Date of Award

Winter 3-1-2025

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Electrical Engineering

First Advisor

Prashanna Bhattarai

Abstract

Accurate fault location is a critical aspect of power system protection, ensuring grid reliability and minimizing downtime. Traditional traveling-wave-based fault location methods face limitations when applied to nonhomogeneous transmission lines due to the reliance on precise segment velocities and propagation time data. This thesis addresses these challenges by proposing a novel algorithm that leverages historical fault data to estimate segment velocities and refine these estimates as more faults occur. The algorithm was rigorously tested using the digital model of an 11-segment, 65.694 km real overhead transmission line and validated using both simulations and hardware tests using commercially available time-domain protective relays. Testing included 57 simulated faults and 57 physical relay tests, with results demonstrating consistent error reductions, with average improvements of up to 140 meters in fault location accuracy. This work highlights the algorithm’s ability to mitigate linear error trends and significantly improve fault location precision across diverse fault scenarios, even with limited training data. The proposed methodology offers a practical, low-cost solution to a longstanding challenge in fault location for nonhomogeneous transmission lines, showcasing strong potential for real-world applications and widespread implementation.

Download

Share

COinS