A computational analysis of similarity relations using helium as a surrogate of hydrogen in semi-confined facilities

Understanding hydrogen dispersion in a semi-confined space is crucial for the safe operation of hydrogen-related systems, such as hydrogen fuel cell vehicles. Analyzing hydrogen release and its subsequent dispersion can be effectively conducted using computational models, which necessitate experimental measurements for validation. For safety reasons, accidental scenarios are frequently replicated using helium as a hydrogen simulant in experiments, despite some concerns about the comparability of helium and hydrogen dispersions. Hence, this study conducts a comprehensive parametric numerical analysis of the similarities between hydrogen and helium dispersion when released in a semi-confined space. The model is based on the Macquarie Dispersion Chamber and the simulation results are first validated against the measurements. The study focuses on the impact of the leakage rate and ventilation velocity. Three release models were analysed: equal concentration (Method A), equal volumetric flow rate (Method B), and equal buoyancy (Method C). This study extends previous research on Method B, used for gas release in open spaces, to environments where flow-wall interactions become important. It also compares Methods A and C for a few scenarios, examining variations in concentration distribution, purge time, and activation time under different conditions. In the absence of ventilation, Method B effectively compares hydrogen and helium flow rates by accounting for volumetric flow rates, regardless of buoyancy effects. However, under increased ventilation velocities, Method A and Method C exhibit superior agreement across sensors, highlighting the necessity of choosing detection methods according to prevailing environmental conditions.