[en] Within the participation in the OECD International Standard Problem No. 46 the degradation phase of the Phebus FPT1 experiment was simulated with the MELCOR 1.8.5 computer code. The experiment provides the opportunity to assess the capability of systems-level severe accident modelling codes in an integral manner, covering core degradation through to the late phase, hydrogen production, fission products release and transport, circuit and containment phenomena, and iodine chemistry. The input model was developed strictly following the recommendations on noding for the reference case simulation provided in the ISP-46 specification report. To be able to assess the capability of MELCOR to model the processes involved in the experiment, first the correct temperature conditions in the bundle have to be achieved. It turned out that the temperature conditions in the bundle are highly dependent on the adequacy of heat transfer modelling. The comparison of simulation results and experimental measurements showed that good agreement of thermal-hydraulic variables in the bundle can be achieved if the radiation inside the bundle and the heat losses through the shroud are correctly considered. (author)

[en] A steam explosion is a fuel coolant interaction process by which the energy of the corium is transferred to water in a time-scale smaller than the time-scale for system pressure relief and induces dynamic loading of surrounding structures. A strong enough steam explosion in a nuclear power plant could jeopardize the containment integrity and so lead to a direct release of radioactive material to the environment. To help finding answers on open questions regarding steam explosion understanding and modelling, the steam explosion simulation code ESE-2 is being developed. In contrast to the developed simulation code ESE-1, where the multiphase flow equations are solved with pressure-based numerical methods (best suited for incompressible flow), in ESE-2 densitybased numerical methods (best suited for compressible flow) are used. Therefore ESE-2 will enable an accurate treatment of the whole steam explosion process, which consists of the premixing, triggering, propagation and expansion phase. In the paper the basic characteristics of the mathematical model and the numerical solution procedure in ESE-2 are described. The essence of the numerical treatment is that the convective terms in the multiphase flow equations are calculated with the AUSM+ scheme, which is very time efficient since no field-by-field wave decomposition is needed, using second order accurate discretization. (author)

[en] Two applications of the neural network methodology in the field of nuclear safety analysis are described. The first one is the 3-D response surface generation by the Back Propagation Method. The results were not satisfactory. The second is the application of the Optimal Statistical Estimator methodology for the generation of 8-D response surface. It was used as a statistical part of the Code, Scaling, Applicability and Uncertainty (CSAU) methodology for the evaluation of Large Break Loss of Coolant Accident. The result was comparable to the one obtained by the ordinary method. (author). 24 refs, 5 figs, 3 tabs

[en] The ESE (Evaluation of Steam Explosions) computer code has been developed to model the interaction of molten core debris with water during the first premixing stage of a steam explosion. A steam explosion is a physical event, which may occur during a severe reactor accident following core meltdown when the molten fuel comes into contact with the coolant water. In this paper the numerical treatment of probabilistic multiphase flow equations on which ESE is based is described. ESE is a general two-dimensional compressible multiphase flow computer code. Each phase in the multiphase flow usually water, steam, melt and air is represented by one flow field with its own local concentration and temperature and is described with its own set of partial differential mass, momentum and energy equations. These transport equations are solved on a staggered in a 2D rectangular or cylindrical co-ordinate system using a high-resolution finite difference method. The pressure equation is solved using the stabilized squared conjugate gradient method (CGSTAB), which converges fast also for high density ratios. The numerical methods used in ESE were precisely tested on a number of carefully chosen cases where the analytical solutions are known. All results are presented in a form of graphs and they clearly show that the applied high-resolution method most exactly reproduces the analytical behavior.(author)

[en] Continuous efforts to ensure the safety of nuclear installations in Slovenia have led to comprehensive analysis of Levels II and III of hypothetic station blackout accident modelled using the tools at our disposal. This paper represents the thermal hydraulic and radionuclide transport part of the overall effort. MARCH3 and VANESA modules of Source Term Code Package were used to analyze four different scenario depending on different reactor coolant pump leak rate (125 gpm and 400 gpm, respectively) and containment design pressure (i.e. 0.309 Mpa and 0.785 Mpa). The final aim of the project was to prepare input into the Level III analyses of the accident. The accident starts by loss of off-site power combined with loss of diesel generators. The turbine driven auxiliary feedwater pump operates additional two hours after the inception of the accident. The results are given in form of graphs displaying reactor coolant system and containment parameters. (author)

[en] The solidification influence modeling in fuel-coolant interaction codes is strongly related to the modeling of the temperature profile inside the melt droplets and to the modeling of the mechanical effect of the formed crust on the fine fragmentation process. A purpose of the study was to enable solidification influence modeling in codes with an Eulerian formulation of the droplet field. Therefore additional transport quantities based on the most important melt droplet features regarding the steam explosion phenomenon were derived. This enables a more accurate prediction of the amount of droplets participating in the fine fragmentation process during the explosion phase. Also the potential effect of the proposed modeling was assessed. The simulations supported the key role of the solidification in the steam explosion phenomenon. (author)

[en] Steam explosion experiments have revealed that there are important differences of behaviour between simulant and prototypical melts (the efficiency of steam explosions with prototypical melts is about one order of magnitude lower than with alumina melts), and that also with prototypical melts the fuel coolant interactions (FCI) depend on the composition of corium (eutectic corium may explode spontaneously, whereas non-eutectic corium never did). To explore the ability of global FCI codes to adequately simulate the explosion phase if the premixture conditions at triggering would be known and to establish the influence of droplets freezing on the steam explosion, a number of explosion phase simulations of KROTOS experiment K-53 were performed with the code MC3D, starting from different predefined premixture and droplets conditions. In the performed parametric analysis the premixture radius and the corium droplets cut-off diameter were varied, assuming that corium droplets with a diameter smaller than the cut-off diameter can not participate in the explosion due to solidification. It was assumed that the phases in the premixture are distributed homogeneously. The phase fractions were determined based on experimental measurements of the average void fraction and the mass of released fuel. The analysis showed that there is an important influence of droplets freezing on the strength of the steam explosion, since already a small variation of the cut-off diameter significantly changed the calculated pressure impulses. Therefore the influence of droplets freezing on the steam explosion should be considered in global FCI codes. With the best-fit cut-off diameter and premixture radius a quite good agreement of simulation results with experimental measurements could be obtained. This is an indication that at least in principle the explosion phase can be reasonably well predicted if the premixture conditions at triggering are adequately determined and if we could judge based on the droplet solidification conditions whether the droplet can effectively participate in the steam explosion or not. (author)

[en] This paper describes the improved version of the general two-dimensional, multiphase flow code ESE. The ESE code has been developed to model the mixing process and interaction of molten core debris with water. In case of a steam explosion, a trigger may produce locally enhanced heat transfer and pressurization and may evolve into a shock propagating through the coarse mixture. The propagation phase of the interaction is not modeled by the code; however, the ESE provides for initial condition evolution in time. The indication of the amount of well-mixed melt at the time of the trigger occurrence can be deduced based on the code's results. The objective of this work is to present the advantages of the high-resolution method applied to a particular set of partial differential equations and to incorporate these advantages into a code that was conceived using less traveled paths, namely, ensemble averaging and use of available data in probabilistic density functions describing momentum and energy cofluctuation tensors

[en] A steam explosion may occur, during a severe reactor accident, when the molten core comes into contact with the coolant water. A strong enough steam explosion in a nuclear power plant could jeopardize the containment integrity and so lead to a direct release of radioactive material to the environment. To resolve the open issues in steam explosion understanding and modeling, the OECD program SERENA phase 2 was launched at the end of year 2007, focusing on reactor applications. To verify the progress made in the understanding and modeling of fuel coolant interaction key phenomena for reactor applications a reactor exercise has been performed. In this paper the BWR ex-vessel steam explosion study, which was carried out with the MC3D code in conditions of the SERENA reactor exercise for the BWR case, is presented and discussed. The premixing simulations were performed with two different jet breakup modeling approaches and the explosion was triggered also at the expected most challenging time. For the most challenging case, at the cavity wall the highest calculated pressure was ∼20 MPa and the highest pressure impulse was ∼90 kPa.s. (authors)

[en] A steam explosion is an important nuclear safety issue in case of a severe reactor accident because it could induce dynamic loading on surrounding structures, leading potentially to an early release of radioactive material into the environment. Studies of the steam explosion consequences have to be based on experimental research programs like TROI (Test for Real cOrium Interaction with water) and FCI (fuel coolant interaction) codes like MC3D. In this work the TROI-13 FCI experiment was analysed with the MC3D code. The TROI-13 experiment resulted in a spontaneous steam explosion. The premixing simulation was performed to determine the initial conditions for the steam explosion. A number of steam explosion simulations were performed, changing the mass of melt droplets and position of triggering. The results showed that there is an important influence of the participating mass of melt droplets on the pressure impulse. To determine the participating mass, the processes of melt droplets creation and droplets solidification should be properly taken into account. (author)