Model-based optimisation of gas turbine maintenance

The optimal performance of maintenance actions contributes to the economically competitive operation of gas turbines. For the optimisation, the condition of the gas turbine components and the cost effects of possible degradation must be known. This paper presents the necessary measured data and the condition parameters to evaluate the thermodynamic condition of a gas turbine in relation to the selected faults. The special requirements and application principles for a steady-state gas turbine model to be used for cost flow estimation are also addressed. The resulting information from condition monitoring is applied to a model-based optimisation of gas-turbine maintenance with the focus on two example cases: 1) the timing for compressor on-line and off-line (crank) washing, and 2) the timing for the inlet filter change. For both example cases, general formulation of the minimising cost functions, together with optimisation principles are presented.