<p>Gas turbines are critical power plant components, operating under high-temperature and high-pressure conditions, increasing failure risks. Fault diagnosis of gas turbines is essential to reduce maintenance costs and enhance power plant reliability. Particularly, a fault diagnosis approach that can be applied across diverse power plants offers broader applicability than plant-specific approaches; however, this is challenging as sensor infrastructure differs among the plants. To address this challenge, the combustor outlet temperatures, a common measure in power plants, are utilized as an indicator of gas turbine health status. During normal operation, combustion is stable with a balanced profile of combustor outlet temperature, whereas a fault disrupts the combustion, yielding an imbalanced and declining profile. In this study, we propose the sliding radar chart (SRC) for a broadly applicable gas turbine fault diagnosis using combustor outlet temperatures. The SRC represents values of temperature sensors, arranged around the combustor outlet and collected over time, in radar charts to reflect profile balance. The radar charts are then overlaid within a sliding window to capture temporal variation. The SRC approach was validated using real data from two different power plants and compared with approaches that employed the combustor outlet temperature as a vector form or as a radar chart without the use of the sliding window. The SRC approach achieved the highest performance in fault diagnosis across both plants, rarely misclassifying faults as normal. This study contributes to the stability of power plant operations by enabling sensitive fault diagnosis using the commonly measured combustor outlet temperature.</p>