Delta I based approach to detect high impedance faults in traction systems

Traditionally traction system protection relies on wide-area disconnection with sectional re-energization and selective fault isolation. The former offers faster initial isolation but with prolonged service restoration. The latter however minimizes disruption but may have slower clearing times. Modern traction systems are increasingly converter-fed, presenting challenges for conventional distance protection, particularly in detecting high-impedance faults (HIFs). HIFs, characterised by low fault currents and arcing, are difficult to distinguish from normal operations using traditional overcurrent, ground fault and distance relays. Existing advanced HIF detection methods, including wavelet packet transform, harmonic/interharmonic analysis and AI-based techniques, often have limitations. This work presents a novel delta-based backup protection application for traction systems. This application detects sudden changes in current magnitude, incorporates reactance, harmonic and phase angle checks for fault discrimination and uses a minimum voltage check for enhanced reliability. A novel directional determination method is also implemented based on delta changes in sequence active and reactive powers. Validation using field-generated disturbance records (DR) in both MATLAB and an intelligent electronic device (IED) demonstrates the application's accurate fault detection and trip signal issuance, confirming its effectiveness as a backup to distance protection.