The widespread integration of renewable energy sources, particularly utility-scale photovoltaic (PV) plants, has fundamentally altered the characteristics of electrical power systems. The heart of a grid-connected PV system is the solar inverter, a power electronic interface that converts the direct current (DC) from the PV array into alternating current (AC) suitable for the grid. While this enables the utilization of solar energy, the transient behavior of solar inverters during grid faults differs significantly from that of traditional synchronous generators. One critical aspect demanding in-depth analysis is the harmonic content of the fault current supplied by these inverters, as it can have profound implications for the reliability and security of grid protection schemes, such as transformer differential protection.
This article delves into the phenomenon of third-harmonic current generation in solar inverter output during unbalanced grid faults. We will derive the mechanism, quantify its magnitude under various control strategies, and critically assess its potential to cause maloperation of protection relays, proposing effective countermeasures.
1. Asymmetric Faults and Negative-Sequence Excitation
When an asymmetric fault (e.g., single-line-to-ground, line-to-line) occurs on the grid, negative-sequence components appear in the voltage and, consequently, in the current at the point of common coupling (PCC). For a solar inverter operating with a standard vector control scheme in a synchronous (dq) reference frame, this negative-sequence voltage manifests as a 100 Hz (2ω₁, where ω₁ is the fundamental grid angular frequency) oscillatory component. The impact of this oscillation propagates through the inverter’s control loops, ultimately affecting the DC-link voltage.
The instantaneous active power P injected by the inverter under unbalanced conditions contains a constant term and a double-frequency oscillatory component:
$$ P = P_0 + P_{c2} \cos(2\omega_1 t) + P_{s2} \sin(2\omega_1 t) $$
Ignoring losses, the power balance between the DC input from the PV array (Parray) and the AC output determines the DC-link voltage (udc) dynamics across the capacitance C:
$$ C \frac{du_{dc}}{dt} = \frac{P_{array} – P}{u_{dc}} $$
Assuming the average output power P0 matches Parray, the solution to this equation reveals that the double-frequency power oscillation Pc2, Ps2 forces a corresponding 100 Hz ripple on the DC-link voltage:
$$ u_{dc} = \sqrt{ \frac{1}{C} \left[ -\frac{P_{c2} \sin(2\omega_1 t)}{\omega_1} + \frac{P_{s2}\cos(2\omega_1 t)}{\omega_1} \right] + u^2_{dc0} } \approx U_{dc0} + U_{dc\_2} \cos(2\omega_1 t + \phi_{dc\_2}) $$
where Udc0 is the pre-fault DC voltage, and Udc_2 is the amplitude of the induced second harmonic voltage. This DC-link voltage perturbation is the primary source of the subsequent current harmonics.
2. Generation Mechanism of Third Harmonic Current
The standard control structure of a grid-following solar inverter involves an outer DC voltage or power control loop that sets the reference for the d-axis current (id*), responsible for active power flow. The q-axis current reference (iq*) is set by reactive power or voltage support requirements, such as Low Voltage Ride-Through (LVRT) mandates. The DC-link voltage ripple Udc_2 is fed into the outer voltage controller, which typically uses a PI regulator (with gains kup, kui). The output of this controller, id*, will therefore also contain a 100 Hz component. This oscillatory reference is then processed by the inner current control loop (with PI gains kip+, kii+ for positive sequence).
The transfer function from the DC-link ripple to the resulting d-axis voltage reference command can be analyzed in the frequency domain. The 100 Hz component propagating through these cascaded PI controllers results in a voltage reference signal containing not only the original 100 Hz but also other frequency components due to the derivative action of the controllers. A detailed derivation shows that the resulting three-phase voltage reference in the stationary (abc) frame contains terms at the fundamental frequency and, critically, at the third harmonic frequency (150 Hz or 3ω₁).
$$ u^{*}_{a\_2} = M_{dc\_2} \left[ \cos(\omega_1 t + \phi’_{dc\_2}) + \cos(3\omega_1 t + \phi’_{dc\_2}) \right] $$
where Mdc_2 is a magnitude factor dependent on Udc_2, the PI controller gains, and ω₁. This third-harmonic voltage reference, when imposed by the inverter’s pulse-width modulation (PWM), drives a third-harmonic current through the grid impedance. The amplitude of this harmonic current is directly proportional to the amplitude of the DC-link voltage ripple Udc_2 and the controller gains.
$$ I_{3h} \propto U_{dc\_2} \cdot \sqrt{ \frac{k_{up}^2\left((k_{ip}^+)^2 + \frac{k_{ui}^2 (k_{ii}^+)^2}{16\omega_1^4}\right) }{4\omega_1^2} + \frac{k_{ui}^2 (k_{ip}^+)^2}{4\omega_1^2} } $$
This mechanism is inherent to the control structure under unbalanced voltage conditions and persists as long as the fault remains uncleared.
3. Influence of Control Strategies and Current Limiting
The magnitude of the third-harmonic current is not constant; it is heavily influenced by the inverter’s control objectives during faults and its physical current-limiting capability. Modern solar inverters employ various negative-sequence control strategies to manage power oscillations and meet grid codes. These strategies directly affect Pc2 and Ps2, thereby changing Udc_2 and the resultant third harmonic.
| Control Objective | Primary Goal | Effect on Active Power Ripple (Pc2, Ps2) | Relative Third Harmonic Magnitude |
|---|---|---|---|
| I: Suppress Grid Negative-Sequence Current | Improve current quality at PCC | Moderate | Medium |
| II: Suppress Reactive Power Oscillation | Maintain steady reactive power output | Maximized (Active power ripple absorbs imbalance) | Highest |
| III: Suppress Active Power Oscillation | Maintain steady active power output | Minimized | Lowest |
Furthermore, the current reference prioritization scheme under LVRT and the hard current limit (typically 1.1-1.2 p.u. of rated current, IN) play a decisive role. The sequence of satisfying current references is typically: 1) Reactive current for LVRT support, 2) Negative-sequence current per the chosen control objective, 3) Positive-sequence active current. If the remaining current “headroom” after providing mandatory reactive support is insufficient to fully accommodate the oscillatory active current reference, the current limiter will clip the id* waveform, suppressing the 100 Hz component and, consequently, the generated third harmonic.
The following factors lead to significant third-harmonic current output from a solar inverter:
- Moderate Voltage Dip: Very deep dips prioritize large reactive current, leaving little room for active current oscillation. Very shallow dips cause small Udc_2. Moderate dips (e.g., 0.3-0.7 p.u.) often create the worst-case scenario.
- Low Pre-Fault Output: If the inverter is operating at low power before the fault, it has ample current capacity to follow the oscillatory active current reference without hitting the limit.
- Use of Control Objective II: This strategy intentionally maximizes active power ripple to cancel reactive power ripple, leading to large Udc_2.
| Factor | Condition for High 3rd Harmonic | Mechanism |
|---|---|---|
| Fault Severity | Moderate asymmetric fault | Generates substantial negative-sequence voltage without forcing inverter into max reactive current/limit. |
| Pre-Fault Power | Low (< e.g., 0.5 p.u.) | High current headroom allows full tracking of oscillatory active current reference. |
| Negative-Sequence Control | Objective II (Suppress Q-oscillation) | Maximizes the 100 Hz active power ripple and thus Udc_2. |
| Current Limiter Action | Not activated for id* oscillation | The 100 Hz component in id* remains undistorted, fully propagating through control loops. |
4. Impact on Transformer Differential Protection and Mitigation
Transformer differential protection is a fundamental primary protection scheme. A significant challenge is preventing false operation during external faults that cause Current Transformer (CT) saturation. Saturated CTs produce distorted secondary currents with rich harmonic content, which can generate spurious differential current.
To counter this, transformer differential relays commonly employ harmonic restraint/blocking logic. A typical second or third harmonic blocking criterion is:
$$ I_{\phi3} > K_3 \cdot I_{\phi1} $$
where Iφ3 and Iφ1 are the phase’s third harmonic and fundamental currents, respectively, and K3 is a restraint ratio (e.g., 0.15-0.25). If this condition is met, the sensitive low-set percentage differential element (with a low pickup and slope) is blocked, while a higher-set, more robust element remains active to clear severe internal faults even with CT saturation.
The vulnerability arises in systems with high penetration of solar inverter-based generation. During an internal transformer fault, the inverter-fed contribution to the fault current may contain a naturally high third harmonic (e.g., >25% of fundamental) due to the mechanisms described. This can satisfy the harmonic blocking criterion (Iφ3 > K3Iφ1).
The traditional safety net—the high-set differential element—may fail to operate if the overall fault current is low. This scenario is plausible when:
- The fault is within the transformer zone fed by the PV plant.
- The solar inverter provides most of the fault current but is current-limited to ~1.2 p.u.
- The grid-side source is weak or the fault impedance is high, providing a modest contribution.
- The inverter’s fault current has high third harmonic content, blocking the low-set element.
- The total differential current is insufficient to reach the pickup of the high-set element.
This creates a protection blind spot, leaving a genuine internal fault undetected.
Proposed Mitigation Strategies
To address this challenge, solutions can be implemented both within the inverter control and the protection relay logic.
1. Inverter-Side Mitigation (Source Control):
A feedforward compensation term can be added to the inverter’s control loops to counteract the influence of the measured DC-link 100 Hz ripple (Udc_2). By actively canceling this perturbation before it enters the current reference calculation, the generation of the third harmonic voltage reference can be significantly suppressed. This improves the quality of the inverter’s fault current and eliminates the problem at its source.
2. Protection-Side Mitigation (Adaptive Logic):
The relay can be enhanced with additional criteria to distinguish between harmonics caused by CT saturation and those inherent to the source. The blocking logic can be made conditional:
- Fundamental Current Level Check: CT saturation typically occurs only for high-magnitude fault currents. If the phase fundamental current Iφ1 is below a threshold (e.g., 2.0 p.u.), significant third harmonic is unlikely to be from CT saturation.
- Waveform Distortion Analysis: Severe CT saturation produces current waveforms with clear “flat-top” periods or intervals of near-zero current (interruption angles). The relay can detect these signature patterns.
An improved, conditional blocking logic can be formulated as:
Block Low-Set Differential IF: (Iφ3 > K3 · Iφ1) AND [ (Iφ1 > IHigh_Thresh) OR (Waveform shows CT saturation signature) ]
If the high third harmonic is accompanied by low current magnitude and no saturation signature, the low-set differential element remains unblocked, ensuring sensitivity for internal faults fed by inverter-based sources.
| Strategy | Location | Principle | Advantages | Challenges |
|---|---|---|---|---|
| Feedforward Compensation in Control | Solar Inverter | Eliminates the 3rd harmonic at source by canceling DC-link ripple effect. | Fundamental solution, benefits all protection schemes. | Requires modification to inverter firmware/controls, adds complexity. |
| Adaptive Harmonic Blocking in Relay | Protection Device | Discriminates harmonic source (CT saturation vs. inverter) using current magnitude and waveform. | Retrofittable to existing relays, addresses the symptom directly. | Relies on accurate waveform analysis, requires setting additional thresholds. |
5. Analytical and Simulation Insights
The theoretical analysis can be substantiated through simulation of a detailed model of a grid-connected solar inverter. Key parameters include the DC-link capacitance, PI controller gains for voltage and current loops, current limit, and the specific negative-sequence control strategy implemented.
Simulations under various unbalanced fault conditions (e.g., phase-to-ground, phase-to-phase) confirm the derived relationships. They demonstrate that the worst-case third harmonic magnitude, potentially exceeding 40-50% of the fundamental, occurs under the confluence of conditions outlined earlier: moderate voltage dip, low pre-fault power, and the use of negative-sequence control Objective II. The harmonic amplitude can be quantified as:
$$ \text{THD}_{3} (\%) = \frac{I_{3h}}{I_{1}} \times 100\% \approx f(U_{dip}, P_{pre-fault}, \text{Control Objective}, I_{limit}) $$
Simulations further validate the protection system vulnerability. Scenarios can be constructed where an internal transformer fault is fed primarily by a PV plant. The inverter’s current-limited, harmonic-rich output causes the transformer differential protection’s third harmonic blocking criterion to be met, while the total differential current remains below the high-set element threshold, demonstrating the protection failure mode. Re-running the same scenario with either the inverter’s feedforward compensation enabled or the relay’s adaptive logic active shows the fault being correctly cleared, confirming the efficacy of the proposed solutions.
6. Conclusion
The transition to inverter-dominated power systems introduces new dynamic phenomena that must be thoroughly understood to maintain protection reliability. This analysis has detailed the mechanism by which a grid-following solar inverter generates third-harmonic current during unbalanced grid faults. This harmonic stems from the interaction of negative-sequence grid voltages with the inverter’s cascaded control loops, leading to a double-frequency DC-link voltage ripple that is processed into a third-harmonic AC output.
The magnitude of this harmonic is not fixed but is a complex function of the fault conditions, the inverter’s pre-fault operating point, its chosen negative-sequence control objective, and the action of its current limiter. Under specific but plausible grid conditions, this inherent third harmonic can reach levels sufficient to activate the harmonic blocking functions in transformer differential relays. Crucially, unlike harmonics from CT saturation, these inverter-sourced harmonics can exist alongside relatively low fault current magnitudes. This combination can create a protection blind spot, where the sensitive low-set differential element is blocked, and the high-set element fails to operate due to insufficient current, leaving an internal transformer fault uncleared.
Addressing this challenge requires a multi-faceted approach. Proactive measures can be taken within the solar inverter control software, such as implementing feedforward compensation to suppress the harmonic generation. Concurrently, protection engineers can enhance relay logic with adaptive algorithms that distinguish between harmonics from CT saturation and those from the source, using criteria like fundamental current magnitude and waveform distortion signatures. By integrating these insights into both inverter grid-support functions and next-generation protection schemes, the secure and reliable integration of large-scale photovoltaic generation can be further ensured.