A quantitative analysis of the effect of ELMs on H-mode thermal energy confinement in DIII-D

The desire to reach ignition in future tokamaks makes the energy confinement time (τE) a critical parameter. The most promising enhanced (over L-mode) confinement regime is the H-mode, discovered on ASDEX with neutral beam heating, and then confirmed with various auxiliary heating sources on numerous machines. The knowledge of how H-mode τE depends on different parameters is of crucial importance to the performance predictions for next generation devices. Inter-machine H-mode total and thermal energy confinement (τth) scaling, which are being utilized to predict ITER thermal energy confinement, have been created for discharges where the Edge Localized Mode (ELM) instability has not been present. Confinement scaling research has concentrated on this ELM-free H-mode phase mostly owing to the difficulty of characterizing ELM behaviour. To date, long pulse H-mode operation has only been achieved by utilizing ELMs to flush out impurities and prevent radiative collapse of the discharge. Unfortunately, accompanying the ELMs is a decrease of the plasma stored energy due to the expulsion of particles near the edge of the discharge resulting in a reduction of the steep edge electron density gradient. In order to predict ITER's H-mode τth in the presence of ELMs, an estimated 25 % confinement degradation factor has been applied to the ELM-free predictions. Our work, summarized in this paper, indicates that this 25 % reduction factor is too large and instead a value of approximately 15 % would be more appropriate. (author) 5 refs., 3 figs