No abstract

No abstract

No abstract

Heavy-duty gas turbines are usually devised in power plants to generate electrical energy. Sudden failure in any of its parts or subdivisions will result in a decrement of the efficiency of the system or emergency shutdown of the system. The highest risk of failure in these turbines is subjected to the hot gas path (HGP) of the turbine. Due to the existence of uncertainty in diagnosing process or damage growth, in this research, a modified risk-based probabilistic failure analysis model using Bayesian networks (BN) was developed. First, a failure model was developed using the Fault Tree Analysis, and then it is transformed into a BN model. This model is capable of predicting and diagnosing critical components and critical failure modes and mechanisms for each component by updating failure probabilities. Moreover, in order to enhance the application of the proposed model and to identify the risk factors, the sensitivity analysis of the HGP components is presented with applying definitions of importance measures and extend them to BN. The sensitivity analysis and application of its results for making decisions during the system operation will enhance the reliability and safety of the system.

No abstract