A new approach to recognize performance deviation pattern of a micro gas turbine under different degradation factors

The pivotal role of micro gas turbines (MGT) as backup power sources for renewable energy systems and distributed power generation has significantly increased the importance of condition monitoring. Fouling and erosion are two main factors of gas turbine deterioration. Gas path parameter analysis is a popular way to carry out more accurate and robust diagnostics. Due to the high nonlinearity of the MGT dynamics, this analysis is a challenging process that consumes high computational costs. This study proposes a new approach to recognizing the faults’ signatures. A performance adaptation approach is proposed to generate off-design operating conditions for the MGT. The similarity method is applied to consider component degradation. Both clean and faulty operation modeling are validated against available experimental data. The analysis of parameters shows fault detection is obtainable by monitoring shaft speed versus output power; however, any other decision to isolate faulty components or identify fault origin is impossible due to the confusing pattern of variations. To solve the problem, A set of derived variables is formulated using the original parameters. The results show that using shaft speed, Compressor Exit Pressure (CEP), Compressor Exit Temperature (CET), and Exhaust Gas Temperature (EGT) to define three transformed variables enables fault diagnosis under off-design operating conditions.