A fuzzy reliability assessment methodology for city gas stations based on an extended T-S fault tree

Abstract

City gas stations (CGSs) play a crucial role in ensuring a stable and safe supply of natural gas to urban users. However, as the service time of stations increases and the performance of components deteriorates, concerns about the safety and reliability of these station have grown among operators and local government authorities. This paper proposes a fuzzy reliability assessment methodology for CGSs that considers the polymorphism of component faults and the uncertainties associated with fault relationships, failure probabilities, and fault magnitudes. The methodology utilizes T-S fuzzy gates to describe the correlation among events and constructs a T-S fuzzy fault tree for CGSs. Component fault states are represented using fuzzy numbers, and a fuzzy group decision-making approach is introduced to evaluate the current fault magnitude of components. To handle the uncertainty caused by sparse failure sample data, a Bayesian updating estimation method is presented to estimate the failure probabilities of components. Furthermore, T-S fuzzy importance analysis is applied to identify the weak points in the CGS system. The effectiveness of the developed methodology is demonstrated through a case study of reliability analysis of a city gas distribution station. The research findings provide valuable support for optimizing the design and implementing preventive maintenance of CGSs.

Keywords

• Extended T-S fault tree
• Bayesian updating estimation
• Fuzzy group decision-making
• Reliability assessment
• City gas station