Part 4 of 4: Five Billion for Eternity — or 38 Billion Barrels for Forty Years

Series: Part 3 of 4: The Parts of a Revolution — and Who Holds Them Back

The first three parts looked at the physics, the stellarator, and the components that would have to turn plasma into a power plant. The conclusion is not as convenient as a brochure.

Fusion is not finished. The tritium breeding blanket has not been demonstrated. Materials still need realistic neutron tests. High-temperature superconductors are too expensive and too scarce. A power plant is not merely one political declaration away.

But the conclusion is also not as bleak as the old joke — “fusion is always 30 years away” — suggests. The key questions are known. There is research, there are prototypes, and there are private companies. What is missing is a society able to handle long projects without pretending every year that research is either immediately successful or pointless.

Obstacle 1: the “no nuclear” reflex

In Germany, three generations have learned that anything beginning with “nuclear” is dangerous. That is not simply irrational. Fission has real risks and real political costs. Anyone who ignores that history will persuade no one.

But equating fission and fusion is scientifically wrong. A fusion plasma needs extreme conditions. When those conditions disappear, the reaction ends. There is no self-sustaining chain reaction. There are, however, tritium, neutrons, and activated components. Fusion is neither the next meltdown nor a completely consequence-free energy source.

That sober distinction is missing. People are entitled to reject fusion. They should not reject it because they do not know what it is. The fact that parties avoid this educational task is understandable: material activation and fuel cycles fit poorly on an election poster. That does not make the task less necessary.

Obstacle 2: four years against four decades

Imagine you are a research minister with 500 million euros.

You can fund a tritium facility. It may be ready in ten years, if licensing, construction, and testing work. Before then there will be no attractive opening ceremony and probably no personal political benefit.

Or you can fund solar panels on school roofs. Two years later, cameras will be standing in front of a new roof. The project is useful, visible, and politically legible.

Which decision is more likely in a system judged every four years?

This is not a question of individual morality. It is an incentive structure. Democracies reward visible results in short periods. Research and infrastructure often reward the people who are still following their original plan ten years later.

The familiar fusion joke is that it is 30 years away and always has been. The part often omitted is funding. A technology does not become timeless merely because it is difficult. It becomes timeless when it is funded just enough for the next experiment, but not enough to build the full chain.

Obstacle 3: fossil path dependence

Oil and gas are a multi-trillion-dollar business. No secret sabotage plan is required. Path dependence is quieter.

When billions are invested in LNG terminals and pipelines, those facilities need decades of use to justify the investment. That creates a political interest in gas remaining necessary. Not because every participant opposes fusion, but because infrastructure does not like making itself obsolete.

That is the difference between a conspiracy and a system effect. The first needs a plan. The second arises when many reasonable individual decisions point in the same direction.

Obstacle 4: renewables as a new industry

Solar and wind are not enemies of fusion. They are the bridge technologies we need now. They reduce fossil emissions while other systems are being developed. That is a good reason to expand them.

But a bridge is not a house. The renewable industry also has lobby groups, funding channels, and an interest in presenting its technology as the complete end solution. Energy debates therefore focus on wind, solar, storage, and grids. Fusion appears, if at all, as a distant footnote.

That is politically convenient. It is technically incomplete. Dunkelflauten, storage requirements, grid expansion, and land-use conflicts do not disappear because they are left out of a target graphic. Fusion would not replace renewables, but it could later complement them with firm generation. Solar and wind today, fusion perhaps tomorrow: that is not a contradiction but a division of labour across time.

The China factor

China has an advantage in strategic projects that democratic systems cannot simply copy: long-term planning without a change of government, referendum, or public lawsuit stopping every subproject.

China operates the EAST tokamak, is developing the next step known as CFETR, and has expanded its fusion programme over decades. The system can concentrate funding and hold priorities without having to explain every four years why basic research is not finished yet.

This is not an endorsement of Chinese politics. Surveillance, the lack of individual freedom, and human-rights violations are not an acceptable price for faster research. But when the question is who builds the first industrially relevant fusion reactor, planning continuity is an objective advantage.

Democracies should not conclude that they need to become less democratic. They should draw the much more boring conclusion that long-term research needs institutional protection: multi-year budgets, transparent milestones, independent evaluation, and a politics that does not confuse the failure of one project with the failure of an entire technology.

The solar-panel film

Europe did not invent all the foundations of photovoltaics, but it researched, developed, and manufactured early. China later took over large parts of mass production, drove down costs, and built a dominant supply chain. Europe now imports much of its solar equipment.

That is not automatically a disaster. Cheap solar panels help cut emissions. But the story shows that scientific leadership and industrial benefit are two different things.

Fusion dependence would be more strategic. A power plant is not a single module but a system of magnets, materials, tritium technology, cooling, and control. Whoever owns that system competence controls a possible key industry.

Germany has a strong starting position with Wendelstein 7-X and stellarator expertise. Proxima Fusion is trying to turn that expertise into a commercial concept. The agreement with RWE, Bavaria, and IPP shows that something is moving.

An advantage is not a title deed. It ages when research does not become manufacturing, manufacturing does not become demonstration, and demonstration does not become a market.

At the same time, fusion should not be imagined as a national project. The pellet injector for Wendelstein 7-X came from the United States. Research is international, supply chains will be international, and the atmosphere has no borders. If China builds a working plant first and cuts emissions with it, humanity benefits.

That does not mean Germany will automatically benefit. Its own competence is still needed.

What is already happening

The German government adopted a fusion action plan in 2025. More than two billion euros were announced for the legislative period; in May 2026, the Federal Ministry for Economic Affairs referred to around 2.4 billion euros for German participation in European and national fusion projects. Bavaria offered up to 400 million euros for the Proxima programme.

That is substantially more than in previous years. It signals that fusion is no longer confined to research institutes.

Several billion dollars are now also flowing from the private sector. In its overview, ITER refers to more than 40 private fusion companies and investments exceeding seven billion dollars. These sums do not prove that every start-up is right. They do show that fusion is no longer only a state research programme.

The United States uses ARPA-E to fund risky, time-limited programmes. Britain, Japan, and South Korea are expanding their programmes. The question is no longer whether anyone is researching fusion. It is where the first complete industrial chain will emerge.

The calculation

Our estimate in Part 3 was five to ten billion euros over 15 years. That is 300 to 700 million euros a year. It is not a guarantee of a commercial plant. It is the price of keeping a serious option open.

For comparison:

The world burns roughly 38 billion barrels of oil every year. At a rough price of 80 dollars per barrel, that is more than three trillion dollars a year.
Germany spends many billions each year on renewable expansion and the management of the energy crisis.
ITER’s official planning has shifted considerably; deuterium-tritium operation is scheduled for 2039. The project shows how expensive and slow international large-scale research can be.

Five to ten billion euros is not a trivial sum. It is also not unimaginable. Choosing against such an investment is therefore not a neutral decision made by “the market”. It is a priority.

What should happen now

First, Europe needs a credible tritium strategy. A facility needed in ten years has to be planned today.

Second, it needs an industrial base for HTS tapes. Without affordable magnets in sufficient quantity, the smaller reactor remains a beautiful sketch.

Third, blanket and materials research must move faster from simulations into test facilities. The greatest risk does not become smaller because it is described more precisely.

Fourth, the public needs an explanation. Fusion is not “nuclear power without disadvantages”. That would be as false as saying “fusion is simply nuclear power under a new name”. The public has to hold both things at once: the opportunity and the unresolved risks.

Conclusion

Fusion will not automatically solve the energy problem. Some concepts may fail. A commercial reactor may be more expensive and arrive later than expected. Concealing that would not be education. It would be advertising.

But concluding from uncertainty that nothing should be done confuses caution with passivity.

We burn 38 billion barrels of oil every year and know that this dependence is geopolitically, environmentally, and economically costly. At the same time, facilities in Greifswald are working on a possible successor to that logic. Not a finished solution. A real possibility.

Five to ten billion euros over 15 years would be a bet on technological sovereignty. Thirty-eight billion barrels a year is the bet we are already making — on finite resources, stable supply chains, and a climate that will not send us the bill.

Physics has not made the final decision. Society is making one anyway.