1.1 Island Effect Principle
"Islanding" refers to a situation where, after the public power grid stops supplying electricity, some lines in the affected area remain energized due to distributed generation (connected to the grid and transmitting power), creating a self-sufficient power supply island. In this islanded state, the power company loses control over line voltage and frequency, leading to a series of safety hazards, accidents, disputes, personal injury, and equipment damage. Therefore, power companies require grid-connected distributed generation systems to employ anti-islanding detection technology to promptly detect islanding and disconnect the distributed generation devices from the public grid.
1.2 Detection of the Island Effect
In November 2005, my country issued relevant national standards, namely the technical requirements for grid connection of photovoltaic systems. These standards came into effect on January 1, 2006, and February 1, 2006, respectively. The requirements for islanding detection in the standards include: when the grid loses voltage, the anti-islanding protection must complete the disconnection of the photovoltaic system from the grid within 2 seconds; and at least one active and one passive islanding detection method should be used.
Islanding detection methods are mainly divided into two types: passive and active. Passive islanding detection methods determine whether islanding has occurred by detecting whether the inverter's output deviates from the range specified by the grid connection standard (such as voltage, frequency, or phase). Its working principle is simple and easy to implement, but it cannot detect islanding when the inverter's output power is balanced with the local load power. Active islanding detection methods involve controlling the inverter to introduce a certain disturbance in its output power, frequency, or phase. When the grid is operating normally, these disturbances are undetectable due to the grid's balancing effect. Once a grid fault occurs, the disturbance in the inverter's output will quickly accumulate and exceed the range allowed by the grid connection standard, thereby triggering the islanding protection circuit. This method has high detection accuracy and a small non-determination zone (NDZ), but it is more complex to control and reduces the quality of the inverter's output power.
Passive method:
1. Voltage and frequency detection
During the grid-connected operation of a photovoltaic power generation system, in addition to preventing overvoltage, undervoltage, and overfrequency, it is also necessary to ensure that the inverter output voltage is synchronized with the grid. Therefore, continuous power monitoring is required to prevent overvoltage, undervoltage, and overfrequency issues. Passive islanding detection methods, which only require voltage and frequency detection, only need to make judgments and do not require additional detection circuitry. The maximum failure rate of this method is when the load power is balanced. After a grid power outage, the inverter output changes, resulting in missed islanding detections.
2. Phase Detection Method for Inverter Output Voltage Phase Detection: The principle of this method is similar to that of voltage and frequency detection methods. When a grid fault occurs, the output voltage and current of the inverter in the photovoltaic power generation system can be determined by observing the phase changes before and after the grid fault. Since inductive loads are common in the grid, this method is superior to voltage and frequency detection methods. However, when the load is resistive and its impedance characteristics remain constant, this method loses its islanding capability.
3. Harmonic detection
Harmonic detection refers to the method used when a power grid fails and its balancing function ceases, the output current of a photovoltaic power generation system generates a large number of harmonics after passing through a transformer. The state of the system can be determined by analyzing the changes in these harmonics. Experimental research and practical applications show that this method has good performance. However, due to the large number of nonlinear devices in the current power grid, a unified harmonic standard for islanding effect detection is needed.
Active detection:
The working principle of the active frequency drift method is as follows:
① The system controls the inverter to ensure that the frequency of its output voltage has a certain error Δf compared to the frequency of the grid voltage (Δf is within the allowable range of grid connection standards);
② When the power grid is operating normally, due to the corrective effect of the phase-locked loop circuit, the error Δf between the inverter output voltage frequency and the power grid voltage frequency is always within a small range;
③ When a grid fault occurs, the frequency of the inverter output voltage, plus v, will change. In the next power frequency cycle of the inverter, the system will use plus v as a reference and then add a set frequency error Δf to control the frequency of the inverter output voltage, thus further increasing the frequency error between the inverter output voltage and the grid voltage. This process repeats until the frequency of the inverter output voltage exceeds the grid connection standard, thereby triggering the islanding protection circuit to disconnect the inverter from the grid.
The frequency disturbance waveform in the active frequency perturbation method is shown in the figure. The curve in the figure represents the current waveform and its disturbance control signal for one power frequency cycle. The vertical axis represents the per-unit value of the current, and the horizontal axis represents time. Let t be the time period when the voltage is zero in the active frequency perturbation method. The ratio of t to the fundamental voltage half a cycle fraction (rl) is called the disturbance signal (choping fraction, cf).
