Integrated Modeling and Simulations of ITER Burning Plasma

Integrated Modeling and Simulations of ITER Burning Plasma Scenarios C. E. Kessel, R. V. Budny, K. Indireshkumar, D. Meade Princeton Plasma Physics Laboratory IEA Large Tokamak Workshop (W 60) Burning Plasma Physics and Simulation Tarragona, Spain July 4 -5, 2005

Integrated Scenario Development for ITER Advanced Operating Modes • Goals – Provide discharge simulations for self-consistency of plasma configurations – Identify impact of uncertain physics – Identify operating space within device/engineering limitations, and required upgrades – Define auxiliary systems requirements – Examine plasma control issues (equilibrium, stored energy, current profile) – Examine physics/engineering interfaces (PF coils, divertor) – Examine detailed physics modeling outside scope of 1. 5 D simulations, feeding back constraints for 1. 5 D – Provide experiment/theory comparison/verification within an integrated physics simulation

Integrated Modeling of ITER Hybrid Burning Plasma Scenarios • 0 D systems analysis to identify operating space within engineering contraints • 1. 5 D discharge simulations – – • • Energy transport (GLF 23) Heating/CD Free-boundary equilibrium evolution/feedback control Other control; stored energy, f. NI, etc. Energy transport experimental verification Ideal MHD analysis Offline heating/CD source analysis Offline gyrokinetic transport simulations (Budny) Fast particle effects and MHD (Gorelenkov) Particle transport/impurity transport Integrated SOL/divertor modeling Non-ideal MHD, NTM’s

0 D Systems Analysis Identifies Device Constraints for Scenario Simulations • ITER’s Primary Device Limitations That Affect Scenarios – Fusion power vs pulse length ----> heat rejection system • 350 MW for 3000 s • 500 MW for 400 s • 700 MW for 150 s ----> (maximum Pfus cryoplant limits) – Divertor conducted heat load, maximum > 20 MW/m 2, nominal 5 -10 MW/m 2 ----> allowable divertor heat load • Radiation from plasma core and edge, PSOL = (1 - fcorerad) Pinput • Radiation in divertor and around Xpt, Pcond = (1 - fdivrad) PSOL • Radiation distribution in divertor channel, impurities, transients – Volt-second capability ----> PF coil current limits • Approximately 260 -280 V-s – First wall surface heat load limit (not limiting for normal operation) – Duty cycle, tflattop/(tflattop + tdwell) ----> cryoplant for SC coils • Limited to about 25% What device upgrades are required for advanced operating modes, and are they major or minor upgrades?

Pursuing 1. 5 D Integrated Modeling of ITER with TSC/TRANSP Combination • TSC • Predictive • Freeboundary/structures /PF coils/feedback control systems • T, n, j transport with model or data coefficients ( , , D, v) • LSC for LH • Assumed source deposition for NB, EC, and ICRF: typically use off-line analysis to derive these Plasma geometry T, n profiles q profile TSC evolution treated like an experiment Accurate source profiles fed back to TSC • TRANSP • Interpretive • Fixed boundary Eq. Solvers • Monte Carlo NB and heating • SPRUCE/TORIC/CURRAY for ICRF • TORAY for EC • LSC for LH • Fluxes and transport from local conservation; particles, energy, momentum • Fast ions • Neutrals both codes have models for bootstrap current, radiation, sawteeth, ripple loss, pellet fueling, impurities, etc.

1. 5 D ITER Hybrid Simulations Integrate Transport, Heating/CD, and Equilibrium • Density evolution prescribed, magnitude and profile • 2% Be + 2% C + 0. 12% Ar for high Zeff cases • GLF 23 thermal diffusivities, no rotation stabilization, and with rotation stabilization (plasma rotation from TRANSP assuming = i) • Prescribed pedestal height and location amended to GLF 23 thermal diffusivities • Control plasma current, radial position, vertical position and shape • Plasma grown from limited starting point on outboard limiter, early heating required to keep q(0) > 1, keep Pheat < 10 MW • Control on plasma stored energy, PICRF in controller, PNB not in controller since it is supplying NICD

ITER Hybrid with GLF 23 Requires High n/n. Gr, High Tped to Reach N ≈ 3 IP = 12 MA BT = 5. 3 T INI = 7. 75 MA N = 2. 90 n/n. Gr = 0. 93 Wth = 450 MJ H 98 = 1. 56 Tped = 9 ke. V ∆ rampup = 150 V-s Vloop = 0. 042 V Q = 9. 43 P = 100 MW Paux = 53 MW Prad = 27 MW Zeff = 2. 25 q(0) < 1, ≈ 0. 9 r(q=1) = 0. 45 m li(1) = 0. 8 tflattop. V-s ≈ 3000 s t = 170 s t = 1500 s Shape control points

ITER Hybrid Simulation Shows Rapid q(0) Drop, V-s are Low, Long Core Relaxation

ITER Hybrid Scenario Needs High Tped for GLF 23 w/o and w Ex. B ITER expected to have Low vrot (≈ 1/10 vrot. DIII-D) Ti ≈ T e Low n(0)/<n> Present Expts have High vrot Ti > T e n(0)/<n> > 1. 25 Direct extrapolation from present Expts to ITER may be optimistic Continuing analysis with = f x i , higher n(0)/<n>, etc. vrot from TRANSP with = i

Using TRANSP Monte Carlo NB and SPRUCE Full Wave/FP ICRF Analysis to Model ITER Hybrid Sources IP = 12 MA, PNB = 33 MW, PICRF = 20 MW NINB Heating/CD Wth = 300 MJ Wth = 350 MJ INB = 2. 1 MA INB = 1. 8 MA ICRF Heating

Using JSOLVER/BALMSC/PEST 2, … to Analyze Ideal MHD Stability of ITER Hybrid discharges operate in a N window NNTM < Nn=1(no wall) Hybrid discharges have f. NI ≥ 40%, from NBCD on-axis and BS off-axis Hybrid discharges prefer q(0) > 1 or small sawtooth amplitude or possibly small r(q=1) Examine Porcelli sawtooth model in 1. 5 D simulations to determine the sawtooth response to small r(q=1), and local dq/dr and dp/dr

Efforts to Benchmark GLF 23 Transport in DIII-D 104276 Hybrid Discharge t = 1. 5 s TSC free-boundary, discharge simulation DIII-D 104276 data PF coil currents Te, i( ), n( ), v( ) NB data TRANSP Use n( ) directly TSC derives e, I to reproduce Te and Ti L-mode, i-ITB Turn on GLF 23 in place of expt thermal diffusivities Test GLF 23 w/o Ex. B and w EXB shear stabilization H-mode t = 5. 0 s

Energy Transport is at Center of Modeling/Projections for ITER • Whats wrong: – How is comparing to experiment improving our modeling? ? Error are say 20 -30% on Te and Ti profiles, and maybe 10% on stored energy – Multiple models, for example GLF 23 and MMM 95, will give reasonable agreement on any given experiment, how good should this be to believe a projection to ITER (remember ITER-EDA, comparison of transport models in Physics Basis 1999, what has changed? ? ) • What can we do: – Examine the critical features of transport; external rotation, Te/Ti, safety factor, density peaking, etc. and test these – Apply transport models to difficult expts. C-mod ITB, JT 60 -U high P, DIII-D/AUG/JET Hybrid and AT discharges…. – Apply transport model to entire discharge, not a single flattop time-slice – Consider what will be present in burning plasma device

Integrated Modeling of Burning Plasmas • • Integration includes feeding back numerous offline analyses to constrain the core 1. 5 D transport modeling – 0 D analysis for operating space limitations – Ideal and non-ideal MHD analysis – Source modeling benchmarks – Detailed SOL/divertor modeling – Particle and impurity transport (fueling) TSC/TRANSP is being used to improve the 1. 5 D simulations of burning plasma scenarios on ITER --- provides integration of full discharge freeboundary/feedback control and sophisticated source modeling/fast particle treatment Energy transport is at the center of 1. 5 D transport simulations – Project to ITER with consideration of difference from present expts. – Apply theoretical models to difficult experimental cases and for the entire discharge – Pedestal projections need to transition from empirical to theory based In some areas our integrated modeling needs more effort – Particle/impurity transport – SOL/divertor integrated into core evolution – Non-ideal MHD