Advanced Two-Stage Islanding Detection Framework Utilizing Kinetic Energy Rate of Change

Inverter-based generations (IBGs) in distribution networks require robust anti-islanding protection systems (AIPS) to ensure operational integrity. However, conventional AIPS often struggle to detect accidental islanding events in scenarios where loads and generation are perfectly matched. Therefore, to get over this challenge, a novel hybrid anti-islanding protection scheme using Sandia frequency shift (SFS) phase angle transformation and RMS value of absolute frequency deviation at the point of common coupling (PCC) has been proposed in this paper. The proposed approach triggers the Sandia phase angle transformation solely upon suspicion of an accidental islanding occurrence, utilizing specific parameters monitored at the Point of Common Coupling (PCC). The approach outlined in this methodology is expected to enhance the power quality and stability of IBGs substantially. The performance analysis of the proposed hybrid method for the time domain has been carried out in MATLAB/Simulink. Experimental validation was conducted using a power hardware-in-the-loop (PHIL) testbed setup. The effectiveness of the proposed anti-islanding protection system (AIPS) was tested across more than 200 diverse islanding and non-islanding events. The results demonstrate its robustness and cost-effectiveness, with an islanding detection time as low as 180 milliseconds. Note to Practitioners—This paper introduces a new method for detecting unintentional islanding, a potentially hazardous scenario if left undetected. The proposed methodology offers improved robustness, ease of implementation, and higher accuracy than existing techniques. It aims to address power quality issues associated with active islanding detection methods and achieve zero non-detection zones. With a rapid detection time of only 180 milliseconds, the proposed method ensures swift and reliable identification of islanding events. Experimentally tested across more than 200 islanding and non-islanding scenarios, it proves to be highly accurate, making it a promising solution for real-time islanding detection applications.