In recent years, the solar power industry has experienced rapid growth, with photovoltaic (PV) systems becoming a significant contributor to the global energy mix. As solar inverters play a critical role in interfacing PV panels with the grid, their control performance under grid disturbances is paramount for system stability. This is especially true in weak grids, characterized by high impedance and low short-circuit ratios, where issues like voltage imbalances and resonances are common. Low-voltage ride-through (LVRT) capability is a mandatory requirement for grid-connected solar inverters, ensuring they remain connected and provide reactive current support during voltage sags. However, existing standards often lack explicit requirements for active current during LVRT, which can lead to frequency fluctuations and stability problems in weak grids. In this article, we explore advanced control strategies for solar inverters to enhance LVRT performance in weak interconnection grids, focusing on phase-locked loop (PLL) techniques, resonance suppression, and coordinated active-reactive current control.
The integration of solar inverters into weak grids poses unique challenges. Weak grids, often found in remote areas with high PV penetration, have substantial line impedances and transformer leakage inductances. These factors can exacerbate voltage imbalances during faults, leading to negative-sequence components and harmonics in the grid voltage. Traditional control methods for solar inverters, designed for balanced grids, may fail to mitigate these issues, resulting in poor power quality and potential resonance between the inverter and grid. Resonance, influenced by distributed capacitance and parasitic inductance, can cause oscillations and instability. Therefore, it is essential to develop robust control strategies that address both LVRT requirements and weak grid conditions.
From a design perspective, we analyze key aspects of solar inverter control in weak grids. First, we examine phase-locked loop (PLL) methods for unbalanced grid conditions. Common PLLs include the Synchronous Reference Frame PLL (SRF-PLL), Dual Second-Order Generalized Integrator PLL (DSOGI-PLL), and Decoupled Double Synchronous Reference Frame PLL (DDSRF-PLL). The SRF-PLL is suitable for ideal grid conditions but struggles with voltage imbalances. The DSOGI-PLL can handle imbalances but requires significant computational resources. In contrast, the DDSRF-PLL decouples positive- and negative-sequence components, allowing accurate and fast detection of grid voltage amplitude and phase even during faults. For solar inverters in weak grids, the DDSRF-PLL is preferred due to its effectiveness in unbalanced scenarios.
Under unbalanced grid voltages, the grid voltage can be expressed as the sum of positive-sequence, negative-sequence, and zero-sequence components. Since most solar inverters are connected to three-phase three-wire systems, the zero-sequence component is negligible. Using Clarke and Park transformations, the voltage in the stationary αβ-frame can be derived. Let the grid voltage be represented as:
$$ V_S = V_S^{+1} + V_S^{-1} $$
After transformations, the voltage in the rotating dq-frame for positive and negative sequences is given by:
$$ \begin{bmatrix} V_{d}^{+1} \\ V_{q}^{+1} \end{bmatrix} = V_S^{+1} \begin{bmatrix} \cos(\omega t + \phi^{+} – \theta’) \\ -\sin(\omega t + \phi^{+} – \theta’) \end{bmatrix} + V_S^{-1} \begin{bmatrix} \cos(-\omega t + \phi^{-} – \theta’) \\ -\sin(-\omega t + \phi^{-} – \theta’) \end{bmatrix} $$
$$ \begin{bmatrix} V_{d}^{-1} \\ V_{q}^{-1} \end{bmatrix} = V_S^{+1} \begin{bmatrix} \cos(\omega t + \phi^{+} + \theta’) \\ \sin(\omega t + \phi^{+} + \theta’) \end{bmatrix} + V_S^{-1} \begin{bmatrix} \cos(-\omega t + \phi^{-} + \theta’) \\ \sin(-\omega t + \phi^{-} + \theta’) \end{bmatrix} $$
When the PLL is locked, the positive-sequence voltage aligns with the d-axis, leading to coupling terms at twice the grid frequency. The DDSRF-PLL employs cross-decoupling to extract DC components, eliminating these harmonics and enabling precise phase detection. The control framework for DDSRF-PLL is summarized in the following table, highlighting its advantages for solar inverters in weak grids.
| PLL Type | Key Feature | Suitability for Weak Grids |
|---|---|---|
| SRF-PLL | Fast phase detection under balanced conditions | Low – fails during voltage imbalances |
| DSOGI-PLL | Handles imbalances via adaptive filtering | Moderate – high computational load |
| DDSRF-PLL | Decouples positive and negative sequences | High – effective in unbalanced faults |
Resonance suppression is another critical aspect for solar inverters in weak grids. Resonance can occur due to interactions between inverter output filters and grid impedance, leading to instability. Suppression strategies are broadly classified into passive damping and active damping. Passive damping uses physical resistors but incurs power losses, while active damping employs control algorithms to mimic damping behavior. For solar inverters, active damping is preferred due to higher efficiency. Among active damping methods, virtual resistor approaches are common, where a virtual resistor is inserted in the capacitor branch of the LCL filter to damp resonances. The equivalent structure involves adding a damping current component to the current control loop, enhancing stability without additional hardware.
The virtual resistor control algorithm for solar inverters can be illustrated through a block diagram. By incorporating a feedback term based on capacitor current, the system’s damping characteristics are improved. This method is adaptable to varying grid impedances, making it suitable for weak grids where resonance frequencies may shift. The effectiveness of this approach has been validated in various studies, showing reduced harmonic distortion and better performance for solar inverters under LVRT conditions.
Turning to LVRT control strategies, traditional methods for solar inverters often prioritize reactive current support as per standards, with active current set to zero to protect power semiconductor devices like IGBTs from overcurrent. The reactive current reference during LVRT is typically calculated based on voltage dip depth:
$$ I_{q,ref} = K \times (0.9 – V_{dip}) \times I_n $$
where \( I_{q,ref} \) is the reactive current reference, \( K \) is a gain factor (often 1.5 to 2), \( V_{dip} \) is the per-unit voltage dip, and \( I_n \) is the rated current. However, this ignores active current, which can lead to frequency deviations in weak grids due to sudden active power loss. To address this, we propose a coordinated active-reactive current control strategy for solar inverters during LVRT. The active current reference is determined based on the inverter’s overcurrent capability and pre-fault active power. Solar inverters are typically designed to handle 1.1 times rated current, so the active current reference \( I_{d,ref} \) should satisfy:
$$ I_{d,ref} \leq \sqrt{(1.1 I_n)^2 – I_{q,ref}^2} $$
Additionally, the pre-fault active current reference \( I_{d,ref}^* \) is stored, and the final \( I_{d,ref} \) is set as the minimum between \( I_{d,ref}^* \) and the calculated limit. This ensures that solar inverters provide both reactive support and active power during faults, mitigating frequency swings in weak grids.
The control logic for this coordinated strategy involves several steps. First, grid voltages are processed through DDSRF-PLL to obtain positive-sequence voltage components and phase angle. If the voltage dips below 0.9 per unit, LVRT mode is activated, locking the pre-fault active current reference and phase angle. Reactive current reference is computed based on dip depth, and active current reference is adjusted accordingly. The current references are then used in a dual-dq current control loop to generate modulation signals. This approach enhances the stability of solar inverters in weak grids during LVRT events.
To model solar inverters under unbalanced conditions, we use the DDSRF framework. The inverter dynamics in the positive and negative dq-frames are described by:
$$ \begin{aligned}
V_{d}^{+} &= L \frac{di_{d}^{+}}{dt} + e_{d}^{+} – \omega L i_{q}^{+} \\
V_{q}^{+} &= L \frac{di_{q}^{+}}{dt} + e_{q}^{+} + \omega L i_{d}^{+} \\
V_{d}^{-} &= L \frac{di_{d}^{-}}{dt} + e_{d}^{-} + \omega L i_{q}^{-} \\
V_{q}^{-} &= L \frac{di_{q}^{-}}{dt} + e_{q}^{-} – \omega L i_{d}^{-}
\end{aligned} $$
where \( V_{d}^{+}, V_{q}^{+}, V_{d}^{-}, V_{q}^{-} \) are inverter output voltages, \( i_{d}^{+}, i_{q}^{+}, i_{d}^{-}, i_{q}^{-} \) are currents, \( e_{d}^{+}, e_{q}^{+}, e_{d}^{-}, e_{q}^{-} \) are grid voltages, \( L \) is filter inductance, and \( \omega \) is grid frequency. This model allows independent control of positive and negative sequence currents, enabling solar inverters to handle unbalanced faults effectively. Combined with virtual resistor damping, the overall control strategy ensures robust performance in weak grids.
We implemented this control strategy on a hardware platform based on a TMS320F28335 DSP and CPLD, which is commonly used for solar inverter applications. The platform was integrated with a real-time hardware-in-the-loop (HIL) simulation using RT-LAB to validate the approach. The solar inverter parameters in the HIL setup are summarized below.
| Parameter | Value |
|---|---|
| Rated Power | 500 kW |
| Grid Voltage | 315 V (line-to-line) |
| MPPT Voltage | 600 V |
| Switching Frequency | 3 kHz |
| DC-Link Capacitance | 10,080 μF |
| Filter Inductance | 0.5 mH |
| Filter Capacitance | 220 μF (delta-connected) |
The HIL simulation model emulated a weak grid with high impedance and included a voltage dip generator to test LVRT. The solar inverter was operated at 350 kW active power and zero reactive power initially. A three-phase voltage dip to 0.6 per unit was applied for 1.41 seconds. Waveforms captured during the test showed that the solar inverter maintained grid connection, provided reactive current support, and delivered active current per the coordinated strategy, minimizing frequency deviations. The virtual resistor damping effectively suppressed resonances, confirming the strategy’s validity for solar inverters in weak grids.
In conclusion, the integration of solar inverters into weak grids requires advanced control strategies to address LVRT challenges. We have analyzed DDSRF-PLL for accurate phase detection under unbalanced voltages and virtual resistor-based active damping for resonance suppression. Moreover, we proposed a coordinated active-reactive current control method during LVRT, ensuring solar inverters provide both reactive support and active power to stabilize weak grids. The strategy was validated through HIL simulations, demonstrating improved performance for solar inverters in terms of grid stability and power quality. Future work could focus on adaptive tuning for varying grid conditions and integration with energy storage systems to further enhance the resilience of solar inverters in weak interconnection grids.