Most discussion of fusion focuses on machines: magnets, lasers, plasma temperatures. Yet the choice of fuel shapes a fusion company’s cost base, its regulatory profile and its supply risk just as much. Three fuel pairs dominate the debate: deuterium–tritium (D–T), deuterium–helium-3 (D–He3) and proton–boron-11 (p–B11).
The baseline: deuterium–tritium
ITER describes D–T as the only fusion reaction accessible with current technology (ITER). The economics of its two ingredients could hardly be more different:
- Deuterium is plentiful. Seawater contains about 33 grams per cubic metre (ITER).
- Tritium is radioactive with a half-life of 12.3 years and occurs in nature only in trace amounts. Today’s supply comes mainly as a by-product of CANDU-type fission reactors, which ITER put at roughly 100 grams per year for a 600 MW reactor (ITER).
The business implication is that a D–T power plant has to breed most of its own tritium from lithium inside the machine. That adds engineering scope, but the fuel cost itself is not the main concern; the breeding blanket is.
Helium-3: easier neutrons, harder supply
In D–T fusion about 80% of the energy is carried by neutrons, which drives materials damage, activation of components and shielding requirements (Kulcinski & Santarius, University of Wisconsin Fusion Technology Institute). D–He3 shifts much of the energy to charged particles instead. Helion has said it will use D–He3 for commercial operations; its Polaris prototype reached plasma temperatures of 150 million °C in early 2026, and the company says it will keep raising temperatures to operate reliably with that fuel (Helion, February 2026).
The constraint is supply. Helium-3 makes up only about 0.000137% of atmospheric helium (CIAAW, helium). That scarcity is why lunar extraction keeps appearing in the conversation. Interlune has announced that the U.S. Department of Energy Isotope Program agreed to buy three litres of lunar helium-3 for delivery no later than April 2029 (Interlune). Three litres is a pilot quantity, not a fuel supply chain; it is best read as a test of whether the logistics can work at all.
Boron-11: abundant fuel, demanding physics
Proton–boron-11 is the most “aneutronic” of the three options. Boron-11 is the dominant natural isotope of boron, with an abundance of roughly 80% (CIAAW, boron), so raw-material supply is not the issue. The challenge is physics: the same Wisconsin analysis notes that the advanced fuels require higher plasma temperatures and better confinement than D–T (UWFDM-1150). That applies to D–He3 as well, but most strongly to p–B11.
TAE Technologies is the best-known company on this path. In November 2025 it reported forming a field-reversed configuration plasma using only neutral-beam injection in its Norm device, and said this allows it to skip a planned intermediate machine and move towards its first power-plant design, Da Vinci (TAE). TAE says Norm is being upgraded to reach 100 million °C, which it describes as a record for a steady-state FRC.
How fuel choice changes the business case
| Fuel pair | Fuel availability | Main engineering burden | Typical commercial risk |
|---|---|---|---|
| D–T | Deuterium abundant; tritium must be bred on site | Breeding blanket, neutron shielding, materials | Tritium self-sufficiency and component lifetime |
| D–He3 | Helium-3 very scarce on Earth | Higher plasma temperatures | Securing fuel at scale |
| p–B11 | Boron abundant | Far more demanding plasma conditions | Physics risk before engineering risk |
Table: BusinessCraft summary based on the sources linked in this article.
For investors and suppliers, the practical reading is that D–T developers carry engineering and supply-chain risk, while alternative-fuel developers carry more physics risk and, for helium-3, a resource risk that no single company controls. None of the three routes has yet produced net electricity.
Key figures
| Figure | Value | Source |
|---|---|---|
| Deuterium in seawater | ≈ 33 g per m³ | ITER |
| Tritium half-life | 12.3 years | ITER |
| Share of D–T energy carried by neutrons | ≈ 80% | UWFDM-1150 |
| Helium-3 share of atmospheric helium | ≈ 0.000137% | CIAAW |
| Boron-11 share of natural boron | ≈ 80% | CIAAW |
| Helion Polaris plasma temperature (early 2026) | 150 million °C | Helion |
Figures as published by the linked sources; table compiled by BusinessCraft.
Related reading
- Aneutronic Fusion Under the Microscope: Five Tests p–B11 and D–He3 Still Have to Pass
- Tritium, Lithium and Helium-3: Mapping the Fuel Supply Chain Fusion Will Need
- Fusion in 2026 by the Numbers: Records, Money and the Distance Still to Go
This article is general information, not investment advice.
Sources
- ITER Newsline: Tritium – changing lead into gold
- Kulcinski & Santarius, New Opportunities for Fusion in the 21st Century – Advanced Fuels (UWFDM-1150, 2000)
- Helion: Polaris milestones (13 Feb 2026)
- CIAAW: helium isotopic abundance
- Interlune: DOE Isotope Program helium-3 purchase (7 May 2025)
- CIAAW: boron isotopic abundance
- TAE Technologies: TAE shortens device roadmap (17 Nov 2025)




