Edge pedestal physics and its implications for ITER

1
Edge pedestal physics and its implications for
ITER
Y.Kamada1, A.W.Leonard2, G.Bateman3, M.Becoulet4,
C.S.Chang5, T.Eich6, T.E.Evans2, R.J.Groebner2,
P.N.Guzdar7, L.D.Horton6, A.Hubbard8, J.W.
Hughes8, K.Ida9, G.Janeschitz10, K.Kamiya1,
A.Kirk11, A.H.Kritz3, A.Loarte12, J.S.Lonnroth21,
C.F.Maggi6, R.Maingi13, H.Meyer11,
V.Mukhovatov14, T.Onjun15, M.Osipenko16,
T.H.Osborne2, N.Oyama1, G.W.Pacher17,
H.D.Pacher18, A.Y.Pankin3, V.Parail11,
A.R.Polevoi14, T.Rognlien19, G.Saibene12,
R.Sartori12, M.Shimada14, P.B.Snyder2,
M.Sugihara14, W.Suttrop6, H.Urano1, M.R.Wade2,
H.R.Wilson20, X.Q.Xu19, M.Yoshida1, and the ITPA
Pedestal Edge Physics Topical Group 1 Japan
Atomic Energy Agency, 2 General Atomics, 3 Lehigh
Univ.,4 Association Euratom-CEA, 5 New York
Univ., 6 Association Euratom-IPP, 7 Univ.
Maryland, 8 MIT Science and Fusion Center, 9NIFS,
10 FZK-PL-Fusion, 11 Association Euratom-UKAEA,
12 EFDA-CSU, 13 Oak Ridge National Laboratory, 14
ITER International Team, 15 Thammasart Univ., 16
Kurchatov Institute, 17Hydro-Quebec, 18INRS, 19
Lawrence Livermore National Laboratory, 20 Univ.
of York, 21 Association EURATOM-Tekes,
Remarkable progress has been achieved by
integrating experimental results obtained in
single- and inter-machine experiments (C-Mod,
AUG, DIII-D, JET, JFT-2M, JT-60U, MAST and NSTX)
with theoretical progress.
2
Edge Pedestal Key area determining integrated
performance of ITER
Edge Pedestal determines
core confinement as boundary condition b-limit
through p(r)j(r) heat/particle pulse to Div.

• Present experiments indicate ELM heat flux could
  be a problem for ITER.
• ITER plasma performance is determined by the
  pedestal height.
• How the Pedestal Structure is determined?
• How the type I ELM cycle evolves?

gt How control?
3
Outline
ITPA-PEP

• Identification of the processes determining the
  pedestal structure
• Understanding of the type I ELM cycle
• Development evaluation of small / no ELM
  regimes
• Type I ELM mitigation techniques
• Development of integrated prediction codes.
• Summary

4
Parameter Linkage Determined
ITPA-PEP
5
Pedestal height determines Q in ITER
ITPA-PEP
ITER fusion gain predicted by theory based
transport models depends strongly on Tped (pped).
Reason of uncertainty is large scatter in Tped
V. Mukhovatov, PPCF 45 (2003) A235
6
Temperature Width plasma transport Density
width plasma neutral transport
ITPA-PEP
Density width DIII-DJET neutral penetration
explains. AUGC-mod width constant.
Temperature Width determined by the magnetic
structure and non-dimensional parameters. (
Multi-machine comparison exp.)
Fenstaermacher (NF2005)
7
Plasma rotation ripple affect pedestal height
ITPA-PEP

• JET JT-60U comparison with matched 'absolute
  parameters
• pedestal pressure JET gt JT60U gt possible
  reason TF ripple.
• Ferritic steel tile installation in JT-60U
• Both co-directed shift of Vt reduction of
  filed ripple improves pedestal.
• (wider pedestal width higher pedestal
  pressure)

Urano EX5-1
Saibene (NFsubmitted)
Thermal ion transport enhanced by ripple has been
proposed (Parail THP8-5)
8
Type I ELM Trigger Crash identified
ITPA-PEP
Type I ELM Trigger The P-B model has been
confirmed on a number of tokamaks. ELM crash
dynamics 2-3D structure poloidal asymmetry of
erosion inside the separatrix and helical
filament structure expanding into the SOL.
Type I P-B critical JET, AUG, C-Mod, JT-60U,
MAST, NSTX
Saarelma (POCF2005)etc.
9
ELM Energy Release dependence clarified
ITPA-PEP

• ?WELM/Wped
• increases with decreasing ?, (multi-machine)
• 15- 20 at the expected ? in ITER.
• tends to increase with triangularity(AUG),
• with increasing co-directed rotation(JT-60U).
• ? -dependence of the efflux is large for
  conductive loss
• and small for the convective loss (JET, DIII-D)

Energy release at an ELM is carried partly by the
filaments lt20. Main loss channel has not
been identified. One possibility is that the
filaments tear the closed flux surfaces allowing
parallel transport.
Loarte (PPCF 2003), Kamiya (PPCF 2006)
10
ELM crash Transport recovers quickly
ITPA-PEP

• Structure of the edge Er shear is suddenly
• broken by the ELM crash (DIII-D)
• After an ELM crash, recovery of the pedestal
• rotation profile takes place faster than
• recovery of the pedestal pressure (DIII-D
  JT-60U).
• Then the edge pressure recovers in the time
  scale of the inter-ELM transport
• inter-ELM close to neoclassical (JT60U),
  still anormalous remains(AUG)

Er Well Flattened at ELM
Wade (PoP 2005)
11
Peeling-Ballooning Model explaines ELMs
Successfully
ITPA-PEP
Nonlinear explosive evolution of the filaments
reproduced numerically by using the 3D
electromagnetic two-fluid code BOUT (Snyder, PoP
2005)
Type I ELM onset The P-B model has been
confirmed in many tokamaks A non-linear
analytic theory, valid early in the evolution of
a ballooning mode, predicts that filamentary
structures should grow explosively. A number of
codes support this general result. Sufficient
edge current density is required to cause the
filaments to be ejected outwards towards the wall
(otherwise they erupt inwards, towards the plasma
core) (H. Wilson TH4-1Rb)
12
Small/no ELM regimes accessibility identified ,
reproduced by inter-machine comparison
ITPA-PEP

• DWELM/Wped lt5 .
• All small/no ELMregimes
• reproduced in multiple devices
• Except for Grassy and type V,
• the edge fluctuations
• enhance particle transport,
• and the pedestal pressure is
• below the type I ELM limit.

Oyama (PPCF 2006)
13
Small/no ELM Regimes need to be extended to ITER
regime
ITPA-PEP
Only Grassy ELM and QH-mode achieved at n close
to ITER. Better understandings of the effects of
n, plasma shape and driving mechanisms of the
edge fluctuations are needed.
Plasma rotation seems to be important. DIII-D
etc. CTR rotation produces QH mode. JT-60U
grassy ELM freq. increases linearly up to 1400 Hz
with CTR rotation. Even at no-rotation
400 Hz
Oyama (PPCF, submitted)
Oyama (PPCF 2006)
14
ELM control with pellet pace making
ITPA-PEP
AUG fELM fpellet, DWELM decreases with fELM.
(Lang NF 2004)

• ? Natural ELM frequency in ITER
• fELM ?WELM 0.4 Ploss ? Ploss?80MW, ?WELM
  20 MJ, ? fELM 1.6 Hz
• critical value for sublimation of CFC
• ?T?W / S / ?0.5 ? 40 MJ m-2s-0.5
• ? fELMp?1.6?(2-3) ?3-5Hz

(Polevoi NF 2005)
Issue confinement degradation 1015
reduction when fELM is increased by a factor of
2-3 xfELM, (AUG)
15
ELM control with Resonant Magnetic Perturbation
ITPA-PEP
DIII-D elimination of Type I ELMs at ITER's n
by applying external field. RMP increases
particle transport. Issuers compatibility of
operation at high ne
Evans (PRL 2004)
For ITER Required ergodization for ELM
suppression can be realized with In-vessel coil
(20kA), Ex-vessel coil (150kA) or external
coils (400kA). Effect of generated island in
core and impact on engineering design need
further study.
Becoulet IT P1-29
16
Progress of Integrated Modeling
ITPA-PEP
The modeling capability for integrating core
transport, pedestal (NC PB), SOL and divertor
regions has achieved remarkable progress.
LEHIGH-JETTO, ICPS, JETTO, TOPICS-IB Accurate
simulations underway kinetic effects, finite
gyroradius effects TEMPEST particle distribution
functions are represented as continuous functions
in 4D / 5D with full toroidal geometry. (Xu TH
P6-23) XGC the turbulence suppression after the
H-mode transition can be sustained by
neoclassical sheared flow alone. (Chang TH P6-14)
ITER baseline ELMing Q16.6 with Tped4.9keV
LEHIGH -JETTO
Onjun (PoP2005)
17
Summary
ITPA-PEP
1) The complex parameter linkages in pedestal
have been identified. Dped is determined by
both plasma processes and neutral transport.
largest issue
prediction of the pedestal width in ITER 2)
Type-I ELM onset explained successfully by the
P-B modes. Evolution of the type I ELM cycle
(crash and recovery) revealed Explosive
evolution predicted theoretically and reproduced
numerically.
Issue Change of surface current across an ELM
ELM
Energy Loss mechanisms 3) Small and no-ELM
regimes reproduced in multiple devices, and
accessibility to these regimes has been
identified and categorized.
Issue extend to ITER regime 4)
Rotation plays important roles in determining
pedestal structure and ELMs.
Issue rotation controlability 5)
Modeling capability integrating the core,
pedestal and SOL regions has achieved
remarkable progress.
Issue Pedestal width, and jedge(t)
across an ELM crash 6) Based on these results,
the pedestal height required for ITER has been
evaluated, the ELMing ITER base line scenario has
been simulated, type I ELM mitigation methods
have been evaluated for ITER.
Issue Confinement degradation
island formation