[3B3] Gas path analysis of a two-shaft gas turbine engine by utilising multi-operating point conditions and bank of Kalman filters
S Abyaneh
Mapna Group, Iran
One of the main purposes of gas path analysis (GPA) in gas turbine industry is to evaluate the health conditions of gas turbine engine components. In this study, a GPA module has been developed to compare all health parameters of a compressor of a specific gas turbine engine, such as flow capacity, efficiency and pressure drop, with respect to the reference values from a healthy engine. Health parameters’ deviations from their reference values (due to engine degradation) are then evaluated based on measured gas path parameters in multiple operating points. The GPA module operates as a developed computer-based simulator model that receives operating point parameters as inputs. The performance behaviour of a healthy engine at the specified operating conditions is generated by the simulator as the model outputs. A set of these outputs is compared with the corresponding actual turbine outputs received by sensors embedded on the turbine. In case of differences between the simulator outputs and actual sensor data, model health parameters are changed by an estimator whose algorithm is based on an unscented Kalman filter (UKF), so that the evaluated parameters and their corresponding sensor data coincide. It has been proved that a certain number of health parameters could be ignored in order to optimise the GPA module’s overall performance. The developed GPA module encompasses different scenarios for the engine model, in each of which the number of health parameters varies with a single Kalman filter (KF) estimator attributed to each model. The combination of KFs constitute a bank of KF. The results show the effectiveness of bank of KF utilisation in health parameter fault estimation for some case studies performed on the studied gas turbine. It has also been shown that the fault estimation algorithm is enhanced by employing multiple operating points. The results have been verified utilising a GPA module developed and implemented on a gas turbine engine.
Key words: gas path analysis, gas turbine, fault diagnosis, health parameters.
Key words: gas path analysis, gas turbine, fault diagnosis, health parameters.
