A novel model-based diagnostics for identifying component degradations in gas turbines for power generation

Abstract

Improving the accuracy of gas turbine performance diagnosis is important for reducing maintenance costs. Conventional model-based diagnostics use either compressor map adaptation based on correction curves or compressor map scaling based on an assumed ratio of fouling factors. However, they usually do not reflect the characteristics of the actual gas turbine. In this study, adaptation of both the compressor and turbine maps was carried out. The ratio of fouling factors was determined by quantifying the actual fouling factors so that the novel diagnostics could be applied to any gas turbine. The novelty of the proposed diagnostics is that it identifies component degradations individually, particularly distinguishing between compressor and turbine degradations using only measurement data. To demonstrate the capabilities of the proposed diagnostics, the method was applied to a 180-MW-class gas turbine operated over a long period with off-line washing. According to the results, the compressor air flow rate before off-line washing was reduced by 3.1 kg/s. Consequently, the power and efficiency of the gas turbine decreased by 3.5 MW and 0.43%p, respectively. The diagnosis results after off-line washing showed that the air flow rate and efficiency of the compressor were fully recovered, but the turbine efficiency was still degraded by 0.49%p. As a result, it was found that the gas turbine power and efficiency were not recovered by 1.2 MW and 0.33%p. The results showed that the component degradations could be identified individually. The novel diagnostics is expected to be used in a variety of strategies to reduce maintenance costs.