ELM Control and Pacing

Edge-localized modes are periodic edge bursts; control aims to make them small and frequent rather than large and rare.

What ELMs are

An edge-localized mode (ELM) is a repetitive instability at the plasma edge in high-confinement operation. Each ELM expels a burst of particles and energy from the edge pedestal. Large, infrequent ELMs deposit heat on plasma-facing components in short pulses that can exceed material limits, so they must be managed.

Two control philosophies

The first approach is suppression: change the edge so ELMs do not occur at all, typically using resonant magnetic perturbations from small coils. The second is pacing: deliberately trigger many small ELMs before a large one can build, spreading the same exhaust over more, gentler events. Pacing is often done with rapid pellet injection or small vertical position kicks.

Pacing as feedback

ELM pacing can be run as a control loop. A magnetic or radiation signal detects each ELM; if the natural ELM frequency drops below a target, the controller injects a pacing pellet to trigger one on demand. Holding the frequency high keeps each ELM small because less energy accumulates in the pedestal between events.

Coupling to other controls

ELM control interacts with fueling, because pellets used for pacing also add particles, and with divertor control, because the exhaust ELMs produce must be handled downstream. It also couples to the perturbation coils, which are shared with error-field correction, so actuator allocation must arbitrate between these uses.

In the Kronos program

The Hyperion breeder uses negative triangularity, a shape associated with access to good confinement without the strong edge pedestal that drives large ELMs. This shape choice reduces reliance on active ELM control, though the control system retains pacing and detection capability as a safeguard in scenarios where edge conditions approach the ELM regime.