Dynamic Hazard Identification on Solid Oxide Fuel Cell system using Bayesian Networks

Accidents are expected when operating Solid Oxide Fuel Cell (SOFC) unit system in a plant due to its complexity and conditions. SOFC which consists of risky components such as combustor, reformer, heaters stack poses risk fire explosion especially high temperature. In addition, other factors failure rate components, quantity materials, gas leakage chemical characteristics involved further increase the risks an alarming level. reality, these conditions evolving depending on current situation make it challenging determining actual risks. Since technology is still emerging not widely used, studies hazard identification very minimal. The present work develops new evolution framework for mapped into Bayesian network model using open-source software programs, GeNie bring dynamics identifying hazards. This allow all potential hazards be updated real-time ensure safe implementation plant. With this, contributing can estimated with higher precisions reduce accidents probability. Sensitivity analysis also carried out determine how input parameters influencing identified Results showed probability occurring approximately 21% 7% respectively. Operating (temperature pressure) main causes Higher temperature increases Fire from 17% while pressure Explosion 18%. occurrence final dynamically served guideline safer SOFC.