This paper gives an overview of the release behaviour of fission products and fuel elements in the Phébus FPT4 debris bed test, based on the analysis of non destructive examinations performed on the test section after the experiment (radiography, tomography, g-scanning), on the analysis of on-line gamma spectrometry measurements performed in the circuit and the containment and on the analysis of post test measurements performed on upper plenum deposits and filter-cartridge deposits. A global analysis of the data is presented in order to discuss the observed release of elements of interest. A comparison with releases observed in other Phébus tests performed with fuel bundles is presented.
Results indicate that releases of noble gases (Xe, Kr) and high-volatile fission products (Cs, I) were nearly complete and comparable to those obtained during Phébus tests performed with a fuel bundle geometry (FPT1, FPT2). Volatile fission products such as Mo, Te, Rb, Sb were released significantly as in previous tests. Ba integral release (30 % of its initial inventory) is more comparable to that observed during analytical tests, such as VERCORS tests, than that observed during FPT1 and FPT2 (1 % of its initial inventory). This higher Ba volatility in FPT4 is attributed to the absence of some metal oxides such as Fe or Cr oxides and the delayed formation of UO2-ZrO2 mixtures relative to other Phébus FP tests; these reducing Ba oxide vapour pressures. Release of Ru was comparable to that observed during FPT1 and FPT2 (1 to 2 % of its initial inventory). However, as in other Phébus tests, Ru distribution in the fuelled section after the test suggests significant Ru volatility followed by fast re-deposition in the fuelled section. The same released fraction was measured for all lanthanides (Eu, Nd, Ce, La, Gd) and fuel elements (U, Pu, Am, Cm) suggesting a release mechanism involving fuel particles which deposited onto plenum surfaces above the degraded fuel. The total mass released during the test and recovered above the degraded fuelled section is estimated to be of the order of 50 g which shows a low release scenario was obtained during the test, explained by the test section blockage by the degraded fuel (relocalised or molten material) which induced a steam by-pass through the shroud containing the fuel.
(1) : IRSN
(2) : CEA