Review of differential protection schemes for power transformers in medium-voltage distribution networks

Power transformers are essential components of medium-voltage (MV) distribution networks, particularly in 33 kV substations, where they enable voltage transformation, maintain power quality, and ensure reliable electricity delivery. Due to their critical operational role, transformer failures can significantly affect system stability, service continuity, and equipment safety. However, transformers are highly vulnerable to both internal and external disturbances such as winding short circuits, insulation degradation, overloading, overexcitation, magnetizing inrush current, and current transformer (CT) saturation, all of which can compromise protection accuracy and operational reliability. Among transformer protection methods, differential protection remains the most widely used technique for internal fault detection because of its high sensitivity, fast response, and selective fault isolation capability. Over time, transformer protection systems have evolved from conventional electromechanical relays to numerical and intelligent relay-based systems, improving signal processing, fault discrimination, and communication performance. Despite these developments, challenges such as false tripping during low-harmonic inrush conditions, CT mismatch, saturation effects, harmonic distortion, and limited adaptability under fluctuating grid conditions still remain. This review critically examines transformer operating principles, fault behavior, protection hierarchy, differential protection schemes, relay technologies, and emerging intelligent methods involving adaptive algorithms, signal processing, and machine learning. Using a PRISMA-based systematic literature review, key trends, limitations, and research gaps were identified. The study emphasizes the urgent need for adaptive, robust, and cost-effective differential protection solutions specifically designed for 33 kV MV transformers, particularly in developing power systems with weak grid conditions and infrastructure limitations.