The Availability Gate
The burner's projected uptime falls 30 to 100 times short of hyperscale Tier III reliability, so it is not a sole source for a data center today.
Uptime is its own gate
One burner housing, MetroVolt, is aimed at data centers, which measure reliability in fractions of a percent of downtime. Hyperscale Tier III operation targets an availability of about 0.99982. The burner's projected availability is roughly 0.86 to 0.995 — short of that standard by 30 to 100 times in downtime terms. As specified, it could not serve as a sole Tier III power source.
Reading the numbers honestly
- 0.99982 availability means downtime measured in a handful of minutes per year.
- 0.86 to 0.995 means downtime measured in hours to weeks per year — a different regime entirely.
- The 30–100x figure is the ratio of the downtime gap, stated as a reliability fact rather than a cost.
What the gate does and does not say
The gate does not say the burner is useless; it says the burner, at its projected availability, is not a standalone Tier III source. A role alongside other supply is a different claim than a role replacing it, and the honest statement distinguishes them. This is a reliability engineering fact, kept clear of any economic translation.
Why it belongs with the others
Availability sits alongside the plug-coil, operating-regime, and fuel-supply gates because a data-center generator must clear all four to matter. Publishing the uptime gap keeps the MetroVolt case honest about what it would take to be relied upon. See the honest gates.
A data center needs near-perfect uptime. We state exactly how far the current design is from it.