For MetroVolt to close, its plasma must hold energy 2.28 times better than the standard H-mode scaling law (IPB98(y,2)) predicts for its size and field — 2.38 on the 85:15 banking leg. That multiplier, H98, is the design's single most important open question, and we treat it that way.
The requirement is published without a hedge: H98 = 2.28 at the Mode-D operating baseline and 2.38 on the 85:15 banking leg (self-consistent profile basis, S76/KX-1D). The Mode-C-era figures — 1.84 hot-ion, 1.91 at 80:20, on a pinned stored-energy basis the Phase-V audit retired — are kept as the labeled history of how the requirement moved when the physics was corrected.
Why is it plausible? Negative-triangularity experiments already report confinement at or above H-mode-like levels without ELMs, and MetroVolt's high-β, high-density spherical geometry is the regime where NT core turbulence behaves best. The claim is extrapolated, gated, and assigned to a decision gate (S18) with deposited gyrokinetic decks waiting on HPC time.
Every fusion venture is betting on something. We name our bet, price it, and publish the experiment that settles it. If NT confinement lands where the evidence points, MetroVolt closes with margin; the requirement band, the fallback posture, and the adverse cases are all in the open record.
| Required H98 | 2.28 baseline · 2.38 banking leg (Mode D) |
| Mode-C-era figures | 1.84 / 1.91 (labeled history) |
| Demonstrated NT band | 1.0–1.2 sustained (≈1.5 transient) |
| Adjudicating gate | S18 (nonlinear gyrokinetics) |
| Evidence base | DIII-D / TCV NT experiments |