Advanced divertor configurations for DEMO

In order to demonstrate an attractive operational regime, DEMO will need to operate with high power crossing the separatrix (Psep >= 150 MW), low peak power flux at the targets ( 5-10 MW/m2) and target temperature sufficiently low to limit W sputtering (T < 5 eV). The last two conditions will require most likely that plasma detachment should be achieved at all target plates. The baseline power exhaust strategy for DEMO extrapolates the ITER solution, employing a Single Null divertor (SN) and relying on extrinsic impurities injection in order to enhance SOL radiation and limit the power reaching the solid target via the basic plasma conduction/advection channels. However, it is expected that the DEMO SOL physics should be more challenging than in the case of ITER, because of the larger Psep/R ratio (~ 19 MW/m for DEMO vs. ~ 17 MW/m for ITER) and the smaller power decay length (1 - 3 mm estimated for DEMO at the midplane, vs. ~ 5 mm for ITER), which makes uncertain the success of the baseline power exhaust strategy. Advanced configurations such as double null (DN), X-divertor (XD) and Super-X divertor (SX) are considered as a back- up solution for the power exhaust problem in DEMO, in case the single-null (SN) baseline strategy, already chosen for ITER, does not extrapolate favorably to the DEMO reactor conditions. These alternative solutions aim at lowering the target temperature and increasing the power deposition area by means of a longer connection length (XD, SX), proper flaring of the magnetic field lines in front of the targets (XD), placing the outer target at the outermost possible position to increase the toroidal revolution length (SX), or by increasing the number of targets available for power exhausts (DN). A previous comparative study of advanced divertor configurations with Ar impurity seeding led with the SOLPS code showed the clear advantage of XD and SX over the SN in reaching low target temperature and target power densities; it also pointed out how, independently of the configuration considered, it was not possible to obtain acceptable target conditions unless the plasma density was sufficiently high and a considerable amount of Ar impurity was present in the divertor. This work extends the previous study in several aspects: we increase the number of considered impurities, by analyzing in details the effect of Xe, N and Kr in addition to Ar. The analysis of the SX configuration is now performed with reference to the newest equilibria, made available from 2017. First results confirm the significant advantage of advanced divertor configurations over the baseline SN, while it appears that a combination of more than one radiator is preferable over the single-impurity solution previously considered. In particular, a combination of lower-Z and higher-Z radiators (e.g. N and Xe) allows some flexibility to try setting independently the radiation level in the SOL and the confined plasma. Finally, the DN configuration has been added to the pool of geometries analyzed with the SOLPS code: we discuss the preliminary results currently available.