Massive Gas Injection Systems for Disruption Mitigation on the DIII-D Tokamak

Injection of massive quantities of deuterium or noble gases (>10
²²
molecules) has proven to be very effective at mitigating the deleterious effects of disruptions in the DIII-D tokamak. Both the heat load to the divertor and the halo current forces to the first wall were reduced by more than a factor of four. Total electron densities (free and bound) of ~10
²¹
m
^-3
have been achieved, close to densities required to prevent avalanche multiplication of runaway electron currents during the fast plasma current ramp down that injection produces. Two gas injection configurations used in experiments in the period 2001-2005 will be described. Both configurations use a fast operating (open/close~0.5 ms) solenoid valve with an orifice diameter of 4 mm. The maximum flow rate in helium is 5times10
⁴
Pa m
³
/s at a reservoir pressure of 7 MPa. A new valve with an orifice diameter of 20 mm will be tested when DIII-D resumes operation in 2006. The new valve, with a theoretical flow of 25 times the original valve, has flow capabilities that approach those required for ITER. Calculations show that a set of four such valves can reach the no-avalanche density in ITER in ~0.25 t
_CQ
where t
_CQ
, ~30 ms, is the 'fastest-possible' plasma current quench time. While there are still-open questions about the mechanisms whereby injected neutral gas and free and bound electrons penetrate (or are mixed) into the core of a high-temperature plasma, it is clear that massive gas injection can easily and reliably provide sufficient mass flow to minimize disruption damage in large reactor-scale tokamaks. The relatively long current decay times inherent in such devices will make it feasible for injection valves to be located external to the blanket modules, in regions of relatively low magnetic field. It will be necessary, however, to provide a relatively high-conductance 'injection tube' path from the valve to the plasma surface. Trade offs of many system design factors including gas flow rise time, neutron streaming and damage to the valve seat and actuator and valve and driver reliability and in-situ maintainability will have to be made to determine the most suitable configuration. The principal open physics question is what is the mixing mechanism which transports the impurities to the plasma interior and how fast and effective will it be in a reactor-scale plasma