Computing Library › AI Plasma Control
AI Plasma Control

Disruption Mitigation

When a disruption cannot be avoided, mitigation spreads the energy release and suppresses runaway electrons to protect the machine.

The role of mitigation

Mitigation is the last line of plasma-side defense. It does not save the discharge; it accepts the loss and acts to minimize damage. Its goals are to radiate the stored thermal energy over a large area rather than a small spot, to reduce the mechanical forces of the current quench, and to prevent or dissipate runaway electron beams.

Massive material injection

Kronos motion — lego machine

The main technique is to inject a large quantity of impurity material into the plasma very quickly, either as a gas jet or as a shattered frozen pellet. The injected material radiates the stored energy nearly isotropically, converting a localized heat dump into a distributed radiative one, and raises the density to suppress runaways.

The runaway problem

During the rapid current quench, the induced electric field can accelerate electrons to high energy, forming a runaway beam that, if it strikes the wall, damages it deeply. Mitigation aims to keep density high enough to collide these electrons out, or to deconfine the beam before it localizes. This is one of the hardest problems in disruption handling.

Timing

Mitigation must fire within milliseconds of the decision, which means the decision itself must be made by fast automated logic, not a human. This is why disruption prediction and mitigation are tightly coupled: the predictor's warning triggers the injection with enough lead time for the material to act.

In the Kronos program

The Hyperion breeder carries a mitigation capability as the fallback behind avoidance. Because avoidance is prioritized and the machine is designed to run away from disruptive boundaries, mitigation is intended as a rare last resort. Its triggering logic and effectiveness are studied in simulation ahead of hardware operation.