The Phébus FP programme is investigating key phenomena involved in LWR severe accident (SA) sequences, through a series of in-pile integral experiments. The facility provides prototypic reactor conditions which allow the study of basic phenomena governing the releases, transport, deposition and retention of the fission products (FP). The phenomena studied take place in the core region, in the primary system components and in the containment building; the processes involved in these studies are thermal-hydraulics, physics, chemistry and radioactivity which are closely coupled. The fourth experiment in the series, namely FPT-2, was performed in October 2000. It involved a bundle of 20 irradiated fuel rods, with a silver-indium-cadmium in the central position. The low injected steam flow rate led to a steam starvation period during the cladding oxidation phase, in contrast with the first two Phébus-FP tests (FPT-0 and FPT-1). The fuel degradation was important during the heat-up phase, up to the formation of a well-developed molten pool. The FP release peaked during the second half of the steam starvation period, except for molybdenum. It was then rather continuous, up to the molten pool formation, for volatile elements. Low volatile FP and silver had a maximal release during the relocation events. New features have been observed for the transport of iodine in the primary circuit. thanks to gamma scanning of sampling lines with a temperature gradient from 700°C to 150°C. From the position of the deposition peaks in the lines, it is clear that the iodine vapour in the hot leg was not cesium iodide during the early zircaloy oxidation, the steam starvation and the late fuel relocation phases. During the steam-rich phase following the steam starvation, two vapour species were coexisting, one being probably cesium iodide. As in the first two Phébus tests, most of the Cs was already condensed at the hot leg temperature (700°C°). This confirms that cesium hydroxide is not the likely transported species at medium and low temperature. Other important results concerning iodine chemistry in the containment have been obtained and were presented during the 2002 CSARP meeting.
The third experiment in the series, namely FPT-4, was performed in July 1999. It involved a pre-fabricated debris bed, made of irradiated fuel fragments and oxidised cladding shards. The experiment dealt with the release of low volatile FP and actinides in solid state and during the melt progression. In the contrary to Phebus bundle tests, the release of barium was significant. This goes in the direction of the proposed interpretation of barium release : it is expected that interactions of fuel with zircaloy and/or stainless steel reduces significantly the vapour pressure of barium species. As the interaction of fuel with zirconia only occurred during the last part of the transient during the FPT-4 test, barium could be released from solid fuel before any liquefaction occurs. Another point of interest is that uranium and other actinides (plutonium, americium and curium), lanthanides and zirconium (FP) had the same fractional release. The underlying phenomenon has to be investigated.
The present Phébus-FP programme will be completed by realising the FPT-3 test dealing with the impact of B4C on fuel degradation and source term. IRSN is actively preparing a follow-up programme, named STLOC (Source Term and Loss Of Coolant). This programme aims at addressing pending issues on severe accidents and design basis loss of coolant accidents. For SAs, it is intended to treat:
- the air ingress issue,
- the consequences of reflooding a degrading core on fuel degradation, hydrogen production, RCS pressurisation and source term,
- the influence of high burn-up and MOX fuels.
For LOCA, the questions are linked to the use of high burn-up and mixed oxide fuels, together with modern cladding alloys. Fuel relocation in the clad balloons has already been observed in previous experiments. The STLOC programme will investigate the time of fuel relocation, the accumulation of fuel in the balloons, the impact on safety criteria (peak cladding temperature and equivalent cladding reacted). It will also provide information on flow blockage and quenching in bundle geometry and representative conditions, with special emphasis on the size of flow blockage at high burn-up, the possible loss of rod integrity at quenching, the reduction of residual cladding ductility after quenching.
A tentative test matrix for a first series of five STLOC tests has been elaborated, including three (or two) SA experiments and two (or three) LOCA tests. The first experiment will be devoted to air ingress. Reactor calculations are underway to determine the test scenario. They investigate the amount of air that may enter a reactor pressure vessel after failure of its lower head, together with the amount and state of the residual fuel in the vessel.