Speaker
Description
ITER is planned to start operation using radiofrequency (RF) only, with a high fraction of heating power using electron cyclotron resonance heating (ECRH). Typically, most machines heat the plasma either with neutral beam injection (NBI) only or with a combination of RF and NBI heating. Between ECRH and NBI, the particle sources and the heating distribution between ions and electrons are different. These differences can influence the pedestal transport and stability, and the pedestal structure in H-mode. The goal of the work is to compare the pedestal structure and stability in TCV between plasmas heated only with ECRH and plasmas heated only with NBI, and to understand if and how the heating scheme affects the pedestal performance. Experiments were conducted in TCV with $I_p=170 \mathrm{kA}$ and $B_t=1.44\mathrm{T}$, for a high triangularity shape, without baffles and without seeding. At similar gas rate and absorbed power, it is found that NBI heating leads to larger pedestal top density $n_e^{\mathrm{ped}}$ and similar separatrix density $n_e^{\mathrm{sep}}$ compared to ECRH heating. Therefore, the ECRH pulse has larger $n_e^{\mathrm{sep}}/n_e^{\mathrm{ped}}$ than the NBI pulse, leading to an outward shift of the pressure pedestal. In consequence, the ECRH pulse has lower temperature and higher resistivity at the position of the maximum gradient compared to the ECRH pulse. Calculating the stability boundary with ideal and resistive MHD, it is found that the NBI pulse is well described by both ideal and resistive MHD, while the ECRH pulse is far from the ideal MHD stability boundary and well described by resistive MHD. These results are extended to the rest of the dataset, in which ECRH pulses have typically lower $n_e^{\mathrm{ped}}$ and similar $n_e^{\mathrm{sep}}$ compared to NBI pulses. The increase in $n_e^{\mathrm{sep}}/n_e^{\mathrm{ped}}$ is correlated with an outward shift of the pressure pedestal and an increase in the resistivity at the position of the maximum pressure gradient. MHD stability analysis show that ideal MHD is sufficient to describe the low $n_e^{\mathrm{sep}}/n_e^{\mathrm{ped}}$ NBI-heated pulses, while resistive MHD is necessary to describe the high $n_e^{\mathrm{sep}}/n_e^{\mathrm{ped}}$ ECRH-heated pulses, due to the higher resistivity.