The widespread integration of large-scale photovoltaic (PV) power generation into modern power grids presents significant challenges to system stability. A critical requirement for grid-connected solar inverters is the capability to remain connected and support the network during transient voltage dips, a function known as Low-Voltage Ride-Through (LVRT). The sudden disconnection of a major solar inverter during a grid fault can exacerbate power imbalances, hindering system recovery. Therefore, developing advanced LVRT control strategies for solar inverters is paramount for maintaining grid reliability. This article delves into a sophisticated control strategy designed for two-stage solar inverters, focusing on rapid power balance and coordinated active/reactive current injection to fulfill LVRT requirements without additional hardware.
The typical solar inverter architecture can be categorized into single-stage and two-stage topologies. The two-stage solar inverter, comprising a front-end DC-DC booster (typically a Boost converter) and a rear-end DC-AC inverter, offers advantages like higher efficiency over a wider input voltage range. Our analysis centers on this prevalent configuration. During normal operation, the Boost converter performs Maximum Power Point Tracking (MPPT) to extract the maximum available power from the PV array, while the grid-side inverter regulates the DC-link voltage and injects synchronized active power into the grid.
The fundamental challenge for a two-stage solar inverter during an LVRT event arises from the instantaneous power imbalance. When a grid voltage sag occurs, the power that can be delivered to the grid ($P_{out}$) is drastically reduced due to voltage limitations and current caps. However, if the MPPT algorithm remains active, the PV array continues to generate power near its maximum ($P_{pv}$). This surplus power ($\Delta P$) charges the DC-link capacitor, causing a potentially hazardous rise in the DC-link voltage ($u_{dc}$), as described by:
$$
\Delta P = P_{pv} – P_{out} = \frac{1}{2} C \frac{(u’_{dc})^2 – (u_{dc})^2}{\Delta t}
$$
where $C$ is the DC-link capacitance, and $u_{dc}$ and $u’_{dc}$ are the DC-link voltages before and after the fault, respectively. Simultaneously, the inverter’s current control loop, in an attempt to regulate the soaring DC-link voltage, may drive the output current beyond the solar inverter’s safe operating limits. Consequently, an effective LVRT strategy for a two-stage solar inverter must swiftly mitigate this power imbalance to protect the hardware while ensuring continuous grid connection.
The proposed integrated control strategy tackles the LVRT challenge by modifying the control objectives of both the Boost converter and the grid-side inverter upon fault detection. The core innovation lies in the seamless transition of the Boost converter’s role from MPPT to DC-link voltage stabilizer.
Enhanced Boost Converter Control for LVRT
Under normal conditions, the Boost converter in a solar inverter operates with an MPPT algorithm, such as Perturb and Observe (P&O), adjusting the PV array voltage ($u_{pv}$) to maximize power harvest. During an LVRT event, this objective must change immediately. The proposed strategy introduces a DC-link voltage control outer loop that overrides the MPPT reference. The control law is given by:
$$
u_{pv\_ref} = u_{max} + \left[ K_{p\_B} + \frac{K_{i\_B}}{s} \right] (u_{dc\_ref} – u_{dc})
$$
where $u_{max}$ is the pre-fault maximum power point voltage (fed forward), $u_{pv\_ref}$ is the new reference for the PV array voltage, $u_{dc\_ref}$ and $u_{dc}$ are the reference and measured DC-link voltages, and $K_{p\_B}$ and $K_{i\_B}$ are the PI controller gains. When $u_{dc}$ rises due to the fault, this controller outputs a positive signal, increasing $u_{pv\_ref}$ above $u_{max}$. Operating on the right side of the P-V curve (higher voltage, lower power) allows for a faster and more significant reduction in $P_{pv}$ compared to operating on the left side, effectively curtailing the PV power to match the reduced grid injection capability. This active power curtailment by the front-end converter is the first pillar of stabilizing the solar inverter during faults.
Coordinated Current Control for Grid Support
While the Boost converter manages the DC-link stability, the grid-side inverter must fulfill the reactive current injection mandates of LVRT grid codes. The standard current control in the synchronous ($dq$) reference frame is described by:
$$
\begin{aligned}
u_d &= \left( K_{p\_i} + \frac{K_{i\_i}}{s} \right)(i^*_d – i_d) – \omega L i_q + e_d \\
u_q &= \left( K_{p\_i} + \frac{K_{i\_i}}{s} \right)(i^*_q – i_q) + \omega L i_d + e_q
\end{aligned}
$$
where $i_d$ and $i_q$ are the active and reactive currents, $e_d$ and $e_q$ are the grid voltage components, and $L$ is the filter inductance. Normally, $i^*_q=0$ for unity power factor, and $i^*_d$ comes from the DC-link voltage PI controller. During LVRT, this reference generation is modified. The active current reference $i^*_{d\_fault}$ is dynamically reduced based on the voltage sag depth $k = U_{fault}/U_{rated}$, while the reactive current reference $i^*_{q\_fault}$ is maximized within the inverter’s current limit $I_{max}$.
The current references are calculated as follows:
$$
i^*_{d\_fault}(t) = \begin{cases}
i^*_d(pre-fault) & t_0 \leq t < t_1 \\
\max( i^*_{d\_fault}(t-\Delta t) – i_R \cdot \Delta t, \quad k \cdot I_{max} ) & t_1 \leq t < t_3
\end{cases}
$$
$$
i^*_{q\_fault} = \sqrt{I^2_{max} – (i^*_{d\_fault})^2}
$$
where $t_0$ is the fault inception time, $t_1 = t_0 + \Delta T_{hold}$ (a short hold period to avoid transients), $i_R$ is a defined ramp rate, and $t_3$ is the voltage recovery time. This strategy ensures the solar inverter’s total current magnitude does not exceed $I_{max}$ while prioritizing reactive power injection to support grid voltage recovery, as mandated by modern grid codes for solar inverters.
Simulation Analysis and Verification
A detailed simulation model of a 100 kW two-stage solar inverter was developed to validate the proposed LVRT strategy. Key system parameters are summarized in Table 1.
| Parameter | Value | Unit |
|---|---|---|
| Rated Power | 100 | kW |
| Grid Voltage (L-N, RMS) | 220 | V |
| DC-Link Voltage Reference ($u_{dc\_ref}$) | 800 | V |
| DC-Link Capacitance ($C$) | 6000 | μF |
| Inverter-side Inductor ($L_1$) | 0.4 | mH |
| Grid-side Inductor ($L_2$) | 0.05 | mH |
| Filter Capacitor ($C_f$) | 10 | μF |
| Inverter Current Limit ($I_{max}$) | 1.1 * 200 | A |
A severe three-phase symmetrical voltage sag to 0.2 per unit (p.u.) was applied at t=0.3s, lasting for 100ms. The dynamic response of the solar inverter under the proposed control is captured in multiple key waveforms.
The most critical result is the DC-link voltage behavior. As shown conceptually in Table 2, the proposed strategy successfully limits the DC-link overvoltage peak. Without LVRT control, the surplus energy would cause $u_{dc}$ to rise uncontrollably, likely triggering protective shutdowns in the solar inverter. With the proposed Boost control intervention, $u_{dc}$ experiences a small, manageable rise (e.g., to ~861.5V from a 800V reference) and is quickly stabilized as the PV array power is curtailed.
| Performance Metric | Without LVRT Control | With Proposed Control |
|---|---|---|
| DC-Link Voltage Peak | Exceeds safety limit, causing trip | Limited to safe value (e.g., ~1.08 p.u.) |
| Grid Current During Fault | Uncontrolled increase, potential overcurrent | Controlled within limit ($\leq I_{max}$) |
| Active Power ($P$) Injection | Forced high, exacerbating imbalance | Automatically reduced per voltage dip |
| Reactive Power ($Q$) Injection | Zero (Unity Power Factor) | Maximized within current capability |
| Grid Voltage Support | None | Provided, aids system recovery |
| Post-Fault Recovery | N/A (System trips) | Smooth and stable |
Concurrently, the grid-side current control performs as designed. The three-phase output currents remain sinusoidal and within the safe limit. The $dq$-current decomposition reveals the coordinated control: the active current $i_d$ is ramped down according to the sag depth, while the reactive current $i_q$ is increased to its maximum permissible value. This reactive current injection provides crucial dynamic voltage support at the Point of Common Coupling (PCC). A comparative simulation confirms that the PCC voltage during the fault is higher with reactive support from the solar inverter than without it, demonstrating the tangible grid-support benefit of the strategy.
The power balance is effectively managed. The PV array output power $P_{pv}$, initially at 100 kW, is rapidly reduced by the Boost converter to approximately 20 kW, aligning with the reduced grid injection capacity. This swift power curtailment is the key to preventing DC-link overvoltage. Upon voltage recovery at t=0.4s, both the Boost and inverter controls smoothly transition back to normal MPPT and unit power factor operation, with stable transients.
Mathematical Foundation of Power Balance
The efficacy of the strategy can be further understood through the power balance dynamics. The power at the DC-link node is governed by:
$$
P_{pv} = P_{dc\_link} = P_{inv} + P_{cap}
$$
where $P_{inv}$ is the power processed by the inverter (equal to $P_{out}$ minus losses), and $P_{cap}$ is the power absorbed by the capacitor. $P_{cap}$ is related to the voltage change:
$$
P_{cap} = \frac{d}{dt} \left( \frac{1}{2} C u_{dc}^2 \right) = C u_{dc} \frac{du_{dc}}{dt}
$$
During a fault, if $P_{pv}$ remains constant and $P_{inv}$ suddenly drops, then $P_{cap} > 0$, leading to $du_{dc}/dt > 0$. The proposed control law for the Boost converter effectively reduces $P_{pv}$ by shifting the PV operating point. The control aims to achieve $P_{pv} \approx P_{inv}$, making $P_{cap} \approx 0$ and thus $du_{dc}/dt \approx 0$, stabilizing the DC-link voltage. This closed-loop action is what allows the solar inverter to ride through the fault without auxiliary circuits.
Conclusion
This article has presented a comprehensive and effective low-voltage ride-through control strategy for two-stage solar inverters. The strategy hinges on two coordinated actions: firstly, the fast curtailment of photovoltaic power via a DC-link voltage control loop integrated into the Boost converter stage, which prevents dangerous overvoltage on the DC bus; secondly, the intelligent current reference generation in the grid-side inverter that prioritizes reactive current injection within the safe operating current limit of the solar inverter to provide dynamic grid voltage support. The proposed method achieves full LVRT compliance without the need for additional hardware such as braking choppers or energy storage devices, enhancing the solar inverter’s robustness and grid-support functionality. Simulation results from a detailed 100 kW model confirm that the strategy ensures stable operation during deep voltage sags, maintains currents within limits, and contributes positively to grid stability during faults. This control approach represents a significant step forward in the design of grid-friendly and resilient solar inverters for the modern power system.