Core-melts at Kurchatov

The concept of arresting the progression of a severe accident in a light water reactor vessel through external cooling has been recognised as a useful accident management measure. This idea has considerable appeal because it avoids the ex-vessel progression of the severe accident and the consequent major threats to containment integrity.

If the retention of a melted core in-vessel becomes an important safety objective for present or future nuclear power plants, care has to be taken in evaluating the various phenomena related to its feasibility. Since prediction of the relevant phenomena has to be performed for the prototypical accident conditions, the applicability of the measured data, or of the correlations derived from the measurable data, has to be established, and the uncertainties determined. Most uncertainties are introduced by the aspects of the experiments which are not representative of the reactor situation, the main one being the use of simulant fluids for the simulant melt pool experiments. The Rasplav Project was designed specifically to address this concern.

RASPLAV PROJECT OBJECTIVES

The principal objectives of the first phase of the Rasplav Project were as follows:

• Develop a technology for high temperature experiments with prototypical materials of reactor core melt which will allow heating and maintaining a major part of the 200 kg loading at a temperature above the liquidus temperature (over 2600°C).

• Conduct confirmatory large-scale experiments to investigate the behaviour of the melt of core prototypical materials in the reactor vessel lower head, as well as to determine any differences in the behaviour of the melt as compared with the simulation liquids.

• Conduct supporting small-scale experiments to determine the behaviour of the melt at high temperatures as well as the main thermo-physical properties required for experiment analyses.

• Conduct salt experiments to study heat transfer processes of molten core material and to justify the choice of the procedure of large-scale experiments.

• Develop an analytical model and computer program for analysing and interpreting the results of confirmatory and supplementary experiments and the pre- and post-test analyses.

FIRST PHASE – MAIN RESULTS

The above principal objectives of the first phase were achieved. A series of small scale and supporting experiments, which had been conducted while studying the feasibility of large scale experiments, enabled the solution of major technical problems. Technology for conducting a large scale experiment under controlled conditions was developed. This technology is based on the application of the technique of side-wall heating of corium and a protective layer of materials which prevents undesirable interactions of corium with the structural elements. Computation aids enabling pre- and post-test analyses were also developed.

A number of physical properties of corium of different composition were measured up to 2850°C. These properties include viscosity, heat conductivity, electric conductivity, and melting temperature. The values of most of the above properties at such high temperatures were obtained for the first time.

Two large scale (200 kg of corium) experiments were carried out and important data on the differences in the behaviour of corium melt and simulating liquids in the range of temperatures exceeding 2550°C under controlled conditions were obtained. The main thermal and physical parameters of the processes at the facilities were controlled and recorded during the operation of measurement systems. The results point to a significant difference between the behaviour of corium and that of simulation liquids. Liquid stratification and a variation with height of the relative distribution of U and Zr were observed experimentally for different compositions of corium.

Five series of molten salt experiments were carried out in the first phase of the project investigating the use of various salts as corium simulators. The results obtained have improved the understanding of the effect of heating techniques and crust formation on the process of natural convection. These results enabled the verification of computation models at high Rayleigh number (in this case a modified Rayleigh number for thermal convection phenomena, expressing the ratio of the bouyancy forces arising from thermal expansion to the viscous resisting forces), as well as the effect of different heating techniques, the presence of a transient zone, melting and resolidification of the crust.

Analytical and software tools were developed for experimental investigations. The models developed take into account the most important processes, such as natural circulation of the melt, formation of crusts and their impact on the dynamics, heat behaviour of structural materials, etc. A 3D approach allows the consideration of structural peculiarities of the experimental facilities, including real geometry and heating techniques. Several additional models, including a model of the so-called mushy zone (a mixture of solid and liquid phases), were incorporated into the calculation code. The calculation program was verified using the data of corium experiments as well as experiments with simulating liquids, including experiments which were carried out at the salt facility within the scope of the project. The results of the above experiments were described with a sufficient degree of accuracy to allow the computation programs that had been developed to be applied intensively to prepare large scale experiments.

The chemistry investigations performed identified the presence of carbon in the corium mixture of the first two tests as a possible catalyst for the stratification (separation) process. Additional materials-chemistry research identified the presence of the miscibility gap in the ternary phase diagram of the UZrO system as a possible reason for the melt stratification process. In this context, the corium composition is an important element, since its location in the phase diagram would determine whether it is within the miscibility gap and would separate into two liquids of different composition and different density. An additional uncertainty is that the boundaries of the miscibility gap region in the phase diagram are not known.

SECOND PHASE – OBJECTIVES

The phenomena found in the first phase specific to corium melt behaviour led to a second phase of the project directed at clarifying the physical mechanisms which govern corium melt behaviour, and in particular, to establish conditions under which stratification occurs, as well as to study mass transfer mechanisms. The database on the properties of materials was also to be considerably expanded, including coverage of surface tension.

The second phase of the project runs from July 1997 to June 2000. Two further 200 kg scale molten corium tests have been carried out in this phase, along with many small and medium scale tests. On 6 July 1999 the final large-scale test was successfully carried out. During the test, core material was heated to over 2500°C, and was kept in a steady state at this temperature for about three hours. The post-test examination will as usual consist of sectioning the solidified material and performing metallographic examination to gain information yielded by the test.

The chemistry of the UZrO corium mixtures is also affected by the presence of additives. Clearly, the prototypic accident corium melt in BWRs and VVERs could contain carbon from the B4C control rods and also steel components. Similarly the PWR corium melt could contain additives from control rod materials, clad (eg tin and tin compounds) and stainless steel components if they all do not vaporise at the prevailing temperatures of the corium melt. Certainly, the corium melt for both PWRs and BWRs would contain substantial quantities of low-volatile fission compound products (eg La2O3, CeO2, BaO etc) which may also affect the chemistry of the corium melt. Some work has already been done in the second phase of the project on the effect of such additives, but further work would be needed fully to understand this. Consideration is also now being given to the effect and significance of the distribution of fission products in molten pool, and to further work in this area.

Work until the end of the project in June 2000 will involve further post-test metallurgical examination and numerical analysis, and assessment of the significance and application of the results.

The question of what special information was derived from the Rasplav corium experiments (and the supporting corium experiments) which would provide confidence in performing evaluations for prototypic accident conditions can be answered:

• Rasplav experiments showed that corium melt pool convection may occur, similar to that which occurs at the equivalent Rayleigh number in facilities employing simulant materials.

• Rasplav experiments showed that there may be possible melt stratification, which would affect the natural circulation in the pool and change the thermal loadings on the vessel wall, thus affecting the margins for the in-vessel melt retention accident management concept.

• The existence of melt stratification for prototypical conditions, however, cannot yet be predicted or certified. The data obtained are not sufficient to unravel the complex sub-issues of melt chemical behaviour as a function of corium composition and as a function of the various additives that may be present in prototypical corium.

A conference will be held at the end of 2000 to present the results.
Related Articles
Kurchatov starts waste removal