Reliability of Force Cooled Superconducting Magnets for Fusion

The latest superconducting magnets (SM) for fusion are mostly force-cooled, mainly because it allows reliable electrical insulation of the coils using vacuum pressure impregnation (VPI). SM of this type have many leads, feeders and coolant tubes, located in cryostat vacuum, which must sustain high voltages, induced on them by fast current changes. However vacuum loss can spoil their insulation. A few such cases occurred during the T-15 tokamak coils testing, initially having bare leads relying upon vacuum. But its loss generated a coil quench, a protecting current dump at high voltage, followed by breakdown and arc. Even leads insulation by Teflon and fiberglass tape wrap proved to be insufficient. Nevertheless, similar tape wrap insulation of leads and feeders (ILF) was used in EAST, KSTAR, SST-1 and W-7X. So far, seven breakdowns occurred during their coil tests at operating voltage
$\sim<$
3 kV. Breakdowns never initiated in the coils, but always on their leads, feeders and sensor lines, indicating that their insulation made by tape wrap were too weak. Instead of ILF improvement some projects undertake Paschen tests. These are planned as the baseline for ITER too. But these tests are valid for the coil with open insulated surface, but are not appropriate for the final tests, when insulation should not be exposed to vacuum. Up to now ILF final tests have been done in all devices at 10–21 kV, but only in good vacuum in spite of the fact that such tests could not guarantee safe operation in case of vacuum loss. We propose to increase ILF strength to the same level, as in the coils, using vacuum-tight grounded stainless steel casings filled up by VPI over magnet leads. This will provide reliable and easily testable solid insulation. Besides, casings would exclude He leaks, providing the second vacuum tight barrier over the ILF. Thus it would increase the magnet reliability and would make it possible to avoid the needs of all single coils test.