explain the observed D/W pressure increase.

The energy resolved neutron imaging system RADEN, installed at the Japan Proton Accelerator Complex (J-PARC), utilizes short-pulsed neutrons in the energy range from meV to keV by means of the time-of-flight method. The wide neutron energy range makes it possible to adjust neutron attenuation of a sample to suit a particular measurement by selecting the neutron energy range. The Core Material Melting and Relocation (CMMR) experiments have been performed to investigate core melt accidents in the Fukushima-Daiichi (1F) Nuclear Power Plant. Some amount of boride was found to be contained in simulated melted core material. The authors show the status of the development of a method utilizing energy-resolved neutrons, and demonstrate some measurements of boron-containing samples, including simulated melted core material.

With the current modeling assumptions in MELCOR, the best estimate conditions for RPV pressure history of Unit 3 suggested that 6 Safety Relief Valves (SRVs) could have remained open when the major core slumping is assumed to occur at ca. 45:20 h (ca. 12:00, March 13) with 50 to 80 tons of water inventory in the lower plenum. The current analysis also suggested that the efficiency of the AWI to the reactor core could have been only 15 percent as of reported by TEPCO estimated amount of water discharged by fire engines with the current modeling conditions if debris dryout was assumed to have occurred at around ca. 54:00 h (20:40 h, March 13th). As for lower head failure, there is still large uncertainty in predicting lower head failure time with Larson-Miller creep rupture model in the current MELCOR modeling.

Some radioactive material samples were taken inside the PCV of Fukushima Daiichi Units 1-3. TEPCO is working for sample collection, transport, and detailed examinations at hot cells using SEM, TEM and ICP-MS. Uranium-bearing micro-particles were analysed using SEM and Emit was confirmed that particles derived from fuel can be categorized into two groups. One is believed to have been separated from hot corium by hydrodynamic processes, while the other showed characteristics suggesting evaporation from hot corium and condensation afterward. The former particles might have compositional features common to the major part of fuel debris. Both groups possibly cause alpha contamination (actinide behavior) in the reactor building.

Difference in core energy between Unit 2 and Unit 3 is evaluated and its effects on coolability of debris relocated to lower plenum and debris-relocation behavior to pedestal are discussed.

Temperature distribution and composition of atmosphere (H$_{2}$/H$_{2}$O ratio) in the accident of Fukushima Daiichi Nuclear Power Plant No.2 were evaluated using RELAP/SCDAPSIM code. Furthermore, distruption test of stainless steel specimen by metal melt assumeed to be generated in core melting phase in the accident was performed. From these results and the result of another plasma heating test using simulated fuel budle performed by JAEA, relocation behavior of melt in the accident was evaluated.

It is necessary to maximize the knowledge with: (1) In-depth data analysis of 1F plant data, (2) Well-targeted experiments addressing the BWR-specific uncertainties, and (3) Evaluation of accident progression behavior based on integration of all the available information. In-depth analysis of Unit 2 and Unit 3 plant data was conducted as step (1). This step provided outlines of accident progression behavior in these units. This information is then reflected into SA code analyses. A series of plasma heating experiments using a test section simulating a part of BWR core were conducted to get insights for the above step (2). Core material relocation behavior in the high-temperature range up to ceramic fuel melting was observed in these tests. The above step (3) consisted of an evaluation of core thermal energy for Unit 2 and Unit 3.

In the accidents of the Fukushima Dai-ichi Nuclear Power Plant (FDNPP), in contrast that core degradations are thought to begin before depressurization in Unit 1 and 3, the core of Unit 2 was presumed to be intact at the time of depressurization and the degradation is thought to begin when the liquid level was close to or below the bottom of the core. The authors analyzed a temperature increase behavior during the core degradation of Unit 2 based on simulations with RELAP/SCDAPSIM computer code.

A large uncertainty is present in understanding of BWR accident progression behavior. A series of tests were conducted to provide experimental data that address uncertainties. As a result, useful information on core state just before slumping was obtained. This information will help us to comprehend core degradation in BWRs like those in the Fukushima Dai-ichi Nuclear Power Plant (1F).

Several contaminant samples from 1F1 primary containment vessel are carefully observed by transmission electron microscope to find micro-scale uranium-containing particles in their chemical forms of U-rich c-(U,Zr)O$_{2}$ and Zr-rich t-(Zr,U)O$_{2}$, which are also observed in fuel debris of TMI-2 and Chernobyl NPP-4.

Investigation of in-reactor cesium chemical behavior in TEPCO's Fukushima Daiichi Nuclear Power Station accident was performed. Especially, analysis of samples from containment in Fukushima Daiichi NPS was evaluated.

In this study aiming at contribution for safe decommissioning of Fukushima-Daiichi NPP, CFD (Computational Fluid Dynamics) method was applied and temperature distribution of Unit 3 was reproduced. This temperature distribution was then compared with the measured data obtained by TEPCO so that debris distribution can be estimated. Combined application of optimized tools and CFD method to resolve inverse problem determining best suited thermal balance within the containment vessel is a characteristic of this study.

The PCV pressure measurement of Fukushima-Daiichi NPP Unit 3 showed cyclic pressure up and down. Although some early pressure changes of them can be regarded as consequence of venting actions, later changes are not likely due to venting nor significant leakage through the PCV boundary. In this study, preliminary correction for pressure measurement data was applied based on assumed consistency among different pressure measurement data. With this correction, subtle pressure differences among RPV, D/W and S/C became visible and information related to accident progression was obtained. With this knowledge, possible mechanism of PCV pressure decrease by itself will be reported.

An experimental program using non-transfer type plasma heating is underway in JAEA in order to address large uncertainty in core-material-relocation behavior in severe accidents of BWRs. Based on preparatory tests, an experiment with simulated fuel rods, channel box, control blade and lower support structure was carried out. After the heating, control blade and channel box were mostly gone but major part of the fuel columns remained standing. Downward relocation of molten materials probably consisting mainly of metals was confirmed in the lower support structure region. These results together with data from planned X-ray CT measurement on the heat teat piece will provide effective data for core material relocation behavior with the BWR design conditions.

An experimental program using non-transfer type plasma heating is underway in JAEA in order to address large uncertainty in core-material-relocation behavior in severe accidents of BWRs. Based on preparatory tests, two experiments with simulated fuel rods, channel box, control blade and lower support structure were carried out. After the heating, control blade and channel box were mostly gone but major part of the fuel columns remained standing. Downward relocation of molten materials probably consisting mainly of metals was confirmed in the lower support structure region. These results together with data from planned X-ray CT measurement on the heat teat pieces will provide effective data for core material relocation behavior with the BWR design conditions.

There is a large uncertainty in core material relocation behavior with the BWR design condition during severe accident. In order to address this issue, a new high-temperature experimental technology using non-transfer type plasma torch is under development. Authors have conducted preparatory tests with small scale test pieces and realized melting of simulated core materials. It was also confirmed that computed tomography with X ray as non-destructive means provide precise information on material distribution after the transient. Characteristics of materials after the transient was also studied by various measurement means. These results are showing an excellent perspective for applicability of this new technology to the intended experimental program.

As the results of good collaboration between Kazakhstan and Japan in EAGLE-1and 2, it was shown that there exists a solution to the recriticality issue of SFR, which has been one of the major safety issues for more than a half century from the beginning of the SFR development. Experimental techniques and facilities have been developed for the SFR severe accident study. Since 2014, JAEA participates the ASTRID program in which severe accident study is one of important issues. The EAGLE-1 and 2 data will be also used as an essential part of the severe accident study for ASTRID. EAGLE-3 was just started from beginning of 2015. Points of experiments moved into the later phase of core damage process, i.e., material relocation and cooling after achieving neutronic shutdown. A number of out-of-pile tests and in-pile tests are planned in coming five years.

Experimental needs for clarification of core-material relocation behavior during BWR severe accident are studied based on existing experimental database. It is understood that relocation of molten core materials from the original core region into the lower plenum is dependent on the structural characteristics of BWR. Existing experiments are quite limited for this behavior and future experiments are necessary for better understanding. Possible experiments to fill the experimental needs are proposed.

The event progression of core disruption in UTOP (Unprotected Transient Overpower), which was judged to be an important core disruption category in risk assessment of JOYO, was analyzed. It was confirmed that power burst did not occur in the initial phase of the accident and was not likely to occur in the successive phase of disruption extension.

In-vessel retention of molten core fuel with the use of debris trays in a reactor lower plenum is being studied as a mitigation measure against core disruptive accident for sodium-cooled fast reactors. If the molten core fuel is finely fragmented before coming at the debris trays, fuel coolability on the debris trays can be enhanced. In the present study, the length for molten jet break-up due to fragmentation was measured with out-of-pile experiments in which about 10 kg of molten alumina was injected into a sodium pool.

CDF caused by ATWS (Anticipated Transient Without Scram) event and PLOHS (Protected Loss of Heat Sink) event are majority of CDF in probabilistic safety study at Joyo. Evaluation of these values by reactor dynamic characteristics analysis and natural convection analysis are conducted with a detailed success standard. It was found from evaluation result that CDF are quite low and a dominant core damage accident is UTOP (Unprotected Transient Over Power) event.

The event tree of the initiating phase in ULOF, which was selected as a representative unprotected event, was developed. The most up-to-date knowledge was investigated as a technical basis for the level 2 PSA.

To develop a core damage evaluation technology (level-2 PSA) for sodium-cooled fast reactor, the transition phase analysis of a ULOF event that was a representative event in unprotected events was carried out to identify dominant factors influencing event progression. Technical basis for develop event tree was summarized systematically with previous findings.

Concerning safety characteristics of FBR core with hydride control materials, representative accidents for design basis and those beyond the design basis were selected and transient responses were evaluated. Based on this result, it was concluded that no unfavorable response was expected for this option.

The preparation of technical basis to develop phenomenological event tree and determine the probability at each branch has been developed. In order to extract dominant factor affecting the event progression, parametric analysis of initiating phase in ULOF (unprotected loss-of-flow) event was performed using SIMMER-III code.

The preparation of technical basis to develop phenomenological event tree and determine the probability at each branch has been developed. In order to extract dominant factor affecting the event progression, parametric analysis of transition phase in ULOF (unprotected loss-of-flow) event was performed using SIMMER-III code.

To develop a core damage evaluation technology (level-2 PSA) in sodium-cooled fast reactors, a new analysis method is developed for core-material relocation phase and internal containment vessel event. This study also develop technical basis necessary for the level-2 PSA.

The present study has confirmed that three-dimensional safety analysis code SIMMER-IV could realistically evaluate a three-dimensional motion of degraded core materials and appropriately simulate an actual fuel assembly configuration that had not been addressed by conventional two-dimensional analyses. It has been shown that a cooing effect of control rod guide tubes has a possibility to supress recriticality occurrence due to fuel compaction.

First this paper describes the design requirements in the FaCT project, especially for the safety design. The results of the EAGLE project are briefly described, and their implication in the safety design/assessment of the SFR investigated in the FaCT Project is shown. In the experiments in the EAGLE project, molten pool of fuel-steel mixture or alumina was successfully formed in the test section and discharged downward through the internal duct structure. These results show that the molten material in the core is discharged due to the pressure in the core shortly after the failure of the internal duct in the early stage of molten material formation and that the discharged material could be quenched as fragmented debris if sufficient amount of coolant is available. This implies that there could be a design solution for the safety design requirements in the FaCT Project.

In the EAGLE program, several in-pile and out-of-pile tests have been conducted by August 2006, under a co-operation between JAEA and NNC/RK. The main objectives of these tests are; (1) to demonstrate effectiveness of special design concepts to eliminate the re-criticality issue in the course of CDAs of SFRs, and (2) to acquire basic information on early-phase relocation of molten-core materials toward cold regions surrounding the core, which would be applicable to various core design concepts. As the final step of this program, integral in-pile tests simulating realistic accident conditions were conducted. Geometry of the test apparatus adopted in these tests is corresponding to a typical special design concept equipped with a "discharge duct" within each fuel sub-assembly. In these tests, fuel-steel mixture pool was successfully realized and discharge of the pool materials through the duct was observed.

The EAGLE experimental program is dedicated to show experimental evidences supporting a safety logic eliminating the recriticality issue in the core disruptive accidents (CDAs) of sodium-cooled fast breeder reactors. Two kinds of neutron detectors (one was placed in the test section and another was placed around IGR driver core for power monitoring use) were analyzed to get prospect that the data include information of molten-fuel motion in the in-pile test.

The EAGLE experimental program is dedicated to show experimental evidences supporting a safety logic eliminating the recriticality issue in the core disruptive accidents (CDAs) of sodium-cooled fast breeder reactors. In order to confirm an inherent nature of early fuel escape from the core region, both in-pile (using IGR) and out-of-pile experiments have been performed in the program. This presentation shows the preliminary interpretation of the second integral experiment, in which fuel discharge through a duct-type escape path (initially filled with sodium) was investigated using about 8kg of molten fuel. Energy insertion in this second experiment was smaller than that in the first experiment. The duct-wall failure timing was a little delayed compared with that in the first experiment, and the fuel discharged through the duct intermittently.

In order to develop the core damage evaluation technology (level 2 PSA) for sodium-cooled fast reactors, we develop the new analysis codes of post accident material relocation phase and of ex-vessel events, and we develop the technical bases that is necessary for level 2 PSA. In this presentation, summary and scope of the entire study is introduced as a part of the 4 series presentations.

The SIMMER-III code which introduced the models treating a diffusion-limited vaporization/condensation behavior and a large-bubble interface has been applied to the analyses of existing reactor safety test and reactor case, so that the characteristics of their models in evaluating reactor safety analysis were grasped. This study also revealed that the effect of non-condensable gas is not so significant in the situation of bubble expansion as rapid as reactor condition.

Japan is a country far away from Europe and the US. The news of the accident occurred in this country was immediately reached to countries throughout the world and the image of the explosion was repeated on TV screens. This accident aroused various discussions on nuclear policy. The responses of countries divided with the US and France maintaining the stance of promoting nuclear power and Germany and Italy clarifying its policy to withdraw from nuclear energy. This report provides information on the responses to the accident taken by the US and other states.

Dynamic behaviors of solid-particle dominant multiphase flows were investigated to model the mobility of core materials in a low-energy disrupted core of a liquid metal fast reactor. Two series of experiments were performed, that is dam-break experiments and bubble visualization experiments. Verification of fluid-dynamics models used in the fast reactor safety analysis code SIMMER-III was also conducted based on the numerical simulations of these experiments. The experimental analyses show that SIMMER-III can, to some extent, represent effects of solid particle interaction on multiphase flow behaviors by adjusting model parameters of the particle jamming model. Further improvement of SIMMER-III with more generalized models is necessary to appropriately simulate interactions between solid particles in a wider range of flow conditions.

In CDA of LMFBR, molten core materials would discharge from the core region through the coolant paths. Rapid vaporization of the coolant by mixing of the molten core materials provides effective evacuation of the liquid coolant from the path and reduces significantly possibility of core-material freezing and blockage formation inside the paths. This characteristic enhances early discharge of molten-core materials and reduces possibility of severe re-criticality events. In this study, melt discharge experiments were conducted with a coolant channel simulating the discharge path with an enhanced length of the path compared with that of the realistic design structure. An alloy and water were used as simulant of the molten fuel and sodium respectively. This series of experiments showed that the discharge path can be entirely voided by vaporization of a part of the coolant at the initial melt discharge phase, followed by vapor expansion toward the end of the coolant channel. Furthermore, it was revealed that the condition where coolant void expansion started can be defined by melt-coolant sensible heats ratio and the heated height of the coolant. The heat balance evaluation during the coolant void expansion phase shows that the film condensation heat transfer should be considered. The coolant-void-expansion behavior in the discharge path of the realistic design condition was estimated based on an application of this knowledge to existing experiments with molten oxide and sodium.

In the first licensing procedure of the prototype FBR "Monju", the event sequence of ULOF (Unprotected Loss of Flow) was analyzed and the estimated mechanical energy was about 380 MJ as an isentropic expansion potential to atmospheric pressure. The prototype FBR has been stopped for more than 10 years since the sodium leakage accident in the secondary loop in 1995. The neutronic characteristics of reactor core changed as a consequence of radioactive decay of fissile Plutonium during this shutdown period. In order to assess the effect of this neutronic characteristics change to the mechanical energy release in ULOF, the event sequence of ULOF was analyzed reflecting the current knowledge, which was obtained by safety studies after the first licensing of the prototype reactor. It was shown that the evaluated mechanical energy release became smaller than 380 MJ, even with the change of neutronic characteristics.

Dynamic behaviors of solid particle beds in a liquid pool against pressure transients were investigated to model the mobility of core materials in a low-energy disrupted core of a liquid metal fast reactor. A series of experiments was performed with a particle bed of different heights, comprising different monotype solid particles, where variable initial pressures of the originally pressurized nitrogen gas were adopted as the pressure source. Computational simulations of the experiments were performed using SIMMER-III, a fast reactor safety analysis code. Experimental analyses using the SIMMER-III code show that physical models and methods used in the code can reasonably represent the transient behaviors of multiphase flows with rich solid phase as observed in the experiments. The validation of several key models of SIMMER-III was also discussed for treating transient behaviors of the solid-particle phase in multiphase flows.

For the transition phase analysis of core disruptive accidents, the development of a three-dimensional reactor safety analysis code, SIMMER-IV, has been carried out based on the technology of the two-dimensional SIMMER-III code. The world first application of SIMMER-IV to a small-sized sodium-cooled fast reactor was also attempted to clarify event progression in the early stage of the transition phase. This SIMMER-IV calculation was compared to the two-dimensional case calculated by SIMMER-III, neglecting the presence of control rod guide tubes. The present analysis with the three-dimensional representation suggested that the conventional scenario leading to rather early high-mobility fuel-pool formation is unrealistic and the degraded core tends to keep low mobility in the early stage of transition phase.

拡散律速による蒸発／凝縮挙動および大気泡界面を扱うモデルを導入したSIMMER-IIIコードを既存の炉心安全性試験および実機解析に適用し、安全評価上のモデルの特性を把握した。また、実機条件のように急速に蒸気泡が成長する状況では、非凝縮性ガスによる凝縮抑制効果は著しい影響を与えないことが示された。

It is one of important problems for more reliable reactor safety evaluation to improve numerical simulation techniques for involved thermal-hydraulic phenomena of multiphase, multicomponent flows in core disruptive accidents. In the present cooperative research, physical model development and experimental investigation were performed for transient condensation phenomena of a vapor bubble with noncondensable gas to improve applicability of a fast-reactor safety analysis code for the phase-transition phenomena in multicomponent systems. In addition, basic validity of the developed models was demonstrated through the experimental analysis, and then applicability of the fast-reactor safety analysis code was discussed for bubble condensation behaviors under rector conditions.

During the course of core disruptive accidents in liquid-metal fast reactors, a boiling pool of molten fuel/steel mixture could be formed. The stability of this boiling-pool, for which in-pile experimental data with real reactor materials are very limited, plays an important role in the determination of the accident scenarios. In the TPA2 test of the CABRI-RAFT program (from 1996 to 2002), the fuel-to-steel heat transfer characteristic governing the pool behavior was investigated as a joint study with the French 'Institut de Radioprotection et de Surete Nucleaire' (IRSN). This test was performed in the CABRI reactor in 2001 using a test capsule that contains fresh 12.3\% enriched UO$_{2}$ pellets with embedded stainless steel balls. Following a pre-heating phase, the capsule was submitted to a transient overpower resulting in fuel melting and steel vaporization. The steel vapor-pressure build-up observed during the transient was quite weak, suggesting the presence of a strong mechanism to limit the fuel-to-steel heat transfer. The detailed experimental data evaluation suggested a phenomenon that the steel vaporization at the surface of steel ball blanketed the steel from molten fuel. This vapor blanketing seems to be a mechanism reducing the fuel-to-steel heat transfer. An analysis with the SIMMER-III code, a multi-component multi-phase thermal-hydraulics code, was performed in this study. This code simulation could well reproduce the pressure buildup and boiling pool behavior which occurred in the test by applying specifically reduced heat transfer coefficients.

The objective of the EAGLE project is to confirm a possible scenario in the postulated core disruptive accident of sodium cooled fast reactors, in which the molten fuel discharging from the core region in the early stage of the accident with its inherent mechanisms would prevent energetic re-criticalities. In order to obtain necessary experimental data, in-pile and out-of-pile experiments utilizing facilities of National Nuclear Center in the Republic of Kazakhstan were planed and are currently carried out. This document reports progress and results of the out-of-pile experiments which consist of a part of the EAGLE program. In the out-of-pile program so far, a series of experiments has been carried out aiming at establishment of basic necessary technology which consists of melting of the fuel-simulating material using the induction heating, transferring it into the test section and measuring the related phenomena during the experiments. Following knowledge and results have been obtained so far: - Using uranium dioxide and alumina as candidate materials for fuel simulant, basic experimental technology has been established, and fundamental data of molten material discharge were obtained under the condition without sodium. - With uranium dioxide, certain efforts, such as providing carbonized metal coating on the crucible inner surface, have been made to prevent chemical reaction between uranium dioxide and the graphite crucible so as to obtain molten uranium dioxide without a great deal of unfavorable impurities which prevent reasonable simulation of the real fuel behavior. It was concluded, however, that present techniques did not allow molten uranium dioxide with sufficient grade. - Alumina, which has well-known thermophysical properties, was evaluated to have adequate characteristics in terms of simulating real molten fuel behavior. Through implementation of the experiments, it was confirmed that molten alumina with sufficient purity and characteristics could be g

多成分多相流の熱流動現象に対する高速炉安全解析コードSIMMER-IIIの適用性を検証するため，非凝縮性ガスを伴う比較的大きな蒸気泡の過渡凝縮挙動について実験的知見を得るとともに，実験解析により同コードの妥当性を確認した。

Focusing on the cover layer materials (as the Radon Barrier Materials), which could have the effect to restrain the radon from scattering into the air and the effect of the radiation shielding, we produced the radon barrier materials with crude bentonite on an experimental basis, using the rotary type comprehensive unit for grinding and mixing, through which we carried out the evaluation of the characteristics thereof.

This report shows the results of the study on countermeasures for the elimination of re-criticality issue for the sodium cooled reactors, which was conducted in 2003 as a part of the feasibility study phase II for the commercialization of fast reactors. A sort of analytical studies related to the in-vessel retention capability under the unprotected loss of flow condition was conducted for the large scale and medium scale sodium cooled reactors, aiming at establishing some promising concepts to resolve the re-criticality issue keeping consistency with the basic concept of the core and plant design. Major conclusions are as follows. ABLE concept, which is proposed as a measure to enhance the fuel discharge capability in the early transition phase, needs much time to initiate fuel discharge than wrapper tube failure. Therefore it is currently concluded that it is difficult to show clear perspective. A modified version of FAIDUS which has less drawbacks on the core and cycle performance and related R\&Ds than original FAIDUS was proposed for further study. In-place retention and cooling in the core region is important from view point of reduction of R\&D loads conceming post accident material relocation and cooling at the bottom of the reactor vessel. A possibility of which the in-vessel retention can be achieved by quantitatively clarifying the effect of the superior cooling potential of sodium was shown. Based on the currently available information related to FAIDUS and ABLE, possible candidates of experimental studies were shown. An initiating phase analysis for the metallic fuel core with 550$^{\circ}$C of core outlet temperature and 8{\$} of sodium void worth resulted in mild consequence without prompt criticality. Although there is still large uncertainty in the early transition phase, it might be possible to avoid severe re-criticality. And it was shown that power excursion due to molten fuel sloshing might be milder than that of MOX fuel case.

It is one of important problems for more reliable reactor safety evaluation to improve numerical simulation techniques for involved therma-hydraulic phenomena of multiphase, multicomponent flows in core disruptive accidents. In the present cooperative research, physical model development and experimental investigation were conducted for transient condensation phenomena of a vapor bubble with noncondensable gases to improve applicability of a fast-reactor safety analysis code for the phase-transition phenomena in multicomponent systems. This fiscal year experiments using steam mixed with nitrogen gas were performed for the transient bubble condensation phenomena, and then experimental data were obtained for relatively large-scale bubble behavior. In addition, experimental analyses was performed by the fast-reactor safety analysis code and its validity was discussed.

In the CABRI-2 program, 12 tests were performed under various transient conditions covering a wide range of accident scenarios using two types of pre-irradiated Fast Breeder Reactor (FBR) fuel pins with different smear densities and burn-ups. For each fuel, a non-failure-transient test was performed and it provided basic information such as fuel thermal condition, fuel swelling and gas release. From the failure tests, information on failure mode, failure time and axial location was obtained. Based on this information, failure conditions such as fuel enthalpy and cladding temperature were evaluated. These failure conditions were compared with the CABRI-1 tests in which different fuels as well as different transient conditions were used. This comparison, together with supporting information available from existing in-pile and out-of-pile experiments, allowed an effective understanding on failure mechanisms depending on fuel and transient conditions. It is concluded that Pellet Cladding

It is one of the important problems for more reliable reactor safety evaluation to improve numerical simulation techniques for involved thermal-hydraulic phenomena of multiphase, multicomponent flows in core disruptive accidents.In the present joint research, physical model development and experimental investigation were conducted for transient condensation phenomena of a vapor bubble with noncondensable gases to improve applicability of a fast-reactor safety analysis code for the phase-transition phenomena in multicomponent systems.In this fiscal year, preliminary experiments using noncondensable gas were performed for the transient bubble condensation phenomena, and then basic data were obtained for large-scale bubble behavior without condensation.In addition, a multiple-scale flow-regime model treating large-scale bubbles was newly proposed for the fast-reactor safety analysis code and applied to the analysis of the preliminary experiments successfully.

The LTX test was performed using a SCRABIX pin in March 2000 in the framework of the CABRI RAFT Program to investigate the pin failure me-chanism, in pin fuel motion and post-failure relocation behavior under a simulated TUCOP accident in LMFR. The transient of the test was initiated by a coolant flow reduction and a structured TOP was tri- ggered when coolantaverage temperature at TFC reached a predefined value to keep the channel in the subcooled condition. Pin failure occured rather early, before the initiation of any significant fuel melting.This early failure of the cladding was presumably caused by a local cladding heatup and stress concentration arising from the exce- ssive pin bending. Rapid gas release upon the cladding failure led to the voiding of coolant channel, followed by a molten fuel ejection andgradual axial relocation in the test channel. An effort was made to interpret the experimental results of the LTX test using the SAS4A co-de. Although the ori

The CABRI-RAFT LTX test aims at a study on the fuel-pin-failure mecha-nism, in-pin fuel motion and post-failure fuel relocation with an an- nular fuel which was pre-irradiated up to peak burm-up of 6.4 at.\%. The transient test conditions similar to those of the LT4 test were selexted in the LTX test using the same type of fuel pin, allowing an effective direct comparison between the two tests. In contrast to the LT4 testwhich showed a large PCMI-mitigation potential of the annular fuel-pin design, early pin failure occurred in the LTX test when fuel dpes not seem to have molten.In order to clarify the fuel pin failure mechanism, interpretation of the LTX test up topin failure is per- formed in this study, through an experimental data evaluation and a PAPAS-2S-code analysis. THE PAPAS-2A code simulates reasonably the fu-el thermal conditions such as transient fuel-pin heat-up anf fuel mel-ting. The present detailed data evaluation shows that the earlier cla-dding failure compared

The CABRI-RAFT RB1 and RB2 tests were aiming at a study on impact of fuel pin failure under an overpower condition leading to fuel melting. Using a special technique, combination of through-cladding failure and fuel melting was realized.In the RB1 rest, fuel ejection was prevented under a limited fuel melting condition. On the other hand, significant fuel melting was applied in the RB2 test so as to get the fuel ejection, thereby obtaining information on the fuel ejection behavior. Interpretation for these tests through the detailed experimental data evaluation and the PAPAS-2A code analysis is performed in this study. Through this study, it is indicated that molten fuel ejection can be prevented with the low smear density fuel as far as the fuel melting is not large for a slit-type cladding defect ct. Fuel ejection becomes possible in the case of significant fuel with a very thin solid fuel shell surronding the molten fuel cavity. However, the r

The LT4 test was performed in the CABRI-FAST in-pile experiment program carried out in 1992$\sim$1995. The objectives of this test were to study the fuel pin failure mechanism and to observe the transient fuel motion within the pin and in the coolant channel. The objectives of the present study are to clarify phenomena taking place in the experiment through data evaluation and SAS4A code analysis. Various experimental data have been analyzed with a help of SAS4A code calculation to interpret fuel pin failure mechanism and post-failure material relocation behavior. Through this study, the rapid fissile elongation up to the fuel pin failure was recognized to have potential to delay the failure by about 50 ms, and Probable effect of plenum gas to enhance dispersive fuel relocation has been recognized. And it was confirmed that SAS4A can reasonably simulate rapid molten-fuel ejection from failed fuel pin, rapid fuel relocation within the coolant channel assisted by the Plenum-gas and the fuel freezing in the last Part of transient.

In the CABRI-FAST LT1 test, simulating a ULOF (Unprotected Loss of Flow) accident of LMFBR, pin failure took place rather early during the transient. No fuel melting is expected at this failure because the energy injection was too low and a rapid gas-release-like response leading to coolant-channel voiding was observed. This channel voiding was followed by a gradual fuel breakup and axial relocation. With an aid of SAS4A analysis, interpretation of this test was performed. Although the original SAS4A model was not well fitted to this type of early pin failure, the global behavior after the pin failure was reasonably simulated with temporary modifications. Through this study, gas release behavior from the failed fuel pin and its effect on further transient were well understood. It was also demonstrated that the SAS4A code has a potential to simulate the post-failure behavior initiated by a very early pin failure provided that necessary model modification is given.

None

None

Based on the RB1 test result in the CABRI-RAFT Program, it was agreed between the partners to perform the RB2 test which aims at observation of molten fuel ejection into the coolant channel at further fuel melting and at confirmation of coolability of ejected fuel. In this study, a preliminary post-test calculation for the RB1 test is performed first to reflect the fuel thermal condition expected for the pins with the special artificial defect preparation. Pre-test calculations for the RB2 test are then performed based on the results of this RB1 calculation. Power and coolant flow histories as well as the axial location of defect were selected as parameters in this study and a set of test condition is proposed which is believed to be most suitable to fulfill the test objectives.

Existing data of in-pile ramp-type transient-overpower tests (slow TOPs hereafter), such as those of the CABRI-2 and CABRI-FAST tests, the EBR-II TOPI-1E test and the former TREAT tests, were extensively surveyed and this led to a global interpretation which provided a consistency among the tests. Through this study, a basic fuel pin failure mechanism was comprehended and it was confirmed that fuel pins with low to intermediate smear density have a very high failure threshold with significant mitigation effects against fuel-cladding mechanical interactions. Such high failure threshold of low to intermediate smear density fuel is considered to be attributed to the following three effects: (1)absorption of fuel thermal expansion and fuel swelling by void space (porosity or cracks) within the fuel, (2)mitigation of fuel swelling by an early gas escape into the free volume, and (3)mitigation of molten cavity pressurization upon fuel melting. These effects were refrected to the analytical model of the transient fuel behavior code PAPAS-2S. Application of this improved PAPAS-2S model to representative slow TOP tests provided results consistent with the test data, suggesting that the above-mentioned consideration is valid.

None

None

None

None