In order to investigate the mechanisms of steam explosion which may occur during a severe accident of light water reactors (LWRs), the MISTEE facility was developed at Royal Institute Technology (KTH) to visualize the micro interactions of steam explosion when a single molten droplet was falling into a water pool. For preparation of a molten droplet, an aerodynamic levitation system was proposed to prevent the droplet from falling out of the crucible during heating in an induction furnace by injecting argon gas through a purging line connected to the bottom nozzle of the crucible. To support the design of such levitation system, a numerical simulation of the aerodynamic levitation system was performed using the CFD code ANASYS FLUENT v16.2. The problem was simplified as adiabatic two-phase flow dynamics in a 2-D axisymmetric geometry. The VOF method is employed to track the interface of two phases (liquid metal and argon gas), and the SST k-omega model was chosen for turbulence. Various characteristics of droplet dynamics in incorporated with argon gas flowrates through the crucible were examined in the numerical simulation. The simulation results suggested there exists an optimal range of argon gas flowrate for levitating a coherent shape of droplet in the crucible. The wall adhesion had a considerable effect on initiating the levitation of the droplet, which means the properties of the inside surface of the crucible may play an important role in the levitation and discharge of the droplet. Proof-of-concept tests were carried out on the prototype of the design, and it was confirmed that the levitation system was able to fulfill its function, i.e., to keep the droplet in the crucible during its melting process, although the actual argon gas flowrates for levitation was higher than the predicted ones, probably due to the leakage of flow path and heat transfer which were not considered in the simulation. Generally speaking, the numerical simulation did not only help understand the hydrodynamic characteristics of the levitation system, but also provided insights on operational control and improvement of the system.

Accidental contact between hot melt and cold water poses fatal hazard in several industries. Vapor explosion during melt-water contact in nuclear power plant accident can result in catastrophic containment failure. The fast transient phenomena as vapor explosion is not comprehensively understood despite several advances in research. It is not clear why certain parameters of melt and water exhibit differences in fragmentation behavior. To examine the influential parameters, we perform a series of experiments. The interactions between melt and water is visualized by high-speed video and X-ray radiograph.

During a severe accident scenario of nuclear power plants, a steam explosion may occur when a substantial amount of molten core materials is rapidly ejected into a volatile coolant (water) pool, forming so-called Fuel-Coolant Interactions (FCI). The steam explosion poses a serious threat to the containment integrity. It is therefore important to understand and suppress the risk of steam explosions. The present study is concerned with mechanism of steam explosion on effect of coolant composition, i.e., to investigate how seawater impacts steam explosion energetics. For this purpose, a set of preliminary experiments were performed on spontaneous steam explosion by delivering a single molten tin droplet into a cold water pool at different levels of salinity on MISTEE facility. This paper presents the comparative results of the experimental data, including the influences of the salinity on probability of spontaneous explosion occurrence, explosion depth underwater and available thermal energy of droplet for explosion, as well as fragmentation (particle size distribution of debris % by mass). As the reference of the comparisons, the steam explosion characteristics from experiments in deionized water were employed. We experimentally observed that probability of spontaneous explosion occurrence increased in seawater, and more thermal energy of droplet was available for explosion when a droplet was self-triggered. The seawater at high salinity (35.16 g/kg) appeared remarkable enhancement on fragmentation. More experimental data are still needed to reveal more details and to develop a model for better understanding and prediction for the effects. The present data was helpful for prudential assessment on the seawater effects when it was employed as ultimate emergency cooling if NPPs located on sea coasts encounter Fukushima-like accidents.

Zirconium, which is used as the cladding material of nuclear fuel rods in LWRs, can react with steam in the case of a core meltdown accident. This results in the release of hydrogen which poses a significant risk of hydrogen explosion. Oxidation of Zr occurs either during the core degradation when the steam flows over the hot fuel rod surfaces or during an FCI when molten corium falls into a water pool (e.g. in the lower head). An experimental study was performed at the MISTEE-OX facility at KTH to observe and quantify the oxidation of molten zirconium droplets in a water pool. During the experimental runs, single droplets of molten zirconium were discharged into a subcooled water pool and the dynamic events were recorded using a high-speed camera. The bubble dynamics indicate a process of cyclic oxidation and hydrogen release from the rear periphery of a droplet while it is quenched in the water. The debris (solidified state of the droplet) after each run was collected for compositional and microstructural analysis via SEM/EDS. The obtained data were employed to estimate the oxidation fractions of the droplets and the results revealed several interesting insights into the oxidation phenomenon of the Zr melt. The water subcooling was observed to have a significant influence on the oxidation: the degree of oxidation decreased with increase in the water subcooling. Furthermore, the degree of oxidation was also influenced by the depth into the debris, forming compounds whose oxygen content decreases from the outer surface towards the core of the debris. Therefore, the qualitative and quantitative results presented in this paper are important in the context of developing a phenomenological understanding of the oxidation behaviour of zirconium melt during the FCI as well as to improve and validate the currently available models implemented in the state-of-art steam explosion codes.

Nordic type boiling water reactors (BWRs) adopt reactor cavity flooding as a severe accident mitigation strategy (SAMS) to achieve core melt fragmentation and long-term cooling of decay heat generating core debris. The qualification of this SAMS needs to address two main severe accident issues: debris bed coolability and steam explosion. 

Since the coolability of a debris bed is determined by the bed’s  properties including debris particle’s size distribution and morphology as well as the bed’s configuration and inhomogeneity, it is important to investigate the mechanisms of melt jet breakup and resulting fragmentation in water which affect debris bed’s properties. Hence, the first part of this thesis is concerned with characterization of melt jet breakup and resulting debris particles.  A series of jet breakup experiments have been conducted in small scale with simulant binary oxide melt mixtures of WO₃-Bi₂O₃, WO₃-ZrO₂ and Wood's metal. The experiments reveal significant influence of melt superheat, water subcooling, melt jet diameter and material properties on debris size and morphology. Specifically, transition in debris size and morphology is found to occur at a specific water subcooling range. The difference in debris properties at varied melt release conditions is attributed to the competition between liquid melt hydrodynamic fragmentation and thermomechanical fracture of quenched particles.

The second part of this thesis work is dedicated to provide a new understanding of steam explosion (SE) with the support of small-scale experiments at the level of droplets. Self- and externally-triggered SE experiments are conducted with simulant binary oxide melt mixtures in the temperature range of 1100 to 1500°C. The dynamics of steam explosion process is recorded using a sophisticated simultaneous visualization system of videography and X-ray radiography. Further, the influence of melt composition on steam explosion is summoned.  The results reveal that a droplet of eutectic composition is more explosive than a droplet of non-eutectic composition since latter may form a mushy zone which thereby limits the amount of melt actively participating in a steam explosion. To reduce the temperature difference between simulant melt and corium, investigation was extended to perform high temperature (˃2000°C) melt experiments. For this purpose, steam explosion of a molten Al₂O₃ droplet was investigated, and the experimental results confirmed that Al₂O₃ melt can undergo spontaneously triggered steam explosion at a high melt superheat and high subcooling. Within the context the effects of melt superheat and water subcooling were obtained.

The third part of this thesis is concerned with the oxidation of metallic melt representing unmixable metallic liquid of molten corium, which interactions with water can be spatially and chronologically separated from the oxidic corium FCI. The objective of the study  is to provide new insights into the characteristics of oxidation of Zr droplet falling in a water pool through a series of small-scale experiments. The dynamics of droplet and bubbles were recorded by high-speed cameras, and the spatial distributions of the elements in the quenched droplet (debris) were acquired by Energy- Dispersive X-Ray Spectroscopy (EDS). The results have shown noticeable influence of generated hydrogen and oxidation heat on droplet behavior and cooling rate. Water subcooling had significant influence on oxidation kinetics, and the oxygen content of the solidified particle increased with decreasing subcooling. Incomplete oxidation of Zr happened before melt crystallization and cooling down in all experiments.  

In the frame of the European Commission FP7 SAFEST project, IRSN proposed to experimentally investigate the steam explosion triggering mechanisms of a superheated alumina droplet falling into water, through a set of experiments in the Micro Interactions in Steam Explosion Energetics facility (MISTEE) at KTH. Since thermal fragmentation is considered to be a likely process for the triggering of Steam Explosions in the KROTOS tests (performed at CEA) with alumina, the ability of a single droplet of such material to undergo thermally induced fine fragmentation is studied on the MISTEE facility with a close-up visualization. A series of experiments were conducted, where droplets of molten alumina were discharged into a water pool and potentially exposed to a small pressure wave. The intense interactions were recorded with a high-speed camera along with the pressure in the droplet vicinity. The ability of alumina to undergo thermal fragmentation is expected to be firstly contingent on the stability of the vapour film enshrouding the melt droplet. The water and melt temperatures may then play a crucial role on the vapour film stability, and therefore on the observation of a steam explosion. Indeed, under high to moderate water sub-cooling conditions, experimental observations indicate that fine fragmentation of the melt can occur when the droplet is exposed to even a weak pressure wave, in the range of 0.15 MPa. In contrast, melt fine fragmentation is suppressed at low water sub-cooling conditions (less than 30 °C), where the formation of a thick vapour film (and large wake) is observed, and which is probably too stable to be destabilized by the weak pressure wave. The effect of the melt temperature on thermal fragmentation is also assessed. This parameter influences the solidification of the droplet and the strength of the explosion as it determines the available heat energy. In the present conditions, fine fragmentation of melt occurred even at quite low melt superheat (≈60 °C). For a high melt superheat (above 200 °C) a very energetic spontaneous steam explosion was observed. A physical analysis on the debris particles acquired indicates a mass median diameter of ≈100 µm, comparable to the one observed in the KROTOS alumina experiments. The MISTEE experimental results are finally used to assess the heat and mass transfer modelling of the coolant during the fragmentation process in the FCI code MC3D.

This part of the study investigates the performance of HFC refrigeration systems for supermarkets and compares the performance with alternative CO2 trans-critical solutions. The investigated HFC system solutions are typical in supermarkets in Sweden. The analysis in this study is based on field measurements which were carried out in three supermarkets in Sweden. The results are compared to the findings from Part I of this study where five CO2 trans-critical systems were analyzed. Using the field measurements, low and medium temperature level cooling demands and COP’s are calculated for five-minute intervals, filtered and averaged to monthly values. The different refrigeration systems are made comparable by looking at the different COP’s versus condensing temperatures. The field measurement analysis is combined with theoretical modelling where the annual energy use of the HFC and CO2 trans-critical refrigeration systems is calculated. Comparing the field measurement and modelling results of COP’s for HFC and CO2 systems, the new CO2 systems have higher total COP than HFC systems for outdoor temperatures lower than about 24 C. The modelling is used to calculate the annual energy use of HFC and new CO2 system in an average size supermarket in Stockholm, new CO2 systems use about 20% less energy than a typical HFC system. The detailed analysis done in this study (Part I and Part II) proves that new CO2 trans-critical refrigeration systems are more energy efficient solutions for supermarkets than typical HFC systems in Sweden.

Experiments were carried out to investigate the characteristics of jet breakup and debris formation after melt jets fall into a subcooled water pool, which may occur in industrial processes such as the interactions of molten corium with coolant during a severe accident of light water reactors. A high-speed visualization system developed previously at KTH was used to capture the jet fragmentation phenomenon. Molten metal (Woods metal or tin) and mixture of binary oxides (WO₃-Bi₂O₃ or WO₃-ZrO₂) were employed separately as melt materials to address different breakup mechanisms (e.g., hydrodynamic vs. thermodynamic fragmentation) and material effect. Moreover, the parameters related to melt and water conditions, including superheat, jet diameter and velocity of melt as well as subcooling of water, were scrutinized for their influences on jet fragmentation characteristics. The experimental data were acquired on the melt jet fragmentation patterns, breakup length, droplet size spectrum, droplet breakup and solidification as well as debris morphology, which can be useful for validation of the codes used for the prediction of debris formation phenomena.

Breakup of melt jet and formation of a porous debris bed at the base-mat of a flooded reactor cavity is expected during the late stages of a severe accident in light water reactors. Debris bed coolability is determined by the bed properties including particle size, morphology, bed height and shape as well as decay heat. Therefore understanding of the debris formation phenomena is important for assessment of debris bed coolability. A series of experiments was conducted in MISTEE-jet facility by discharging binary-oxide mixtures of WO3-Bi2O3 and WO3-ZrO2 into water in order to investigate properties of resulting debris. The effect of water subcooling, nozzle diameter and melt superheat was addressed in the tests. Experimental results reveal significant influence of water subcooling and melt superheat on debris size and morphology. Significant differences in size and morphology of the debris at different melt release conditions is attributed to the competition between hydrodynamic fragmentation of liquid melt and thermal fracture of the solidifying melt droplets. The particle fracture rate increases with increased sub cooling. Further the results are compared with the data from larger scale experiments to discern the effects of spatial scales. The present work provides data that can be useful for validation of the codes used for the prediction of debris formation phenomena.

During a hypothetical severe accident of a light water reactor (LWR), molten corium could fall in the form of jet into a water pool. The jet fragmentation is crucial process during fuel coolant interactions (FCI) which fragment into droplets and disperse in the coolant, and it may cause a steam explosion. This paper deals with a study of computational fluid dynamics on the melt jet falling into a water pool in order to get qualitative and quantitative understanding of initial premixing phase of FCI. The preliminary objectives to pursue are modeling of jet fragmentation and estimation of the jet breakup length. A commercial CFD code ANSYS FLUENT 14.0 is used for the 2D numerical analysis employing Volume of Fluid (VOF) method. The problem and simulation conditions are similar to the jet breakup tests carried out at KTH (Manickam et al., 2014). Initially, a fragmentation/breakup pattern of the jet is discussed. Further, the effect of jet diameter and the jet injection velocity on the jet breakup length is studied, with a wide range of ambient Weber number (We_a) from 1.25 to 1280. The numerical results compared with the experimental data are in a reasonable agreement. The impacts of physical properties of melt (density, viscosity and surface tension) on the jet breakup lengths are also investigated and presented. Finally the droplet size distributions are obtained based on the simulation results. These preliminary data may be helpful to substantiate the understanding of the phenomena during melt jet interactions with coolant.