| description abstract | Gas turbines (GT), like other prime movers, experience wear and tear over time, resulting in decreases in available power and efficiency. Further decreases in power and efficiency can result from erosion and fouling caused by the airborne impurities the engine breathes in. To counteract these decreases in power and efficiency, it is a standard procedure to “wash†the engine from time to time. In compressor stations on gas transmission systems, engine washes are performed offline and are scheduled in such intervals to optimize the maintenance procedure. This optimization requires accurate prediction of the performance degradation of the engine over time. A previous paper demonstrated a methodology for evaluating various components of the GT gas path, in particular, the air compressor side of the engine since it is most prone to fouling and degradation. This methodology combines gas path analysis (GPA) to evaluate the thermodynamic parameters over the engine cycle followed by parameter estimation based on the Bayesian errorinvariable model (EVM) to filter the data of possible noise due to measurement errors. The methodology quantifies the engineperformance degradation over time, and indicates the effectiveness of each engine wash. In the present paper, the methodology was extended to assess both recoverable and unrecoverable degradations of five GT engines employed on TransCanada's pipeline system in Canada. These engines are: three GE LM2500+, one RR RB21124G, and one GE LM1600 GTs. Hourly data were collected over the past 4 years, and engine health parameters were extracted to delineate the respective engine degradations. The impacts of engine loading, site air quality conditions, and site elevation on engineaircompressor isentropic efficiency are compared between the five engines. | |
