Design study of blanket structure based on a water-cooled solid breeder for DEMO

Someya, Yoji; Tobita, Kenji; Uto, Hiroyasu; Tokunaga, Shinsuke; Hoshino, Kazuo; Asakura, Nobuyuki; Nakamura, Makoto; Sakamoto, Yoshiteru

Blanket concept with simplified interior for mass production has been developed with a mixed bed of LiTiO and BeTi pebbles, a coolant condition of 15.5 MPa and 290-325C and cooling tubes only without any partitions. A neutronics analysis ensured the blanket concept meets a self-sufficient supply of tritium. However, this concept is vulnerable to the inner pressure. A plant availability for DEMO may drop to a lower value, because a potential of resume operations after an accident such as a coolant leakage in blanket is not considered. The blanket design will be revisited for the availability. Considering the continuity with the ITER-TBM option of Japan and the engineering feasibility of fabrication, our design study focuses on a water-cooled solid breeding blanket using the mixed pebbles bed. A breakage of the blanket casing should be avoided not to contaminate the plasma chamber with water and breeding materials. A water-cooled solid blanket with inner pressure tightness is estimated by the ANSYS code. As a results, the pressure tightness of 8 MPa (water vapor pressure at 300C) can be compatible with the self-sufficient production of tritium when the blanket is as thick as about 0.9 m and the ribs are arranged in the radial direction. Therefore, the blanket concept with pressure tightness of 8 MPa is adopted with depressurization system as which a tritium recovery system such as helium purge-gas line is posteriorly arranged in blanket to serve. On the other hand, a handling of decay heat is a serious problem at an accident such as LOCA. Coolant flow is divided into the blanket to secure heat removal for the safety. Finally, the blanket segmentation with the shape and dimension of blanket and routing of coolant flow has also been proposed. Moreover, overall TBR is estimated with torus configuration based in the segmentation using three-dimensional MCNP calculation.