Risk Retirement
The build is organized to retire the largest physics and engineering risks in a defined order, with the first-of-a-kind unit as the primary risk-retirement instrument.
Retiring risk in order
Rather than assuming the design is correct, the program identifies its largest uncertainties and retires them in sequence. The first-of-a-kind (FOAK) unit is the main instrument for this: its purpose is to convert the biggest simulation-based assumptions into measured results.
The dominant risks
- Confinement scaling: will the plasma reach Q_sci 3.076? (retired at G1)
- High-field magnets: is 16.84 T real at scale under load? (retired at G2)
- Breeding: does the blanket net-breed tritium in real 3-D geometry? (retired at G3)
- Integration: do the subsystems work together? (retired in commissioning)
- Build execution: can the machine be built in ~3.5 years? (retired by FOAK itself)
Why FOAK is the instrument
Some risks cannot be retired by more simulation, they require the built machine. FOAK is heavily instrumented and reported against pre-registered predictions precisely so that each measured result closes an open question. A FOAK that finds problems has done its job.
Open reconciliations
The program tracks specific open items honestly, including the reconciliation between design-level and net three-dimensional tritium breeding ratio, and details of the magnet materials. These are named, not hidden, and are mapped to the gates that will settle them.
Risk retirement describes a machine not yet built. Its honesty is the point: the program states what is unproven and where each unknown gets resolved.