The global energy landscape is undergoing a profound transformation, driven by the urgent need to mitigate climate change and ensure sustainable development. Among the various renewable energy sources, solar photovoltaic (PV) power generation has emerged as a cornerstone technology due to its abundance, scalability, and rapidly decreasing costs. Large-scale grid-connected PV plants are becoming increasingly common, contributing significant portions of electricity to power systems worldwide. However, this high penetration level introduces new challenges for grid stability and operation. Unlike traditional synchronous generators, PV systems interface with the grid through power electronic converters, primarily solar inverters, which have different dynamic characteristics and responses to grid disturbances.
One of the most critical challenges is maintaining grid stability during and after faults that cause voltage dips. Historically, distributed generators were required to disconnect immediately when grid voltage fell outside normal ranges to protect the equipment and prevent instability. This practice, however, can exacerbate the situation during major faults by causing a sudden loss of a large amount of generation, potentially leading to a cascading failure. Therefore, modern grid codes mandate that PV power plants, similar to wind farms, must possess Low Voltage Ride-Through (LVRT) capability. LVRT requires the solar inverter to remain connected to the grid and support voltage recovery during specified periods of voltage depression, only disconnecting if the fault persists beyond a defined time limit or if the voltage drops below a very severe threshold.
The core of the LVRT challenge lies in the solar inverter control. During a voltage sag, the ability to inject active power into the grid is inherently limited by the reduced grid voltage. If the inverter continues to try to deliver pre-fault levels of active power, it will lead to excessive current, potentially damaging the semiconductor switches and triggering protection shutdowns. Furthermore, the imbalance between the DC power generated by the PV panels and the AC power injected into the faulty grid causes the DC-link voltage to rise dangerously. An effective LVRT strategy must therefore address three key objectives: 1) prevent overcurrent in the inverter switches, 2) stabilize the DC-link voltage, and 3) provide dynamic reactive power support to help the grid voltage recover. This article delves into a sophisticated control strategy for solar inverters based on reactive current injection and control mode switching, offering a robust solution for reliable grid integration.
System Configuration and Mathematical Modeling
A standard two-stage three-phase grid-connected PV system topology is considered, as it is prevalent in medium to high-power applications. The system comprises a PV array, a DC-DC boost converter, a DC-AC voltage source inverter (VSI), an L-filter, and the utility grid. The boost converter performs Maximum Power Point Tracking (MPPT) to optimize energy harvest from the PV panels and elevates the DC voltage to a level suitable for the inverter. The VSI, the central component of the solar inverter, converts the DC power into synchronized AC power for grid injection.
The dynamics of the three-phase grid-side inverter in the synchronous rotating dq-reference frame, aligned with the grid voltage vector (usually along the d-axis), are described by the following differential equations. This transformation simplifies control by converting AC quantities into DC values.
$$L\frac{di_d}{dt} = u_{cd} – e_d – Ri_d + \omega L i_q$$
$$L\frac{di_q}{dt} = u_{cq} – e_q – Ri_q – \omega L i_d$$
$$C_{dc}\frac{du_{dc}}{dt} = i_{pv} – i_{dc} = i_{pv} – \frac{3}{2}(s_d i_d + s_q i_q)$$
Where:
- $i_d$, $i_q$ are the d- and q-axis components of the grid current.
- $u_{cd}$, $u_{cq}$ are the d- and q-axis components of the inverter output voltage.
- $e_d$, $e_q$ are the d- and q-axis components of the grid voltage (with $e_q = 0$ in grid-voltage-oriented control).
- $R$, $L$ are the resistance and inductance of the grid filter.
- $\omega$ is the grid angular frequency.
- $C_{dc}$ is the DC-link capacitance.
- $u_{dc}$ is the DC-link voltage.
- $i_{pv}$ is the current from the DC-DC boost stage.
- $s_d$, $s_q$ are the switching functions in the dq-frame.
In this model, the active power $P$ and reactive power $Q$ are proportional to the d-axis and q-axis currents, respectively, when the d-axis is aligned with the grid voltage:
$$P = \frac{3}{2} e_d i_d$$
$$Q = -\frac{3}{2} e_d i_q$$
This decoupling is fundamental for independent control of active and reactive power output from the solar inverter.
Normal Operation Control Strategy
Under normal grid conditions, the primary control objectives for the solar inverter are to extract maximum power from the PV array and to inject this power into the grid with high quality (low harmonic distortion) and a specified power factor, typically unity.
1. Boost Converter Control (MPPT):
The DC-DC boost converter operates with Perturb and Observe (P&O) algorithm to maintain the PV array at its Maximum Power Point (MPP). The algorithm periodically perturbs the operating voltage and observes the resulting change in power. The duty cycle $D$ of the boost switch is adjusted to move the operating point toward the MPP. The input-output voltage relationship is given by:
$$U_{dc} = \frac{U_{pv}}{1-D}$$
where $U_{pv}$ is the PV array voltage and $U_{dc}$ is the boosted DC-link voltage.
2. Grid-Side Inverter Control (Dual-Loop):
The inverter typically employs a cascaded dual-loop control structure.
- Outer Voltage Loop: A PI controller regulates the DC-link voltage $u_{dc}$ to a fixed reference $u_{dc}^*$. The output of this controller is the reference for the active current $i_d^*$. This loop ensures power balance between the DC input from the panels and the AC output to the grid.
- Inner Current Loop: Two fast-acting PI controllers regulate the d-axis and q-axis currents ($i_d$, $i_q$) to follow their references ($i_d^*$, $i_q^*$). To achieve decoupled control, cross-coupling terms ($\omega L i_q$, $\omega L i_d$) and grid voltage feedforward ($e_d$, $e_q$) are added to the controller outputs. The resulting voltage commands $u_{cd}^*$ and $u_{cq}^*$ are:
$$u_{cd}^* = \left(K_{p\_id} + \frac{K_{i\_id}}{s}\right)(i_d^* – i_d) – \omega L i_q + e_d$$
$$u_{cq}^* = \left(K_{p\_iq} + \frac{K_{i\_iq}}{s}\right)(i_q^* – i_q) + \omega L i_d + e_q$$
These are then transformed back to the abc-frame and used by the Pulse Width Modulation (PWM) block to generate switching signals for the solar inverter.
Normally, $i_q^*$ is set to zero for unity power factor operation.
Proposed LVRT Control Strategy: Reactive Current Injection and Mode Switching
The proposed strategy is activated upon detection of a grid voltage dip below a predefined threshold (e.g., 0.9 pu). It involves coordinated switching of the control modes for both the boost converter and the grid-side inverter to fulfill LVRT requirements without additional hardware like chopper resistors or energy storage.
1. Inverter Control Mode Switch (Dual-Loop to Single-Loop):
During a fault, maintaining a constant DC-link voltage via the outer voltage loop is not the priority. The priority shifts to current control: limiting the absolute current magnitude to safe levels and injecting reactive current for grid support. Therefore, the outer DC-voltage loop is deactivated. The reference for the active current $i_d^*$ is no longer provided by the voltage PI controller. Instead, it is calculated dynamically based on the permissible maximum current $I_{max}$ (often 1.0-1.2 pu of the rated current) and the required reactive current reference $i_q^*$:
$$i_d^* = \sqrt{I_{max}^2 – (i_q^*)^2}$$
This ensures the total current magnitude does not exceed the solar inverter‘s rating. The system thus switches from voltage-oriented dual-loop control to a current-controlled single-loop strategy, where the inner current loops directly track the new $i_d^*$ and $i_q^*$ references.
2. Reactive Current Reference Generation:
The reactive current reference $i_q^*$ is generated according to the grid code requirements, which typically stipulate a specific amount of reactive current injection proportional to the voltage dip depth. A common piecewise linear function, compliant with standards like the Chinese GB/T 19964-2012, can be defined as:
$$
i_q^* =
\begin{cases}
1.05 \cdot I_N & \text{for } U_T < 0.2 \\
k \cdot (0.9 – U_T) \cdot I_N & \text{for } 0.2 \leq U_T \leq 0.9 \\
0 & \text{for } U_T > 0.9
\end{cases}
$$
Where $U_T$ is the per-unit value of the positive-sequence grid voltage at the Point of Common Coupling (PCC), $I_N$ is the rated current of the solar inverter, and $k$ is a gain factor (e.g., 1.5 as per some standards). This proportional injection provides dynamic voltage support, helping to “push” the grid voltage back towards its nominal value.
3. Boost Converter Control Mode Switch (MPPT to Voltage Regulation):
With the inverter injecting less active power, the power balance on the DC-link is disrupted. The PV array, still operating at MPPT, would produce more power than the inverter can sink, causing $u_{dc}$ to rise. To prevent this overvoltage, the boost converter is switched from MPPT mode to a DC voltage regulation mode. It now controls the PV array voltage $U_{pv}$ to a lower reference value, effectively moving the PV operating point away from the MPP and reducing the generated power to match what the inverter can deliver. If the voltage dip is extremely deep (e.g., $U_T < 0.2$), the boost converter may be shut down entirely, allowing the PV array voltage to float near its open-circuit voltage, minimizing power generation.
The following table summarizes the control mode transitions:
| Component | Normal Operation Mode | LVRT Operation Mode | Key Objective during LVRT |
|---|---|---|---|
| Grid-Side Inverter | Dual-Loop Control (Voltage Outer + Current Inner) | Single-Loop Current Control | Limit current magnitude; Inject reactive power per grid code. |
| Boost Converter | MPPT Control | DC Voltage Regulation (or shutdown for deep faults) | Regulate DC-link voltage by reducing PV power extraction. |
Simulation Analysis and Validation
The proposed LVRT strategy was validated through detailed simulation using a professional electromagnetic transient tool. A model of a 1 MW PV system with the described topology and control algorithms was developed. A severe symmetrical three-phase fault was simulated, causing the grid voltage at the PCC to drop to 0.5 pu for a duration of 500 ms.
The simulation results clearly demonstrate the effectiveness of the strategy:
- Grid Current: Upon fault inception, the control mode seamlessly switches. The three-phase grid currents remain sinusoidal and are effectively limited to a safe amplitude, avoiding dangerous overcurrent conditions in the solar inverter switches.
- Active and Reactive Power: The active power $P$ injection drops significantly due to the reduced voltage and the current limitation logic ($i_d^*$ calculation). Simultaneously, a substantial amount of reactive power $Q$ is injected according to the $i_q^*$ profile, providing crucial dynamic voltage support to the weakened grid.
- DC-Link Voltage: The switch of the boost converter to voltage regulation mode successfully stabilizes the DC-link voltage $u_{dc}$. A small, manageable transient is observed, but the voltage is quickly regulated back to its reference, preventing any tripping due to DC overvoltage.
- Voltage Recovery: After fault clearance, the control modes switch back to normal operation (MPPT and dual-loop control) smoothly. The system regains stable operation without any oscillations or instability, demonstrating successful LVRT.
The transient behavior of key system variables can be described by analyzing the modified system equations during LVRT. With the outer loop disabled, the DC-link voltage dynamics become:
$$C_{dc}\frac{du_{dc}}{dt} = i_{pv}(U_{pv\_reg}) – \frac{3}{2} e_d i_d$$
where $i_{pv}(U_{pv\_reg})$ is the PV current at the regulated (reduced) voltage set by the boost converter. The controller ensures $i_{pv}$ adjusts to balance the right side of the equation, stabilizing $u_{dc}$. The current dynamics are governed by the fast inner loop PI controllers tracking the new references $i_d^*$ and $i_q^*$, ensuring rapid and accurate reactive support.
Comparative Analysis and Discussion
The proposed strategy offers distinct advantages over other common LVRT methods for solar inverters:
| LVRT Method | Key Principle | Advantages | Disadvantages |
|---|---|---|---|
| DC Chopper (Passive) | Dissipates excess DC power in a resistor. | Simple, reliable, fast overvoltage suppression. | Wastes energy, generates heat, requires large hardware. |
| Battery Storage Support | Absorbs excess power or provides deficit. | Maintains full power delivery/support. | High cost, complex energy management, reduced lifespan. |
| Advanced MPPT Limiting | Actively moves PV operating point off MPP. | No extra hardware, manages DC power. | Slower response, requires precise PV model. |
| Proposed Strategy (Control Switching) | Switches inverter/boost control modes for reactive support and DC balance. | No added hardware, provides grid support, manages DC-link, cost-effective. | Relies on software robustness, requires accurate grid sensing. |
The control mode switching strategy is particularly attractive because it leverages the existing hardware and control capabilities of a modern two-stage solar inverter. It transforms the inverter from a simple power source into an active grid supporter during disturbances, aligning with the future vision of “grid-forming” or “grid-supporting” inverters. The challenge lies in the seamless and bumpless transition between control modes, which requires careful design of the switching logic and potential use of anti-windup techniques in the PI controllers.
Conclusion
As photovoltaic penetration continues to rise, the ability of PV power plants to support the grid during faults is no longer optional but a mandatory requirement. This article has presented a comprehensive and effective Low Voltage Ride-Through strategy for two-stage grid-connected solar inverters. The core of the strategy is the coordinated switching of control modes: the grid-side inverter transitions from a standard dual-loop control to a reactive-power-focused single-loop current control, while the front-end boost converter switches from MPPT to DC voltage regulation. This approach actively limits the output current to safe levels, provides dynamic reactive current injection proportional to the voltage dip to aid grid recovery, and maintains the DC-link voltage stability by reducing the power extracted from the PV array.
The strategy successfully meets modern grid code requirements without the need for additional hardware such as braking choppers or storage elements, making it a highly cost-effective and practical solution. Simulation studies confirm its efficacy in maintaining stable operation, protecting the solar inverter hardware, and fulfilling its role as a supportive asset to the power grid during transient faults. Future work may explore the integration of this strategy with advanced grid-feeding and grid-forming capabilities, further enhancing the value and resilience of inverter-based renewable energy resources.