No abstract

No abstract

The challenges for a reliable operation of electrical power system have increased due to the presence of multi-distributed generation units (DGs) in the distribution systems in order to meet the increase of the load demand. Detection of unintentional islanding situation is very important as nondetection of islanding situation could result in a cascaded failure of the system. If the islanding situation remains undetected, the instability in the islanded part can lead to a complete failure of the electrical power system. This paper introduces a new passive scheme for islanding detection, which is suitable for multi-distributed generation units based on rotating machines. The proposed method is based on the measurements of the system voltage and frequency to compute two indices called the islanding index and harmonics index. The islanding index is the main index used to discriminate and identify the islanding situation. However, the harmonics index in conjunction with a strategy called speed reduction strategy assists the islanding index to discriminate between islanding situation in case of a close power match and system disturbances. The simulation studies were conducted in MATLAB/SIMULINK environment, and various cases have been considered, such as normal operation, islanding operation, sudden load change, DG tripping, separation of some DG units, faults, etc. The novelty of the proposed strategy is that it provides fast detection and has zero nondetection zone compared with the existing detection methods. Moreover, the proposed strategy has no effect on the power quality, and the maximum detection time is almost 350 ms at a close power match. The results indicate that the proposed scheme is successful in discrimination of the islanding conditions from other grid disturbances, revealing its great potential to be able to detect islanding events. Finally, the proposed method is applied only for rotating machine based DGs, such as wind turbines. Wind farms'