Skip to content
Technology How it works Breeder — Hyperion Burner — Aegis Burner — MetroVolt AI-Native Architecture Magnets Fuel cycle Safety Roadmap
Solutions AI & Data Centers Defense & Government Grid & Baseload Neutron Detection Quantum
Learn Technical Library
Proof Publications Whitepapers Technical Library Open Science & Reproducibility The Honest Gates
Company About / Mission Leadership Environment Health & Safety Investors Careers Press Contact
3D Model
Hyperion › Materials & Components
Materials & Components

Blanket Tritium Extraction

Bred tritium must be recovered from the blanket faster than it is lost or it never reaches the product stream — a driver of net breeding.

Bred is not the same as recovered

Breeding tritium is only half the job; it must be extracted from the blanket and delivered to the tritium plant before it decays, permeates away, or is trapped in materials. Extraction efficiency and speed directly reduce the effective breeding ratio, because tritium that is bred but lost never becomes product.

Bred tritiumPurge /carrierExtraction unitTritium plantStorage /productTritium recovery from blanket to product

Where tritium goes wrong

Coupling to the open TBR question

Extraction losses are part of why net tritium breeding falls below the idealized local value. A blanket that breeds at a local TBR of 1.8 but recovers tritium poorly can still leave the machine short of self-sufficiency. The breeder counts these losses in the honest gap between local and net breeding rather than assuming perfect recovery.

Design priorities

The extraction system is sized to keep the in-blanket tritium inventory low, both to reduce losses and to limit the tritium at risk in the machine. Low inventory plus fast recovery is the goal, and it is designed together with the coolant and breeder-form choices.

Inventory as a safety term

Sizing extraction to keep the in-blanket tritium inventory low serves breeding and safety at once: less held tritium is less lost to decay and less at risk in any incident. The extraction rate is therefore judged against both the net-breeding requirement and the safety-case source term, not one alone.

This page documents a design and simulation study, not a built machine. Construction begins Q2 2027; first-of-a-kind first tritium is targeted near 2030. Figures are computed, reproducible targets, not measurements.

Content reviewed August 2026 · design-and-simulation stage