In the realm of renewable energy integration, solar photovoltaic systems have garnered significant attention due to their environmental friendliness and low maintenance costs. As a critical interface between photovoltaic arrays and the utility grid, the utility interactive inverter plays a pivotal role. Traditional transformerless utility interactive inverters, while offering high efficiency and compact size, are plagued by the persistent issue of leakage current. This leakage current can lead to electromagnetic interference, additional power losses, and safety hazards. In this paper, we propose a novel single-phase transformerless utility interactive inverter designed to effectively suppress leakage current while eliminating circulating currents between its constituent units. The proposed utility interactive inverter employs a dual-unit structure, where the main unit operates with unipolar modulation and the auxiliary unit with bipolar modulation, sharing a common DC bus. This configuration not only enhances performance but also simplifies control, making it a promising candidate for modern grid-connected applications.
The topology of the proposed utility interactive inverter is illustrated conceptually below. It consists of two primary units: a main unit and an auxiliary unit. The main unit is based on the HERIC (Highly Efficient and Reliable Inverter Concept) topology, which incorporates six power switches (V1 to V6) with anti-parallel diodes (VD1 to VD6) and two filter inductors (L1 and L2). This unit is responsible for delivering the bulk of the electrical energy to the grid. The auxiliary unit adopts a standard H-bridge topology, comprising four power switches (V7 to V10) with anti-parallel diodes (VD7 to VD10) and two filter inductors (L3 and L4), where L3 equals L4. Both units are connected in parallel to a common DC bus, and their AC outputs are coupled to the grid. The grounding point of the grid is denoted as G. This arrangement allows the auxiliary unit to provide a low-impedance path for common-mode voltage components, thereby mitigating leakage currents. The utility interactive inverter leverages this dual-frequency operation—with the main unit switching at a lower frequency to reduce losses and the auxiliary unit at a higher frequency to handle ripple—to achieve optimal performance.
To understand the operational principles of this utility interactive inverter, we first analyze the circulating currents that can arise between the main and auxiliary units due to their shared DC bus. Circulating currents, if unchecked, can degrade efficiency and cause instability. The proposed modulation strategy ensures that these currents are eliminated. Specifically, the main unit utilizes unipolar pulse-width modulation (UP-PWM), while the auxiliary unit employs bipolar pulse-width modulation (BP-PWM). By examining all possible switching states under this combined modulation, we identify eight distinct states, as summarized in Table 1. Each state defines the on/off status of the key switches in both units, influencing the current paths and voltages.
| State | V1, V4 | V2, V3 | V5 | V6 | V7, V10 | V8, V9 | Description |
|---|---|---|---|---|---|---|---|
| 1 | On | Off | Off | Off | Off | Off | Power delivery phase with positive grid voltage |
| 2 | Off | On | Off | Off | Off | Off | Power delivery phase with negative grid voltage |
| 3 | On | Off | Off | Off | On | Off | Combined operation with auxiliary unit active |
| 4 | Off | On | Off | Off | Off | On | Combined operation with auxiliary unit complementary |
| 5 | Off | Off | On | Off | Off | Off | Freewheeling state through main unit diodes |
| 6 | Off | Off | Off | On | Off | Off | Freewheeling state alternative path |
| 7 | Off | Off | On | Off | On | Off | Freewheeling with auxiliary unit support |
| 8 | Off | Off | Off | On | Off | On | Freewheeling with auxiliary unit complementary |
From these states, we derive the differential expressions for the main unit filter inductor currents, i.e., $$i_{L1}$$ and $$i_{L2}$$. The analysis shows that under all switching states, the circulating current between the units is nullified because the current derivatives satisfy $$di_{L1}/dt = di_{L2}/dt$$. This is mathematically represented in Table 2, which summarizes the current derivatives for each state. The equivalence of these derivatives across states confirms that no net circulating current flows, ensuring efficient operation of the utility interactive inverter.
| Switching State | $$di_{L1}/dt$$ | $$di_{L2}/dt$$ |
|---|---|---|
| 1 | $$(U_{dc} – U_g) / (L_1 + L_2)$$ | $$(U_{dc} – U_g) / (L_1 + L_2)$$ |
| 2 | $$(-U_{dc} – U_g) / (L_1 + L_2)$$ | $$(-U_{dc} – U_g) / (L_1 + L_2)$$ |
| 3 | $$(U_{dc} L_{13}/(L_{13}+L_{24}) – U_g L_{14}/(L_{14}+L_{23})) / L_1$$ | $$(U_{dc} L_{13}/(L_{13}+L_{24}) – U_g L_{14}/(L_{14}+L_{23})) / L_1$$ |
| 4 | $$(-U_{dc} L_{24}/(L_{13}+L_{24}) – U_g L_{23}/(L_{14}+L_{23})) / L_1$$ | $$(-U_{dc} L_{24}/(L_{13}+L_{24}) – U_g L_{23}/(L_{14}+L_{23})) / L_1$$ |
| 5 | $$(U_{dc} + U_g) L_{11}/(L_{11}+L_{22}) / L_1$$ | $$(U_{dc} + U_g) L_{11}/(L_{11}+L_{22}) / L_1$$ |
| 6 | $$-U_g L_{12}/(L_{12}+L_{34}) / L_1$$ | $$-U_g L_{12}/(L_{12}+L_{34}) / L_1$$ |
| 7 | $$-U_g L_{12}/(L_{12}+L_{34}) / L_1$$ | $$-U_g L_{12}/(L_{12}+L_{34}) / L_1$$ |
| 8 | $$-U_g L_{12}/(L_{12}+L_{34}) / L_1$$ | $$-U_g L_{12}/(L_{12}+L_{34}) / L_1$$ |
Next, we delve into the leakage current analysis, which is crucial for transformerless utility interactive inverters. Leakage current primarily arises due to common-mode voltage variations across the parasitic capacitance between the photovoltaic panels and ground. In the proposed utility interactive inverter, the auxiliary unit, operating with BP-PWM, provides a low-impedance shunt path for harmonic components of the common-mode voltage. This effectively diverts leakage currents away from the ground path. To quantify this, we model the leakage current equivalent circuit, as shown in Figure 2 (conceptually described). The common-mode voltage $$U_{CM}$$ is defined as the average of the voltages from the DC bus positive (P) and negative (N) terminals to ground (G):
$$ U_{CM} = \frac{U_{CP} + U_{CN}}{2} $$
where $$U_{CP}$$ and $$U_{CN}$$ are the voltages from P and N to G, respectively. This voltage can be decomposed into a DC component $$U_{dc}/2$$ and an AC component $$U_{GN}$$. The fundamental-frequency leakage current amplitude is given by:
$$ I_{LEAK\_F} = \frac{0.5 U_{GN}}{2\pi f_g C_{leak}} $$
Here, $$f_g$$ is the grid frequency (50 Hz), and $$C_{leak}$$ is the parasitic capacitance (e.g., 100 nF). For harmonic leakage currents, we consider the high-frequency components induced by switching actions. The transfer function from the high-frequency common-mode voltage source $$U_{hc}$$ to the leakage current $$i_{LEAK}$$ is derived as:
$$ G_{leak-1}(s) = \frac{G_{la}(s)}{G_{pn}(s) G_{la}(s) + G_{lp}(s) [G_{la}(s) + G_{lp}(s)]} $$
with $$G_{pn}(s) = C_{leak} s$$, $$G_{la}(s) = L_{ea} s + R_{la}$$, and $$G_{lp}(s) = L_{ep} s + R_{lp}$$, where $$L_{ea}$$, $$L_{ep}$$, $$R_{la}$$, and $$R_{lp}$$ are equivalent inductances and resistances of the units. In contrast, a conventional HERIC utility interactive inverter has a simpler transfer function:
$$ G_{LEAK-2}(s) = \frac{1}{G_{pn}(s) + G_{lp}(s)} $$
By comparing the Bode plots of $$G_{leak-1}(s)$$ and $$G_{LEAK-2}(s)$$ for typical parameters (e.g., $$U_{dc} = 200 V$$, $$L_1 = L_2 = 1.5 mH$$, $$L_3 = L_4 = 0.4 mH$$, switching frequencies of 2 kHz for main unit and 60 kHz for auxiliary unit), we observe that the proposed utility interactive inverter achieves significantly lower gain at frequencies below 10 kHz, indicating superior leakage current suppression. For instance, at frequencies under 4 kHz, the gain is about 13 dB lower than that of the conventional HERIC inverter, highlighting the efficacy of our design.
The control strategy for this utility interactive inverter is designed to be simple and cost-effective, utilizing a single control chip for both units. The system comprises two current loops: one for the main unit and one for the auxiliary unit. The main unit current loop ensures that the output current $$i_g$$ tracks the grid current reference $$i_g^*$$, which is generated based on a phase-locked loop (PLL) and a power reference. The PLL, implemented via a second-order generalized integrator, extracts the grid phase $$\theta_g$$. The reference current is:
$$ i_g^* = I_M^* \cos(\theta_g) $$
where $$I_M^*$$ is the amplitude reference. The main current regulator ACRP is a proportional-resonant (PR) controller, outputting a voltage $$U_{ACRP}$$. The total control voltage for the main unit is the sum of $$U_{ACRP}$$ and a feedforward compensation term $$U_{COMP}$$ derived from the grid voltage.
The auxiliary unit current loop aims to maintain its average output current at zero, ensuring that power flows only through the main unit. Its regulator ACRA is a multi-resonant PR controller. A key feature is the feedforward compensation $$U_{COMA}$$, which cancels the switching ripple from the main unit. This compensation is calculated as:
$$ U_{COMA} = \frac{L_3 + L_4}{U_{dc} (L_1 + L_2)} \left[ U_g + (L_1 + L_2) \frac{di_g}{dt} – U_{PE} \right] + U_{offset} $$
Here, $$U_{PE}$$ is the estimated voltage of the main unit’s switching states, easily obtained from the shared control signals without additional sensors. This integrated control approach reduces hardware complexity and cost, making the utility interactive inverter practical for widespread deployment.
To validate the proposed utility interactive inverter, we constructed a 1 kW experimental prototype with the following parameters: main unit filter inductors $$L_1 = L_2 = 1.5 mH$$, main switching frequency $$f_P = 2 kHz$$, auxiliary unit filter inductors $$L_3 = L_4 = 0.4 mH$$, auxiliary switching frequency $$f_A = 60 kHz$$, grid voltage $$U_g = 150 V$$, grid frequency $$f_g = 50 Hz$$, and parasitic capacitance $$C_{leak} = 100 nF$$. The power devices were selected appropriately: AUIRGP54070D0 for the main unit and IMW65R027M1H for the auxiliary unit. The experimental results confirm the theoretical analyses.
First, the leakage current waveforms were measured. The common-mode voltage $$U_{CN}$$ exhibited an amplitude of approximately 78 V, and the fundamental leakage current $$I_{LEAK}$$ had an amplitude of about 2.45 mA, aligning with the calculation from Equation (2). The total harmonic distortion (THD) of the leakage current was found to be 0.74%, indicating that high-frequency harmonics were effectively suppressed. This low THD underscores the utility interactive inverter’s ability to minimize leakage currents, enhancing safety and compliance with grid standards.
Second, the filter inductor currents were examined. The currents $$i_{L1}$$ and $$i_{L2}$$ of the main unit showed identical shapes without any divergence, confirming the absence of circulating currents. Similarly, the auxiliary unit currents $$i_{L3}$$ and $$i_{L4}$$ remained balanced. These observations validate the modulation strategy and control design, proving that the utility interactive inverter operates efficiently without internal power circulation.
Furthermore, we evaluated the overall efficiency and power quality. The utility interactive inverter achieved a peak efficiency of over 98% at rated power, attributed to the low switching losses in the main unit and effective ripple cancellation by the auxiliary unit. The grid current THD was below 3%, meeting typical grid codes for harmonic distortion. The dual-frequency operation allowed the main unit to handle bulk power at low frequency, while the auxiliary unit managed high-frequency components, optimizing the trade-off between switching losses and filtering performance.
In conclusion, we have presented a novel single-phase transformerless utility interactive inverter that addresses key challenges in grid-connected photovoltaic systems. By combining a HERIC-based main unit with unipolar modulation and an H-bridge auxiliary unit with bipolar modulation, this utility interactive inverter eliminates circulating currents and suppresses leakage currents effectively. The analytical models, supported by experimental results from a 1 kW prototype, demonstrate its superiority over conventional designs in terms of leakage current reduction and operational stability. The control strategy, leveraging a single chip and feedforward compensation, simplifies implementation and reduces costs. Future work may explore scalability to higher power levels, integration with energy storage, and adaptation to three-phase systems. This utility interactive inverter represents a significant step forward in the development of efficient, safe, and cost-effective transformerless inverters for renewable energy integration.
To further illustrate the operational principles, we can summarize the key mathematical expressions and parameters in additional tables. For instance, Table 3 lists the system parameters used in the analysis and experiment, while Table 4 compares the performance metrics with a conventional HERIC utility interactive inverter. These tables provide a concise reference for understanding the design trade-offs and advantages.
| Parameter | Symbol | Value | Unit |
|---|---|---|---|
| DC Bus Voltage | $$U_{dc}$$ | 200 | V |
| Grid Voltage | $$U_g$$ | 150 | V |
| Grid Frequency | $$f_g$$ | 50 | Hz |
| Main Unit Inductance | $$L_1, L_2$$ | 1.5 | mH |
| Auxiliary Unit Inductance | $$L_3, L_4$$ | 0.4 | mH |
| Main Switching Frequency | $$f_P$$ | 2 | kHz |
| Auxiliary Switching Frequency | $$f_A$$ | 60 | kHz |
| Parasitic Capacitance | $$C_{leak}$$ | 100 | nF |
| Rated Power | $$P_N$$ | 1 | kW |
| Metric | Proposed Utility Interactive Inverter | Conventional HERIC Utility Interactive Inverter |
|---|---|---|
| Leakage Current THD | 0.74% | Typically >5% |
| Circulating Current | Eliminated | Present, requires damping |
| Efficiency at Rated Power | 98.2% | 97.5% |
| Control Complexity | Moderate (single chip) | Low to Moderate |
| Cost Implication | Reduced due to shared bus | Higher if isolation needed |
| Grid Current THD | <3% | ~3-5% |
The proposed utility interactive inverter also offers inherent advantages in terms of reliability and maintenance. By avoiding transformers, it reduces weight and volume, facilitating installation in space-constrained environments. The dual-unit architecture provides redundancy in some aspects; for example, if the auxiliary unit fails, the main unit can still operate at reduced performance, ensuring continuous power delivery. This robustness is essential for utility interactive inverters in critical applications such as microgrids or off-grid systems.
From a theoretical perspective, the utility interactive inverter’s operation can be further analyzed using state-space models or frequency-domain techniques. For instance, the dynamics of the main unit current loop can be described by a second-order system with the transfer function:
$$ H_{main}(s) = \frac{K_p s + K_r}{s^2 + 2\zeta\omega_n s + \omega_n^2} $$
where $$K_p$$ and $$K_r$$ are proportional and resonant gains, $$\zeta$$ is the damping ratio, and $$\omega_n$$ is the natural frequency. Similarly, the auxiliary unit’s impact on ripple reduction can be quantified by the attenuation factor:
$$ A_{ripple} = 20 \log_{10} \left( \frac{V_{ripple\_without}}{V_{ripple\_with}} \right) $$
which, in our experiments, exceeded 20 dB for switching harmonics above 1 kHz. These analytical tools help optimize the design for specific grid conditions and load profiles.
In summary, this utility interactive inverter represents a holistic solution to the challenges faced by transformerless grid-connected systems. Its innovative topology, combined with advanced control, sets a new benchmark for performance and cost-effectiveness. As renewable energy penetration increases, such utility interactive inverters will play a crucial role in ensuring grid stability and power quality. We encourage further research into adaptive control algorithms and wide-bandgap semiconductor integration to push the boundaries of efficiency and power density. The journey toward a sustainable energy future relies on continuous innovation in power electronics, and this utility interactive inverter is a testament to that progress.