China Just Broke Ground on Fusion Nobody Else Will Touch

Frontier Signal · Deep Dive · Energy & Investing · 7 min read · Wednesday, 30 September 2026

Every fusion company racing for a working reactor this decade, Commonwealth Fusion, Tokamak Energy, ITER, is chasing the same reaction: deuterium and tritium at around 150 million degrees. It is one of the hardest engineering problems in industrial history. By fusion standards it is also the easier version.

There is a harder reaction that produces no stray neutron at all, and for decades almost nobody has been willing to build toward it, because it needs plasma roughly twenty to thirty times hotter.

On 15 September, China's ENN Group broke ground on Helong-2, a platform built to chase exactly that: hydrogen fused with boron.

3+ BILLION DEGREES REQUIRED · 20 TO 30X HOTTER THAN D-T FUSION · FIRST ELECTRICITY TARGETED FOR 2030
Temperature figures are independently published, not vendor-estimated. Nothing has fused at ENN yet. This is a groundbreaking, not a working reactor.

Construction started on 15 September 2026 in the Langfang Economic and Technological Development Zone, Hebei Province, across roughly 120 acres with more than ten core systems under way. China Construction First Group is the contractor.

ENN's timeline calls for equipment installation by the end of June 2027, commissioning finished by the end of that year, and first hydrogen-boron electricity by 2030. Chairman Yu Jianchao called Helong-2, the company's third-generation device, a key engineering platform for a fusion smart power plant.

It is designed for 3 million amperes of plasma current, about three times the 1 million amperes its predecessor Xuanlong-50U has demonstrated on hydrogen-boron fuel, figures ENN reported and ITER repeated when approving its membership in an international physics collaboration.

Ordinary reactors, including every prominent Western project, fuse deuterium and tritium. That is relatively easy to ignite, but it fires off a fast neutron every time, and those neutrons slowly wreck the inner walls.

Fuse hydrogen with boron-11 instead and you get three helium nuclei and no neutron. Nothing is irradiated the same way, and because the energy leaves as charged particles rather than neutral neutrons, there is a real path to making electricity directly instead of boiling water for a turbine.

In plain English, it trades a much harder ignition problem for a much simpler waste problem. Getting there means holding plasma above roughly 3 billion degrees, because of a radiation loss called bremsstrahlung that some physicists think will always cool the plasma faster than it can be reheated, and others argue is not the dealbreaker it looks like. That disagreement, not the construction schedule, is the real open question.

Betting that a stubborn physical limit can be engineered around is not unique to fusion. We found a similar wager behind the claim that a 165-year-old law of thermal physics had been broken to make heat programmable.

Every other well-funded company sidesteps the problem rather than solving it. Commonwealth Fusion, Tokamak Energy and ITER all build toward deuterium-tritium. Helion uses deuterium and helium-3, lower-neutron but nowhere near hydrogen-boron's demands, and already has a power agreement with Microsoft. Tokamak Energy uses the same spherical-tokamak shape ENN does, and still chose the easier fuel.

The one serious Western competitor on this fuel is TAE Technologies, which has raised more than 1.3 billion dollars from backers including Google and Chevron and is building a machine called Copernicus. In 2023, TAE and Japan's National Institute for Fusion Science published the first confirmed detection of hydrogen-boron fusion products in a magnetically confined plasma, at a temperature far too low for useful power. It remains the only independently published evidence that the reaction can be measured at all.

ENN Fusion is private, valued at 10.6 billion yuan after a July 2026 Pre-A round led by Loongson Venture Capital, CAS Star and Matrix Partners China. ENN Energy Holdings (HKEX: 2688), a sibling under the same parent, trades in Hong Kong, but it is a gas-distribution utility and no source discloses any financial link to the fusion spending. Buying it as fusion exposure means buying a pipeline company that happens to share a parent.

TAE is the direct competitive read-through, and it is private too. If either reaches net hydrogen-boron energy first, it validates the whole aneutronic bet and puts real pressure on the other.

The broader signal is where Chinese capital is flowing: roughly 7.3 billion yuan into fusion in the first half of 2026, reported as 233 times the prior year, alongside a state-backed vehicle, China Fusion Energy Co., capitalised separately at 15 billion yuan. ENN is the highest-valued private entrant in that boom, not the only one. It is one of several capital-intensive energy moonshots chasing a physics-limited payoff, alongside bets like beaming solar power down to Earth from orbit.

Commonwealth Fusion and Tokamak Energy are the clearest comparisons on architecture and funding scale, but both chase deuterium-tritium, so they are peers in ambition rather than in physics. Marvel Fusion, a German company using lasers rather than magnets on the same fuel, is the more useful comparison for whether the fuel or the reactor architecture is the harder problem.

Nobody has fused hydrogen and boron at ENN. Every milestone reported so far, roughly 1 million amperes of plasma current and electron temperatures near 100 million degrees on the predecessor machine, describes running the right fuel at real confinement conditions, not achieving a reaction. A hundred million degrees is still about thirty times short of what the reaction needs, and ENN's own roadmap places an actual hydrogen-boron reaction as a 2026 target on that smaller machine.

The central physics risk, whether bremsstrahlung makes net energy achievable at all, is unresolved. It is disputed inside the field, including between rival companies' physicists, and a construction schedule does not change that.

We also found no independent researcher or rival publicly assessing the 2030 timeline, so the record is company statements and straightforward news coverage, not outside scrutiny. And despite one aggregator headline calling this the world's first hydrogen-boron reactor, ENN's own chairman said world-leading, which is the framing used here.

If bremsstrahlung turns out to be the dealbreaker its skeptics say it is, does any amount of engineering on Helong-2 matter?

That is what the whole bet rests on, and it is a physics dispute, not an engineering one. No construction schedule resolves it.

On 15 September 2026 it broke ground on Helong-2, a fusion research platform in Hebei Province, China, aimed at producing electricity from hydrogen-boron fusion by 2030. Nothing is built, and no fusion reaction has occurred on this device.

It produces only helium and no stray neutron, unlike the deuterium-tritium reaction nearly every other major project uses, but it needs plasma above 3 billion degrees, roughly twenty to thirty times hotter.

TAE Technologies and Japan's National Institute for Fusion Science published the first confirmed detection of hydrogen-boron fusion products in 2023, at a temperature far too low for useful power. No source shows ENN has measured an equivalent result.

TAE Technologies, a private US company backed by Google and Chevron, using a different reactor design on the same fuel. Most others, including Commonwealth Fusion and Tokamak Energy, pursue the easier deuterium-tritium reaction.

Not directly. ENN Fusion is private. Its sibling ENN Energy Holdings trades in Hong Kong but is a separate gas-distribution business with no disclosed link to the fusion research. TAE is also private.

Construction facts, dates and company figures are as reported by ENN and the outlets below. Temperature requirements are from independent published sources. Targets and timelines are labelled as such.

Frontier Signal explains frontier technology in plain English. Company-reported and vendor-reported claims should be independently verified. This is general information, not investment advice.