In modern power systems, the integration of renewable energy sources like solar power is crucial for sustainable development. Solar inverters, which convert direct current (DC) from photovoltaic (PV) arrays into alternating current (AC) for grid connection, play a pivotal role in ensuring grid stability. Among various configurations, two-stage solar inverters, consisting of a Boost converter and a DC-AC inverter, are widely used due to their efficiency and flexibility. However, grid faults, particularly voltage sags, pose significant challenges to the reliable operation of solar inverters. Low-voltage ride-through (LVRT) capability is essential to prevent solar inverters from disconnecting during voltage dips, thereby maintaining grid stability. This article presents a comprehensive LVRT control strategy for two-stage solar inverters that integrates power coordination between active and reactive currents. The strategy focuses on modifying the Boost converter control to stabilize DC-link voltage and adjusting inverter current references to provide reactive support during faults. We delve into the theoretical analysis, control design, and simulation validation, emphasizing the importance of solar inverter resilience in grid-connected applications.
The two-stage solar inverter topology comprises a Boost circuit for DC voltage elevation and an inverter for grid synchronization. Typically, the Boost stage employs maximum power point tracking (MPPT) to optimize PV array output, while the inverter regulates grid current using vector control. Under normal conditions, the solar inverter operates at unity power factor, injecting active power into the grid. However, during voltage sags, the power imbalance between the PV array and the grid can lead to DC-link overvoltage and inverter overcurrent, potentially causing tripping. To address this, we propose an LVRT strategy that dynamically adjusts the Boost control mode and coordinates active and reactive current injection. This approach enhances the solar inverter’s fault tolerance without additional hardware, ensuring compliance with grid codes that mandate LVRT and reactive power support.
The conventional control of a two-stage solar inverter involves separate loops for the Boost converter and the inverter. For the Boost stage, MPPT algorithms like perturb and observe (P&O) are used to extract maximum power from the PV array. The inverter side utilizes a synchronous reference frame (SRF) based current control, with an outer voltage loop to regulate the DC-link voltage. The current references are given as:
$$
i_d^* = k_{p,v} (u_{dc,ref} – u_{dc}) + k_{i,v} \int (u_{dc,ref} – u_{dc}) dt
$$
$$
i_q^* = 0
$$
where $i_d^*$ and $i_q^*$ are the d-axis and q-axis current references, $k_{p,v}$ and $k_{i,v}$ are proportional and integral gains of the voltage controller, $u_{dc,ref}$ is the DC-link voltage reference, and $u_{dc}$ is the actual DC-link voltage. In this setup, the solar inverter solely injects active power. During voltage sags, however, this control fails to manage the power surplus, necessitating an LVRT strategy.
When a grid voltage dip occurs, the output power of the solar inverter decreases due to voltage reduction, while the PV array continues to generate power at the MPPT point. The power imbalance $\Delta P$ is expressed as:
$$
\Delta P = P_{pv} – P_{out} = \frac{1}{2} C \left( u_{dc}’^2 – u_{dc}^2 \right) / \Delta t
$$
where $P_{pv}$ is the PV array output power, $P_{out}$ is the grid-injected power, $C$ is the DC-link capacitance, $u_{dc}$ and $u_{dc}’$ are the DC-link voltages before and after the dip, and $\Delta t$ is the duration. This imbalance causes $u_{dc}$ to rise, potentially exceeding safe limits. Moreover, the inverter current may increase beyond rated values due to the voltage control loop. To mitigate this, we redesign the Boost control to rapidly reduce $P_{pv}$ by shifting the PV array operating point away from the maximum power point (MPP).
The proposed LVRT strategy for the two-stage solar inverter involves two key modifications: Boost converter control adaptation and coordinated current reference generation. First, during voltage sags (detected when grid voltage falls below 90% of nominal), the Boost control switches from MPPT to DC-link voltage regulation. A feedforward term based on the pre-fault MPP voltage is introduced to accelerate power adjustment. The control law becomes:
$$
u_{pv,ref} = u_{max} + \left[ k_{p,b} (u_{dc,ref} – u_{dc}) + k_{i,b} \int (u_{dc,ref} – u_{dc}) dt \right]
$$
where $u_{pv,ref}$ is the PV array voltage reference, $u_{max}$ is the pre-fault MPP voltage, and $k_{p,b}$ and $k_{i,b}$ are controller gains. This increases the PV array voltage, moving the operating point to the right of the MPP on the P-V curve, where power sensitivity to voltage is higher. This allows quick reduction of $P_{pv}$ to match $P_{out}$, stabilizing $u_{dc}$. The Boost control transition is summarized in Table 1.
| Operation Mode | Control Objective | Key Equation |
|---|---|---|
| Normal | MPPT | $u_{pv,ref} = \text{P&O Algorithm}$ |
| LVRT (Fault) | DC-Link Voltage Regulation | $u_{pv,ref} = u_{max} + \text{PI}(u_{dc,ref} – u_{dc})$ |
Second, the inverter current references are modified to provide reactive power support during faults. Let $k = u_f / u_N$ define the voltage dip depth, where $u_f$ is the fault voltage and $u_N$ is the nominal voltage. The current references are calculated as:
$$
i_{d,fault}^* = \begin{cases}
i_{d0}^* & \text{for } t_0 \leq t < t_1 \\
i_{max} – i_R (t – t_1) & \text{for } t_1 \leq t < t_2, \text{ if } i_{d,fault}^* \geq k i_{max} \\
k i_{max} & \text{otherwise}
\end{cases}
$$
$$
i_{q,fault}^* = \sqrt{i_{max}^2 – (i_{d,fault}^*)^2}
$$
where $i_{d0}^*$ is the pre-fault d-axis current, $i_{max}$ is the maximum allowable current (e.g., 1.1 times rated current), $i_R$ is the decay rate, and $t_0$, $t_1$, $t_2$ are time instants defining the transition. This ensures that the solar inverter remains within current limits while supplying reactive current to support grid voltage recovery. The overall current control structure during LVRT is shown in Figure 1, with the inverter output voltages given by:
$$
u_d = \left( k_{ip} + \frac{k_{ii}}{s} \right) (i_d^* – i_d) – \omega L i_q + e_d
$$
$$
u_q = \left( k_{ip} + \frac{k_{ii}}{s} \right) (i_q^* – i_q) + \omega L i_d + e_q
$$
where $k_{ip}$ and $k_{ii}$ are current controller gains, $\omega$ is grid frequency, $L$ is filter inductance, and $e_d$, $e_q$ are grid voltage components.
To validate the proposed LVRT strategy for the two-stage solar inverter, we developed a simulation model in PSCAD/EMTDC for a 100 kW system. The parameters are listed in Table 2. The solar inverter was tested under a symmetric voltage dip to 0.2 per unit (p.u.) for 0.1 seconds. The simulation results demonstrate the effectiveness of the control strategy in maintaining stability and providing reactive support.
| Parameter | Value |
|---|---|
| Rated Power | 100 kW |
| DC-Link Voltage Reference | 800 V |
| DC-Link Capacitance | 6000 µF |
| Grid Voltage (Nominal) | 220 V (rms) |
| Grid Frequency | 50 Hz |
| Inverter Rated Current | 200 A |
| LCL Filter Inductances | L1 = 0.4 mH, L2 = 0.05 mH |
| LCL Filter Capacitance | 10 µF |
| Damping Resistance | 1 Ω |
During the voltage dip, the Boost control successfully reduced the PV array output power from 100 kW to approximately 20 kW within 6 ms, preventing DC-link overvoltage. The DC-link voltage peaked at 861.5 V, well within safe limits. The inverter currents remained below the maximum threshold, with no overcurrent events. Additionally, the solar inverter provided reactive current according to the dip depth, enhancing grid voltage support. Compared to conventional methods, this strategy offers smoother current transitions and better power quality. The dynamic response is summarized by the power balance equation:
$$
P_{pv} = P_{out} + P_{loss} + \frac{d}{dt} \left( \frac{1}{2} C u_{dc}^2 \right)
$$
where $P_{loss}$ accounts for converter losses. During faults, the fast power adjustment minimizes the residual term, ensuring stability.
The integration of such advanced solar inverters in real-world applications, like hybrid systems with battery storage, underscores the importance of robust LVRT capabilities. The image above illustrates a commercial hybrid solar inverter setup, highlighting the hardware context where our control strategy can be implemented. In practice, solar inverters must adapt to varying grid conditions, and our approach offers a software-based solution that enhances reliability without costly hardware additions.
Further analysis of the solar inverter performance under asymmetric faults reveals the versatility of the strategy. For unbalanced voltage sags, the control can be extended using positive and negative sequence decomposition. The current references can be modified as:
$$
i_d^* = i_{d,p}^* + i_{d,n}^*, \quad i_q^* = i_{q,p}^* + i_{q,n}^*
$$
where subscripts $p$ and $n$ denote positive and negative sequences. This allows the solar inverter to mitigate negative sequence currents, reducing grid imbalance. The Boost control remains unchanged, as it responds to the overall DC-link voltage. Simulation studies for asymmetric dips show similar effectiveness in stabilizing the solar inverter, with reactive support tailored to sequence components.
The coordination between active and reactive power in the solar inverter during LVRT is crucial for grid code compliance. Standards often require reactive current injection proportional to voltage deviation. Our strategy aligns with this by setting $i_{q,fault}^*$ based on $k$. For deep sags (e.g., $k < 0.5$), the solar inverter prioritizes reactive support, with $i_d^*$ reduced to near zero. This maximizes the grid voltage boost while preventing overcurrent. The control law can be generalized as:
$$
i_d^* = \min \left( i_{max} \sqrt{1 – \alpha^2}, k i_{max} \right), \quad i_q^* = \alpha i_{max}
$$
where $\alpha$ is a factor dependent on grid requirements. This flexibility makes the solar inverter adaptable to various regulatory frameworks.
In terms of implementation, the proposed LVRT strategy for the solar inverter requires minimal computational overhead. The Boost control switch is triggered by voltage monitoring, and the current reference generator uses simple arithmetic operations. This suits digital signal processors (DSPs) commonly used in solar inverter controls. Additionally, the strategy can be combined with fault detection algorithms to improve response time. For instance, using phase-locked loops (PLLs) for rapid voltage dip detection ensures the solar inverter reacts within milliseconds, enhancing LVRT performance.
To quantify the benefits, we compare our solar inverter strategy with conventional LVRT approaches that use hardware like chopper circuits or energy storage. Table 3 summarizes the key advantages. Our method reduces cost and complexity while maintaining comparable performance. This is particularly relevant for large-scale solar farms where multiple solar inverters are deployed, and cost-effectiveness is critical.
| Method | Hardware Requirement | Cost | Performance |
|---|---|---|---|
| Proposed Control | None (software-only) | Low | High (stable voltage, reactive support) |
| Chopper Circuit | Resistor and switch | Medium | Medium (voltage stabilization, no reactive support) |
| Energy Storage | Batteries or capacitors | High | High (comprehensive power management) |
The solar inverter’s role in grid stability extends beyond LVRT. With the proliferation of distributed generation, solar inverters can provide ancillary services like frequency regulation and harmonic compensation. Our control strategy lays the foundation for such multifunctional capabilities. By integrating advanced algorithms, the solar inverter can become a grid-supporting asset, contributing to overall power quality. For example, during normal operation, the solar inverter can operate in var control mode, adjusting $i_q^*$ to regulate local voltage. This adaptability underscores the importance of sophisticated control in modern solar inverters.
In conclusion, the proposed LVRT control strategy for two-stage solar inverters effectively addresses voltage sag challenges through coordinated power management. By modifying the Boost control to regulate DC-link voltage and adjusting inverter current references for reactive support, the solar inverter achieves robust fault ride-through without additional hardware. Simulation results validate the strategy’s ability to prevent overvoltage and overcurrent while enhancing grid voltage recovery. This approach aligns with grid codes and promotes the integration of solar power into resilient power systems. Future work could explore integration with energy storage for extended fault tolerance and real-time optimization using machine learning. As solar inverter technology evolves, such control innovations will be pivotal in achieving a sustainable energy future.
From a broader perspective, the development of intelligent solar inverters is essential for smart grids. The ability to ride through low-voltage events and provide reactive support makes solar inverters key players in grid stability. Our strategy demonstrates that through careful control design, solar inverters can exceed basic functionality, offering value-added services. This not only benefits grid operators but also enhances the economic viability of solar installations. As research progresses, we anticipate further advancements in solar inverter capabilities, driven by control algorithms like the one presented here.
Throughout this article, we have emphasized the significance of solar inverters in modern power networks. By repeatedly focusing on the solar inverter’s functions and improvements, we highlight its central role. The proposed LVRT strategy is a step toward more reliable and grid-friendly solar inverters, paving the way for higher penetration of renewable energy. As the energy transition accelerates, such contributions will be crucial in building a resilient and sustainable power infrastructure.