Advanced Gas Turbine Control Logic Using Black Box Models for Enhancing Operational Flexibility and Stability

Abstract

In recent years, the importance of operational flexibility has increased for gas turbines that can stably operate under various operation conditions. This study proposes advanced control logic using black box models based on an artificial neural network. The goals are to enhance the operational flexibility by increasing the ramp rate and to enhance the operational stability by overcoming the limitation of conventional schedule-based control. By applying the advanced control logic, the turbine inlet temperature (TIT) and turbine exhaust temperature (TET) can be maintained at the optimal values, resulting in efficiency improvement by 0.35%. Furthermore, the maximum deviation of the rotational speed was reduced from 0.22% to 0.061%, and the maximum variations of TIT and TET were reduced by 15–20 °C during the fluctuation of the gas turbine’s power output. In conclusion, high-efficiency operation and a reduction in the degradation of the high-temperature parts can be expected through optimal operations of gas turbines by applying the proposed advanced control logic in a harsh operating environment. Moreover, fast load following can be achieved to meet the recent requirements of the operation environment of gas turbines by improving the ramp rate from 30 to 60 MW/min.