Anti-islanding protection is a mandatory safety feature in every grid-tied solar inverter — yet its underlying mechanisms are often poorly understood by EPCs and system integrators. This article explains what islanding is, how passive and active detection methods work, which global standards apply, and how Sungrow's SG-CX and SH series implement and certify anti-islanding across multiple markets.
What Is Islanding and Why Is It Dangerous?
Islanding occurs when a grid-tied solar inverter continues to generate and supply power to a section of the local network after the utility grid has been disconnected — whether due to a planned outage, a fault, or the operation of an upstream protection relay. The inverter, still energized by the PV array, maintains voltage and frequency on what it believes is a live grid segment. In reality, that segment is isolated from the main utility system, effectively forming a small "island."
The consequences are serious. Utility field workers operating on a line they believe to be de-energized face electrocution risk from an inverter-fed island they cannot predict. Automatic reclosers — protection devices that briefly disconnect and reconnect a faulted line to clear transient faults — may attempt to re-energize a line that is out of phase with the island, causing a violent transient that damages equipment on both sides of the recloser. Phase mismatch between the island and the returning utility can destroy motors, transformers, and sensitive loads connected downstream.
For these reasons, virtually every grid code worldwide mandates that a grid-tied inverter detect the loss of the utility reference and disconnect within a defined time — most commonly 2 seconds of grid loss, though Italy's CEI 0-21 requires disconnection within 0.1 seconds under specific frequency conditions. Compliance is verified by type-test labs and confirmed during the grid connection approval process.
Passive Anti-Islanding Detection
Passive detection methods work by continuously monitoring grid parameters and tripping the inverter when any parameter moves outside its normal operating window. No deliberate perturbation of the inverter's output is required — the inverter simply observes and reacts. The key passive parameters are:
- Over/Under Voltage (OUV): If the island's load does not perfectly match the inverter's output power, voltage will drift. The inverter monitors AC voltage and trips if it moves outside a defined band (e.g., 85–110% of nominal).
- Over/Under Frequency (OUF): Similarly, a power imbalance in the island causes frequency to drift. The inverter trips on frequency deviation (e.g., outside 47.5–51.5 Hz for a 50 Hz system).
- Rate of Change of Frequency (ROCOF): Rather than waiting for frequency to drift to the trip threshold, ROCOF detects the rate of change — a fast-changing frequency signals a sudden loss of the grid reference. ROCOF relays are widely used in European markets (VDE-AR-N 4105, G99).
- Vector Shift: Detects a sudden jump in the phase angle of the voltage waveform, which occurs when the grid disconnects and the islanded system loses its phase reference. Particularly effective at detecting large grid events.
Passive methods are computationally simple, add no noise to the grid, and respond quickly when a significant power imbalance exists. Their weakness is the non-detection zone (NDZ): if the island's load happens to closely match the inverter's active and reactive power output at the moment of disconnection, voltage and frequency may stay within the normal operating band indefinitely. The NDZ problem is the primary motivation for adding active detection methods alongside passive ones.
Active Anti-Islanding Detection
Active methods deliberately introduce a small, controlled perturbation into the inverter's output. When the inverter is grid-connected, the grid's large apparent impedance absorbs the perturbation without measurable effect. When islanded, the perturbation accumulates — driving voltage or frequency out of the normal band — and the inverter detects its own growing deviation and trips. Common active techniques include:
- Active Frequency Drift (AFD / AFDPF): The inverter introduces a slight positive or negative bias to the output frequency, pushing an islanded system's frequency toward the trip threshold faster than natural drift alone.
- Slip Mode Frequency Shift (SMS): The inverter's phase-locked loop is biased so that, under islanded conditions, frequency deviates in a predictable direction, accelerating detection. SMS is one of the most widely used active methods in European and Australian certified inverters, including Sungrow's SG-CX series.
- Reactive Power Variation (RPV): The inverter injects a small reactive power pulse; under islanded conditions the resulting voltage deviation exceeds detection thresholds.
- Sandia Frequency Shift (SFS): An enhanced AFD variant in which the frequency bias is proportional to the observed frequency deviation — providing positive feedback that drives islanded frequency rapidly to the trip threshold while remaining virtually imperceptible on a live grid.
Active methods effectively eliminate the NDZ and are required by most current grid codes alongside passive protection. The trade-off is a very small degree of power quality perturbation — measured in fractions of a percent of rated output — which is well within the limits defined by IEC 62116 and imperceptible in practice.
Global Anti-Islanding Standards
Anti-islanding requirements are set by regional grid codes and supported by international test standards. The table below summarizes the most relevant frameworks for solar EPCs working across multiple markets.
| Standard | Region | Max Trip Time | Key Requirement |
|---|---|---|---|
| IEEE 1547-2018 | USA | 2 seconds | OUV, OUF plus active method; performance categories A/B/C |
| VDE-AR-N 4105 | Germany | 5 seconds | ROCOF plus vector shift; specific voltage/frequency trip windows |
| AS/NZS 4777.2 | Australia / NZ | 2 seconds | OUV, OUF, active method; volt-VAr response integrated |
| IEC 62116 | International | 2 seconds | Test procedure defining NDZ verification for any detection method |
| G99 / G100 | UK | 2 seconds | OUV, OUF, ROCOF; loss of mains protection mandatory above 16 A |
| CEI 0-21 | Italy | 0.1 seconds | Strict frequency limits; fastest mandatory trip time globally |
IEC 62116 is the test procedure standard rather than a grid code — it defines how any anti-islanding method must be validated in a type-test environment using a resonant RLC load to simulate worst-case NDZ conditions. Inverters certified to IEC 62116 have demonstrated that their chosen combination of passive and active methods detects islanding within the standard's time limit even when generation and load are balanced. This certificate is typically required as part of a G99 or AS/NZS 4777 grid connection application.
How Sungrow Inverters Implement Anti-Islanding
Sungrow's SG-CX series commercial string inverters and SH series hybrid inverters implement anti-islanding through a layered approach that combines passive OUV/OUF/ROCOF monitoring with active SMS or AFD perturbation. The specific combination depends on the market version and the applicable grid code. Market-specific firmware variants are pre-configured with the correct trip thresholds, ROCOF sensitivity, and active method parameters for the target country — EPCs do not need to manually configure these at commissioning unless a DNO requests non-standard settings.
Certification coverage for Sungrow's current string and hybrid platforms includes IEC 62116, IEEE 1547-2018, VDE-AR-N 4105, and AS/NZS 4777.2, with country-specific type-test certificates available on request through Econo Solar. Because anti-islanding firmware is tied to the market version of a unit, it is important to confirm that the correct country variant is ordered at the quotation stage — retrofitting a firmware variant after delivery is possible in some cases but requires factory authorization.
For utility-scale central and string inverter deployments, Sungrow's plant controller can supplement inverter-level anti-islanding with additional ROCOF and vector-shift relay protection at the point of connection, providing a defence-in-depth architecture that satisfies strict interconnection requirements. Settings are configurable via iSolarCloud or Modbus TCP, and can be adjusted by authorized personnel with the appropriate access credentials.
Commissioning and Testing Anti-Islanding Protection
Verifying anti-islanding behavior during site acceptance testing (SAT) is a grid connection requirement in most jurisdictions and a practical necessity before commissioning sign-off. The standard verification approach referenced by IEC 62116 uses a grid simulator or a controlled relay trip test to confirm that the inverter disconnects within the code-mandated time window under near-balanced load conditions.
The practical SAT procedure for most projects involves three steps: first, confirm the inverter's firmware version matches the type-tested and approved variant; second, perform a relay trip test by opening the upstream grid breaker while the inverter is generating at a representative output level (typically 50–100% of rated power), and verify the inverter issues an anti-islanding trip within the required time; third, document the trip time, the method used, and the firmware version in the commissioning report for submission to the DNO or network operator.
For projects where the DNO requires witnessed testing — common in the UK under G99 and in Australia under AS/NZS 4777 connection agreements — Econo Solar can provide the type-test certificates and commissioning documentation templates that network operators expect. Ensuring this paperwork is prepared before the SAT date avoids the most common cause of commissioning delays on medium and large commercial PV projects.