Fusion research: the long bet

01What fusion is

Nuclear fusion is the process in which light atomic nuclei merge to form a heavier one. A small amount of mass disappears in the process and is released as energy, in line with Einstein's relation between mass and energy. In stars, gravity squeezes hydrogen hard enough for fusion to happen continuously.

On Earth, the most promising reaction combines two heavy forms of hydrogen: deuterium, which can be extracted from water, and tritium, which is rare and has to be produced. The result is a helium nucleus and a fast neutron that carries most of the energy away.

D + T → He-4 + n + 17.6 MeV

Fusion is the opposite of fission, the splitting of heavy nuclei such as uranium in today's nuclear power stations. A fusion reaction cannot run away: if conditions are disturbed, it simply stops. It produces no long-lived high-level waste of the kind fission does, although the reactor walls become radioactive through neutron bombardment and must be managed. These features explain the political appeal, including in Germany, which ended fission power generation in 2023.

02Plasma and confinement

For nuclei to fuse, they must collide fast enough to overcome the electric repulsion between them. That requires temperatures of more than 100 million degrees Celsius, several times hotter than the core of the Sun. At such temperatures matter becomes a plasma: a gas in which electrons are stripped from their nuclei, leaving a soup of charged particles.

No material container can hold anything that hot. The central engineering question of fusion is therefore confinement: how to keep the plasma dense, hot and stable long enough for enough reactions to occur. Two broad answers exist, magnetic confinement and inertial confinement.

03Magnetic fusion: tokamaks and stellarators

Because plasma is made of charged particles, magnetic fields can steer it. Magnetic fusion devices bend the field lines into a closed ring, a doughnut shape known as a torus, so that the particles circle without touching the walls.

Tokamak

The tokamak, developed in the Soviet Union in the 1950s and 1960s, is the most studied design. It combines external coils with a strong electric current driven through the plasma itself. Tokamaks are comparatively simple to build but the plasma current makes them prone to instabilities, and they operate in pulses.

Stellarator

The stellarator avoids the plasma current by twisting the magnetic field with intricately shaped coils. It is harder to design and build, but in principle suited to continuous operation. Germany is a world leader here: Wendelstein 7-X at the Max Planck Institute for Plasma Physics in Greifswald is the world's largest stellarator of its type, built to show that the concept can work at power-plant-relevant scales.

04Laser or inertial fusion

Inertial fusion takes the opposite approach. Instead of holding a plasma for a long time, it compresses a tiny pellet of fuel so fast and so hard, typically with powerful lasers, that fusion happens in a fraction of a second before the pellet flies apart. In December 2022 the US National Ignition Facility reported, for the first time, that the fusion energy released from a target exceeded the laser energy delivered to it.

A power plant on this principle would need to repeat such shots many times a second, with lasers far more efficient than today's research lasers. The Fusion Action Plan reflects the growing interest in this route: a substantial part of its funding is earmarked for a new laser-fusion research infrastructure.

05ITER and the European framework

The largest fusion project in the world is ITER, a tokamak under construction at Cadarache in southern France. It is an international collaboration in which the European Union, which hosts the project, works with partners including China, India, Japan, South Korea, Russia and the United States. ITER is designed to demonstrate a burning plasma that produces far more fusion power than the heating power put in, but it will not generate electricity.

German institutes contribute to ITER through the European fusion programme. National policy therefore always operates in a European frame; the wider picture is on the European technology policy page.

06The hard problems

Scientific progress in fusion has been real, but a power plant requires solving several problems at once.

• Net energy. A device must release far more energy than the whole plant consumes, not just more than reaches the fuel. No experiment has yet achieved net electricity.
• Materials. The fast neutrons from fusion damage the walls of the reactor over time. Materials that survive years of bombardment are still being developed and tested.
• Tritium. Tritium hardly occurs in nature. A power plant would have to breed its own fuel, typically by letting neutrons strike lithium in the reactor walls, and handle a radioactive gas safely. This is the "fuel cycle" problem.
• Engineering and cost. Superconducting magnets, high-power lasers, heat extraction and remote maintenance must all work reliably and affordably.

These problems explain why the Fusion Action Plan names a hub for fuel cycle and materials alongside its hubs for magnetic and laser fusion.

07The Fusion Action Plan

The High-Tech Agenda, adopted in July 2025, listed "fusion and climate-neutral energy generation" among its six key technologies and promised a fusion action plan as the next step towards a fusion power plant in Germany. The cabinet adopted the "Aktionsplan Fusion" on 1 October 2025.

Its central figure is more than €2 billion by 2029. That comprises around €1.7 billion of public funding in the current legislative period plus up to €755 million from the special fund for infrastructure for a new laser-fusion research facility. For comparison, federal fusion funding had previously been about €150 million a year. The plan is organised in eight fields of action, foresees hubs for magnetic fusion, laser fusion, and the fuel cycle and materials by the first quarter of 2026, and announces an update of the Radiation Protection Act, the legal framework a future plant would have to meet.

A week later, in the Bundestag's government question time on 8 October 2025, Bär addressed the plan. According to the Bundestag's summary of the session, she said that by 2029 some €2 billion would go into fusion research so that the world's first fusion power plant could be built in Germany.

In May 2026 a fusion roadmap was among the first six technology roadmaps that Bär presented to the Bundestag research committee for public consultation.

08Competition and critics

Germany is not alone. The United States, the United Kingdom, China and Japan all fund national fusion programmes, and private investment in fusion companies has grown sharply over the past decade. German start-ups pursue both main approaches: companies such as Proxima Fusion work on stellarator designs, while Marvel Fusion and Focused Energy pursue laser-based fusion. The plan's emphasis on hubs and infrastructure is intended to keep such firms, and the expertise around them, in Germany.

Analysis  Scepticism about fusion is as old as the field. Commercial fusion power has repeatedly been forecast for a few decades ahead, and many physicists and energy analysts warn against treating it as a contribution to near-term climate targets. Even optimistic timelines put grid-connected plants in the 2040s or later. A "world's first" power plant in Germany is therefore an aspiration whose success cannot be judged within this or the next legislative period.

There is also an opportunity-cost argument: money committed to a long-term technology is not available for research with quicker returns. Supporters reply that the sums are small compared with energy-system spending, that Germany already has world-class institutes, and that the industrial spin-offs, in magnets, lasers and materials, are valuable regardless. The overall budget context is on the science and research funding page.