Abstract To achieve decarbonization in power-generating gas turbines, the technology of mixing hydrogen with natural gas is garnering significant attention. However, when blending natural gas with hydrogen, the altered combustion characteristics can lead to combustion instability in gas turbine combustors. Although fuel staging can effectively suppress combustion instability for can-type combustors, further research on mitigation strategies for hydrogen cofiring and their predictive methods is required. This study involves hydrogen cofiring experiments using a full-scale can-type combustor. Moreover, the resulting suppression of combustion instability is analyzed through fuel staging by utilizing three-dimensional (3D) computational fluid dynamics (CFD) and one-dimensional (1D) thermo-acoustic analysis. The experiments used a full-scale industrial can-type combustor with a five-around-one nozzle configuration. Hydrogen was blended with natural gas up to a volume fraction of 30%, maintaining a constant thermal power. Fuel staging was applied by controlling two out of five outer nozzles (ONs) along with the remaining three. Before the 1D thermo-acoustic analysis, the internal flame structure of the combustor was examined through 3D CFD analysis. Based on the results, a multi-input multi-output (MIMO) system was constructed for 1D thermo-acoustic analysis of the can-type combustor. The application of time delays derived from 3D CFD analysis to the 1D model revealed that differences in flame time delays across the nozzles cause combustion instability suppression observed in fuel staging.
Read full abstract