Correlation-based single-ended traveling wave fault location methods: A key settings parametric sensitivity analysis

Abstract

This paper presents a sensitivity analysis of the key parametric settings used to computationally implementing transmission line single-ended correlation-based traveling wave fault location (SEC-TWFL) algorithms. To estimate the short-circuit distance, these techniques compare the signal transient patterns generated by the first incident traveling wave (TW) that reach the monitored terminal with those created by consecutive reflections that come from the fault point. To do so, a diversity of filtering processes, correlation strategies and transient observation windows can be adopted, resulting in different fault location performances. The Alternative Transients Program (ATP) is used to generate a wide variety of fault records, which are then evaluated with different SEC-TWFL methods varying the used correlation functions, filtering techniques and observation window lengths. From the performed studies, the advantages and limitations of the evaluated computational implementation strategies are addressed, and new parametric settings recommendations are proposed, demonstrating that very accurate fault location estimations can be obtained when specific implementation strategies for each of the evaluated methods are considered. • New setting recommendations are proposed to be applied in SEC-TWFL methods. • The transient signal mean value removal, as typically reported, is not suitable. • The observation window length directly affects the fault location performance. • The window length comprising only the first TW as usually reported is not indicated.