(TH/P3-2) Runaway Electron Generation during Plasma Shutdown by Killer Pellet Injection

K. Gal1), H. Smith2), T. Fülöp2), P. Helander3)
 
1) KFKI-RMKI, Associaton EURATOM, Budapest, Hungary
2) Department of Radio and Space Science, Chalmers University of Technology, Sweden
3) EURATOM/UKAEA Fusion Association, Culham Science Centre, United Kingdom

Abstract.  Disruptions in high-current tokamaks can be accompanied by runaway electron generation. As the plasma cools down quickly in the thermal quench of a disruption, a large toroidal electric field is induced which accelerates some electrons to relativistic energies which severely damage the first wall on impact. To mitigate disruption problems it has been proposed that ``killer" pellets could be injected into the plasma in order to safely terminate the discharge. Killer pellets enhance radiative energy loss and thereby lead to rapid cooling and shutdown of the discharge. However, pellets may also cause increased runaway production, as has been observed in several tokamaks. During pellet injection there are two competing effects that may affect runaway generation: the pellet increases the electron density and therefore suppresses acceleration of runaway electrons because of higher collisional friction, but the pellet also increases the plasma resistivity due to cooling and higher Z, leading to an increased toroidal electric field and more runaway acceleration. There are three runaway generation mechanisms: the ``Dreicer" mechanism, caused by Fokker-Planck diffusion into the runaway region of velocity space; a "burst" of runaway production caused by the rapid cooling of the bulk plasma; and the ``avalanche" mechanism, where existing runaways produce new ones through knock-on collisions with thermal electrons. In high-current tokamaks, the role of the two first mechanisms is to provide a ``seed" for the runaway avalanche. In the present work, two aspects of the runaway dynamics in connection with killer pellet induced fast plasma shutdown are considered. First we give a criterion for whether runaway bursting is more important than Dreicer production, by solving the kinetic equation in a cooling plasma and estimating the number of electrons in the runaway region. Second, we determine the post-disruption runaway current profile by solving the equation for runaway production coupled to an equation for the evolution of the toroidal electric field. To provide the evolution of the background plasma density and temperature we rely upon a pellet ablation code. In this way we can investigate the effect of varying pellet size and composition.

Full paper available (PDF)