On the Use of Fault Tree Analysis to Capture Dynamic and Multistate Aspects in the Analysis of Hydrogen Systems

Hydrogen systems are being deployed to reduce fossil fuel dependence and support cleaner energy. Large-scale upscaling of hydrogen technology demands a strong understanding of reliability and risk. Fault tree analysis is a way to support this understanding, as a deductive technique utilized to assess failure paths and probabilities. However, traditional fault trees are often criticized for being static, having limited ability to model dynamic behaviors and multistate events, and relying on fixed probabilities for quantification. This paper explores alternatives to the traditional method and considers their relevance and attractiveness when assessing the reliability and risk of hydrogen systems. The focus is on how to capture dynamic behavior and multistate aspects in the technique. The applicability of the dynamic fault tree and multistate fault tree techniques was studied using three examples of a hydrogen storage explosion due to overpressure. The findings indicate that the two techniques add relevant aspects but also introduce complexities and uncertainties. To improve decision-making in this context, expanding the treatment of uncertainties is recommended to achieve more informed decision support. It is suggested that the strength of knowledge analysis be incorporated into the analyses.