The Three Products
Tritium (~4 kg/yr class), helium-3 (~1.97 kg/yr from tritium decay), and 14 MeV neutrons — one machine, three strategic outputs.
One neutron source, three outputs
Hyperion's three products all trace back to a single physical fact: D–T fusion emits a 14.1 MeV neutron per reaction. At 85.0 MW of fusion power the source term is large, and the machine is engineered to convert that neutron budget into material outputs.
1. Tritium
Bred in a lithium-6-bearing blanket via ⁶Li(n,α)T, with neutron multiplication from beryllium or lead. Target output is a ~4 kg/yr class. Tritium is the primary product and the reason the machine is called a breeder.
2. Helium-3
Not fused directly; it accumulates as tritium beta-decays (³H → ³He, half-life 12.3 yr). About 5.47% of a stored tritium inventory converts per year, giving a coproduct near 1.97 kg/yr under the reference inventory model.
3. 14 MeV neutrons
The uncollided fast-neutron flux is offered as an irradiation service. A fusion-representative 14 MeV spectrum is the correct environment for qualifying structural materials, and no fission reactor reproduces it.
- Tritium — fusion fuel and neutron-target feedstock
- Helium-3 — quantum dilution refrigeration, neutron detection
- 14 MeV neutrons — materials qualification, activation
Two timescales, one source
Tritium and neutron flux are produced in real time while the plasma runs; helium-3 is produced on the slow clock of the 12.3-year tritium decay half-life, accumulating in storage whether or not the machine is running. The foundry therefore has a fast product loop and a slow one, and the two are coupled because helium-3 grows from the tritium the machine chooses to hold.
All three streams are contingent on the same physics. The tritium and helium-3 figures depend on the net tritium breeding ratio clearing self-sufficiency at the 1.8 target — an open reconciliation — and the neutron-service capacity depends on how the fixed neutron budget is allocated between breeding and irradiation. The figures are honest design targets, not demonstrated yields.
This page describes a design-and-simulation study, not a built machine. Construction begins Q2 2027; first-of-a-kind (FOAK) first tritium is targeted near 2030. No net-gain claim is made before FOAK.