Fuel recovery experiments with isotopic plasma wall changeover during long discharges in Tore Supra

Tritium (T) retention constitutes an outstanding constraint for ITER. It has been proposed that the end of the discharge could be used for reducing the amount of tritium trapped in the device by switching to He or H2 injection during the ∼200 s of plasmas following the burning phase (power and plasma current ramp down). Thanks to the long discharge capabilities of Tore Supra, long pulse experiments (> mn) have been carried out to evaluate the effectiveness of such a scenario in reducing the tritium inventory during plasma operations. Starting with the device operated only in D2, series of changeover experiments from D2 to He and from D2 to H2 have been carried out in Tore Supra. The results demonstrate that with He the amount of D recovered after 130 s is limited to 0.8 × 1022 D whilst no further gain is foreseen. From these experiments, it is demonstrated that He injection will not contribute to the drop of the tritium inventory in the vessel. In contrast, with H2 injection the amount of D recovered after 250 s is ∼4.2 × 1022 D with no limitation observed in the amount that could be removed from the vessel. The higher efficiency in removing D from the vessel by H2 injection compared to He is attributed to the H charge-exchange (CX) flux (four to six times larger than the He CX flux) allowing for a significantly stronger plasma wall interaction with carbon deposition and layer areas. In Tore Supra, since most of the D retention through co-deposition with eroded material (C) takes place in these areas, H plasmas result in a better removal efficiency of D(T) from these regions. These experimental observations are supported by the results obtained using the EIRENE code for evaluating both the ion and CX fluxes for He and H plasmas. Finally, the consequences of removing D(T) from the vessel for the next discharges are unfavourable for both the He and H2 removal methods. Indeed, in both cases, twice the amount of D(T) removed through the isotope exchange has to be re-injected since co-deposition of the re-injected D(T) will also take place in addition to the plasma wall isotope exchange. In these conditions, the low efficiency of the H2 gas injection for controlling the plasma isotopic ratio inhibits a recovery of the initial plasma isotopic ratio over a time scale in the range of 200 s.