Search

Detecting High Impedance Faults (HIFs) in Distribution Systems (DSs) still requires effective solutions. HIFs can cause damage, such as shock hazards and bushfires. However, conventional protection cannot identify HIFs due to their low fault current, and most existing solutions neglect the multiple operating scenarios of the DSs. Thus, this paper proposes a new self-adaptive approach for HIF detection, considering the conductor fall period in a system with distributed generation, a condition usually overlooked in existing studies. The algorithm is based only on the fundamental component and third harmonic extracted from the neutral current using the Stockwell Transform. The approach was thoroughly evaluated, considering several scenarios, such as changes in the system loading, background noise, events that can cause false detection, and variations in the distributed generator power output. The tests considered HIFs signals from real field measurements and generated by a simulation model. The results showed that the algorithm is promising, with high detection rates for all the evaluated scenarios, which included over 400,000 HIF and 6,000 non-HIF cases. Additionally, a comparison proved the superiority of the proposed approach over existing methodologies.

Conductor Rupture (CR) is a critical problem in Distribution Systems (DSs), as it is challenging to identify and distinguish it from other events, which becomes even more complex when considering systems with Distributed Generation (DG). Accordingly, this short communication aims to analyze new metrics based on voltage and current negative sequence components in a DS with DG to differentiate CR from other events. The analysis comprises noisy and non-noisy signals, different statistics, and rupture locations using clustering and supervised algorithms. Overall, this study can help researchers develop increasingly effective CR detection methods, acting before the cables touch the ground, avoiding damage to living beings and bushfires. • Use of clustering and supervised algorithms to detect conductor rupture. • Exploring new metrics derived from voltage and current negative sequence components. • Analysis considering the presence of distributed generation and noisy signals. • Evaluation of the ability to distinguish conductor rupture from other events. [ABSTRACT FROM AUTHOR]

Detecting High Impedance Faults (HIFs) in Distribution Systems (DSs) still requires effective solutions. HIFs can cause damage, such as shock hazards and bushfires. However, conventional protection cannot identify HIFs due to their low fault current, and most existing solutions neglect the multiple operating scenarios of the DSs. Thus, this paper proposes a new self-adaptive approach for HIF detection, considering the conductor fall period in a system with distributed generation, a condition usually overlooked in existing studies. The algorithm is based only on the fundamental component and third harmonic extracted from the neutral current using the Stockwell Transform. The approach was thoroughly evaluated, considering several scenarios, such as changes in the system loading, background noise, events that can cause false detection, and variations in the distributed generator power output. The tests considered HIFs signals from real field measurements and generated by a simulation model. The results showed that the algorithm is promising, with high detection rates for all the evaluated scenarios, which included over 400,000 HIF and 6,000 non-HIF cases. Additionally, a comparison proved the superiority of the proposed approach over existing methodologies. • The substation transformer neutral current is the only measurement needed. • The method is easily replicable and uses only two metrics. • The method's adaptive thresholds enable using it in different system conditions. • The method did not falsely detect the tested short and long-lasting transients. • The technique was tested in wider scenarios than existing methods in the literature. [ABSTRACT FROM AUTHOR]