Simulation of pellet ELM triggering in low-collisionality, ITER-like discharges

Unmitigated heat loads from edge localized modes (ELMs) are intolerable in ITER. Therefore, ITER operation relies on multiple approaches to control ELM heat fluxes. One is pellet ELM pacing. Predicting the performance of it is critical in ITER, which is expected to operate in a regime with low-collisionality, peeling-limited pedestals. The hypothesis for pellet ELM triggering is that the local pressure increase in the expanding pellet cloud pushes the equilibrium over the ballooning pedestal stability limit. This suggests that the distance of the equilibrium's operational point from the ballooning branch of the pedestal stability boundary could play a critical role. M3D-C1 simulations of DIII-D low-collisionality discharges are used to determine the pellet size threshold for ELM triggering using linear simulations. 3D nonlinear M3D-C1 simulations, which are more realistic but computationally expensive, confirm that a pellet larger than the threshold size successfully triggers an ELM. The pellet size threshold is determined for multiple times during a DIII-D discharge and compared against ELM triggering events in the experiment. In all cases the simulated threshold separates events of successfully triggered ELMs from failed ones. Further linear simulations confirm that growth rates are significantly larger when an equilibrium operates closer to the ballooning branch compared to the original equilibrium for the same pellet size. However, the same pellet fails to trigger an ELM for an equilibrium that operates further away from the ballooning branch. This result suggests that pellet ELM triggering in ITER could require large pellets, which makes ELM pacing mass flow rates challenging for divertor operation.