Battery Storage-Based Novel Hybrid Islanding Detection Technique Using Lissajous Pattern Estimation

Abstract

The reliability, uncertainty, and sensitivity of local islanding detection method for distributed generation (DG) system under diversified real-time scenarios and nonlinear loading conditions are not greatly explored. In this article, a new type hybrid islanding detection technique (IDT) is suggested, which reduces uncertainty and simultaneously improves reliability and sensitivity under different situations. In the technique, one detection parameter is defined, which is calculated from estimation of major and minor axes of Lissajous pattern (LP). Here, the evaluation of the parameter is done in two stages to avoid erroneous detection. In the first stage, the initiation of voltage transient is detected by LP. In the second stage, the active control via battery energy storage system (BESS) distinguishes nuisance tripping case from the islanding state. The extra active power injection/absorption via BESS also helps to identify islanding state at very near to zero power condition, i.e., sensitivity is enhanced. The detection time is found within 160 ms for different test cases, which is well below that of the acceptable upper value (2 s) specified in standard. The uncertainties in detection and threshold setting due to variations of harmonics and other operating conditions are removed in the proposed case, which is a distinct advantage. The Typhoon-HIL tool is used to carry out various case studies in real-time. Here, superior performances of proposed approach are observed in comparison with other state of the arts reported in the literature. © 1963-2012 IEEE.