massive gas injection (MGI) is a disruption mitigation technique in which a large quantity of noble gas is fired into a tokamak plasma so that its stored energy radiates over the whole wall instead of landing on a small area.
A high-pressure valve holding neon, argon or helium opens onto the plasma edge and delivers, in a few milliseconds, far more particles than the 📝plasma contains. The gas cools the edge and collapses the temperature profile, triggering the thermal quench on the operator's terms rather than the instability's: the stored thermal energy leaves as 📝impurity radiation spread across the entire 📝first wall, rather than as a concentrated heat pulse onto a few 📝plasma-facing components. The injected impurities also raise the electron density, restoring the collisional drag that suppresses 📝runaway electrons — a role it shares with the passive 📝runaway electron mitigation coil, which addresses the same threat by a different mechanism.
Mitigation is a compromise rather than a cure. Radiating the thermal energy cools the plasma and shortens the current quench, and a faster current quench induces larger 📝eddy current forces in the structure; too slow a quench leaves the plasma against the wall long enough for 📝halo currents to build. Every machine has a window between the two, and the injected quantity is tuned to land inside it.
High-pressure gas injection was first demonstrated as a 📝disruption mitigation method on 📝DIII-D in 2002 by 📝Dennis Whyte and co-authors, and was developed further on 📝ASDEX Upgrade, 📝JET and 📝Alcator C-Mod. Its limitation is penetration: neutral gas assimilates poorly and stalls in the outer plasma, and the shortfall grows with machine size, which is why 📝ITER selected shattered pellet injection instead. MGI should not be confused with 📝pellet injection, which delivers frozen fuel for density control, or with 📝impurity seeding, which trickles impurity continuously into the 📝divertor during normal operation; MGI is a single emergency discharge. 📝SPARC's 2020 physics basis plans massive material injection to reduce disruption loading, and the 2026 📝ARC strategy sets a target of one mitigated disruption per day behind it — both projections for machines that have not run.