Nuclear fusion is the craft of confining a deuterium-tritium plasma long enough and hot enough for nuclei to fuse, then capturing the 14 MeV neutrons and the heat in fusion reactor materials that do not yet exist as a qualified power-plant catalogue. The work splits across tokamaks and stellarators under magnetic confinement, inertial confinement on laser-driven capsules, fusion fuels and tritium breeding, plasma-facing materials, neutron-resistant materials, fusion plasma physics and fusion safety. ITER is designed as the world's largest tokamak to produce 500 MW of fusion power from 50 MW of heating (Q=10) and to test tritium-breeding module concepts; its current baseline targets Start of Research Operation in 2034 and deuterium-tritium operation in 2039 . That schedule, and the stellarator and ignition results running in parallel, is why the same job title now hides incompatible machines.
Challenges in Nuclear Fusion Recruiting
Tokamaks awaiting a complete divertor before research operation
The tokamak remains the configuration most fusion reactors are still designed around, and ITER's revised baseline is a reminder that the machine is only as real as the exhaust hardware inside it. The 2016 plan chased a 2025 first plasma that would have been a 100 kA hydrogen test because the divertor, shield blocks and other in-vessel components would not be there; the new plan installs those systems first and starts research operation in 2034 with hydrogen and deuterium-deuterium plasmas, full magnetic energy in 2036, and deuterium-tritium in 2039 . ITER's largest magnet, the combined toroidal-field set, has a cold mass above 6,000 tonnes, 12 tesla, and 41 GJ of stored energy . A physicist who has run a medium-sized divertor-less campaign, a magnet engineer who has never integrated 4 K coil tests, and a remote-handling designer who has not lived with cassette replacement are not interchangeable with the people who will bring that complete machine up. Hiring against "tokamak experience" without naming whether the divertor, first wall and disruption mitigation system were installed is staffing a paper device.
Stellarators as magnetic confinement without a driven plasma current
Stellarators are the other magnetic confinement family, and they do not hire as a slower tokamak. The rotational transform is written into a three-dimensional coil set, so the configuration is inherently steady-state and does not rely on a driven plasma current; EUROfusion and IPP present that as the plant-relevant property tokamaks still have to manufacture with current drive . On 22 May 2025 Wendelstein 7-X held a peak triple product for 43 seconds with continuous pellet fueling, electron-cyclotron heating above 20 million °C, and an energy turnover of 1.8 GJ over 360-second pulses, putting long-pulse performance on a par with JET despite a plasma volume about one third as large . The coil designer, the island-divertor exhaust physicist, and the gyrotron operator are different seats from a tokamak scenario owner. A CV that lists "magnetic fusion" and a tokamak code without a stellarator configuration, coil tolerance or detachment state is advertising the wrong machine.
Inertial confinement behind hohlraum capsules not magnetic coils
Inertial confinement is a third labour market that still shares the word fusion. On NIF, up to 192 beams enter a hohlraum, x-rays ablate a capsule, and the fuel is compressed faster than it can fly apart; ignition on 5 December 2022 produced 3.15 MJ from 2.05 MJ on target, and an 7 April 2025 shot yielded 8.6 MJ from 2.08 MJ at 456 TW peak power, a target gain of 4.13 . That is capsule metrology, laser-plasma instability control and stockpile-relevant high-energy-density physics, not a superconducting magnet or a tritium loop rated for 400-second burns. ITER's Q=10 goal is 500 MW for 400 seconds in a magnetic vessel; NIF's gain is megajoules in nanoseconds . Private inertial-fusion ventures hire from the laser and target community. Mixing those CVs with tokamak plant engineers because both say "fusion" produces a shortlist that cannot own either pulse.
Fusion fuels that wait on tritium breeding blankets
Deuterium-tritium is the laboratory reaction of choice, and tritium is the fuel the plant does not get from the atmosphere. ITER's later mission is to test mock-up in-vessel breeding blankets; Fusion for Energy is explicit that ITER will not be tritium self-sufficient, while DEMO must regenerate tritium in a blanket that also extracts heat for electricity . Europe's Test Blanket Module work is welding EUROFER97 structures for helium-cooled pebble-bed and water-cooled lithium-lead concepts under nuclear-pressure-equipment rules . A tritium-plant engineer who has run isotope separation, a blanket designer who owns lithium-lead corrosion and permeation barriers, and a fuelling physicist who has injected pellets are not one hire. World tritium inventory is already spoken for by ITER and a handful of other users; a candidate who has handled fission fission-product chemistry without hydrogen-isotope accountancy has not closed a fusion fuel cycle. The brief has to name which fusion fuels path, which blanket concept, and which inventory the seat is allowed to hold.
Plasma-facing materials under divertor heat and tritium inventory
Plasma-facing materials are the armour that sees the scrape-off layer, not a generic high-heat-flux coating. ITER's new baseline replaces the beryllium first wall with tungsten because tungsten is the DEMO-relevant choice, and the Start of Research Operation phase will already carry a divertor so that heat loads can be tested before tritium arrives . Wendelstein 7-X ran long pulses under fully detached divertor conditions while raising energy turnover to 1.8 GJ, which is an exhaust result as much as a core-physics result . Tritium retained in those surfaces is a fusion safety limit, not a housekeeping metric: EUROfusion's DEMO safety work notes that ITER's chronic atmospheric tritium release is held to about 1 g per year and accidental release to a few grams per event . A tungsten-armour manufacturer, a detachment-control physicist, and a tritium-retention modeller share a wall and do not share a qualification. Hiring that screens for "PFC experience" without the armour, the detachment state and the inventory number will discover the gap when the first-wall change-out, not the equilibrium, is on the critical path.
Neutron-resistant materials fission irradiations cannot qualify
Fusion reactor materials behind the first wall have to survive 14 MeV neutrons that fission spectra do not reproduce. EUROfusion states that those energies cannot easily be mimicked in fission-based experiments, which is why IFMIF-DONES is being built in Granada: a 125 mA, 40 MeV deuteron beam into a 25 mm lithium curtain at 15 m/s, producing a fusion-relevant neutron field to irradiate samples . Baseline EU DEMO materials are EUROFER for structures, tungsten for armour, and CuCrZr as a heat sink; the High Flux Test Module is tailored to EUROFER at up to 50 dpa in three years between 250 °C and 550 °C, while tungsten and copper transmutation are not a perfect match . Neutron-resistant materials work is therefore spectrum-specific: helium-to-dpa ratio, transmutation to rhenium and osmium in tungsten, and the small-specimen mechanical tests that will feed a DEMO design code. A reactor-pressure-vessel metallurgist who has run LWR surveillance capsules has irradiation discipline and still has not qualified a breeding-blanket steel in a 14 MeV field. Programmes that treat fusion materials as "nuclear materials with a divertor" will staff the wrong irradiation case.
Fusion plasma physics evidence a shot log actually owns
Fusion plasma physics is owned on a named campaign, not on a code list. The questions that separate a contributor from a spectator are which pulse, which heating mix, which density peaking, and which exhaust state the candidate was responsible for when the diagnostic disagreed with the model. On a tokamak that includes whether a disruption mitigation system was in the loop; on a stellarator, whether the configuration and pellet train actually held the gradient; on NIF, whether the candidate owned the capsule, the hohlraum or the implosion symmetry . Fusion safety sits in the same verification: ITER has been a French Basic Nuclear Installation since 2012, and the demonstration goal is control of the plasma with negligible environmental consequence, which means inventory, dust, magnet energy and tritium, not a fission-core source term . The cost of a miss is a campaign that cannot restart until the person who can read the shot log is in the room, a breeding-blanket weld that nuclear-pressure rules will not accept, or a materials irradiation plan that qualifies the wrong spectrum. An accurate brief names the confinement scheme, the device, the pulse length and the evidence the seat must produce. Evaluation then tests that evidence and leaves the adjacent skill — conventional cryogenics, laser optics, fission radiochemistry — labelled as a gap, not as a substitute.
References
- ITER: In a Few Lines — ITER Organization. (accessed 2026-09-27)
- Summary of Presentation by Pietro Barabaschi, ITER Director-General (new baseline) — ITER Organization. (accessed 2026-09-27)
- Wendelstein 7-X sets World record for long plasma triple product — EUROfusion / Max Planck Institute for Plasma Physics. (accessed 2026-09-27)
- Achieving Fusion Ignition — Lawrence Livermore National Laboratory, National Ignition Facility. (accessed 2026-09-27)
- Europe ready to prove the fabrication of Test Blanket Modules — EUROfusion / Fusion for Energy. (accessed 2026-09-27)
- International materials facility IFMIF-DONES starts construction phase — EUROfusion. (accessed 2026-09-27)
- EU DEMO Safety and Balance of Plant design and operating requirements — EUROfusion. (accessed 2026-09-27)
- The IFMIF-DONES Irradiation Modules — Nuclear Fusion (IOP Publishing). (accessed 2026-09-27)
