Predictive models are developed for the temperature at the top at the edge of type 1 ELMy (edge localized mode) H-mode (high-confinement mode) plasmas. Theory-motivated models are used for the pedestal width and pressure gradient, while the pedestal density is obtained from experimental data in this study. The pedestal pressure gradient is assumed to be limited by the ballooning mode instability and is expressed in terms of the magnetic shear and geometrical factors. The effect of the bootstrap current, which reduces the magnetic shear in the steep pressure gradient region at the edge of the H-mode plasma, is included in the determination of the magnetic shear. Approaches for calculating the magnetic shear, combined with proposed models for the pedestal width, are used to determine the pedestal temperature. The computed pedestal temperatures are compared with more than 500 measured pedestal temperatures for type 1 ELMy H-mode discharges in four tokamaks. Some of the uncertainties in these results are discussed, and directions for future work to improve edge pedestal scalings are described.

The GLF23 and multi-mode core transport models are used along with models for the H-mode pedestal to predict the fusion performance for the International Thermonuclear Experimental Reactor, Fusion Ignition Research Experiment, and IGNITOR tokamak designs. Simulations using combinations of core and pedestal models have also been compared with experimental data for H-mode profiles in DIII-D, JET, and Alcator C-Mod. Power-independent (ballooning mode limit) and power-dependent pedestal scalings lead to very different predictions when used with the core models. Although the two drift-wave transport models reproduce the core profiles in a wide variety of tokamak discharges, they differ in their projections to burning plasma experiments for the same pedestal parameters. Differences in the core transport models in their response to the ion temperature gradient (i.e. their stiffness) and impact of the power dependence of the H-mode pedestal on fusion performance predictions are discussed.

The values of Q=(fusion power)/(auxiliary heating power) predicted for ITER by three different methods, i.e., transport model based on empirical confinement scaling, dimensionless scaling technique, and theory-based transport models are compared. The energy confinement time given by the ITERH-98(y,2) scaling for an inductive scenario with plasma current of 15 MA and plasma density 15% below the Greenwald value is 3.6 s with one technical standard deviation of ±14%. These data are translated into a Q interval of [7-13] at the auxiliary heating power P aux = 40 MW and [7 Ð 28] at the minimum heating power satisfying a good confinement ELMy H-mode. Predictions of dimensionless scalings and theory-based transport models such as Weiland, MMM and IFS/PPPL overlap with the empirical scaling predictions within the margins of uncertainty.

No abstract

Models for the height of the pedestal at the edge of H-mode plasmas (Onjun T et al 2002 Phys. Plasmas 9 5018) are used together with the Multi-Mode core transport model (Bateman G et al 1998 Phys. Plasmas 5 1793) in the BALDUR integrated predictive modelling code to predict the performance of the ITER (Aymar A et al 2002 Plasma Phys. Control. Fusion 44 519), FIRE (Meade D M et al 2001 Fusion Technol. 39 336), and IGNITOR (Coppi B et al 2001 Nucl. Fusion 41 1253) fusion reactor designs. The simulation protocol used in this paper is tested by comparing predicted temperature and density profiles against experimental data from 33 H-mode discharges in the JET (Rebut P H et al 1985 Nucl. Fusion 25 1011) and DIII-D (Luxon J L et al 1985 Fusion Technol. 8 441) tokamaks. The sensitivities of the predictions are evaluated for the burning plasma experimental designs by using variations of the pedestal temperature model that are one standard deviation above and below the standard model. Simulations of the fusion reactor designs are carried out for scans in which the plasma density and auxiliary heating power are varied.