Current tritium production might not be enough for all future fusion research reactors. Different approaches for tritium production have been studied in the past, one of which was tritium production using the accelerator-driven subcritical systems. This idea was dismissed in the 1990s as uneconomical when compared to using existing commercial light water reactors. This paper presents changes to the basic idea, mainly the use of a molten spallation target and molten lithium breeding volume. This advanced design is described, optimized for tritium yield using the MCNP 6.2.0 code, and compared between different accelerators.

The optimized configuration consists of a 1-GeV, 200-mA proton accelerator, a molten Pb-Bi eutectic spallation target with a length of 60 cm and a diameter of 75 cm, and molten lithium breeding volume with dimensions of 500 cm in length and 900 cm in diameter. As calculated, the annual production of the proposed accelerator-driven tritium production system could be as high as 350 g of tritium with the optimized configuration.