AbstractNew data on hydrogen plasma isotopes retention in beryllium and tungsten are now becoming available from various laboratories for conditions similar to those expected in the International Thermonuclear Experimental Reactor (ITER) where previous data were either missing or largely scattered. Together with a significant advancement in understanding, they have warranted a revisitation of the previous estimates of tritium inventory in ITER, with beryllium as the plasma facing material for the first-wall components, and tungsten in the divertor with some carbon-fibre-composites clad areas, near the strike points. Based on these analyses, it is shown that the area of primary concern, with respect to tritium inventory, remains codeposition with carbon and possibly beryllium on the divertor surfaces. Here, modelling of ITER divertor conditions continues to show potentially large codeposition rates which are confirmed by tokamak findings. Contrary to the tritium residing deep in the bulk of materials, this surface tritium represents a safety hazard as it can be easily mobilised in the event of an accident. It could, however, be possibly removed and recovered. It is concluded that active and efficient methods to remove the codeposited layers are needed in ITER and periodic conditioning/cleaning would be required to control the tritium inventory and avoid exhausting the available fuel supply. Some methods which could possibly be used for in-situ cleaning are briefly discussed in conjunction with the research and development work required to extrapolate their applicability to ITER.
TFTR tritium inventory analysis