Graceful Degradation & Partial Power

When capacity is reduced, the installation should lose the least-important loads, not go dark - degrading gracefully to a protected core.

Fail soft, not hard

A resilient system does not have a single cliff between full power and none. Graceful degradation means that as generation is lost — a unit trips, maintenance takes another offline, or fuel must be conserved — the installation steps down through defined levels, shedding the least-critical loads first and always protecting the mission core.

Degradation levels

The plant defines discrete states: full service, reduced service with deferrable loads shed, essential service with only important and critical loads, and survival mode with only life-safety and mission-critical loads. The microgrid controller moves between states automatically as available generation changes, and restores upward as capacity returns.

• Full → reduced → essential → survival, defined in advance
• Deferrable loads shed first, mission core last
• Automatic step-down and step-up by the controller
• Each level is a stable operating state, not a scramble

Why it complements redundancy

Redundancy aims to prevent capacity loss; graceful degradation manages it when prevention is not enough. Because the single-unit availability gate (0.86–0.995) guarantees occasional shortfalls, the plant is designed to spend those shortfalls on the least-important loads. The two mechanisms together are what make the availability story honest and workable.

Degradation logic is configured per installation to its mission, and because each level is a stable operating point the plant can hold a reduced state for as long as conditions require rather than treating partial power as a passing emergency. Each degradation level is engineered as a stable operating point with its own protection settings, so the installation can dwell there indefinitely if needed rather than treating reduced power as a transient emergency.