Dynamic probabilistic risk assessment (DPRA) is a systematic and comprehensive methodology that has been used and refined over the past two decades to evaluate the risks associated with complex systems such as nuclear power plants, space missions, chemical plants, and military systems. A critical step in DPRA is generating risk scenarios which are used to enumerate and assess the probability of different outcomes. The classical approach to generating risk scenarios is not, however, sufficient to deal with the complexity of the above-mentioned systems.

The primary contribution of this dissertation is in offering a new method for capturing different types of engineering knowledge and using them to automatically generate risk scenarios, presented in the form of generalized event sequence diagrams, for dynamic systems. This new method, as well as several important applications, is described in detail. The most important application is within a new framework for DPRA in which the risk simulation environment is guided to explore more interesting scenarios such as low-probability/high-consequence scenarios. Another application considered is the use of the method to enhance the process of risk-based design.