Runaway electron beam suppression using impurity flushing and large magnetohydrodynamic instabilities

Runaway electrons are considered to be the most challenging consequence of a disruption for future tokamaks like ITER. If runaway generation cannot be avoided, it is mandatory to have a second line of defense to suppress a fully formed runaway beam. Shattered Pellet Injection (SPI) is presently the baseline disruption and runaway mitigation method for ITER. Experiments using SPI on the JET tokamak showed that high-Z material such as neon or argon injected into a runaway beam with a low-density companion plasma accelerated the decay of the runaway current but did not prevent the final impact on the wall. Beams with high-density companion plasmas were found to be insensitive to SPI as they were with Massive Gas Injection (MGI).Conversely, the injection of a D2 SPI into a runaway beam led to an increase of the runaway current followed by a complete and abrupt dissipation of runaways. This dissipation did not generate any measurable heat loads on the plasma-facing components and was found to be related to two key ingredients. First, a large MHD instability possibly linked to a hollow current profile led to the loss of runaways on a large area through stochastization. This was supported by JOREK simulations which capture the experimental time scale of the termination event and the broad runaway deposition. The second ingredient was found to be the purity of the companion plasma. The D2 SPI flushes high-Z material used to trigger the disruption (argon) out of the companion plasma. If the high-Z concentration is low enough when the beam is dissipated by the large MHD burst, no runaway regeneration and therefore no conversion of runaway magnetic and kinetic energy occurs. The “D2 effect” was also found to be still efficient in conditions where the plasma moved vertically, mimicking the situation foreseen in ITER. The extrapolability of such a promising mitigation scheme for ITER will be discussed.