Wendelstein Modular Coil
Fifty hand-shaped, niobium-titanium pretzels — each one engineered to a millimetre, holding hydrogen plasma in a magnetic cage with no axis of symmetry.
Generates the inherently three-dimensional magnetic confinement field of a stellarator without any toroidal plasma current — the design feature that makes stellarators free of the disruption physics that worries tokamak engineers. Wendelstein 7-X uses 50 modular coils (in five-fold symmetry, ten unique shapes) plus 20 planar coils to produce a confinement field optimised for low neoclassical transport, good fast-particle confinement, and steady-state operation.
Each modular coil is a saddle-shaped superconducting magnet wound from cable-in-conduit conductor (CICC) of niobium-titanium (NbTi) cooled to 3.4–3.8 K with supercritical helium. The conductor sits inside a cast stainless-steel coil case fabricated to ±1 mm of its design shape — extraordinary tolerance for a non-planar object 3.5 m across weighing several tonnes. The coils carry up to 17.6 kA, generating a magnetic field of 3 T on axis. Five-fold rotational symmetry means there are only 10 unique coil shapes; the assembly contains 10 copies of each. The non-planar geometry, computed by stellarator-optimisation codes (most famously by Allen Boozer, Per Helander, and the team at IPP Greifswald), is what produces the equilibrium magnetic surfaces that confine the plasma. Even tiny deviations from design — coil-shape errors, assembly misalignment, support-structure deformation — destroy the optimisation, so each coil's geometry is mapped at room temperature with photogrammetry, the deformations under cooldown and electromagnetic loading are simulated, and shimming is applied during assembly. The successful first-plasma operation of W7-X in 2015, with confinement properties matching design, was the first demonstration that stellarator optimisation was a real engineering art rather than a theoretical aspiration.
In plain terms
A tokamak's magnetic cage is like a doughnut wrapped in identical bracelets — easy to imagine. A stellarator's cage is like the same doughnut wrapped in fifty different pretzels, each twisted into a unique shape so that, when you turn them all on at once, the magnetic field lines weave themselves into a perfect twisty bottle that holds the plasma. Each pretzel must be made within a millimetre of its design, or the bottle leaks.
Also called: W7-X coil, stellarator modular coil, twisted superconducting coil.
Why it matters
- Largest superconducting stellarator ever built; demonstrated that stellarator-optimisation theory works in practice.
- Steady-state operation possible without inducing plasma current — eliminates disruptions, the leading risk for tokamak fusion.
- Modular construction enables relatively compact reactor concepts (HELIAS, ARC-class stellarators).
- Coil-fabrication techniques developed for W7-X feed directly into commercial stellarator efforts (Type One Energy, Proxima, Renaissance Fusion).
Where it lives
- Wendelstein 7-X (Greifswald, Germany — operational)
- LHD (Large Helical Device, Toki, Japan — superconducting heliotron)
- Type One Energy 'Infinity Two' design (commercial follow-on)
Key numbers
- Modular coils: 50 (Ten unique shapes, five-fold symmetry)
- Conductor: NbTi cable-in-conduit (Supercritical-helium cooled to ~3.5 K)
- Operating current: up to 17.6 kA per coil
- Magnetic field on axis: 3 T
- Stored energy: ~600 MJ
- Coil-shape tolerance: ±1 mm
Surprising facts
- Each W7-X modular coil weighs 5–6 tonnes and was fabricated to ±1 mm of its design shape — three orders of magnitude better than the precision of a typical building structure of similar size.
- The 50 modular coils carry currents totalling 850 kA and store roughly 600 MJ of magnetic energy — the kinetic-energy equivalent of a fully loaded freight train at 100 km/h.
- After 16 years of construction (1996–2014), W7-X achieved its design plasma in its very first ten-second discharge — a triumph of computational engineering at unprecedented scale.
- Stellarators eliminate plasma-current-driven disruptions entirely — a fundamentally different risk profile from tokamaks for any future power reactor.
Open questions
- Better-than-expected fast-ion confinement: W7-X's optimised geometry was designed to give roughly tokamak-comparable fast-ion confinement; experiments have measured losses that are noticeably lower than predicted by leading codes. The discrepancy is a welcome surprise, but it means the optimisation worked better than its own model — and the residual gap to first-principles theory is not yet fully understood.
- Mode-locking at integer rotational transform: Stellarator plasmas exhibit subtle mode-locking when the rotational transform crosses simple rational numbers (e.g., 5/5, 5/6). Effects on energy confinement near these resonances are reproducible but not yet fully integrated into reduced transport models used for reactor design.
A short history
Lyman Spitzer's original 1951 stellarator concept was eclipsed by tokamaks in the 1970s. The Wendelstein line at Garching/Greifswald sustained the stellarator programme through W7-A, W7-AS, and finally W7-X. The W7-X modular-coil design, optimised by Boozer, Helander and colleagues, was finalised in the early 1990s; coil fabrication began in 1996. After delays driven by coil-tolerance requirements and assembly precision, W7-X achieved first plasma in December 2015 and produced its design-quality optimised plasma in 2018. Subsequent operation has confirmed neoclassical transport reductions and low fast-ion losses predicted by the optimisation. Since 2022, multiple commercial efforts (Type One Energy, Proxima, Renaissance Fusion) have built on the W7-X coil-fabrication legacy.
Sources
- Klinger et al., Nucl. Fusion (2019) — overview of W7-X first operation phase (https://iopscience.iop.org/article/10.1088/1741-4326/ab03a7)
- Wolf, Plasma Phys. Control. Fusion (2019) — W7-X stellarator results (https://iopscience.iop.org/article/10.1088/1361-6587/ab03d7)
- Max Planck Institute for Plasma Physics — W7-X overview (https://www.ipp.mpg.de/w7x)
- Boozer, Phys. Plasmas (1998) — quasi-symmetry framework (https://aip.scitation.org/doi/10.1063/1.872947)