In-vessel retention: on going activity on in-vessel core and debris coolability

European Review Meeting on Severe Accident Research (ERMSAR 2005), Aix-en-Provence, 14-16 Novembre 2005

I. Lindholm (1), Kresna Atkhen (2), M. Bürger (3), F. Fichot (4), S.H. Park (5).

The motivation to investigate reflooding is the possibility of retardation of melting and attack on the reactor pressure vessel or the containment basemat by core debris or even termination of the accident. Particulate debris bed can form in the core region during core degradation by shattering as result of flooding events, in the lower head as a result melt jet fragmentation in the water pool and similarly in water-filled containment cavities following pressure vessel failure.
Classical analyses were based on the dryout heat flux yielding the limit of coolability of laterally uniform debris beds under conditions with top flooding conditions. In this situation the counter-current flow due to inflow of water through the upper region of the bed with accumulated steam may strongly limit water access. In mound-like bed configurations, as may be expected in the lower plenum and reactor cavity, multidimensional effects should be accounted for. As also indicated by experiments, the 1-D application of classical dryout heat flux correlations underestimates coolability in such situations.
Specific experiments on constitutive laws (separate effects tests) are required as well as integral experiments exploring 2D/3D features of behaviour. DEBRIS experiments at IKE are presently emphasising local constitutive laws under boil-off and quenching modes, with various flow patterns in co- and counter-current flow of steam and water. In the STYX experiments at VTT, dryout heat flux measurements under top flooding have been performed with laterally uniform mixtures of particles. Differing from most classical investigations, alumina sands with irregular particle shapes and size distributions oriented at the FARO experiments (JRC Ispra) have been used. The POMECO experiments at KTH also used sands with relatively small particle sizes. In the COMECO facility (KTH), downcomer effects have also been studied in pools of melt. Major outcomes of the experiments at present are :

• significant differences in bottom versus top flooding modes,
• the necessity to explicitly take into account interfacial friction between steam and water in modelling, underlined by results from the DEBRIS experiments,
• significantly reduced coolability with mixtures of irregularly shaped particles in STYX experiments,
• indication that a different determination of averaged particle sizes has to be applied than usually (e.g. number averaged versus surface averaged or even mass averaged),
• improved coolability due to downcomer effects in POMECO/COMECO, investigated especially with smaller particle sizes.

Multidimensional analyses of coolability are required to really estimate the cooling potentials and options. The basic limitation to coolability lies in water access and steam removal. Friction laws determining this, depend on the phase volumes of water and steam. The WABE code of IKE provides a 2D model, IC/CAT (IRSN) and MC3D-REPO (CEA)codes in addition 3D options. Concerning the constitutive laws, basic derivations of heat transfer coefficients and effective thermal properties are being developed and applied in the frame of IC/CAT and MC3D-REPO whereas in the WABE code simplified, empirically oriented approaches are used. Analytical efforts considering SARNET are :

• Joint interpretations about the DEBRIS experiments with the aim to elaborate adequate modelling for interfacial friction, as well as to check the importance of this for safety analyses.
• Extended calculations and comparisons on reflooding of hot particulate debris, with respect to experiments as well as for reactor conditions considering hot debris in the lower plenum of the RPV with WABE and IC/CAT codes as well as other available codes.
• Calculations on the SILFIDE experiments have been performed with WABE and MC3D. Further analyses are required.

Based on the already performed and planned calculations of experiments as well as the joint interpretations, additional experimental checking and joint planning for new experiments will be done. Furthermore, recommendations for the ASTEC module will be derived on the basis of the joint activities in SARNET.

(1) : VTT
(2) : EdF
(3) : IKE
(4) : IRSN
(5) : KTH