Research on detection methods for the islanding effect in photovoltaic inverters

Thread Content

This post was last edited by chenjinfeng on 2020-2-7 at 17:13. With the development of the world economy, environmental and energy issues have become increasingly prominent. Solar photovoltaic power generation, which offers clean and environmentally friendly methods of power production, is receiving growing attention. The electricity generated by solar panels must ultimately be fed into the power grid, and the islanding effect represents a specific problem that photovoltaic inverters face after being connected to the grid; therefore, islanding detection is necessary. Methods for detecting islanding effects can generally be divided into two categories: passive and active. To address the shortcomings of existing methods for detecting islands, this study proposes a method based on the analysis of reactive power relationships to detect islands, and a simulation model is developed to verify the effectiveness of this detection method. 　 Island effect ； Island detection ； Simulation ； 　 Currently, global energy resources are becoming increasingly scarce, and environmental pollution is worsening. In particular, our country is also facing a severe shortage of electricity. As an ideal and environmentally friendly method of power generation, solar photovoltaic power generation boasts significant advantages, and its future prospects are very promising. The direct current electricity generated by solar power generation is inverted by a photovoltaic inverter and then fed into the power grid; during this process, some new problems arise, among which the most serious one is the islanding effect. Therefore, in addition to the protection measures for voltage, current, and frequency monitoring, protective measures against the islanding effect must also be implemented during this process. This paper provides an introduction to the islanding effect, proposes a detection method for the islanding effect that analyzes the relationship between frequency and reactive power, and verifies the effectiveness of this islanding detection method through simulations. 　 1 Overview of the islanding effect 　1.1 Islanding effect 　The islanding effect refers to the situation where, in the event of a power grid outage, a photovoltaic grid-connected system fails to detect this outage in time and disconnect from the grid, resulting in the photovoltaic power generation system together with the electrical loads surrounding it forming an island that is difficult for the power company to manage. 　 1.2 Principle of the islanding effect When photovoltaic power generation is connected to the grid and the system is operating normally, that is, when the circuit breaker is closed, both the photovoltaic system and the grid supply power to the loads simultaneously ； When the power grid loses power due to electrical faults or misoperations, that is, when the circuit breaker opens, the power outage in the grid is not detected in time by the photovoltaic grid-connected system, which continues to supply power to the loads. As a result, a separate island is formed between the photovoltaic grid-connected system and the loads, enabling self-sufficient power supply. 　　2 Methods for detecting islanding effects The methods used to detect islanding can generally be divided into two categories: passive detection methods and active detection methods. 　　2.1 Passive Detection Methods Generally, passive detection methods include over/under voltage detection, high/low frequency detection, as well as detection of phase jumps and voltage harmonics. 　　2.1.1 Phase shift detection Phase shift detection refers to the situation where a photovoltaic power generation system is connected to the grid and is operating properly; in such a case, the current I output by the system has the same frequency as the voltage U of the grid, meaning that the power factor is 1. After the power grid is disconnected, an islanding phenomenon occurs, meaning that the photovoltaic power generation system supplies power to the loads on its own. At this time, the voltage Ua at point a depends on the load impedance Z and the output current I. The frequency and phase of the current are locked within the system by means of a phase-locked loop, enabling the current to be synchronized with the voltage at the zero-crossing points. Outside the zero-crossing point, the current leads or lags the voltage, causing a sudden change in the voltage phase. Therefore, the method of phase jump detection can be used to determine whether an island has formed. 　　2.1.2 Voltage Harmonic Detection Voltage harmonic detection is a method based on the fact that when a photovoltaic grid-connected system is disconnected from the power grid, the load impedance in the grid is much larger than the impedance of the grid network itself. As a result, significant harmonics are generated at point a when current flows into the load. An islanding phenomenon occurs when an abrupt increase in the harmonic content of the voltage at point a is detected. However, due to factors such as nonlinear loads, the grid voltage contains a high level of harmonics, and it is difficult to determine the threshold for detecting these harmonics; therefore, this method has significant limitations. 　　2.1.3 Passive islanding detection method using as the detection indicator 　　All three of the above passive islanding detection methods detect the occurrence of islanding by monitoring changes in electrical quantity. However, when an islanding phenomenon occurs, if the power load matches the power output of the photovoltaic power generation system, the changes in the frequency or voltage output by the inverter are minimal and remain within the normal range; in such cases, the protection circuit is unable to detect the occurrence of islanding. 　　When an islanding phenomenon occurs, changes in the power supply or load can cause significant variations in power and frequency, making the system highly unstable. Additionally, the rate of change increases as well. Therefore, a new method for detecting islanding has been proposed; this method relies on monitoring these changes to determine whether an islanding effect is present. Even when the frequency or voltage changes are minimal, this detection method can accurately and quickly identify islands. The detection time is less than 2 seconds, and it does not affect power quality, indicating that this method has the advantages of speed, accuracy, and effectiveness. 　　2.2 Common Active Detection Methods The common active detection methods include Slip Mode Frequency Shift (SMS), Active Frequency Drift (AFD), current disturbance methods, and output power perturbation methods. The active detection method involves applying frequency, current, or phase disturbance signals at the output of the grid-connected inverter, while simultaneously monitoring the effect of these disturbance signals on the line voltage. When the system is operating in island mode, the disturbances caused by interference signals are reflected and accumulate in the voltage of the lines; therefore, the occurrence of an island condition can be detected by monitoring changes in voltage. This paper focuses on analyzing the output power perturbation method, which enables rapid detection of an islanding condition and subsequent shutdown of the system when it occurs. The detection of output power disturbances in grid-connected inverters is used to control the photovoltaic grid-connected system to periodically generate active and reactive power disturbances. When the grid-connected inverter is operating in island mode, such disturbances cause significant changes in the system’s frequency or voltage, exceeding the preset thresholds, thereby allowing the detection of islanding. 　　3 Conclusion 　In summary, the methods for detecting the islanding effect can generally be divided into two categories: passive detection methods and active detection methods. This paper introduces some islanding detection methods, briefly discusses their advantages and disadvantages, and simulates the method for detecting the relationship between reactive power and frequency, proving that this method is effective and feasible. In islanding detection research over the past few years, composite detection methods that combine traditional active and passive approaches, as well as improved versions of the traditional AFD detection method, have played an important role. There have also been attempts to apply novel intelligent detection methods to islanding detection. However, all these methods still have certain shortcomings; in particular, there are issues such as dead zones and problems affecting power quality, or they may not be suitable for distributed power systems. These issues will be the focus of further research.