Experimental measurements from DIII-D indicate that the energy distribution of avalanched runaway electrons (RE) is strongly skewed toward low energy, in disagreement with the exponential law from quasi-static avalanche theory. A new theoretical understanding is emerging which relates this observation with the frequently observed enhanced dissipation when E ~ Ecrit, and the companion observation that the RE current can be held constant out to the ohmic coil limit only by an applied field strength E/Ecrit > 5-10. The idea is that REs are formed under transient conditions where high electric field, E/Ecrit ~ 60, lasts for a short time, causing an accumulation of low energy RE which suffer pitch angle scattering in the ensuing plateau phase where the E field is much lower. A time-dependent solution of the Fokker Planck equation in the high Z limit is used to verify that the mature runaway current decay time on DIII-D can be of the order of 30 ms when E ~ Ecrit. Synchrotron radiation reaction force adds only a small contribution to the current decay. A steady-state analytical solution of the Fokker Planck equation was developed using a perturbation approach valid at E/Ecrit > 5-10 and (Z+1) » 1 to explain how the constant RE current can be achieved at E/Ecrit > 5-10. In going from the low to the high energy end of the spectrum, the model should predict the shift in the pitch-angle distribution from isotropic to highly anisotropic, as demonstrated in DIII-D experiments to be reported at APS.

The EPED pedestal model predicts that, for strongly shaped plasmas above a critical density, the pedestal height solution splits into multiple roots. These roots include the usual H-mode pedestal regime, and also a “Super H-Mode” regime that sits above it at substantially higher pedestal pressure. The Super H regime can be reached by dynamic optimization, starting at low density and then raising the density, to follow an access path in parameter space. Previous experiments on DIII-D gained access to the predicted Super H regime, and found good agreement with EPED pedestal height and width predictions in both the Super H and H regimes. A new DIII-D experiment has begun to explore both pedestal and core optimization, to take advantage of the potential for high performance offered by the very high Super H pedestal. The new experiment has achieved a record value of betaN for operation with a quiescent edge, simultaneously with high confinement (H98~1.4). Analysis is underway and results will be presented at the upcoming IAEA and APS/DPP meetings.

In collaboration with scientists from ASIPP-Hefei, full EFIT kinetic equilibrium reconstructions of a discharge from a DIII-D EAST-demonstration experiment found that magnetic measurements, together with kinetic profile and Motional Stark Effect measurements in a full kinetic equilibrium reconstruction can discriminate between the Sauter model and direct bootstrap current calculations with NEO in these high poloidal beta plasmas. The large edge bootstrap current in these high collisionality and high poloidal beta plasmas allows the use of magnetic measurements to clearly distinguish the Sauter model and NEO calculations. It was found that the Sauter model over-predicts the peak of the edge current density by about 30%, while the first-principles kinetic calculation with the NEO code is in close agreement. These results are consistent with other recent work showing that the Sauter model largely overestimates the edge bootstrap current at high collisionality.

Work on the turbulent transport code rCYCLO (see highlights from October 4, 2013 and July 19 2013 at Theory Weekly Highlights for July 2013 and Theory Weekly Highlights for October 2013) to test and quantify the high turbulence level breakdown of 5D gyrokinetics against the more fundamental 6D cyclokinetics has shown that the high turbulence breakdown of gyrokinetics does not account for the missing cold L-mode edge transport and low-frequency turbulence. Cyclokinetics follows the fast ion cyclotron motion and has no gyro-phase averaging approximation. It was conjectured that the high turbulence level nonlinear coupling of the low-frequency gyrokinetic drift modes to the high-frequency ion cyclotron modes might interrupt and suppress the gyro-averaging, thereby producing a higher level of transport and low-frequency turbulence. Instead it was found that, for stable ion cyclotron modes, cyclokinetic transport is always at least somewhat lower (not higher) than gyrokinetic. Furthermore, for parameters of physical interest, where the cyclotron frequency is significantly faster than the drift frequencies, gyrokinetic simulations are in good agreement with the more fundamental cyclokinetic simulations. Artificially driving the ion cyclotron modes unstable does not change this conclusion. Zhao Deng, a graduate student at Peking University working with Ron Waltz at GA will report these conclusions in an invited APS-DPP2014 invited talk.