Long Term Tritium Trapping in TFTR and JET

Tritium retention in TFTR [Tokamak Fusion Test Reactor] and JET [Joint European Torus] shows striking similarities and contrasts. In TFTR, 5 g of tritium were injected into circular plasmas over a 3.5 year period, mostly by neutral-beam injection. In JET, 35 g were injected into divertor plasmas over a 6 month campaign, mostly by gas puffing. In TFTR, the bumper limiter provided a large source of eroded carbon and a major part of tritium was co-deposited on the limiter and vessel wall. Only a small area of the co-deposit flaked off. In JET, the wall is a net erosion area, and co-deposition occurs principally in shadowed parts of the inner divertor, with heavy flaking. In both machines, the initial tritium retention, after a change from deuterium [D] to tritium [T] gas puffing, is high and is due to isotope exchange with deuterium on plasma-facing surfaces (dynamic inventory). The contribution of co-deposition is lower but cumulative, and is revealed by including periods of D fueling that reversed the T/D isotope exchange. Ion beam analysis of flakes from TFTR showed an atomic D/C ratio of 0.13 on the plasma facing surface, 0.25 on the back surface and 0.11 in the bulk. Data from a JET divertor tile showed a larger D/C ratio with 46% C, 30% D, 20% H and 4% O. Deuterium, tritium, and beryllium profiles have been measured and show a thin less than 50 micron co-deposited layer. Flakes retrieved from the JET vacuum vessel exhibited a high tritium release rate of 2e10 Bq/month/g. BBQ modeling of the effect of lithium on retention in TFTR showed overlapping lithium and tritium implantation and a 1.3x increase in local T retention