Reduction of carbon dioxide emissions in the steel and mining industries can be achieved through the use of hydrogen gas; however, hydrogen infrastructure is susceptible to gas explosions, making effective risk management essential. The development of risk assessment tools, e.g. accident databases and consequence models, for quantitative risk analysis (QRA) is therefore critical for achieving a realistic understanding of explosion risks. This study presents a critical review of hydrogen risk assessment tools used in QRA, with the aim of evaluating their current state and identifying research needs for reliable risk prediction in large-scale hydrogen infrastructure. It is found that accident databases lack sufficient detail on explosion characteristics, particularly in distinguishing between deflagration and detonation events. Consequence models for confined explosions are typically derived from or validated against medium-scale experiments with significant variability, while open atmospheric models rely on a limited experimental basis. In addition, many models are based on assumptions originating from hydrocarbon gas explosions or solid explosives, raising concerns regarding their applicability to hydrogen. The results indicate that improved accident reporting is required to develop reliable frequency data, and that consequence models must be further developed and validated through large-scale, hydrogen-specific experimental campaigns. Addressing these gaps is essential for enabling more accurate QRAs and a realistic assessment of explosion risks in large-scale hydrogen infrastructure.
Andersson et al. (Fri,) studied this question.