The Availability Gate For Grid Service

The design-stage availability of 0.86 to 0.995 trails the 0.99982 a hyperscale Tier III load expects; this is the load-bearing grid gate.

The number that governs the grid case

Of the four honest burner gates, availability bears most directly on grid and data-center service. The design-stage availability estimate for a burner unit is 0.86 to 0.995. A hyperscale Tier III data center is engineered for 0.99982. The difference is 30 to 100 times more downtime than the most demanding load will accept from a single source.

This does not defeat the case, but it defines it. A single burner (MetroVolt) is not, today, a sole Tier III source. It becomes suitable through redundancy, fleet operation, storage, and a grid tie, the same way any single generator is backed up in a high-reliability facility.

Why the gap exists

At the design stage, availability is limited by unproven confinement (the plug regime is 166 to 830 times beyond experiment), by the maturity of the direct-conversion train, and by the absence of operating experience from which to estimate forced-outage rates. Each of these is a target of the test program rather than a settled figure.

How the design responds

The response is architectural: modular units, N-1 redundancy, co-sited storage to bridge trips, and retention of a grid connection. These raise effective availability above any single unit. The gate is stated plainly here because pretending otherwise would misrepresent a design-stage machine.

Design-and-simulation finding, not a built-hardware result.