IMPROVED SAMPLING ALGORITHMS IN RISK-INFORMED SAFETY APPLICATIONS

The Risk-Informed Safety Margin Characterization (RISMC) approach is developing an advanced set of simulation-based methodologies in order to perform Probabilistic Risk Analyses. These methods randomly perturb (by employing sampling algorithms) timing/sequencing of events and uncertain parameters of the physics-based models in order to estimate stochastic outcomes such as off-normal and damage states of the facility. This modeling approach applied to complex systems such as nuclear power plants requires the analyst to perform a series of computationally-expensive simulation runs given a large set of uncertain parameters. One issue is related to the fact that the space of the possible solutions can be sampled only sparsely and this precludes the ability to fully analyze the impact of uncertainties on the system dynamics. This paper describes how we can use novel methods that optimize the information generated by the sampling process by sampling unexplored or risk-significant regions of the issue space; we call this approach adaptive (smart) sampling algorithms. These methods infer the system response using surrogate models constructed from existing samples and predict the best location of the next sample. Thus, it is possible to understand features of the issue space with a smaller number of carefully selected samples. In this paper, we present how it is possible to perform adaptive sampling using the RAVEN statistical tool and highlight the advantages compared to more classical sampling approaches such as Monte-Carlo.