We present an ultrafast neural network (NN) model, QLKNN, which predicts core tokamak transport heat and particle fluxes. QLKNN is a surrogate model based on a database of 300 million flux calculations of the quasilinear gyrokinetic transport model QuaLiKiz. The database covers a wide range of realistic tokamak core parameters. Physical features such as the existence of a critical gradient for the onset of turbulent transport were integrated into the neural network training methodology. We have coupled QLKNN to the tokamak modelling framework JINTRAC and rapid control-oriented tokamak transport solver RAPTOR. The coupled frameworks are demonstrated and validated through application to three JET shots covering a representative spread of H-mode operating space, predicting turbulent transport of energy and particles in the plasma core. JINTRAC-QLKNN and RAPTOR-QLKNN are able to accurately reproduce JINTRAC-QuaLiKiz T i,e and n e profiles, but 3 to 5 orders of magnitude faster. Simulations which take hours are reduced down to only a few tens of seconds. The discrepancy in the final source-driven predicted profiles between QLKNN and QuaLiKiz is on the order 1%-15%. Also the dynamic behaviour was well captured by QLKNN, with differences of only 4%-10% compared to JINTRAC-QuaLiKiz observed at mid-radius, for a study of density buildup following the L-H transition. Deployment of neural network surrogate models in multi-physics integrated tokamak modelling is a promising route towards enabling accurate and fast tokamak scenario optimization, Uncertainty Quantification, and control applications.

A power-balance model, with radiation losses from impurities and neutrals, gives a unified description of the density limit (DL) of the stellarator, the L-mode tokamak, and the reversed field pinch (RFP). The model predicts a Sudo-like scaling for the stellarator, a Greenwald-like scaling, , for the RFP and the ohmic tokamak, a mixed scaling, , for the additionally heated L-mode tokamak. In a previous paper (Zanca et al 2017 Nucl. Fusion 57 056010) the model was compared with ohmic tokamak, RFP and stellarator experiments. Here, we address the issue of the DL dependence on heating power in the L-mode tokamak. Experimental data from high-density disrupted L-mode discharges performed at JET, as well as in other machines, are taken as a term of comparison. The model fits the observed maximum densities better than the pure Greenwald limit.

The 2014–2016 JET results are reviewed in the light of their significance for optimising the ITER research plan for the active and non-active operation. More than 60 h of plasma operation with ITER first wall materials successfully took place since its installation in 2011. New multi-machine scaling of the type I-ELM divertor energy flux density to ITER is supported by first principle modelling. ITER relevant disruption experiments and first principle modelling are reported with a set of three disruption mitigation valves mimicking the ITER setup. Insights of the L–H power threshold in Deuterium and Hydrogen are given, stressing the importance of the magnetic configurations and the recent measurements of fine-scale structures in the edge radial electric. Dimensionless scans of the core and pedestal confinement provide new information to elucidate the importance of the first wall material on the fusion performance. H-mode plasmas at ITER triangularity (H  =  1 at βN ~ 1.8 and n/nGW ~ 0.6) have been sustained at 2 MA during 5 s. The ITER neutronics codes have been validated on high performance experiments. Prospects for the coming D–T campaign and 14 MeV neutron calibration strategy are reviewed.

Since the installation of an ITER-like wall, the JET programme has focused on the consolidation of ITER design choices and the preparation for ITER operation, with a specific emphasis given to the bulk tungsten melt experiment, which has been crucial for the final decision on the material choice for the day-one tungsten divertor in ITER. Integrated scenarios have been progressed with the re-establishment of long-pulse, high-confinement H-modes by optimizing the magnetic configuration and the use of ICRH to avoid tungsten impurity accumulation. Stationary discharges with detached divertor conditions and small edge localized modes have been demonstrated by nitrogen seeding. The differences in confinement and pedestal behaviour before and after the ITER-like wall installation have been better characterized towards the development of high fusion yield scenarios in DT. Post-mortem analyses of the plasma-facing components have confirmed the previously reported low fuel retention obtained by gas balance and shown that the pattern of deposition within the divertor has changed significantly with respect to the JET carbon wall campaigns due to the absence of thermally activated chemical erosion of beryllium in contrast to carbon. Transport to remote areas is almost absent and two orders of magnitude less material is found in the divertor.

Non-inductive plasma current start-up by EC and RF power was carried out on the TST-2 device.Low frequency RF (21 MHz) sustainment was demonstrated, and the obtained high β p spherical tokamak configuration has similar equilibrium values as the EC (2.45 GHz) sustained plasma. Equilibrium analysis revealed the detailed information on three discharge phases: (i) In the initial current formation phase, linearity between the plasma current and the stored energy was confirmed. (ii) In the current jump phase, the initial closed flux surfaces cause a change in the current increasing rate, but the stored energy does not show such a change. (iii) The current sustained plasma is characterised by the fraction of the current inside the last closed flux surface to the total current, and the fraction seems to determine the ratio of the plasma current to the external vertical field. MHD instabilities often terminate the RF sustained plasma, but no such phenomenon was observed in the EC sustained plasma. IntroductionKey issues in spherical tokamak (ST) research are plasma current I p start-up and formation of the ST configuration without the use of a central solenoid (ohmic coil). Successful current generation, ST formation and sustainment have been achieved by injecting RF power (usually in the EC frequency range) to a configuration with a toroidal field and a weak vertical field. This scenario was developed in CDX-U using EC waves [1], and similar experiments were performed in the ST devices: LATE [2,3], TST-2@K [4], TST-2 [5], CPD [6], MAST [7]. A clear transition from open field line configuration to ST configuration, accompanied by a rapid increase in I p (so called current jump), was found in LATE. This phenomenon was also observed in TST-2 and in CPD. Experiments in these devices suggest that a higher vertical field B z and a higher EC power are preferable to achieve a higher I p as long as a current jump occurs. However, the current formation mechanism and the ST formation mechanism are still not clearly understood. Especially, equilibrium reconstruction was performed for only one case [4], and the time evolution and the variation in different operations remain unknown. The mechanisms for ST formation and the features of the plasma should be identified to extrapolate present results to next step ST devices. In TST-2, effects of various operational parameters are studied in [8]. It was found that the sustained current is roughly proportional to the vertical field strength B z , but dependences on other parameters are very weak. On the other hand, the initial current ramp-up rate depends on several parameters, and a scaling law was obtained. A current jump occurs when the initial current reaches a value proportional to B z , and the value is consistent with the condition to form closed flux surfaces. Recently, a low frequency RF source (21MHz) was used, and ST configuration was sustained by a non-EC heating method for the first time [9]. There is a threshold in the RF power, and the threshold for deuterium plasma was lower...