Early fault detection and condition monitoring of transformers

In this thesis, the detection of electrical faults in transformers and condition monitoring methods are discussed. Turn-to-turn short circuit faults, unbalanced supply voltage, and lamination faults, which are frequently encountered in transformers, are investigated in detail, and innovative analysis methods are used for the early detection of these faults. The thesis evaluates both healthy and faulty conditions by analyzing changes in electrical and magnetic parameters obtained using ANSYS® Maxwell software. Phase currents and stray flux distributions were analyzed, revealing that variations in these parameters play a crucial role in fault detection. Fault signals were examined in the frequency domain using the Fast Fourier Transform (FFT) method, which effectively differentiates between healthy and faulty conditions. Additionally, techniques such as space vector analysis, symmetrical component analysis, and phase angle spectrum analysis were utilized to achieve precise fault identification. The results show that electrical faults in transformers can be detected at an early stage, and different types of faults can be distinguished from one another. This thesis provides fast and efficient fault detection methods that contribute to the sustainability of electric transmission and distribution processes.