ITER 2026: What the World's Largest Fusion Experiment Has Achieved So Far

## What ITER Actually Is

ITER — International Thermonuclear Experimental Reactor — is not a power plant. This distinction is fundamental to understanding both its achievements and its critics. ITER's stated goal is to demonstrate a scientific gain ratio (Q) of 10: for every unit of energy input to heat the plasma, 10 units of fusion energy will be produced. It is designed to prove that fusion can produce more energy than it consumes. It is not designed to generate electricity.

This matters enormously in evaluating progress. When ITER is criticized for not generating power, it is being held to a standard it was never designed to meet. When it is celebrated for being "almost there," the goalposts of actual commercial fusion are being obscured. ITER is a physics experiment at industrial scale — the largest and most complex scientific instrument ever constructed.

The experiment is being built in Cadarache, southern France, by a consortium of 35 nations including the EU, US, China, India, Japan, South Korea, and Russia. When it eventually operates, it will contain 840 cubic meters of plasma heated to 150 million degrees Celsius — ten times hotter than the core of the Sun — sustained for 300–500 seconds at a time.

## The Revised Timeline: First Plasma Delayed to 2034

The original ITER schedule projected first plasma — the moment when the machine first ionizes hydrogen gas into plasma, a technically simple milestone that merely proves the magnets and systems are operational — for November 2025. First deuterium-tritium plasma (the actual fusion experiment) was projected for 2035.

In 2022, ITER management announced a revised baseline schedule shifting first plasma to 2034 and full deuterium-tritium operation to 2039. The delay was attributed to damage discovered during vacuum vessel sector assembly, COVID-19-related supply chain disruptions, and the complexity of integrating components manufactured across 35 countries to millimeter tolerances.

In 2024, further scope reviews suggested the actual first plasma date might slip to 2035, with D-T operation potentially in the 2040s. ITER's director-general Pietro Barabaschi acknowledged that the project faces "significant challenges" in its construction phase.

These timeline revisions have become a defining feature of the project's public communication. ITER has experienced schedule slippage in nearly every phase review since the original baseline was set in 2010.

## The Magnet Systems: An Engineering Achievement Already Realized

Whatever the timeline debates, ITER's magnet systems represent a genuine engineering achievement worth acknowledging.

The 18 toroidal field coils that will confine the plasma are the largest superconducting magnets ever built, each weighing 360 tonnes and generating a field of 11.8 Tesla. They use niobium-tin (Nb₃Sn) superconducting strand at 4 Kelvin (-269°C) — just above absolute zero — to carry currents of up to 68,000 amperes without resistance. The total stored magnetic energy in the toroidal field system is 41 gigajoules, equivalent to roughly 10 tonnes of TNT.

Manufacturing these coils required creating entirely new industrial capabilities. The US supplied the central solenoid modules — six stacked modules wound with 43 kilometers of Nb₃Sn conductor each. Japan manufactured nine toroidal field coils. The EU manufactured nine more. Each coil had to meet dimensional tolerances of millimeters over 14-meter lengths, wound and cryogenically impregnated to ensure zero-resistance superconducting performance.

As of 2026, all 18 toroidal field coils have been delivered to the Cadarache site. Their manufacturing represents the largest international industrial collaboration in superconducting magnet history.

## The Vacuum Vessel Assembly Problem

The most significant technical setback in ITER's construction history occurred during vacuum vessel sector assembly in 2022. The vacuum vessel — the 5,000-tonne donut-shaped steel chamber that contains the plasma — is assembled from nine sectors, each manufactured in Korea and South Korea to millimeter tolerances.

When the assembled sectors were measured, deformations were discovered: the thermal cycling during manufacturing had left some sectors slightly warped, with geometric deviations that would cause interference when the full vessel was assembled. The components were not defective — they met their as-manufactured tolerances — but the cumulative tolerances stacked in a way that was not fully predicted.

Repair and correction work has involved custom shimming, precision machining on the Cadarache assembly floor, and revised assembly sequences. The total time cost of this remediation is measured in years, directly contributing to the 2022 schedule revision. ITER management has stated that the vessel sectors are now being successfully assembled following the corrective program.

## Budget: The $22 Billion and Counting Calculation

ITER's cost accounting is notoriously complex. Member nations contribute "in-kind" — manufacturing specific components rather than writing checks — which makes direct dollar comparison difficult. The EU, as host, provides a larger in-kind share. The US, which withdrew from ITER under the Clinton administration and rejoined under George W. Bush, provides specific components and funding.

Various analyses have estimated the total project cost including in-kind contributions at $22–25 billion through initial operations. Some academic analyses using standardized cost accounting methods have placed the full lifecycle cost, including decommissioning, higher still.

For context: the Large Hadron Collider cost approximately $9 billion to build. ITER is the most expensive scientific instrument in history by a substantial margin. Whether that expense is justified depends on how one values the unique plasma physics data ITER will generate at Q=10 scale — data no private fusion company can replicate at their current funding levels.

## What Private Fusion Companies Cannot Replicate

The private fusion sector — Commonwealth Fusion Systems, TAE Technologies, Helion Energy, General Fusion, and a dozen others — collectively raised over $4 billion in private investment through 2025. Several have made credible engineering progress. Commonwealth Fusion's SPARC project, using new high-temperature superconducting magnets capable of 20 Tesla (compared to ITER's 11.8 Tesla), has a compelling physics argument for a much smaller, faster path to Q>1.

Yet these companies consistently make a claim that requires scrutiny: that ITER is obsolete before it operates. The argument runs that ITER's technology was fixed in the 1990s design freeze, and that advances in HTS magnets, plasma control algorithms, and materials science make ITER's approach suboptimal.

What ITER uniquely provides is tritium breeding blanket testing and 400-second plasma burn physics at the Q=10 scale. Tritium is rare and expensive; it must be bred from lithium in reactor blanket modules. How those blankets perform in the actual neutron flux of a sustained DT plasma cannot be adequately simulated or tested at small scale. ITER has dedicated test blanket module ports specifically for this purpose.

The plasma burn physics at Q=10 — how alpha particle heating (the helium nuclei produced by fusion) sustains the plasma's own temperature, known as "burning plasma" — has never been studied experimentally. Burning plasma physics has implications for plasma stability, confinement time, and instability modes that affect every commercial reactor concept. ITER will generate this data regardless of whether it ever produces electricity.

## Why Q=10 Matters for the Physics

The Q ratio quantifies the scientific gain of a fusion reaction. Current record holders (JET tokamak in England, which produced 59 megajoules in a single pulse in 2022) achieved Q values around 0.33 — a third of the energy consumed was returned as fusion. ITER's Q=10 target represents a thirty-fold improvement.

The physics transition from Q<1 to Q>1 is not merely quantitative. At Q>1, alpha particles from fusion reactions — which carry 20% of the fusion energy — begin to significantly heat the plasma themselves. This "alpha heating" changes the plasma's energy balance fundamentally. The plasma begins to sustain its own temperature rather than relying entirely on external heating systems. Studying this transition experimentally, at scale, with long pulse lengths, is what ITER is designed to do.

Every commercial fusion concept — tokamak, stellarator, inertial confinement, magnetized target fusion — benefits from understanding burning plasma behavior. The companies arguing ITER is obsolete still design their reactors based on extrapolations from Q<1 experiments. ITER's Q=10 data will either validate those extrapolations or reveal surprises. In either case, the information is irreplaceable.

## The 2026 Construction Snapshot

As of 2026, ITER's construction status is roughly as follows:

The Tokamak Complex building is structurally complete. The cryostat base and lower sections are installed. Toroidal field coil delivery is complete. The central solenoid module stacking is underway. Vacuum vessel sector assembly continues under the corrective program.

Assembly of the full machine has not yet begun; major components are being staged, inspected, and prepared. The critical path items — vacuum vessel assembly completion, poloidal field coil installation, and integration of the central solenoid — are each multi-year operations.

The project employs over 5,000 people on-site and involves manufacturing work in facilities across all 35 member countries. Whatever its schedule and budget challenges, ITER represents the largest international scientific collaboration in human history. Its eventual operation, whenever it occurs, will answer questions about burning plasma physics that cannot be answered any other way.

ITER 2026: What the World's Largest Fusion Experiment Has Achieved So Far

// COMMENTS

ON THIS PAGE