Islanding Detection Method for Distributed Generation Systems Based on Voltage and Power Variation Analysis in Smart Grids

Abstract

The growing penetration of distributed generation (DG) in modern power distribution networks has significantly improved system reliability, energy efficiency, and the integration of renewable energy resources. However, this integration also introduces critical technical challenges, with islanding representing one of the most important and potentially hazardous operating conditions. Islanding occurs when a portion of the distribution network remains energized by DG units after disconnection from the main grid, which can pose serious risks to personnel safety, equipment integrity, and overall system stability. Therefore, international standards require fast, reliable, and accurate islanding detection methods to ensure secure operation of smart grids. Despite numerous existing approaches, many conventional islanding detection techniques suffer from a large non-detection zone (NDZ), which limits their reliability and effectiveness under certain operating conditions. To address this limitation, this paper proposes a novel and fast islanding detection method based on the combined evaluation of voltage magnitude variations, their dynamic range behavior, and active power fluctuations at the point of common coupling (PCC) of distributed generation units within a smart grid environment. The proposed method improves detection speed and accuracy while significantly reducing the NDZ, thereby enhancing the overall protection performance of both distributed generation units and distribution networks. Extensive simulation results demonstrate that the proposed approach outperforms conventional islanding detection techniques in terms of sensitivity, robustness, and response time