In 1995, noting the inadequacy of the IFMIF for blanket development, an interna­tional team headed by Abdou [23] proposed a high-volume plasma-based neutron source. A tokamak, naturally, was the best choice for a neutron source that

could cover large areas for blanket development. The group considered both superconducting and normal-conducting toroidal field coils, and it was found that coils made of a single turn rather than multiple windings of copper resulted in a smaller device. This is shown in Fig. 9.31. The major radius is only 80 cm and the toroidal field only 2.4 T; yet the plasma current is 10 MA and the neutron wall loading can be as large as 2 MW/m2. The last number is indicative of how well the device can duplicate the damage to materials in a reactor like DEMO. This is done well even though the volume neutron source (VNS) is only 0.5% of ITER in vol­ume, 2% in wall area, and 4% in fusion power produced. Significantly, the group did a risk-benefit analysis comparing the ways to obtain an 80% confidence level for DEMO to have, say, 50% availability, taking into account the mean time between failures and the time for repairs. Needless to say, operating ITER with VNS wins hands down over ITER alone. VNS also uses much less tritium in the process. The incremental cost is small: the total of capital cost and operating cost over the life of the machine is $19.6B for ITER and $24.4B for ITER plus VNS.