A Passive Islanding Detection Method for Neutral point clamped Multilevel Inverter based Distributed Generation using Rate of Change of Frequency Analysis

• Published in: International journal of electrical and computer engineering (Malacca, Malacca) Vol. 8; no. 4; p. 1967
• Main Authors: Reddy, Ch. Rami, Reddy, K. Harinadha
• Format: Journal Article
• Language: English
• Published: Yogyakarta IAES Institute of Advanced Engineering and Science 01.08.2018
• Subjects: Activation; Clean energy; Distributed generation; Electric power distribution; Electric power systems; Energy sources; Inverters; Renewable energy; Renewable resources
• ISSN: 2088-8708; 2088-8708
• DOI: 10.11591/ijece.v8i4.pp1967-1976

Presently renewable energies have taken a special place in the world and most of the Distributed Generations (DGs) used in the interconnected power system are utilized, renewable energy resources. Due to the DG’s advantages, including use of renewable energy such as, clean nature, does not pollute environment and having endless nature the use of these renewable resources to produce electrical energy in the world are increasing in day to day life. One problem with such Distributed generators is an unintentional islanding phenomenon. Islanding occurs when a Distributed Generation continues to energize an isolated part of a power system even after it was disconnected from the main grid, which is surrounded by unpowered lines. Since islanding can cause hazardous conditions for people and equipment which is connected to it. As per IEEE 1547 DG Interconnection standards, islanding should be quickly detected within 2 seconds, by protective relays and inverters that are part of the DG system. In this paper, a new passive method to identify islanding states has been proposed, based on the rate of change of frequency analysis (ROCOF) for a multilevel inverter based solar distributed generation systems. This method is efficient for both connecting DGs to the network with or without the Inverter. This method is more efficient than the existing methods and reducing the Non Detection Zone (NDZ), which is the disadvantage of existing passive methods and also clearly differentiating between the Islanding and Non-islanding events. The simulation results, which are carried on the MATLAB/Simulink environment shows the performance of the proposed method.