Continuity During Maintenance

Planned maintenance is a predictable outage that must be covered by redundancy so the mission never loses power during servicing.

Maintenance is planned downtime

Unlike a fault, maintenance is scheduled — which makes it fully coverable. But a high-field fusion unit needs real maintenance windows: magnets and cryogenics do not warm and cool quickly, and inspection of a machine with an activated first structure takes time. Continuity requires these windows to be invisible to the mission.

How it is covered

• N+1 units so one can be serviced while others carry the load.
• Scheduling maintenance during lower-demand periods where possible.
• Modular design to shorten the maintenance window (lower MTTR).
• Storage to cover the transition as load moves between units.

Why it is part of the availability figure

The 0.86–0.995 availability range already folds in maintenance downtime, not only unplanned faults. Because maintenance is predictable, it is the easiest part of unavailability to cover with redundancy. A well-run installation loses no mission power to servicing, even though each unit spends real time offline for it.

Continuity during maintenance is a straightforward consequence of the redundancy the availability gate already requires. It is worth stating explicitly because it shows the gate is managed, not ignored: expected downtime is designed for, so the mission does not feel it.

The most coverable part of the gate

Because maintenance is scheduled, it is the easiest component of unavailability to cover: an N+1 architecture lets one unit be serviced while others carry the load, storage covers the transition, and modular design shortens the window. The 0.86–0.995 range already folds in maintenance, so a well-run installation loses no mission power to servicing even though each unit spends real time offline.