Optimised ‘hybrid’ scenario H-mode plasmas for W radiation control in JET-ILW
Screening of high-Z (W) impurities from the confined plasma by the temperature gradient at the plasma periphery of fusion-grade H-mode plasmas was demonstrated for the first time in JET with the Be/W wall [Field A R et al, Nucl. Fusion 63(2023) 016028]. Through careful optimisation of the hybrid-scenario, deuterium plasmas with sufficient heating power (≳ 32MW), high enough ion temperature gradients at the H-mode pedestal top was achieved for the collisional, neo-classical convection of the W impurities to be directed outwards, expelling them from the confined plasma. Further experiments have been performed in JET during 2023, including in DT during the DTE3 campaign, to further optimise the peripheral impurity screening in such hybrid-scenario plasmas, as well as of our measurements of the W impurity fluxes between and during edge-localised modes (ELMs) based on fast bolometry measurements. A decrease in plasma current from 2.3MA to 2.1 MA, reduced the electron density and thereby increased the ion temperature at the H-mode pedestal top, resulting in stronger impurity screening behaviour. The optimised scenario could then be successfully transferred to operation in DT by a reduction in toroidal field, in order to compensate the lower L/H-threshold power in DT compared to D plasmas. Here, results of detailed analysis and modelling of the neo-classical W transport in four pulses from these experiments are presented, two in D at 2.3MA and 2.1MA plasma current and a matched pulse pair at 2.1MA in D and DT. It is found that the peripheral W screening behaviour from the bolometry analysis cannot be resolved with the modelling results. The outward convection just inside the pedestal top found in our earlier study could not not reproduced for these more recent pulses, when using the FACIT code [D Fajardo et al 2023 Plasma Phys. Control. Fusion 65 035021] to model the W transport. Possible reasons for this discrepancy are discussed, including potential deficiencies in our measurement technique and/or incompleteness of the neo-classical transport modelling.