Desempenho na detecção de ilhamento de geradores distribuídos baseados em inversores considerando técnicas ativas e funções de suporte de rede

Abstract

Given the growing expansion of distributed generation (DG), particularly from solar photovoltaic sources connected via inverter-based generators (IBGs), the challenge of unintentional islanding emerges. Islanding refers to the operation in which the generating unit supplies an electrically isolated portion of the distribution system to which it is connected. The islanded operation of distributed generators can lead to issues related to safety, commercial, and technical aspects. Thus, it is necessary to disconnect distributed generators under islanding conditions. For this purpose, anti-islanding techniques are employed, which are generally classified into passive and active techniques. In addition to eliminating islanding, other functionalities are increasingly required from distributed generators to ensure and establish minimum safety standards, including ride-through capabilities and grid support requirements. Therefore, it becomes essential to evaluate the impact that multiple IBGs, compliant with ride-through requirements, participating in grid support, and employing active islanding detection techniques, may have on the performance of such detection. This study evaluates the islanding detection performance of IBGs, considering the listed functionalities, using the active techniques GEFS (General Electric Frequency Scheme) and GEVS (General Electric Voltage Scheme), as well as the grid support functions frequency-watt (PF control), volt-watt (PV control), and volt-var (QV control). The analyses were conducted through computational simulations using the Alternative Transients Program (ATP) software. Performance was assessed for protection function groups based on frequency and voltage, namely under/overfrequency and under/overvoltage, respectively. The results were obtained from nine different scenarios and 465 operating conditions, totaling over 4,185 simulations. Although the results of this study are conditioned by the employed modeling, they indicate that the islanding detection performance of IBGs, considering active techniques and grid support functions, achieved values of up to 100%. However, factors such as the maximum allowed response time, the analyzed protection functions, including pick-up settings and time delays, influenced the outcomes.