Parameter dependencies of the separatrix density in low triangularity L-mode and H-mode JET-ILW plasmas

The midplane electron separatrix density, n _e,sep , in JET-ILW L-mode and H-mode low triangularity deuterium fuelled plasmas exhibits a strong explicit dependence on the averaged outer divertor target electron temperature, n _e,sep ∼ T _e,ot ^−1/2 . This dependence is reproduced by analytic reversed two point model (rev-2PM), and arises from parallel pressure balance, as well as the ratio of the power and momentum volumetric loss factors, (1 − f _cooling )/(1 − f _mom-loss ). Quantifying the influence of the (1 − f _cooling ) and (1 − f _mom-loss ) loss factors on n _e,sep has been enabled by measurement estimates of these quantities from L-mode density (fueling) ramps in the outer horizontal, VH(C), and vertical target, VV, divertor configurations. Rev-2PM n _e,sep estimates from the extended H-mode and more limited L-mode datasets are recovered to within ±25% of the measurements, with a scaling factor applied to account for use of $\left\langle {{T_{{\text{e,ot}}}}} \right\rangle $ , an averaged quantity, rather than flux tube resolved target values. Both the (1 − f _cooling ) and (1 − f _mom-loss ) trends and recovery of n _e,sep using the rev-2PM formatting are reproduced in EDGE2D-EIRENE L-mode-like and H-mode-like density scan simulations. The general lack of a divertor configuration effect in the JET-ILW n _e,sep trends can be attributed to a significant influence of main chamber recycling, which has been shown in the EDGE2D-EIRENE results to moderate n _e,sep with respect to changes in divertor neutral leakage imposed by changes in the divertor configuration. The unified n _e,sep vs $\left\langle {{T_{{\text{e,ot}}}}} \right\rangle $ trends can, however, be broken if large modifications to the divertor geometry (e.g. complete removal of the outer divertor baffle structure) are introduced in the model. The more pronounced high-field side high density region formation in the VH(C) configuration with reduced clearance to the separatrix does not appear to have a significant influence on the outer midplane separatrix and pedestal parameters when mapped to $\left\langle {{T_{{\text{e,ot}}}}} \right\rangle $ , although conditions at the inner midplane could not be assessed.