Open Gate: Availability Falls Short of Tier III
Projected plant availability is ~0.86–0.995; hyperscale Tier III data centers require 0.99982 — a 30–100× gap in downtime the burner cannot yet meet alone.
What the numbers say
A first-of-a-kind fusion plant will not start life as an ultra-reliable machine. Our design study projects burner-plant availability in the ~0.86–0.995 range. Hyperscale data centers built to the Uptime Institute Tier III standard require 0.99982 availability — about 1.6 hours of downtime a year. Measured in the quantity that matters, unavailability, the burner is 30–100× short of that bar.
| Availability | Downtime / yr (approx) |
|---|---|
| Burner design study (early) | 0.86 → ~1,200 h |
| Burner design study (mature) | 0.995 → ~44 h |
| Uptime Tier III requirement | 0.99982 → ~1.6 h |
What this means for the data-center case
As a sole or primary source for a Tier III load, the burner does not meet the reliability bar today. The credible near-term role is as a firm generator inside a redundant architecture — paired with storage, grid backup, and N+1 plant redundancy — not as a single point of supply for a hyperscale campus.
Why we publish this
The reliability gap is a fact about first-of-a-kind hardware, and stating it is how we design the redundancy that closes it. A vendor who hides the availability curve is a vendor you cannot plan around.
Sources. Tier III availability (99.982%) per the Uptime Institute Tier Standard; plant availability range from the Kronos design study (DOI 10.5281/zenodo.21746479).