<![CDATA[The primary aim and contribution of this study is the presentation of a non-intrusive early diagnosis method based on finite element simulation (FEM). The focus was on a 1000 kVA distribution transformer based on manufacturing data and field tests conducted in Mosul, Iraq. An accurate two-dimensional model of the transformer was developed using ANSYS Maxwell software, simulating normal operation and various internal fault scenarios (such as single-phase or double-phase short-circuits and ground faults) at varying rates. The resulting changes in magnetic flux distribution, core losses, currents, and voltages were analyzed as indicators to determine the presence, type, and severity of faults. A representation of internal faults in the three-phase transformer windings was performed to detect and diagnose faults early. The results clearly show that small short-circuit faults (up to 1.2% of the windings) are distinguishable by specific changes in transformer parameters. These faults lead to a localized temperature increase and the onset of insulation deterioration. It was also observed that an increase in the fault percentage (5% to 25%) causes a significant increase in magnetic flux and total losses. These effects are significantly exacerbated by ground faults or faults involving two phases. These results confirm that computational analysis provides a powerful tool for proactive monitoring, enabling preventive maintenance scheduling based on initial fault indications. This contributes to extending transformer life, enhancing network reliability, and avoiding costly catastrophic failures. Continuous monitoring and effective ground protection remain critical elements for maintaining transformer safety and efficiency.]]>
