Speaker
Description
In present-day magnetic fusion devices, hydrogen serves as a minority species for Ion Cyclotron Resonant Heating (ICRH). The efficiency of the heating scenarios depends on the hydrogen fraction in the plasma core that should be controlled within a few percents. Plasma-Facing Components (PFCs) release/trap hydrogen into/from the discharge, thereby complicating this control. Hydrogen increases neither the radiation nor the effective charge, but might dilute the fusion fuel, hence degrading the reactivity of burning plasmas.
This contribution systematically surveys the hydrogen content $n_H/(n_H+n_D)$ in WEST deuterium plasmas over six experimental campaigns and presents some techniques to influence it. Using a high-resolution visible spectrometer, we estimate $n_H/(n_H+n_D)$ in the scrape-off layer from the line ratio $H_\alpha/D_\alpha$ at $\lambda$~656nm. Three lines of sight aim at one inboard limiter, the inner and outer lower divertor. Using a residual gas analyzer, we also monitor the composition of the pumped gas. Although, not surprisingly, the various measurements do not fully agree quantitatively, they generally exhibit similar parametric variations.
In the restart phase after a vent, the plasma always appears naturally rich in hydrogen, while fueled with pure deuterium. Outgassing from the PFCs is suspected: ~1week of baking at 170°C did not empty all their hydrogen reservoirs. In 2024, we identified newly-installed semi-inertial bulk tungsten tiles on the inboard limiters as contributors to the hydrogen influx. Glow Discharge Cleaning (GDC) for 1h or more overnight proves efficient at reducing $n_H/(n_H+n_D)$ and was applied systematically before ICRH sessions early in the campaigns. However, GDC acted only temporarily and $n_H/(n_H+n_D)$ subsequently re-increased gradually.
Several weeks of plasma operation brought $n_H/(n_H+n_D)$ below the 10% suitable for ICRH. One could then control the minority fraction by dosing hydrogen puffs pulse by pulse.
Hydrogen re-appeared during long-pulse operation, when large amounts of energy were injected into the plasma. The hydrogen fraction rose both after some time within a discharge and between successive similar pulses, leading to reduced D-D neutron rates (quantitatively consistent with $D^+$ dilution) and eventually a loss of the density control. The hydrogen release decreased over successive experimental sessions with similar plasma scenarios. Outgassing is suspected from semi-inertial remote objects (not clearly identified) only heated by radiation or charge exchange neutrals.
We finally briefly discuss future reactors. In addition to the D-T mix (needing control), the minority species will be $^3He$, the requested minority fraction will be low and the fuelling efficiency will be lower than in present-day machines.