Field-Proven V-Moveout Detector for Real-Time Power Cable Fault Detection using Distributed Acoustic Sensing

Rapid fault localization is essential, yet conventional methods such as reflectometry are limited on long or inaccessible routes. Distributed acoustic sensing (DAS) enables continuous, realtime monitoring over distances up to ~100 km with meter-level accuracy. However, existing approaches often rely on manual interpretation or simple thresholds, leading to false alarms and limited robustness due to scarce fault data.

This work introduces an automated method that detects the V-shaped wavefront pattern caused by bidirectional fault propagation. Using spectral filtering and geometric feature extraction on optical-phase DAS data, the method was validated on nine real HVAC and HVDC faults, correctly identifying eight, with no false positives across 176 hours of data from four field environments. To our knowledge, this is the first multi-project field validation of V-moveoutbased fault localization. The approach demonstrates strong reliability and supports faster fault response. Future work will focus on larger datasets and data sharing to enable site-specific adaptation and machine-learning-based detection within a secure software framework.