Screening of accident scenarios of hydrogen refueling stations based on HyRAM+ and Latin hypercube sampling

Abstract

Due to the boom in hydrogen energy, risk assessment for hydrogen refueling station (HRS), which is as an infrastructure to supply hydrogen energy, is increasingly important. It is also crucial to screen out accident scenarios with low probabilities and low harms. This study theoretically focuses on an accident scenarios screening method for HRS based on Hydrogen Plus Other Alternative Fuels Risk Assessment Models (HyRAM+) and Latin hypercube sampling (LHS). According to the unit division of process flow in HRS, accident scenarios were preliminarily screened by accident consequence frequencies and maximum leakages. By comparing credible factor C and Potential Loss of Life (PLL), accident scenarios were screened secondly to obtain maximum credible accident scenario (MAS). The uncertainties of hydrogen leak, leak isolation, ignition possibilities, and the occupant positions were considered using LHS method for accident consequence frequencies and PLL. The assessment results for one HRS indicate that consequence frequencies of jet fire and explosion may be underestimated by HyRAM+. Moreover, MASs of jet fire and explosion were determined by complete rupture of the pipeline of the hydrogen compressor unit and explosion has a more significant impact on property loss than jet fire.