Date of this Version
In the current decade, technology innovations and cost reduction of inverter-based Distributed Energy Resources (DERs) have led to higher integration of distributed energy storage and photovoltaic (PV) solar power systems. Increasing growth in PV penetration to the distribution system can raise operational and safety concerns especially in case of an unintended islanding.
In general, standards require distributed generators (DGs) to detect islanding from the main grid and cease to energize the local system. Multiple methods have been introduced in the literature to detect these islands reliably and quickly. In order to connect an inverter to distribution system, inverter should pass certain certification tests such as UL 1741 certification test. The anti-islanding test in UL 1741 standard tests only one type of load over a limited range of loading conditions with a single inverter and lumped load and no impedances in between them. The overall goal of this thesis is to determine those parameters to which run-on times (ROTs) are relatively insensitive and thus do not need to be emphasized in certification testing or risk of islanding studies. This thesis presents a generic MATLAB Simulink inverter model and studies sensitivity of anti-islanding tests to parameters such as inverter location, inverter operating point, load location, load type and circuit impedance. Inverters in these studies are equipped with Group 2A and Group 2B anti-islanding methods. The key contributions in this thesis can be summarized as follows:
- A comprehensive review of anti-islanding techniques in the literature.
- An anti-islanding detection model was developed in MATLAB software with at least one method from different groups of anti-islanding methods; the model can be used further for industrial applications and research purposes.
- The result of analyses indicated that the level of phase-phase imbalance, constant-power load, harmonic-current load and irradiance level have a low or negligible impact on anti-islanding and can be omitted from these studies. These findings are expected to lower the cost and improve the speed of these studies, in large distribution systems.
Adviser: Sohrab Asgarpoor