ABSTRACT
In this chapter, the authors introduce the Swiss cheese accident model and explain how rare catastrophic risk events arise and how they might represent, explain, and quantify them. The authors also introduce the idea of bow-tie modeling in the presence of identifiable hazards that threaten the system and make it vulnerable. They cover fault tree analysis and event tree analysis here and show that how Bayesian networks (BN) can be used to model these approaches to risk quantification and explain why BNs are so much more powerful and flexible. They also cover soft systems, causal models, and risk arguments as a way of modeling operational risk in complex situations that demand a less granular approach to modeling and quantification. The authors describe KUUUB factors as a way of embedding subjective judgments about the uncertainty behind their risk estimates to support the generation of operational models with long tail characteristics.
