Experimental study on the effects of boiling during molten jet and coolant interactions

https://doi.org/10.1016/j.anucene.2020.107392Get rights and content

Highlights

  • The effects of boiling on molten jet and coolant interactions were studied experimentally.

  • The process of energetic and mild FCIs were clearly shown by visualization and other characteristic parameters.

  • The hydrodynamic and thermodynamic fragmentation mechanism were analysed under different boiling modes.

Abstract

In nuclear power plants, core melt accidents are often accompanied with fuel–coolant interactions (FCI), which may escalate to steam explosion destroying the integrity of structural components and even the containment under certain conditions. The vapor film and two-phase region formed by boiling around the molten jet at high temperature have a great influence on the fragmentation of the jet and the possibility of steam explosion. In this study, by changing the initial temperature of coolant, a series of experiments on the interactions of molten tin at high temperature with water have been done and several distinct interactions under different boiling conditions including stable film boiling, unstable film boiling and nucleate or transition boiling were observed. These interactions were distinguished by the high-speed photography, liquid level swell, dynamic pressure, water temperature variation and jet breakup length during interactions as well as the morphology and size of debris after interactions synthetically. It was found that energetic jet-water interactions were only possible under unstable film boiling, and nucleate or transition boiling conditions. While under the stable film boiling, the jet was surrounded by a very stable vapor film, so only mild interactions occur. With the decrease of coolant temperature, the boiling mode transferred from stable film boiling to nucleate or transition boiling. Significant differences in jet breakup length as well as size and morphology of the debris were observed. Because less steam was generated, the melt was more likely to action with water than steam, and water was more conducive to the fragmentation caused by interfacial instability due to higher density. In addition, the thermodynamic fragmentation caused by violent boiling gradually played a more important role. Furthermore, the results were compared with existing theories which ensured the effects of boiling during FCI.

Introduction

During a hypothetical severe accident in nuclear reactor, FCI will occur when the core melt is in contact with the coolant in the lower head of the pressure vessel, or in the reactor cavity after the lower head is melted, or when the debris bed formed in the reactor cavity is submerged and flooded again, which may result in steam explosion and damage the core components or pressure vessel, even lead to the failure of containment. Therefore, FCI has drawn substantial attention in the safety analysis of reactor severe accidents.

The general process of FCI phenomenon is the molten melt fragments into small particles increasing contact area and transferring heat to the coolant rapidly to generate steam, possibly leading to a locally pressure. An energetic FCI (steam explosion) occurs if the melt fragments so violently that the local pressure rise time has a time scale smaller than the time scale for inertial pressure relief. A mild FCI is a phenomenon that does not have the shock wave characteristic of a steam explosion and the melt fragments relatively slowly (Chu and Corradini, 1989). Whatever the mode, the mixing zone, the void fraction in the coolant and the melt droplets size formed by jet fragmentation in premixing phase directly determine the steam explosion possibility, the explosion power and the debris bed characteristics (size distribution and morphology of debris particles as well as bed configuration) (Meignen et al., 2014). The jet fragmentation is a very complex process. It is believed that the jet fragmentation is depended on hydrodynamic mechanisms including the boundary layer stripping, the instability of surface waves on molten metal surface such as the Rayleigh–Taylor instability, Kelvin–Helmholtz instability as well as critical Weber number criteria, and thermodynamic mechanisms such as boiling and thermal stress due to surface solidification. The vapor film and two-phase region formed by boiling around the molten jet at high temperature have a great influence on the fragmentation of the jet.

Over the years, the role of boiling on jet fragmentation have been studied widely. Saito et al. (1988) investigated the penetration behaviors using a hot water jet injecting into Freon-11 and liquid nitrogen at low temperature. It is concluded that the generated vapor plays an important role. The vapor generated at the leading edge due to Rayleigh-Taylor instability increases the jet penetration. Bürger et al. (1995) compared the model with experiments under non-boiling as well as film boiling conditions and found that under whatever conditions, the stripping at the melt surface supported by interface instabilities was considered as a major mechanism of the jet breakup. While under non-boiling, the stripping was due to relative flow of melt and water. Under film boiling, the vapor flow around the melt determined the breakup regime. The vapor density and velocity were the key factor. On the contrary, Dinh et al. (1999) performed a systematic study of jet breakup for jet diameters ranging from 2.5 to 25 mm using various pairs of simulant liquids (e.g. water–paraffin, Cerrobend–water, Cerrobend–paraffin). They observed that the macroscopic hydraulic parameters such as initial relative (jet-pool) velocity, the coolant density and viscosity, the transient flow pattern and the voiding of the coolant pool were the major factors which determined the characteristics of jet breakup, rather than the vapor film characteristics. Bang et al., 2003, Bang et al., 2018 experimentally investigated the jet breakup characteristics in non-boiling condition using Wood’s metal to isolate the effect of steam generation. They found the Kelvin–Helmholtz instability was the most probable cause of jet breakup. In addition, it was observed that the free-falling modes of jet in air space had large effects on average debris size. The average debris size in case of the fully-flooded condition was much smaller than that of the partially-flooded condition because the entrained air layer between the melt and water in fully-flooded condition acted like a vapor film in boiling condition. Lu et al. (2016) visually investigated the breakup of different high superheated molten metals on FCI process. For aluminum with much large thermal conductivity, the vapor film generated immediately on the surface as it came into contact with the coolant. And the stable film boiling prevented the breakup. While for lead and bismuth with small thermal conductivity, the vapor film was so thin that they quickly transferred to transition boiling, the intense pressure pulse from the violent boiling broke up the molten metals.

Another concerned issue is under what conditions the energetic FCI happens. Kondo et al. (1995) conducted a series of Wood metal at various initial temperature of melt and water to realize different boiling conditions. The results showed that energetic jet-water interactions were only possible under relatively narrow initial thermal conditions. Abe et al. (2004) studied the quench behavior of molten U-alloy at low temperature. The results suggested that when the temperatures of melt and water were in the so-called thermal interaction zone, the vapor explosion may occur. KROTOS (Huhtiniemi et al., 1999; Huhtiniemi and Magallon, 2001) investigated the vapor explosion of the core melt and alumina/water interactions with or without external triggering at very high temperatures. Experiments showed that spontaneous vapor explosion occurred in alumina at high subcooling, but not at low subcooling as well as core melt. When triggering, the explosive power of alumina was greater than that of core melt. It was possibly because either the vapor void fraction in the mixing area formed by core melt fragmentation was higher than that of alumina, or that the surface of core melt solidified to form a crust, which was not conducive to the triggering and propagation of steam explosion, or both. While the TROI experiments (Song et al., 2002; Song and Kim, 2005) conducted by KAERI recorded a spontaneous steam explosion of the core melt/water interactions. This overturned the consensus that real core melt did not give rise to spontaneous steam explosion. The causes of the spontaneous explosion need further investigation.

Even though lots of studies focusing on FCI have been carried out, many phenomena and mechanisms during FCI still do not reach a consensus due to its complications. As discussed in the preceding sections, boiling or not and boiling modes have a great influence on the fragmentation mechanism and the possibility of steam explosion. In this paper, by changing the water temperature, the interactions between the melt and coolant under different boiling modes were realized to study whether the energetic FCI could occur and the corresponding fragmentation mechanism. The interaction process was described in detail by visualization, dynamic pressure, liquid level swell, coolant temperature change, jet breakup length, debris morphology and size. Low melting point metal Tin (Melting point 232 °C) was used as the simulant melt at high temperature (800 °C). A medium-scale experimental facility was built, which could avoid that the phenomena of large-scale experimental facilities were not easy to be observed. Also the experimental results were somewhat comparable to the actual reactor conditions. It was helpful to understand the steam explosion and fragmentation mechanism under different boiling modes in actual reactors. The detailed experimental results have also been used for verification of a new FCI calculation code.

Section snippets

Experimental apparatus

Fig. 1 shows schematically the experimental apparatus named ‘Test for Interaction of MELt with COolant’ (TIMELCO). It consists of a furnace, a release tube, a fast valve, a reaction vessel, a pressure vessel and instrumentation and control system.

The furnace is used to melt the simulant material to meet the experimental temperature. The highest temperature of the furnace can heat up to 2750 °C, with a mass capacity of 10 kg. The melt temperature is measured by a B-type thermocouple at low

Experimental results and discussions

According to conventional boiling curve, when a hot liquid jet falls into a pool of cold liquid, a vapor film is formed on the surface of the melt to isolate the hot and cold liquid if the temperature difference is high enough. When the interface temperature is higher than the minimum film boiling temperature, a stable film boiling mode will happen. When the interface temperature is near the minimum film boiling temperature, the formed vapor film is unstable, easily collapses during the falling

Conclusions

An experimental study was carried out to assess the effects of boiling on the molten jet and coolant interactions. From a series of tin’s metal jet-water experiments, by changing the initial temperature of coolant, three distinct interactions under different boiling conditions including stable film boiling, unstable film boiling and nucleate or transition boiling were observed.

It was found that energetic jet-water interactions were only possible under the unstable film boiling, and nucleate or

Acknowledgement

This work was financially supported by China Institute of Atomic Energy (2016-DGB-I-KYSC-0015).

Cited by (2)

View full text