Seismic probabilistic risk analysis based on stochastic simulation of accelerograms for nuclear power plants in the UK

This article presents an approach to probabilistically assess the seismic risk of nuclear power plants (NPPs) in the UK. The approach proposed is based on direct stochastic simulation of the seismic input to conduct nonlinear dynamic analysis of a structural model of the NPP analysed. Therefore, it does not require the use of ground motion prediction equations and scaling/matching procedures to define suitable accelerograms as is done in conventional approaches. Additionally, as the structural response is directly calculated, it does not require the use of Monte Carlo-type algorithms to simulate the damage state of the NPP analysed. However, it demands longer use of computer resources as a relatively large number of nonlinear dynamic analyses are needed to perform. The approach is illustrated using an example of a 1000 MW Pressurised Water Reactor building located in a representative UK nuclear site. A comparison of risk assessment is made between the conventional and proposed approaches. Results obtained are reasonable and well constrained by conventional procedures; hence, it can confidently be used by the UK New Build Programme in the next two decades to generate 16 GWe of new nuclear capacity.