Experimental study on the role of the target electron temperature as a key parameter linking recycling to plasma performance in JET-ILW

Changes in global and edge plasma parameters (H (98(y,2)), dimensionless collisionality nu *, core density peaking, separatrix density n (e,sep)) with variations in the D-2 fueling rate and divertor configuration are unified into a single trend when mapped to ⟨T (e,ot)⟩, the spatially averaged spectroscopically derived outer target electron temperature. Dedicated JET with the ITER-like wall (JET-ILW) experiments in combination with an extended JET-ILW database of unseeded low-triangularity H-mode plasmas spanning a wide range of D-2 fueling rates, I (p), B (t) and heating power have demonstrated the importance of ⟨T (e,ot)⟩ as a key physics parameter linking the recycling particle source and detachment with plasma performance. The remarkably robust H (98(y,2)) trend with ⟨T (e,ot)⟩ is connected to a strong inverse correlation between ⟨T (e,ot)⟩, n (e,sep) and nu *, thus directly linking changes in the divertor recycling moderated by ⟨T (e,ot)⟩ with the previously established relationship between nu *, core density peaking and core pressure resulting in a degradation in core plasma performance with decreasing ⟨T (e,ot)⟩ (increasing nu *). A strong inverse correlation between the separatrix to pedestal density ratio, n (e,sep)/n (e,ped), and ⟨T (e,ot)⟩ is also established, with the rise in n (e,sep)/n (e,ped) saturating at ⟨T (e,ot)⟩ > 10 eV. A strong reduction in H (98(y,2)) is observed as ⟨T (e,ot)⟩ is driven from 30 to 10 eV via additional D-2 gas fueling, while the divertor remains attached. Consequently, the pronounced performance degradation in attached divertor conditions has implications for impurity seeding radiative divertor scenarios, in which H (98(y,2)) is already low (similar to 0.7) before impurities are injected into the plasma since moderate gas fueling rates are required to promote high divertor neutral pressure. A favorable pedestal pressure, p (e,ped), dependence on I (p) has also been observed, with an overall increase in p (e,ped) at I (p) = 3.4 MA as ⟨T (e,ot)⟩ is driven down from attached to high-recycling divertor conditions. In contrast, p (e,ped) is reduced with decreasing ⟨T (e,ot)⟩ in the lower I (p) branches. Further work is needed to (i) clarify the potential role of edge opacity on the observed favorable pedestal pressure I (p) scaling; as well as to (ii) project the global and edge plasma performance trends with ⟨T (e,ot)⟩ to reactor-scale devices to improve predictive capability of the coupling between recycling and confined plasma fueling in what are foreseen to be more opaque edge plasma conditions.