Intentional Islanding

Before an island can be formed, the control system must unambiguously detect the loss of grid connection. Passive methods monitor voltage magnitude, frequency, and rate of change of frequency (ROCOF) at the point of common coupling. Active methods inject small perturbations to probe grid impedance. IEEE 1547-2018 mandates that distributed energy resources cease energizing an unintentional island within 2 seconds, making the distinction between intentional and unintentional separation critical for protection coordination.

A viable island requires real-time equilibrium between local generation and load. The islanded region must have sufficient dispatchable generation to meet demand plus a spinning reserve margin (typically 5-10%). Key considerations include:

• Active power balance: Frequency will collapse if load exceeds generation
• Reactive power balance: Voltage will sag or swell without adequate VAR support
• Cold load pickup: Thermostatically controlled loads can surge 2-5x normal levels upon re-energization Failure to maintain this balance triggers under-frequency load shedding or generator tripping.

Reconnecting an island to the main grid requires precise synchronization to avoid equipment damage and transient instability. The synchrocheck relay must verify three conditions before closing the tie breaker:

• Voltage magnitude difference < 5%
• Frequency difference < 0.1 Hz
• Phase angle difference < 10 degrees Modern microgrid controllers use phase-locked loops (PLLs) and GPS-time-synchronized phasor measurements to execute bumpless transfers, minimizing disturbance to sensitive loads.

Islanding fundamentally alters fault current levels and protection coordination. The transition from a grid-connected to islanded mode typically reduces available short-circuit current by an order of magnitude, as inverter-based resources contribute only 1.1-1.5 per unit of rated current. This can blind conventional overcurrent relays. Adaptive protection schemes address this by:

• Switching between grid-connected and islanded settings groups
• Deploying directional overcurrent and differential protection
• Using GOOSE messaging (IEC 61850) for high-speed scheme reconfiguration

Without the inertia of the bulk power system, an islanded grid is far more susceptible to frequency and voltage excursions. Control strategies shift from grid-following to grid-forming operation. Key techniques include:

• Isochronous governor control: A single unit assumes frequency regulation responsibility
• Droop control: Multiple generators share load proportionally based on frequency-power and voltage-reactive power droop curves
• Virtual synchronous machine (VSM) algorithms in battery inverters emulate rotational inertia, slowing ROCOF during transients Typical frequency regulation tolerance tightens to ±0.5 Hz in island mode.

A critical subset of intentional islanding is black start capability—the ability to re-energize a de-energized island without external grid support. This requires:

• At least one black start unit (typically a diesel generator, microturbine, or battery with grid-forming inverter)
• A sequenced cranking path to energize transformers and lines in small load blocks
• Soft-start ramp rates to avoid inrush current tripping protection Black start islands serve as cranking sources for broader system restoration, forming the foundation of utility resilience plans following a complete blackout.