About the data

France's WEST and US DIII-D compare like-for-like plasmas to inform ITER operations

In a 9 November research update, France's CEA explained coordinated WEST–DIII-D experiments that matched normalized plasma conditions despite different machines. The 2025 tests, published in May 2026, offer clues on confinement, H-mode access and ion heating for ITER.

Published Updated Translation reviewed
Share this article
Read the Japanese original

Announcement date:2026-11-09

FUSION INDUSTRY brand logo
FUSION INDUSTRYFUSION INDUSTRYFI's brand logo, not a photograph of either experimental device.

Two tokamaks with very different shapes and wall materials: can their plasma heat-confinement physics be compared on equal terms? On 9 November, France's CEA described coordinated experiments between its WEST device and the US DIII-D operated by General Atomics. The aim is to understand how ITER could combine high performance with long-duration operation.

High pressure for short periods, or tungsten walls for long pulses

DIII-D can reach high normalized plasma pressures using neutral beams and electron-cyclotron heating, typically in shorter pulses. WEST specializes in long pulses using actively cooled tungsten components. A raw comparison of pulse length or temperature misses the differences between the machines.

The researchers therefore aligned plasma shape and several dimensionless parameters, including normalized pressure, collisionality and the ion-to-electron temperature ratio. Despite different device geometries and wall materials, the plasma cores showed similar normalized thermal confinement in the standard low-confinement regime (L-mode).

Why ITER's high-confinement regime matters

One challenge is entering H-mode, where plasma retains heat more effectively. Comparing the machines helps clarify how magnetic field and machine geometry affect the threshold, particularly when available heating power is limited. In predominantly electron-heated plasmas, the ratio of energy-confinement time to electron–ion heat-exchange time also proved relevant to ion heating.

The experiments took place in April–May 2025 and appeared in a peer-reviewed paper in May 2026. CEA's 9 November 2026 update explained those results and the next coordinated tests. Their value lies in a shared dataset for checking predictive models—not just separate performance records. FI's plasma turbulence guide explains the challenge of plasma heat transport.

Sources and primary documents

Information checked as of:2026-11-13
Experiments April–May 2025; paper published May 2026; CEA research update 9 November 2026. Comparable L-mode performance does not establish long-pulse H-mode operation on either machine.

Share this article

Discuss this article