Conclusion

DLE gas turbine is introduced to achieve a low emission operation where a pilot gas fuel valve is added to the fuel system. However, the turbine is susceptible to frequent trips and a dynamic model for operational representation is required. In this modelling study, a DLE gas turbine fuel system model was proposed for operational study using First Principle Data-Driven (FPDD) method. There are three major contributions in this paper to represent DLE gas turbine fuel system in Rowen’s model. First, the actual DLE gas turbine fuel system setup which consists of a pilot and main gas fuel valves was developed according to the gas turbine manual. Second, the main valve and pilot valve models were proposed according to FPDD method using system identification. Finally, the novel simulation model of DLE gas turbine was designed, integrating its comprehensive fuel system model into Rowen’s model using available operational data. To verify the proposed method, the simulation output of the pilot and main fuel valve, engine gas fuel flow and turbine temperature signals were compared with the actual signals from the power plant. The results clearly demonstrated the accuracy of the proposed model with a very low MAE and RMSE. We believe that this DLE model can be applied in condition monitoring, fault diagnostics and trip prediction study. To further improve the utilization of the model, a compressor pressure discharge parameter can be added, and the tripping mechanism of the turbine can be developed.

Acknowledgement

The authors acknowledge the support of Ministry of Higher Education (MOHE) and Universiti Teknologi PETRONAS in carrying out this research through the FRGS 0153AB-L31 grant.