Computing Library › AI Plasma Control
AI Plasma Control

Disruption Avoidance

Avoidance keeps the plasma away from the boundaries where disruptions occur, always preferred over cleaning up after one.

What a disruption is

A disruption is a sudden, uncontrolled loss of plasma confinement in which the stored thermal and magnetic energy is released rapidly. It can deposit heat on surfaces, exert large forces on the structure, and generate high-energy runaway electrons. Preventing disruptions is a central goal of plasma control.

Avoidance versus mitigation

Kronos motion — control room

There are two responses. Avoidance steers the plasma away from the conditions that cause disruptions, keeping it inside a safe operating region. Mitigation accepts that a disruption is unavoidable and acts to make it less damaging. Avoidance is always preferred, because a mitigated disruption still stresses the machine; mitigation is the fallback when avoidance fails.

Proximity monitoring

Avoidance needs a measure of how close the plasma is to a disruptive boundary: the density limit, the beta limit, the safety-factor limit, or a growing locked mode. Real-time proximity indicators, sometimes combined into a single disruptivity estimate, tell the controller when to back off performance to restore margin.

Actions

In the Kronos program

For the Hyperion breeder, disruption avoidance is the first priority of the off-normal layer, which monitors proximity to the density, beta, and stability limits and to locked-mode onset. When margin shrinks, it backs off performance or terminates safely rather than risk a disruption. The mitigation system stands behind it as a last resort. Both are exercised in the flight simulator.