Assessing the reliability of dynamic systems for safety units recovering flare gas, considering covariates

The Flare Gas Recovery System (GR) is a critical component in preventing the release of pollutants into the atmosphere. However, these systems are expensive to install and maintain, so ensuring their reliability and effectiveness during operation is critical task. Two safety system technologies have been developed for GR, the fast opening and closing valve system (OVS) and the closed drum system (CDS), but the dynamic operating conditions and lack of historical data make reliability estimates as a crucial complicated task. To address this issue, we propose a novel approach to develop a system reliability as a response surface based on multiple operating pressure and temperature using a hybrid fault tree and a fuzzy inference system. The result reveals an average 22.4% improvement in reliability for OVS compared to CDS in various operational scenarios. Our proposed method provides an operative technique to assess the reliability of GR security systems considering various operating conditions. Results also can help decision-makers to choose the security technology that best fits their particular application needs, ultimately reducing maintenance of costs while ensuring optimal performance over the long term.