With the rapid expansion of photovoltaic (PV) power generation systems, the impact of grid disturbances, particularly voltage sags, on system stability and power quality has become a critical concern. Voltage sags, often caused by lightning strikes, large motor starts, or short-circuit faults, can severely degrade the performance of grid-connected solar inverters. Without effective control, these events may lead to excessive current distortion, power oscillations, and even disconnection of the inverter from the grid, posing significant risks to grid reliability. Therefore, enhancing the low voltage ride-through (LVRT) capability of solar inverters is paramount for the large-scale integration of PV systems. In this article, I propose a comprehensive LVRT control strategy designed for three-phase grid-connected solar inverters. This strategy aims to limit active current surges, eliminate active power fluctuations, reduce voltage unbalance at the point of common coupling (PCC), and provide reactive power support to the grid during voltage dips. The approach leverages advanced control theory and sequence component analysis to ensure robust operation under unbalanced grid conditions. Through detailed mathematical derivations and extensive simulations, I demonstrate the feasibility and effectiveness of this method in maintaining grid stability while maximizing the utilization of solar inverters.
The integration of solar inverters into modern power grids requires them to adhere to strict grid codes that mandate LVRT capabilities. During a voltage sag, the inverter must remain connected and support the grid by injecting appropriate currents. Traditional control methods often fail to manage the negative-sequence components that arise under unbalanced voltages, leading to oscillatory power and excessive current peaks. My proposed strategy addresses these issues by dynamically adjusting the active and reactive power references based on real-time grid voltage conditions. This not only prevents inverter tripping due to overcurrent but also enhances grid voltage support. The core of this strategy lies in the precise calculation of current references in the positive and negative synchronous reference frames, ensuring that the inverter operates within its thermal limits while contributing to grid stability. In the following sections, I will elaborate on the system structure, control methodology, and simulation results, providing a thorough analysis supported by formulas and tables.
A typical three-phase PV generation system, as considered in my research, consists of a PV array, a DC-AC inverter, an LCL filter, and a control unit. The solar inverters are pivotal in converting DC power from the PV panels to AC power synchronized with the grid. The use of an LCL filter is advantageous due to its superior harmonic attenuation compared to simple L filters. The LCL filter’s third-order low-pass characteristic allows for smaller inductance values, reducing system size and losses while effectively suppressing switching harmonics. The system parameters, such as grid voltage, filter components, and switching frequency, are carefully selected to optimize performance. For instance, in my simulations, I set the grid voltage amplitude to 380 V, the rated current amplitude to 10 A, and the switching frequency to 10 kHz. The LCL filter values include an inverter-side inductance of 1.5 mH, a grid-side inductance of 1.5 mH, and a filter capacitance of 5 μF. These solar inverters must handle not only steady-state operations but also transient events like voltage sags, making the control strategy crucial for reliable operation.
Under unbalanced grid conditions, the voltage at the PCC can be decomposed into positive, negative, and zero sequences. For three-wire systems, the zero-sequence component is negligible as there is no neutral connection. Thus, the three-phase grid voltages can be expressed as:
$$ u_a = u^+ \cos(\omega t + \phi^+) + u^- \cos(\omega t + \phi^-) $$
$$ u_b = u^+ \cos\left(\omega t + \phi^+ – \frac{2\pi}{3}\right) + u^- \cos\left(\omega t + \phi^- + \frac{2\pi}{3}\right) $$
$$ u_c = u^+ \cos\left(\omega t + \phi^+ + \frac{2\pi}{3}\right) + u^- \cos\left(\omega t + \phi^- – \frac{2\pi}{3}\right) $$
where \( u^+ \) and \( u^- \) are the amplitudes of the positive and negative sequence voltages, respectively, and \( \phi^+ \) and \( \phi^- \) are their initial phase angles. The voltage unbalance factor \( \epsilon \) is defined as:
$$ \epsilon = \frac{u^-}{u^+} $$
My control strategy for solar inverters derives current references in the positive and negative sequence synchronous frames to manage power flow during voltage sags. The reference currents are given by:
$$ i^{+,*}_d = \frac{2}{3} \left[ \frac{P^* u^+_d}{(u^+)^2 – (u^-)^2} + \frac{Q^* u^+_q}{(u^+)^2 + (u^-)^2} \right] $$
$$ i^{+,*}_q = \frac{2}{3} \left[ \frac{P^* u^+_q}{(u^+)^2 – (u^-)^2} – \frac{Q^* u^+_d}{(u^+)^2 + (u^-)^2} \right] $$
$$ i^{-,*}_d = \frac{2}{3} \left[ -\frac{P^* u^-_d}{(u^+)^2 – (u^-)^2} + \frac{Q^* u^-_q}{(u^+)^2 + (u^-)^2} \right] $$
$$ i^{-,*}_q = \frac{2}{3} \left[ -\frac{P^* u^-_q}{(u^+)^2 – (u^-)^2} – \frac{Q^* u^-_d}{(u^+)^2 + (u^-)^2} \right] $$
where \( P^* \) and \( Q^* \) are the active and reactive power references, and \( u^+_d, u^+_q, u^-_d, u^-_q \) are the dq-axis components of the positive and negative sequence voltages. The peak phase currents can be derived as:
$$ I_a = \frac{2}{3} \sqrt{ \left[ (1 + \epsilon^2)(u^+)^2 – 2\epsilon (u^+)^2 \cos(\phi) \right] M } $$
$$ I_b = \frac{2}{3} \sqrt{ \left[ (1 + \epsilon^2)(u^+)^2 – 2\epsilon (u^+)^2 \cos\left(\phi – \frac{2\pi}{3}\right) \right] M } $$
$$ I_c = \frac{2}{3} \sqrt{ \left[ (1 + \epsilon^2)(u^+)^2 – 2\epsilon (u^+)^2 \cos\left(\phi + \frac{2\pi}{3}\right) \right] M } $$
$$ M = \left( \frac{P^*}{(1 – \epsilon^2)(u^+)^2} \right)^2 + \left( \frac{Q^*}{(1 + \epsilon^2)(u^+)^2} \right)^2 $$
with \( \phi = \phi^+ – \phi^- \). To prevent overcurrent tripping, the maximum current must not exceed the rated current \( I_{\text{rated}} \) of the solar inverters. Thus, the condition \( I_{\text{max}} \leq I_{\text{rated}} \) must hold. When the PV-generated power \( P_{\text{pv}} \) is high, the active power reference is limited to:
$$ P_{\text{max}} = \frac{3}{2} I_{\text{rated}} (1 – \epsilon) u^+ $$
In this case, \( P^* = P_{\text{max}} \) and \( Q^* = 0 \). Conversely, when \( P_{\text{pv}} \) is low, reactive power can be injected to support the grid. The reactive power reference is computed as:
$$ Q^* = \sqrt{ \left( \frac{3}{2} I_{\text{rated}} (1 + \epsilon) u^+ \right)^2 – \left( \frac{P_{\text{pv}}}{1 – \epsilon^2} \right)^2 } $$
This ensures that the solar inverters operate at their maximum capacity without exceeding current limits. The instantaneous active and reactive powers are then:
$$ P = P^* $$
$$ Q = Q^* + \frac{2\epsilon \cos(2\omega t – \phi)}{1 + \epsilon^2} Q^* – \frac{2\epsilon \sin(2\omega t – \phi)}{1 – \epsilon^2} P^* $$
showing that active power oscillations are eliminated, while reactive power may contain a ripple component. This control strategy effectively enhances the LVRT performance of solar inverters.
To validate the proposed method, I conducted simulations using MATLAB/Simulink. The system parameters are summarized in Table 1. The grid voltage is set to 380 V (line-to-line RMS), and a voltage sag is simulated by reducing the phase-A voltage to 20% of its nominal value at t = 0.3 s, with recovery at t = 0.5 s. The performance of the solar inverters is evaluated under two scenarios: high PV power output and low PV power output.
| Parameter | Value | Unit |
|---|---|---|
| Grid Voltage Amplitude | 380 | V |
| Rated Current Amplitude | 10 | A |
| Grid Frequency | 50 | Hz |
| DC-link Capacitance | 1.9 | mF |
| LCL Inverter-side Inductance | 1.5 | mH |
| LCL Grid-side Inductance | 1.5 | mH |
| LCL Filter Capacitance | 5 | μF |
| Switching Frequency | 10 | kHz |
In the first scenario, with \( P_{\text{pv}} = 3.0 \, \text{kW} \), the solar inverters adjust the active power reference to \( P_{\text{max}} = 2.8 \, \text{kW} \) during the sag to limit the current. The three-phase output currents remain below 10 A, as shown in Table 2, which summarizes the current peaks before, during, and after the sag. The active power remains constant at 2.8 kW without oscillations, while the reactive power is near zero. This demonstrates the ability of the solar inverters to avoid overcurrent while maintaining stable power delivery.
| Phase | Before Sag (A) | During Sag (A) | After Sag (A) |
|---|---|---|---|
| A | 8.5 | 9.8 | 8.5 |
| B | 8.5 | 9.6 | 8.5 |
| C | 8.5 | 9.7 | 8.5 |
In the second scenario, with \( P_{\text{pv}} = 1.5 \, \text{kW} \), the solar inverters inject reactive power to support the grid. The reactive power reference is calculated using Equation (16), resulting in \( Q^* = 2.5 \, \text{kVAR} \) during the sag. The instantaneous powers are plotted in the simulations, showing that active power is maintained at 1.5 kW without ripple, while reactive power has a small oscillation due to the negative sequence component. The current peaks are all within the rated limit, as indicated in Table 3. This highlights the flexibility of the control strategy in utilizing the solar inverters’ capacity for grid support.
| Phase | Before Sag (A) | During Sag (A) | After Sag (A) |
|---|---|---|---|
| A | 5.2 | 9.5 | 5.2 |
| B | 5.2 | 9.3 | 5.2 |
| C | 5.2 | 9.4 | 5.2 |
Furthermore, I evaluated the grid voltage support capability by considering line impedance \( Z = R + jX = 3 + j0.5 \, \Omega \). The PCC voltages under unbalanced conditions (e.g., \( u_a = 0.7 \angle 0^\circ \), \( u_b = 0.7 \angle -127^\circ \), \( u_c = 1 \angle -225^\circ \)) are analyzed. With the proposed control, the voltage unbalance factor decreases from 0.3 to 0.1, and the positive sequence voltage amplitude increases by 15%, as summarized in Table 4. This confirms that the solar inverters can effectively mitigate voltage unbalance and provide reactive support, enhancing grid stability during faults.
| Metric | Without Control | With Proposed Control |
|---|---|---|
| Voltage Unbalance Factor (\( \epsilon \)) | 0.3 | 0.1 |
| Positive Sequence Voltage (V) | 0.8 p.u. | 0.92 p.u. |
| Negative Sequence Voltage (V) | 0.24 p.u. | 0.09 p.u. |
The mathematical foundation of this control strategy relies on the precise decoupling of sequence components. For implementation in digital controllers, I recommend using a dual second-order generalized integrator (DSOGI) for sequence extraction, which provides accurate and fast detection of positive and negative voltages. The block diagram of the control system is illustrated in the simulations, showing the inner current loops and outer power loops. The robustness of solar inverters under various fault conditions is enhanced by this approach, making them suitable for modern grid requirements.
In conclusion, the proposed LVRT control strategy for solar inverters offers a comprehensive solution to address voltage sags in grid-connected PV systems. By dynamically adjusting active and reactive power references, it ensures that the inverter currents remain within safe limits, eliminates active power oscillations, reduces voltage unbalance, and provides reactive support to the grid. The effectiveness is verified through detailed simulations, demonstrating stable operation under both high and low PV power conditions. This strategy not only complies with grid codes but also maximizes the utilization of solar inverters, contributing to the reliability and stability of power systems with high PV penetration. Future work could focus on hardware-in-the-loop testing and integration with energy storage systems to further enhance LVRT capabilities.