Non-linear MHD Simulations of Pellet Triggered ELMs in JET and ASDEX Upgrade tokamaks

ITER operation in its high fusion performance DT scenarios relies on the achievement of the H-mode confinement regime, which is expected to lead to the quasi-periodic triggering of ELMs (Edge Localized Modes). The energy fluxes associated with natural ELMs will produce excessive erosion and/or superficial surface damage on the plasma facing components. Controlled triggering of ELMs by the injection of small p ellets (small deuterium ice bodies) at frequencies significantly exceeding those of uncontrolled ELMs is one of the foreseen schemes to control ELM energy losses and divertor power fluxes in ITER. Although the technique has been demonstrated to decrease ELM size successfully in ASDEX Upgrade [1], JET [2], and DIII-D [3], uncertainties still remain regarding the physics understanding as well as of the consequence of its application, such as localised power loads associated with this technique [4]. Moreover, pellets may fail to trigger ELM for plasma scenarios in all metal wall ASDEX Upgrade which also requires better understanding of the underlying physical processes of the ELM triggering [8] Modelling of ELM triggering by pellet injection for JET (#84690) and ASDEX Upgrade (#29178) discharges has been carried out with the non- linear MHD code JOREK [5, 6]. JOREK allows to determine the energy and particle losses by the pellet triggered ELM. Regarding the JET discharge simulation, the pellet is injected after the natural ELM simulation which allows to compare the power deposition profiles of natural and pellet triggered ELM.