Dynamic system reliability assessment for flare gas recovery safety units subject to covariates

Prevention of pollutant emissions in the atmosphere may be achieved via flare gas recovery system (FGRU) that consists of expensive components and costly repair processes. Hence nominating the most suitable safety system needs to carry out reliability assessment. There are two types of safety system technologies known as fast opening valve system (FOVS) and seal drum system (SDS), which their reliability estimation is a complicated task due to dynamic operating conditions and absence of enough historical data. To compensate them, we proposed a new approach using a hybrid of fault tree and fuzzy inference system to develop the system reliability surface versus a few operating pressures and temperature. The result reveals an average of 22.4% more reliable for FOVS and it is recommended for putting into practice.