helium ash is the helium left in a 📝fusion 📝plasma after the 📝alpha particles born in 📝D-T reactions have given up their energy to the fuel, and it must be pumped out continuously or it accumulates and smothers the burn.
The name is a metaphor and a slightly misleading one: helium bears no resemblance to the residue of a fire, and the "ash" is not a waste product in any chemical sense — it is simply a nucleus with no further use. Each D-T reaction releases a 3.5 MeV alpha that stays magnetically confined and thermalises against the fuel ions and electrons, which is the mechanism of 📝alpha heating and the reason a 📝burning plasma can sustain itself. Once thermalised the helium is inert cargo. It occupies the pressure budget without fusing, so at fixed density every percent of helium displaces 📝deuterium and 📝tritium; it also raises 📝Zeff (effective charge), increasing 📝bremsstrahlung losses. Fuel dilution enters the fusion power quadratically through the fuel density, so a few percent of core helium is a meaningful tax and ten percent is a burn-ending one.
The engineering requirement is therefore a residence time, not a concentration: helium must be transported out of the core, neutralised at a 📝divertor target and removed by the pumps on a timescale of order ten 📝energy confinement times or shorter, which couples core transport, divertor neutral pressure and 📝impurity pumping into a single constraint. Helium ash is distinct from seeded impurities, which are injected deliberately to radiate power, and from wall impurities such as sputtered 📝tungsten, which arrive by accident — helium is the one contaminant a working reactor manufactures for itself, in direct proportion to the power it produces. For 📝ARC the 2026 exhaust study (DOI 10.1017/S002237782610155X) predicts that the divertor's high neutral pressure will hold core helium below two percent; that is a modelled projection for a plant not yet built, and no machine has yet run long enough in D-T to demonstrate steady-state ash control.