Depressurization Accident Research Articles

Experimental study of the depressurization phenomena of supercritical carbon dioxide system

Depressurization accidents pose a significant safety concern for supercritical carbon dioxide (S–CO2) nuclear and power systems. However, due to the limited availability of system-level experiments, clear understandings of the phenomena associated with S–CO2 depressurization accidents are lacking. To address this gap, this paper conducted a series of experiments on a S–CO2 system to investigate the characteristics of depressurization accidents. Several key phenomena, including flash evaporation, transcritical transient heat transfer, fluid-thermo-structure coupling effect, and the fluid field changes in containment vessel, were identified and analyzed. Furthermore, the impact of break size and location, containment vessel, and initial system conditions on depressurization accidents was investigated. The acquired results provide crucial groundwork for establishing a Phenomenon Identification Ranking Table (PIRT) tailored to depressurization accidents in S–CO2 nuclear and power systems. This will significantly enhance the design and safety analysis of these systems.

An innovative neutronic deterministic scheme for moderator high-void effect calculation in PWR partially loaded with MOX assemblies

In some countries, as France, plutonium is used in industrial Pressurized Water Reactors (PWRs), using Mixed Oxides (MOX) fuels. To avoid a reactivity increase during a loss of coolant accident, the plutonium content in a PWR is limited at 12 wt%. This limit, that have been determined in the early 1990 s using a very conservative approach, is a strong constraint on the possibility to stabilize the plutonium inventory. Thus, further studies are now conducted to refine the plutonium limit content in the fuel. One interesting case is the local moderator voiding, which could for instance occurs at the beginning of a core depressurization accident. To study this phenomenon, a 2-steps neutronic deterministic scheme is proposed in this paper, using the APOLLO3® code. It is shown that the choice of the selfshielding method to use at the lattice step has a strong influence on the precision of the calculation. It is shown that, as expected, the subgroup formalism is the more precise to cope with both nominal and voided configurations. However, methods relying on this formalism are numerically costly. The use of a method relying on the finestructure formalism allow to conserve a satisfying estimation of the void effect on the reactivity or the spatial distribution of energy-integrated reaction rates. Unfortunately, we show that those good results are due to compensation of large energetic biases in the nominal case. A combination of Tone’s and finestructure selfshielding methods is presented. This method permit to strongly improve per-energetic-groups precision in the voided configuration compared to the method solely relying on the finestructure formalism, while drastically decreasing the numerical cost and duration compared to the subgroup method. The necessity to perform environed lattice calculation is demonstrated in the case of the voiding of a single assembly. A Local Flux Volume normalization of SPH coefficients is proposed to run equivalence calculation on this heterogeneous geometry. A method of characteristic solver is used for the flux calculation at the lattice step, while a S8 transport calculation is performed at the core level, using pin-by-pin discretization and a 26-groups energy mesh. Finally, the precision of the calculation scheme is assessed by comparison to Monte Carlo calculation ran with TRIPOLI–4® on a two-dimensional clusters of assemblies.

Control method of natural circulation flow by helium gas injection

There is a possibility that the mixture gas of air and helium diffuse into the reactor core at the depressurization accident of a HTGR. In order to investigate a method to control natural circulation flow of air and helium, we have conducted an experiment using an apparatus that simulates the flow path configuration of the HTGR. The experimental apparatus consists of two vertical channels with rectangular cross-sections, a connecting channel at the top of the channels, and a storage vessel at the bottom of the channels. The temperature of the heated and cooled walls and the temperature of the mixture gas in the vertical slots were measured. The concentration of air was also measured by ultrasonic concentration meter. From the results obtained in this experiment, it was confirmed that the natural circulation flow stopped for a while after helium gas injection. We will also describe the results obtained in the three-dimensional numerical analysis.

Graphite dust emission evaluation in an HTGR depressurization accident

This study evaluated graphite dust emissions in an HTGR depressurization accident. A simplified experimental platform based on air at ambient temperature was developed to simulate particle resuspension during a depressurization accident and helium blower startup. Then, computational fluid dynamics models were used to calculate the impact, rebound, and deposition of graphite particles to determine the particle deposition fraction in the containment based on the HTR-PM design parameters. The classical deposition formula was then used to estimate the proportion of particles released into the environment through a releasing pipe. The results show that the mass of particles released increases with decreasing helium temperature. The release mass first increases and then decreases with increasing particle diameter. This analysis shows that about 3–5 g of particles will be released to the environment in a design-basis HTR-PM depressurization accident.

ESTIMATION OF THE RADIOACTIVE SOURCE TERM FROM RDE ACCIDENT POSTULATION

The design process of Experimental Power Reactor (Reaktor Daya Eksperimental/RDE) has been carried out by BATAN for the last five years, adopting HTGR-type reactor with thermal power of 10 MW. RDE is designed with the reference of similar reactor, namely HTR-10. During this process, source term estimation is required to prove the safety of RDE design, as well as to fulfill the concept of As Low As Reasonably Achievable (ALARA) in radiation protection. The source term is affected by the magnitude of the radioactive substances released from the reactor core due to an accident. Conservative accident postulations on the RDE are water ingress and depressurization accidents. Based on these postulations, source term estimation was performed. It follows the mechanistic source term flow, with conservative assumptions for the radioactive release of fuel into the coolant, reactor building, and finally discharged into the environment. Assumptions for the calculation are taken from conservative removable parameters.The result of source term calculation due to the water ingress accident for Xe-133 noble gas is 8.97E+12 Bq, Cs-137 is 3.59E+07 Bq, and I-131 is 4.34E+10 Bq. As for depressurization accident, the source term activity for Xe-133 is 3.90E+13Bq, Cs-137 is 1.56E+07 Bq, and I-131 is 1.89E+10Bq. The source term calculation results obtained in this work shows a higher number compared to the HTR-10 source term used as a reference. The difference is possibly due to the differences in reactor inventory calculations and the more conservative assumptions for source term calculation.Keywords: RDE, HTGR, Radioactive, Source term, accident

Estimation of radioactivity impact for RDE based on HTR-10 hypothetical accident - a case study

BATAN priority activity supported by the Center for Nuclear Reactor Technology and Safety is the conceptual design documents and evaluation of experimental power reactor (RDE). Based on the design established, the radiation safety involving the dispersion of radioactive into the site and environment shall be calculated. The objective of this study is to obtain the radioactivity impact of RDE due to hypothetical accident in Serpong II Nuclear Area (KNS-II) in Puspiptek Area. Postulated hypothetical accidents are water ingress and depressurization accident. The sourceterm input data taken are based on HTR-10 hypothetical accident. Meteorological and environmental data are taken from available data of KNS Serpong. The calculation is carried out using PC-Cosyma code. The highest radioactivity of air dispersion for Kr-87due to depressurization accident is 1.96E+04 Bq/m3 and due to water ingress accident is 1.96E+04 Bq/m3. The highest radioactivity of surface deposition due to depressurization accident for Cs-137 is 1.51E+03 Bq/m2 and due to water ingress accident for I-131 is 1.5E+01 Bq/m2. The impact of RDE radioactivity for hypothetical accidents on the KNS-II site shows lower than BAPETEN regulatory requirements.

Experimental and Numerical Analysis of Mixing Process of Two Component Gases in a Vertical Fluid Layer in a VHTR

When a depressurization accident of a very-high-temperature reactor (VHTR) occurs, air is expected to enter into the reactor pressure vessel from the breach and oxidize in-core graphite structures. Therefore, in order to predict or analyze the air ingress phenomena during a depressurization accident, it is important to develop a method for the prevention of air ingress during an accident. In particular, it is also important to examine the influence of localized natural convection and molecular diffusion on the mixing process from a safety viewpoint. Experiment and numerical analysis using a three-dimensional (3D) computational fluid dynamics code have been carried out to obtain the mixing process of two-component gases and the flow characteristics of localized natural convection. The numerical model consists of a storage tank and a reverse U-shaped vertical rectangular passage. One sidewall of the high-temperature side vertical passage is heated, and the other sidewall is cooled. The low-temperature vertical passage is cooled by ambient air. The storage tank is filled with heavy gas and the reverse U-shaped vertical passage is filled with a light gas. The result obtained from the 3D numerical analysis was in agreement with the experimental result quantitatively. The two component gases were mixed via molecular diffusion and natural convection. After some time elapsed, natural circulation occurred through the reverse U-shaped vertical passage. These flow characteristics are the same as those of phenomena generated in the passage between a permanent reflector and a pressure vessel wall of the VHTR.

Simulations of the dust behavior in the sampling and dust filters in the primary loop of HTR-10

Abstract The radioactive dust in the primary loop of high-temperature gas-cooled reactors (HTGRs) is one of the key challenges to operational safety, particularly in the event of depressurization accidents. To study the behavior of radioactive dust in the primary loop of the HTR-10 reactor, a sampling loop (equipped with a sampling filter) was designed and built in its helium purification system, parallel to the dust filter. Two experiments were conducted using this sampling loop in 2015. The concentrations, particle size distribution of the dust, and typical solid fission and activation products detected in the dust were investigated (Xie et al., 2017). In this paper, we presented the γ spectra of the three filter elements from the second experiment as well as those of the unused filter elements, for comparison. To further analyze the characteristics of the dust motion during the sampling experiment, the computational fluid dynamics (CFD) method was used to simulate the flow field in the experimental system. Subsequently, we analyzed the transport and deposition behaviors of the dust, and discussed the effects of the diameters and flow rates of the particles on the behavior of the dust. Our calculations indicated that different particle deposition characteristics corresponded to different filters. We noticed that for the horizontal sampling filter, small particles were more prone to deposit on the first filter, whereas the large particles were more likely to deposit on the bottom surface. Moreover, as the mass flow increased, the effect of gravity decreased. For the vertical dust filter, we noticed that the distance between the deposition location of the particles and inlet increased, while the residence time decreased as the particle size increased. Moreover, the distance between the deposition location and inlet for the small particles increased, while the incidence of the large particles decreased as the mass flow increased. These simulation results could prove valuable for the experimental research of the radioactive dust of HTR-10.

Particle Deposition by Thermophoresis Under High-Temperature Conditions in a Helium Flow

The continuous generation of graphite dust particles in the core of a high-temperature reactor (HTR) is one of the key challenges of safety during its operation. The graphite dust particles emerge from relative movements between the fuel elements or from contact to the graphitic reflector structure and could be contaminated by diffused fission products from the fuel elements. They are distributed from the reactor core to the entire reactor coolant system. In case of a depressurization accident, a release of the contaminated dust into the confinement is possible. In addition, the contaminated graphite dust can decrease the life cycle of the coolant system due to chemical interactions. On one hand, the knowledge of the behavior of graphite dust particles under HTR conditions using helium as the flow medium is a key factor to develop an effective filter system for the discussed issue. On the other hand, it also provides a possibility to access the activity distribution in the reactor. The behavior can be subdivided into short-term effects like transport, deposition, remobilization and long-term effects like reactions with material surfaces. The Technische Universität Dresden has installed a new high-temperature test facility to study the short-term effects of deposition of graphite dust particles. The flow channel has a length of 5 m and a tube diameter of 0.05 m. With helium as the flow medium, the temperature can be up to 950 °C in the channel center and 120 °C on the sample surface, the Reynolds number can be varied from 150 up to 1000. The particles get dispersed into the accelerated and heated flow medium in the flow channel. Next, the aerosol is passing a 3 m long adiabatic section to ensure homogenous flow conditions. After passing the flow straightener, it enters the optically accessible measurement path made from quartz glass. In particular, this test facility offers the possibility to analyze the influence of the thermophoretic effect separately. For this, an optionally cooled sample can be placed in the measuring area. The thickness of the particle layer on the sample is estimated with a three-dimensional laser scanning microscope. The particle concentration above the sample is measured with an aerosol particle sizer (APS). Particle image velocimetry (PIV) detects the flow-velocity field and provides data to estimate the shear velocity. In combination with the measured temperature-field, all necessary information for the calculation of the particle deposition and particle relaxation times are available. The measurements are compared to results of theoretical works from the literature. The experimental database is relevant especially for computational fluid dynamics (CFD)-developers, for model development, and model verification. A wide range of phenomena like particle separation, local agglomeration of particles with a specific particle mass, and selective remobilization can be explained in this way. Thus, this work contributes to a realistic analysis of nuclear safety.

Process of Air Ingress during a Depressurization Accident of GTHTR300

A depressurization accident is the design-basis accidents of a gas turbine high temperature reactor, GTHTR300, which is JAEA’s design and one of the Very-High-Temperature Reactors (VHTR). When a primary pipe rupture accident occurs, air is expected to enter the reactor core from the breach and oxidize in-core graphite structures. Therefore, it is important to know a mixing process of different kinds of gases in the stable and unstable density stratified fluid layer. In order to predict or analyze the air ingress phenomena during the depressurization accident, we have conducted an experiment to obtain the mixing process of two component gases and the characteristics of natural circulation. The experimental apparatus consists of a storage tank and a reverse U-shaped vertical rectangular passage. One side wall of the high temperature side vertical passage is heated and the other side wall is cooled. The other experimental apparatus consists of a cylindrical double coaxial vessel and a horizontal double coaxial pipe. The outside of the double coaxial vessel is cooled and the inside is heated. The results obtained in this study are as follows. When the primary pipe is connected at the bottom of the reactor pressure vessel, onset time of natural circulation of air is affected by not only molecular diffusion but also localized natural convection. When the wall temperature difference is large, onset time of natural circulation of air is strongly affected by natural convection rather than molecular diffusion. When the primary pipe is connected at the side of the reactor pressure vessel, air will enter the bottom space in the reactor pressure vessel by counter-current flow at the coaxial double pipe break part immediately. Afterward, air will enter the reactor core by localized natural convection and molecular diffusion.

Resuspension models for monolayer and multilayer deposits of graphite dust

Graphite dust that will be generated in a multi-pass pebble-bed HTR (high temperature reactor), for example the Chinese HTR-PM and the South African PBMR, during normal reactor operation will be deposited inside the primary system and will become radioactive due to sorption of fission products. A significant amount of radioactive dust may be resuspended and released from the reactor cooling system in case of a depressurization accident. Therefore, accurate particle resuspension models are required for HTR/PBMR safety analyses.A review of available resuspension models applicable for monolayer and multilayer deposits is presented in this paper. It is demonstrated that for both multilayer and monolayer deposits, the main problem is the lack of data on adhesion forces and particle-to-particle contact forces.For monolayer deposits, a simple resuspension model, based on a moment balance, referred to here as KS-MB, is proposed and compared with several available resuspension models and available experimental data. It is concluded that a key factor in successful resuspension predictions is a good knowledge of the adhesion force distribution for dust particles deposited on rough surfaces. We demonstrate that relatively simple, quasi-static models, such as the KS-MB model, are as useful as the more complicated dynamic models for resuspension calculations in lack of precise data concerning adhesion forces.For multilayer deposits, resuspension modelling is even more complex. Several models exist, but there is no sufficiently extensive validation. Furthermore, the models may even give contradicting trends of the resuspension rates. The KS-MB resuspension model applicable for multilayer deposits is proposed here and validated against available experimental data. It is concluded that important factors are deposit structure as well as adhesion forces and particle-to-particle contact forces. Furthermore, it is not possible to positively identify the trend of resuspension rates in multilayer deposits. The effect of the multiple layers is overwhelmed by uncertainties in the adhesion force definitions. The main recommendation from the current work is that further measurements of adhesion forces, preferably done for the actual materials and conditions (temperatures, pressures) of the analyzed system (for example HTR-PM) are crucial for development of models and accurate prediction of resuspension.

Experiments and numerical analysis of a control method for natural circulation through helium gas injection

This study investigated a control method for natural circulation of air by helium gas injection. A depressurization accident is a design-basis accident of a very high temperature reactor. When a primary pipe rupture accident occurs, air is expected to enter the reactor pressure vessel from the breach. Thus, in-core graphite structures are oxidized. In order to predict and analyze the phenomena of air ingress during a depressurization accident, numerical analysis was carried out using a one-dimensional (1D) analysis code and three-dimensional computational fluid dynamics (3D CFD). An experiment was carried out regarding natural circulation using a circular pipe consisting of a reverse U-shaped channel. The channel consisted of two vertical heated and cooled pipes. The temperature difference between the vertical pipes was maintained at 40–80K, and a small amount of helium gas was injected into the channel. The injected volume of helium was about 3.1–12.5% of the total channel volume. After injecting helium gas, each component gas moved through molecular diffusion and very weak natural circulation. After approximately 1180s, ordinary natural circulation of air was suddenly produced. The numerical results of the 3D CFD code were in good agreement with the experimental results. The numerical results also showed that the natural circulation of air can be controlled by helium gas injection.

Study on heat transfer characteristics of the one side-heated vertical channel with inserted porous materials applied as a vessel cooling system

In the very high temperature reactor (VHTR), which is a next generation nuclear reactor system, ceramics are used as a fuel coating material and graphite is used as a core structural material. Even if a depressurization accident occurs and the reactor power goes up instantly, the temperature of the core will change only slowly. This is because the thermal capacity of the core is so high. Therefore, the VHTR system can passively remove the decay heat of the core by natural convection and radiation from the surface of the reactor pressure vessel. The objectives of this study are to investigate the heat transfer characteristics of natural convection of a one-side heated vertical channel with inserted porous materials of high porosity and also to develop the passive cooling system for the VHTR. An experiment was carried out using a one-side heated vertical rectangular channel. To obtain the heat transfer and fluid flow characteristics of the vertical channel with inserted porous material, we have also carried out a numerical analysis using a commercial Computational Fluid Dynamics (CFD) code. This paper describes the thermal performances of the one-side heated vertical rectangular channel with an inserted copper wire of high porosity.

Particle deposition and resuspension in gas-cooled reactors—Activity overview of the two European research projects THINS and ARCHER

The deposition and resuspension behaviour of radio-contaminated aerosol particles is a key issue for the safety assessment of depressurization accidents of gas-cooled high temperature reactors. Within the framework of two European research projects, namely Thermal Hydraulics of Innovative Nuclear Systems (THINS) and Advanced High-Temperature Reactors for Cogeneration of Heat and Electricity R&D (ARCHER), a series of investigations was performed to investigate the transport, the deposition and the resuspension of aerosol particles in turbulent flows. The experimental and numerical tests can be subdivided into four different parts: (1) Monolayer particle deposition, (2) Monolayer particle resuspension, (3) Multilayer particle deposition and (4) Multilayer particle resuspension. The experimental results provide a new insight into the formation and removal of aerosol particle deposits in turbulent flows and are used for the development and validation of numerical procedures in gas-cooled reactors. Good agreement was found between the numerical and the experimental results.

ICONE23-1930 STUDY ON THE ONE-DIMENSIONAL NATURAL CIRCULATION PHENOMENON ON THE MIXING PROCESS OF TWO COMPONENT GASES IN A VERTICAL FLUID LAYER

A depressurization accident is one of the design-basis accidents of the VHTR. When the pipe rupture accident occurs, air is expected to enter the reactor core from the breach and oxidize in-core graphite structures. In order to predict or analyze the process of air ingress during the depressurization accident of the VHTR, it is very important to develop computer programs and to validate them by experiments. This study is to investigate the effect of one-dimensional natural circulation on the mixing process of two component gases by evaluating the onset time of natural circulation through the apparatus under the stable density stratified fluid layer. The experimental apparatus consists of a reverse U-shaped vertical slot and a storage tank. The left side vertical slot consists of the heated wall and the cooled wall. The right side vertical slot consists of the two cooled walls. These experimental results show that generation time of natural circulation was affected by molecular diffusion and localized natural convection. When the two components of gases have large density ratios and large Gr numbers, the mixing process of two components of gases was affected by more intensively molecular diffusion than localized natural convection when temperature difference was 50K. The mixing process of two component gas was affected by more intensively localized natural convection than molecular diffusion when temperature difference was 70 to 100K. However, two component gases were affected by more intensively molecular diffusion than localized natural convection at small density ratios and small Gr numbers.

11014 鉛直流体層内の2成分気体混合過程に関する数値解析

When a depressurization accident occurs in the Very-High-Temperature Reactor (VHTR), a mixing process of different kind of gases occurs. Then, in order to research mixing process of two component gases and flow characteristics of localized natural convection, we have carried out the numerical analysis using three dimensional CFD code. The numerical model was consisted of a reverse U-shaped vertical slot and a storage tank. The left side slot was consisted of a heated wall and a cooled wall. The right side slot was consisted of two cooled walls. The heavy gas was transported to the slot by the molecular diffusion and localized natural convection. And then, natural circulation generated through the reverse U-shaped slot. As a result of analysis, generation time of natural circulation depended on not only molecular diffusion coefficient but also natural convection.

Core design and safety analyses of 600 MWt, 950 °C high temperature gas-cooled reactor

The conceptual core design study of high temperature gas-cooled reactor (HTGR) is performed. The major specifications are 600MW thermal output, 950°C outlet coolant temperature, prismatic core type, enriched uranium fuel. The decay heat in the core can be removed with only passive measures, for example, natural convection reactor cavity cooling system (RCCS), even if any electricity is not supplied (station blackout). The transient thermal analysis of the depressurization accident in the case the primary coolant decreases to the atmosphere pressure shows that the fuels and the reactor pressure vessel temperatures are kept under their safety limit criteria. The fission product release, Ag-110m and Cs-137 from the fuels under the normal operation is small as to make maintenance of devices in the primary cooling system, such as a gas turbine, without remote maintenance. The HTGRs can achieve the advanced safety features based on their inherent passive safety characteristics.

Study on control method of natural circulation by injection of helium gas

This study is to investigate a control method of natural circulation of air by injection of helium gas. A depressurization accident is one of the design-basis accidents of a very high temperature reactor (VHTR). When the primary pipe rupture accident occurs in the VHTR, air is expected to enter into the reactor pressure vessel from the breach and oxidize in-core graphite structures. Finally, it seems to be probable that the natural circulation flow of air in the reactor pressure vessel produce continuously. In order to predict or analyze the air ingress phenomena during the depressurization accident of the VHTR, therefore, it is important to develop the method for prevention of air ingress during the accident. The experiment has been carried out regarding natural circulation using a circular tube consisting of a reverse U-shaped type. The vertical channel consists of the one side heated tube and the other side cooled tube. The experimental procedure is as follows. Firstly, the apparatus is filled with air and one vertical tube is heated. Then, natural circulation of air will be produced in the channel. After the steady state is established, a very small amount of helium gas injects from the top of the channel. The velocity, temperature of gas, and temperature of the tube wall are measured during the experiment. The analysis also has been carried out regarding natural circulation. The results were obtained as follows. The temperature difference between the vertical pipes was 50–130K, and a small amount of helium gas injected to the channel. The volume of injected helium gas is about 3.5–10% of the total volume of the channel. When the temperature difference between the vertical tubes was kept at 52K, the velocity of natural circulation flow became about 0.12m/s. During a steady state, a small amount of helium gas injected into the channel. Then, the flow velocity of natural circulation suddenly decreased. The volume of injected helium gas is about 3% of the total volume of the channel. The velocity became around zero. After 1500s elapsed, the natural circulation suddenly produced again. The experimental and analytical results show that the natural circulation flow of air can be controlled by injecting of helium gas.

Graphite dust resuspension in a depressurization accident of HTR

Graphite dust has a significant effect on the safety of high temperature gas-cooled reactors. The present study experimentally analyzes the resuspension of graphite dust in a depressurization accident. A series of aerodynamic verification tests in an experimental system with gas flow at the sound speed at the orifice agree well with theoretical results. The results showed that the friction velocity decreases with time during the depressurization while the resuspension fraction increases when augmenting the initial pressure. The experiment results show good agreement with the predictions of the revised Rock ‘n’ Roll model by Biasi et al., indicating that this model accurately predicts the resuspension of graphite dust in a HTR depressurization accident.

Effects of the onset time of natural circulation on safety in an air ingress accident involving a HTGR

An air ingress accident is a major safety issue pertaining to high-temperature gas-cooled reactors (HTGRs). When a helium depressurization accident occurs, air enters the core through broken pipes. The natural circulation (NC) phase of this accident leads to massive air ingress, causing extensive oxidation of the core graphite structure. The oxidation results in three main safety issues: an increase in the core temperature, degradation of the core graphite structure, and explosive CO accumulation in the upper plenum. As a result, the onset time of the natural circulation is one of the key parameters related to reactor safety. Using GAMMA (GAs Multi-component Mixture transient Analysis) we performed a one-dimensional (D) analysis of GT-MHR in the same manner used by the conventional safety tools. A 2D analysis was also done to determine the 2D effects, including the degree of stratified-flow induced-air ingress in the horizontal ducts. The 1D and 2D simulations showed that NC occurs at around 360h and over 500h, respectively (Jin et al., 2011). In an effort to determine the effect of the onset time of NC on safety, we performed a sensitivity study of the NC onset time when it is intentionally controlled. Due to the limitation of the 1D simulation to represent multi-dimensional stratified flows in hot and cold pipes, the 1D and 2D results are quite different. In the 1D calculation, the bottom reflector temperature is very sensitive to the onset time of NC, whereas it is insensitive in the 2D simulation. According to the 2D analysis, however, a strong-back flow from the air in the cavity is observed in the lower part of the reactor core and at the entrance of the broken pipes. As a result, the bottom reflector temperature gradually decreases in the 2D analysis irrespective of the onset time of NC. Based on these calculation results, the possibility of CO detonation and damage to the graphite structure are examined. Our conclusions are that there are large safety margins of the maximum fuel temperature, CO detonation, and damage to the graphite structure and even larger margins considering the more realistic 2D simulation.