Gas Turbine Compressor Washing Economics and Optimization Using Genetic Algorithm

Abstract

Abstract Studies have shown that online compressor washing of gas turbine engines slows down the rate of fouling deterioration during operation. However, for most operators, there is a balancing between the performance improvements obtained and the investment (capital and recurring cost). Washing the engine more frequently to keep the capacity high is a consideration. However, this needs to be addressed with expenditure over the life of the washing equipment rather than a simple cost-benefit analysis. The work presented here is a viability study of online compressor washing for 17 gas turbine engines ranging from 5.3 to 307 MW. It considers the nonlinear cost of the washing equipment related to size categories, as well as nonlinear washing liquid consumption related to the variations in engine mass flows. Importantly, the respective electricity break-even selling price of the respective engines was considered. The results show that for the largest engine, the return of investment (RoI) is 520% and the dynamic payback time of 0.19 years when washing every 72 h. When this is less frequent at a 480-h interval, the investment return and payback are 462% and 0.22 years. The optimization study using a multi-objective genetic algorithm shows that the optimal washing is rather a 95-h interval. For the smallest engine, the investment was the least viable for this type of application.