The Availability Gate & Redundancy

A design-stage availability of 0.86-0.995 is well short of a hyperscale Tier III target of 0.99982, so resilient Aegis siting is inherently redundant.

The number, stated plainly

Availability is the fraction of time a unit is producing power. The design-stage estimate for a single Aegis unit is 0.86 to 0.995. A hyperscale Tier III facility targets 0.99982. The gap is not marketing rounding: at the low end it is roughly 100× more downtime, and even at the high end it is about 30× more. A single Aegis unit is not a Tier III source.

Why redundancy is mandatory

If one unit is down 0.5–14% of the time, the way to reach installation-grade continuity is to run more than one unit and a storage buffer, so that the loss of any single unit does not reach the critical load. This is the N+1 (or N+2) principle. Redundancy converts a moderate single-unit availability into a high system availability, which requires extra units and additional floor space in exchange for the continuity gained.

• Single unit: 0.86–0.995 (design-stage estimate)
• Tier III target: 0.99982 — 30–100× lower downtime
• System answer: N+1 units + storage, not a lone unit
• Result: high system availability from moderate unit availability

Honesty as the design

We state the gate rather than soften it because it drives the architecture. Every resilience page in this section — microgrid, black start, continuity — assumes the single-unit number and builds redundancy and storage around it. The correct claim is that Aegis reaches high system availability through configuration, and that closing the single-unit gap is an explicit development goal, not a present fact.

It is a physics-and-reliability statement about how a redundant configuration converts a modest per-unit figure into an installation-grade one, and about the honest distance still to close before a single unit could stand alone.