Edge-localised modes (ELMs) can produce plasma eruptions resembling small solar flares.

Although tokamaks appear to be one of the most promising approaches to achieving useful thermonuclear fusion, they still face many physical and technical challenges. The results of the first tokamaks, especially the Soviet T-3, which reached plasma temperatures of about 10 million kelvin in 1968, were remarkable. However, it soon became clear that developing a practical fusion power plant would require solving far more complex problems than originally expected.

Plasma Purity

Radiative losses caused by impurities were reduced by improving vacuum conditions, choosing suitable plasma-facing materials, baking and conditioning the vacuum vessel, and developing divertors to remove impurities from the plasma edge.

Turbulence

Researchers succeeded in heating the plasma, but this revealed another serious problem. Experiments on PLT and other tokamaks showed that increasing the heating power could degrade energy confinement in the low-confinement regime. Steep gradients in plasma temperature and density can drive turbulence, which enhances particle and heat transport and allows energy to escape from the confined plasma.

Plasma can exhibit many types of turbulence and instability that degrade confinement or, in severe cases, terminate the discharge prematurely. Such a sudden loss of plasma confinement is called a disruption. During a disruption, plasma thermal energy and current can be lost very rapidly, producing intense heat loads and large electromagnetic forces on the vacuum vessel and other components.

During disruptions, a strong induced electric field can accelerate a small fraction of electrons to relativistic energies, creating so-called runaway electrons. If these strike plasma-facing components, they can cause severe local damage. Such effects were observed, for example, on the TFR tokamak in the 1970s. Although runaway electrons can pose a serious threat to the machine, they do not represent an uncontrolled fusion-energy release because the total amount of plasma in a tokamak is very small.

Edge Localised Modes

Collisions can shift the gyro-orbits of charged particles, causing diffusion across magnetic field lines.

In 1982, an important breakthrough occurred on the ASDEX tokamak, which achieved a regime with much better confinement. This became known as high-confinement mode (H-mode), while the previous regime was called low-confinement mode (L-mode). In H-mode, an edge transport barrier forms, producing steep density and temperature gradients and reducing turbulent transport. Energy confinement times can typically be about two to three times longer than in L-mode.

The improvement brought by H-mode also revealed a new problem: edge-localised modes (ELMs). These are quasi-periodic instabilities of the plasma edge that can eject bursts of particles and energy towards the divertor and vessel walls. Large ELMs could produce unacceptable transient heat loads in a fusion reactor. For this reason, researchers are developing methods to suppress or mitigate ELMs and are also studying naturally ELM-free high-confinement regimes such as Quiescent H-mode.

ELMs can contribute to particle and impurity exhaust, but large ELMs are undesirable because of their intense transient heat loads. Current research therefore aims either to replace large ELMs with smaller, more frequent events or to suppress them altogether while maintaining adequate impurity control. Techniques include resonant magnetic perturbations, pellet pacing and naturally ELM-free operating regimes.

Plasma Volume

Scientific visualisation of a Rayleigh-Taylor instability, illustrating how instabilities can mix different regions of a plasma.

Larger plasma volumes generally offer more favourable energy-confinement scaling and make it easier to achieve reactor-relevant plasma conditions. A future fusion power plant must also provide substantial space between the plasma and the superconducting magnets for the vacuum vessel, breeding blanket, neutron shielding and cooling systems.

Magnetic Field

Achieving strong and stable magnetic confinement also presents major engineering challenges. Copper magnets suffer from significant resistive heating and power consumption, so many large tokamaks use superconducting magnets. Niobium–titanium is suitable for moderate fields, while niobium–tin is used where higher magnetic fields are required, despite its brittleness. Low-temperature superconductors such as NbTi and Nb₃Sn are typically operated at around 4 K. High-temperature superconductors such as REBCO can operate at significantly higher temperatures and retain high current density in very strong magnetic fields, making them attractive for future compact fusion devices.

Pulse Regime

Long-pulse or steady-state operation is highly desirable for a fusion power plant. Conventional inductive tokamak operation is inherently pulsed because the plasma current is driven by the changing magnetic flux of the central solenoid. Non-inductive current-drive methods using radio-frequency waves or neutral beams, together with the self-generated bootstrap current, are therefore being developed to extend discharge duration and enable steady-state operation.

Major progress has been made in extending tokamak discharges: modern superconducting devices have sustained plasmas for many minutes, and EAST has maintained an H-mode edge for more than 1,000 seconds. Nevertheless, no tokamak has yet reached scientific breakeven, where the fusion power produced equals the external heating power supplied to the plasma. Research therefore continues to improve confinement, plasma stability and long-pulse operation. Scientists have developed sophisticated models of how plasma is expected to behave in the much larger volume of ITER. However, as fusion research has repeatedly shown, experiments may still reveal unexpected plasma behaviour and entirely new operating regimes.