Quantification of the probability of containment failure caused by an in-vessel steam explosion for the Sizewell B PWR

Issues on containment integrity during severe accidents

“…An in-vessel steam explosion may also induce RPV failure. Prediction of triggering of an in-vessel steam explosion and its yield has large uncertainty and therefore it calls for a wide distribution of the causative parameters [9]. The melt failure mechanism, for example, formation of a slug or its absence, has significant effect on the power of the explosion.…”

“…Severity of consequences depends largely on the melt release states from RPY. Major processes constituting various ex-vessel phenomena are containment thermohydraulics, fission products transport, and melt dispersion [1,[7][8][9][10][11][12][13][14][15][16][17][20][21][22].…”

“…Variability in structural modeling/analysis is also to be accounted. Air leakage A. In-Vessel Phenomena In-vessel phenomena for PWR can be described under general topics of thermohydraulics, oxidation of core materials, core relocation and failure of reactor pressure vessel (RPV) [7][8][9][10][11].…”

Lower head integrity under steam explosion loads

“…They found that explosions with yields above 1 GJ produced lower head failure. Fuel masses involved were from a few to some tens of tons, pressures were in the kilobar range, and failures occurred at 1 to 2 ms. Turland et al (1995), in their a-mode failure assessment for Sizewell, decided against taking credit for explosion venting due to lower head failure. They noted that the lower head static capability in Sizewell is 650 bar, that there is no evidence of the kilobar pressures found from ideal thermodynamic models, and that experiments and more realistic modeling suggest that for 1 GJ explosions, peak pressures will more probably be in the range 400 -800 bar.…”

“…They noted that the lower head static capability in Sizewell is 650 bar, that there is no evidence of the kilobar pressures found from ideal thermodynamic models, and that experiments and more realistic modeling suggest that for 1 GJ explosions, peak pressures will more probably be in the range 400 -800 bar. Both Theofanous et al (1987) and Turland et al (1995) had indicated the need to re-examine these results, once the capability to properly represent the dynamic explosion loads became available. This time has come.…”