A Hybrid Reliability Assessment Framework for Green Hydrogen Production Systems under Data Scarcity

The rapid expansion of green hydrogen production has intensified the need for reliable and resilient system architectures capable of operating under highly uncertain conditions. Despite increasing attention to economic and environmental aspects, the operational reliability of green hydrogen plants remains insufficiently addressed in the literature, largely due to the scarcity of historical failure data. This work proposes a hybrid reliability assessment framework that integrates qualitative knowledge from the literature with stochastic modeling techniques to evaluate the operational reliability of solar-based green hydrogen production systems. The methodology combines Failure Mode and Effects Analysis (FMEA) for systematic identification and classification of failure mechanisms, Fault Tree Analysis (FTA) to model causal relationships leading to system unavailability, and Monte Carlo simulation to propagate uncertainty throughout the system. In the absence of deterministic failure data, failure probabilities are modeled using Weibull distributions, whose parameters are inferred from qualitative FMEA inputs and physical analogies, while uncertainty is explicitly represented through stochastic sampling. The framework is applied to a representative photovoltaic-powered PEM electrolyzer system, focusing on the top event defined as unplanned shutdown or capacity reduction. The proposed approach demonstrates how dispersed qualitative information can be transformed into quantitative reliability indicators, even under severe data scarcity. The framework provides actionable insights for asset management, maintenance prioritization, and early-stage design decisions, supporting the development of more robust and reliable green hydrogen production plants.

Keywords: Green hydrogen, Reliability analysis, Fault Tree Analysis, Failure Mode and Effects Analysis, Monte Carlo simulation, Data scarcity, Renewable energy, Photovoltaic-electrolyzer integration, System availability, Risk Analysis, Stochastic modelling..