Energy Reports (Apr 2022)
PFD analysis of LNG fuel gas supply system for improving combined-cycle power plant safety
Abstract
In order to determine an unavailability of each safety function subsystem that influences on an overall unavailability of entire protection system for safety analysis and enhancement, a modeling is required. This article presents a safety system modeling to evaluate the detected–undetected failures and repairable behaviors of each safety function subsystem in the studied fuel gas supply system (FGSS), which is utilized in a combined-cycle power plant. The proposed modeling deals with probability of failure on demand (PFD) analysis for calculation of the unavailability of three safety function subsystems, which are gas analytical subsystem (GAS), fuel gas controller subsystem (FGC), and shutdown valve subsystem (SDV). The interested GAS includes two different groups of gas analyzers in 2-out-of-2 (2oo2) voting scheme, when using three gas analyzers in each group in 2-out-of-3 and diagnosis (2oo3D) voting scheme. The interested FGC consists of two controllers in 2oo2 voting scheme, when the structure of each controller is in 1-out-of-1 and diagnosis (1oo1D) voting scheme. The interested SDV consists of two shutdown valves, which are installed in 1-out-of-2 (1oo2) voting scheme. The reliability block diagram is utilized for representative of combination of all three safety function subsystems, and the fault tree is utilized for calculation of PFD unavailability of each safety function subsystem. As the results obtained from the proposed safety function modeling, the influence of three safety function subsystems on the total unavailability of the FGSS is described. In addition, an example of the proposed modeling application to provide two guidelines for effective engineering design and operation is also presented. The first is a guideline to implement an effective human machine interface (HMI) at an operator workstation for real-time monitoring of the studied FGSS in the presence of failures that can be automatically detected by device diagnosis. The latter is a guideline to maintain and troubleshoot the failures that can be revealed by proof test. Based on diagnostic information and proof test results, operation and maintenance savings can be achieved.
