In the realm of photovoltaic power generation, transformerless grid-connected systems have gained significant traction due to their high efficiency, compact size, and cost-effectiveness. These solar inverters are pivotal in converting DC power from solar panels into AC power for grid integration. However, a critical challenge associated with non-isolated solar inverters is the leakage current that arises from the parasitic capacitance between the photovoltaic panels and ground. This leakage current can compromise grid quality, stability, and safety, prompting stringent regulations such as the German standard VDE-0126-1-1, which limits peak leakage current to less than 300 mA and RMS values to under 30 mA. Addressing this issue is essential for the widespread adoption of transformerless solar inverters in modern energy systems.
Leakage current in solar inverters primarily stems from common-mode voltage fluctuations across the parasitic capacitances of PV arrays. In three-phase three-level solar inverters, the common-mode voltage is derived from the inverter output voltages relative to the DC-link midpoint. Mathematically, the common-mode voltage \( u_{cm} \) for a three-phase system is defined as:
$$ u_{cm} = \frac{1}{3}(u_{ao} + u_{bo} + u_{co}) $$
where \( u_{ao} \), \( u_{bo} \), and \( u_{co} \) are the phase-to-midpoint voltages. These voltages contain high-frequency components due to switching actions, which excite the parasitic capacitance \( C_{PV} \), leading to significant leakage currents. The common-mode equivalent circuit can be modeled as shown in the analysis, where the leakage current \( i_{lg} \) flows through the ground resistance \( R_g \) and the parasitic capacitance. The equation governing this circuit is:
$$ u_{cm} = \frac{L}{3} \frac{di_{lg}}{dt} + i_{lg} R_g + \frac{1}{2C_{PV}} \int i_{lg} \, dt $$
This model highlights that the high-frequency harmonics in \( u_{cm} \) are the primary contributors to leakage current, necessitating effective filtering strategies in solar inverters.
Various methods have been proposed to suppress leakage current in transformerless solar inverters, categorized into four approaches: employing novel inverter topologies, designing new modulation strategies, utilizing three-phase four-wire inverter topologies, and constructing common-mode filter loops. Among these, the common-mode filter loop method is particularly attractive for solar inverters as it does not require modifications to the main circuit or modulation schemes, thereby maintaining simplicity and cost-efficiency. This article focuses on a neutral-point capacitance-based technique for three-phase three-level solar inverters, which introduces a filter capacitor between the AC and DC neutral points to form an LC filter that attenuates high-frequency common-mode components without affecting differential-mode operations.
The proposed leakage current suppression method for solar inverters involves connecting the AC-side neutral point (n) and the DC-side neutral point (o) through a capacitor \( C_f \) and a damping resistor \( R_d \). This configuration creates a common-mode LC filter that effectively suppresses high-frequency voltages across the parasitic capacitance. The common-mode equivalent circuit with this filter is derived as follows. The voltage equations for the modified topology are:
$$ u_{cm} = \frac{L}{3} \frac{d(i_a + i_b + i_c)}{dt} + i_{lg} R_g + \frac{1}{2C_{PV}} \int i_{lg} \, dt $$
$$ u_{cm} = \frac{L}{3} \frac{d(i_a + i_b + i_c)}{dt} + i_f R_d + \frac{1}{C_f} \int i_f \, dt $$
$$ i_a + i_b + i_c = i_f + i_{lg} $$
where \( i_f \) is the current through the filter branch. Solving these equations yields the common-mode transfer function \( G(s) \) between \( u_{cm} \) and the voltage \( u_{no} \) (between points n and o):
$$ G(s) = \frac{u_{no}}{u_{cm}} = \frac{\left( \frac{1}{s C_f} + R_d \right) \parallel \left( \frac{1}{2s C_{PV}} + R_g \right)}{\left( \frac{1}{s C_f} + R_d \right) \parallel \left( \frac{1}{2s C_{PV}} + R_g \right) + \frac{sL}{3}} $$
Assuming \( C_{PV} \) is much smaller than \( C_f \), the transfer function simplifies to:
$$ G(s) = \frac{1 + C_f R_d s}{1 + C_f R_d s + \frac{L C_f}{3} s^2} $$
This represents a second-order low-pass filter with resonant frequency \( \omega_r \) and damping coefficient \( \xi \):
$$ \omega_r = \sqrt{\frac{3}{L C_f}}, \quad \xi = \frac{R_d C_f}{2} \omega_r $$
The cutoff frequency \( \omega_c \) is given by \( \omega_c = \sqrt{2} \omega_r \). Proper design of \( C_f \) and \( R_d \) ensures attenuation of high-frequency common-mode voltages, thereby reducing leakage current in solar inverters. The differential-mode circuit remains unaffected, preserving the grid-tied performance of the solar inverter.
Parameter design for the common-mode filter in solar inverters is critical for optimal leakage current suppression. The design process involves selecting the cutoff frequency relative to the switching frequency, choosing an appropriate damping coefficient to minimize resonance peaks, and calculating component values. For instance, setting the cutoff frequency to one-fourth of the switching frequency \( \omega_s \) provides effective high-frequency attenuation. A damping coefficient of 0.170 is often selected to balance resonance suppression and filter performance. The filter parameters can be summarized in the following table, which outlines typical values for a three-phase three-level solar inverter system.
| Parameter | Symbol | Typical Value | Unit |
|---|---|---|---|
| DC Input Voltage | \( U_{dc} \) | 540 | V |
| Switching Frequency | \( f_s \) | 10 | kHz |
| Output Filter Inductance | \( L \) | 2 | mH |
| Parasitic Capacitance | \( C_{PV} \) | 150 | nF |
| Filter Capacitance | \( C_f \) | 12 | μF |
| Damping Resistance | \( R_d \) | 1.9 | Ω |
| Damping Coefficient | \( \xi \) | 0.170 | – |
| Ground Resistance | \( R_g \) | 1 | Ω |
Simulation studies are essential for validating the effectiveness of leakage current suppression in solar inverters. Using software like PSIM, a three-phase three-level solar inverter model was built with the proposed neutral-point capacitance filter. The simulation parameters align with the table above. Key results include the voltage \( u_{no} \) and leakage current waveforms. In the traditional topology without the filter, \( u_{no} \) exhibits high-frequency harmonics, leading to a peak leakage current of 1.7 A and an RMS value of 485.4 mA, exceeding regulatory limits. With the proposed filter, \( u_{no} \) shows significant attenuation of high-frequency components, reducing the peak leakage current to 97.2 mA and the RMS value to 25.1 mA, compliant with VDE-0126-1-1. The performance metrics can be expressed through formulas. For example, the leakage current reduction ratio \( \eta \) is defined as:
$$ \eta = \frac{I_{lg,\text{traditional}} – I_{lg,\text{proposed}}}{I_{lg,\text{traditional}}} \times 100\% $$
where \( I_{lg} \) represents RMS leakage current. In this case, \( \eta \approx 94.8\% \), demonstrating high efficacy. Further analysis using Fast Fourier Transform (FFT) on \( u_{no} \) reveals that the traditional topology has harmonic amplitudes of 27 V at the switching frequency and its multiples, while the proposed topology suppresses these to negligible levels, with dominant low-frequency harmonics at 3rd, 9th, etc., orders. The impact on grid current quality is minimal; the total harmonic distortion (THD) of grid currents increases slightly from about 1.6% to 1.8%, still meeting grid standards. This underscores the robustness of solar inverters incorporating this filter.
Experimental verification reinforces the simulation findings. A prototype solar inverter was developed using a DSP and FPGA for control, with IGBTs for switching. The parameters match the simulation setup. Experimental waveforms for \( u_{no} \) and leakage current show that the traditional topology yields a peak leakage current of 1.9 A and RMS of 510 mA, whereas the proposed topology reduces these to 107 mA peak and 27 mA RMS. The voltage \( u_{no} \) in the proposed topology is free of high-frequency noise, confirming effective filtering. Grid current THD measures 1.93% with a power factor of 0.99, indicating satisfactory grid integration. Dynamic tests also demonstrate rapid suppression of leakage current during transients. These results validate the practicality of the neutral-point capacitance method for real-world solar inverters.
The advantages of this leakage current suppression technique for solar inverters are multifold. It does not require changes to the modulation strategy or main circuit topology, preserving the simplicity and efficiency of transformerless solar inverters. The filter components are passive and inexpensive, adding minimal cost. Moreover, the design is scalable to various solar inverter ratings. Compared to other methods, such as adding extra switches or complex modulation, this approach offers a balanced trade-off between performance and complexity. However, considerations include the need for precise parameter design to avoid resonance and ensure stability. The damping resistor \( R_d \) plays a crucial role in mitigating resonance peaks, as reflected in the transfer function analysis. Future work could explore adaptive tuning of \( C_f \) and \( R_d \) for varying operating conditions in solar inverters.
In conclusion, the neutral-point capacitance-based method provides an effective solution for leakage current suppression in transformerless three-phase three-level solar inverters. By forming a common-mode LC filter between the AC and DC neutral points, it attenuates high-frequency common-mode voltages that drive leakage currents, achieving compliance with safety standards like VDE-0126-1-1. The design methodology, involving cutoff frequency selection and damping coefficient optimization, ensures robust performance without affecting differential-mode operations. Simulation and experimental results confirm significant leakage current reduction while maintaining grid current quality. This technique enhances the reliability and acceptability of solar inverters in grid-connected photovoltaic systems, contributing to the advancement of renewable energy technologies. As solar inverters evolve, integrating such suppression methods will be key to achieving higher efficiency and safety in sustainable power networks.
The broader implications for solar inverters include improved grid stability and reduced electromagnetic interference. By addressing leakage current issues, transformerless solar inverters can be deployed more widely, lowering installation costs and increasing energy yield. Ongoing research may focus on integrating this filter with advanced topologies like multilevel solar inverters or hybrid systems combining storage. Ultimately, innovations in leakage current suppression will drive the growth of solar energy, making solar inverters more integral to global energy transitions. The mathematical models and design principles discussed herein serve as a foundation for further optimization in solar inverter applications.