Improved modelling and probabilistic assessment of in-vessel steam explosions

This paper describes the implementation of a new approach to estimating the probability of damaging, in-vessel, steam explosions, which significantly reduces the problem of properly accounting for uncertainty in the physics. The program comprises separation of macrophysics from microphysics, greatly improved modelling of the macrophysics, probabilistic quantification of microphysical uncertainties, and the propagation of probability through the models by a modified L.H.S. method. The approach is suitable for a wide range of similar problems