No abstract

No abstract

Recently, the utilization of micro gas turbines in smart grids are rising that makes the part-load operation principal situation of the engine service. This leads to faster life consumption that increases the importance of the diagnostics process. Gas path analysis is an effective method for gas turbine diagnostics. Complex dynamics of gas turbine induces challenging conditions to perform applicable gas path analysis. This study aims to facilitate MGT gas path diagnostics through reducing the number of monitoring parameters and preparation a pattern for engine level and component level health assessment in both full and part load operation of a recuperated micro gas turbine. To attain this goal a model is proposed to simulate MGT off-design performance which is validated against experimental data in healthy and degraded operation modes. Fouling in compressor, turbine and recuperator and erosion in compressor and turbine as the most common degradations in the gas turbine are considered. The fault simulation is performed by changing the health parameters of gas path components. According to the result investigation, a matrix comprises deviation contours of four parameters, Power, fuel flow, compressor discharge pressure, and exhaust gas temperature is presented and analyzed. The analysis shows that monitoring these parameters makes it possible to perform engine level and component level diagnostics through evaluating a binary code (generated by mentioned parameter variations) against the fault effects pattern in different load fractions and fault severities. The simulation also showed that the most power drop occurred under the compressor fouling by about 8.7% while the most reduction in thermal efficiency is observed under recuperator fouling by about 7.84%. Furthermore, the investigation showed the maximum decrease in the surge margin induced by the compressor fouling during the lower part-load operation by about 45.7% while in the higher loads created by the turbine fouling by about 14%.