ELM Suppression by Edge Ergodisation

1
ELM Suppression by Edge Ergodisation

• P. Thomas, M. Bécoulet , M.Lipa, E. Nardon, P.
  Ghendrih, A. Grosman G. Huysmans, - CEA Cadarache
• R. Moyer, T.Evans General Atomic
• K.H. Finken KFZ Juelich
• V. Chuyanov, G.Federici, R. Tivey ITER Garching
• S. Hotchin, L. Last UKAEA Culham
• M.Dentan EFDA-CSU Culham
• many others who are indirectly(as yet!)
  implicated.

2
Plan of Presentation

• Scientific rationale
• DIIID experiment with I-coils
• Programmatic logic
• ELM suppression for ITER
• An ITER prototype RMP system for JET
• The project programme

3
Scientific rationale
Large type I ELMs (Edge Localised Modes) in ITER
will

• seriously reduce the lifetime of the divertor
  target either by causing intolerably high
  sputtering rate of a graphite target or surface
  melting of tungsten
• cause considerable operational difficulty in
  advanced scenarios, through coupling between core
  and edge
• interrupt the coupling of ICRH and LHCD and
• reduce the effectiveness of the divertor.

4
How can edge ergodisation eliminate ELMs?

• Magnetic perturbation, resonant with q at edge
• ? ergodisation
• ? increased edge transport.
• ? decreased ?p
• ? edge MHD stable
• ? ELM suppression

5
How should we pursue the idea?

• ELM control necessary for ITER
• Edge ergodisation only method that suppresses
  ELMs and retains confinement quality

• Understand and expand DIIID results
• Prepare a feasible, conceptual design for ITER
• Prepare a design for an ITER prototype to be
  installed on JET

6
The DIII-D I-coils (Internal MHD coil) provide a
resonant perturbation to test this idea

• I-coils installed for MHD investigations.
• n3 symmetry with spectrum m9-12
• Inside vacuum chamber very close to plasma (
  a/6).

7
The I-coils, being close to the plasma, have a
strong resonant component at the edge.
zeros in ?Br
.but the interior perturbations are of the same
order.
8
The initial experiments suppressed type I ELMs
whilst maintaining bN.H 4.2

• Edge ergodisation is only proposal for ELM
  suppression.

9
The toroidal rotation velocity is strongly
reduced during ELM suppression

• Core v? drops by factor 3
• No sign of increased susceptibility to NTMs.
• Momentum input evidently enough.
• What of ITER?

10
What do we need to understand at DIIID?

• It is not completely evident why the ELM
  suppression is occurring...........
• No obvious signs of edge profile erosion, execept
  with 60 phasing.
• Our interpretation is that we just touch the
  threshold for ergodisation.
• Therefore do experiments with much larger dB/BT
• Direct coupling to edge modes?
• Does the toroidal braking matter?
• Can we be assured that v? will always be
  sufficient to prevent RMP induced NTMs(even in
  ITER)?
• Relationship to type II ELMs?
• Experiment at GA last week?

11
Opportunities for coil emplacment in ITER are
limited.

• Can either use the PF coil positions like the
  error field correction system or the port plugs.
• All possible positions are 1-2 minor radii from
  plasma.

12
The port-plug coils looked attractive..
16 turns
1.8 x1.4 m
G.Federicci R.Tivey V.Chuyanov

• Coaxial actively cooled conductor winding
  embedded in port plug walls.

13
.but no easy way to cancel core perturbation.
large compensation coils
port-plug coils

• The resonant components are nearly independent
  of coil size because of equilibrium geometry
  around separatrix.
• Unwanted components not so very different to
  DIIID if it is remembered that island width
  dB0.5

14
..and edge spectrum intrinsically less good

• The two spread-out zeros for simple dipolar coil
  systems reduce the high-m components
• The expansion of ? near separatrix is only
  thing that rescues us

15
Where to install coils on JET?
Available 100-200kAt needed again 2a from
plasma
16
Can devise coil sets with flat resonant
perturbation profiles.

• Solutions exist, like the above, with island
  widths in core lt2cm and reasonable Chirikov
  parameter at edge.
• It looks as if we will have to live with core
  perturbations!

17
Engineering for Coils

• 10-20s at max current, inertially cooled design,
  using cables.
• gt of the order of 2Meuro
• Same technology as the error field correction
  coils.
• gt relatively easy installation (flexible
  cable)

Shims to set compression of cables
Railway company cable - 3kV
Insulating and restraining elastomer padding
10kV
Aluminium case or clamps
100kAt winding
18
Power Supplies

• Use old PRFA 3kA/3kV/20s
• Needs some repair work.
• Cables from north wing to machine.
• Route is well known.
• Need new 33kV transformer this is also required
  for other reasons.

19
Planning for 2005 conceptual design activity

• Outline planning, assuming end 2008 installation

20
Planning for 2005 conceptual design activity
21
Interfaces will require careful consideration

• Neutral beam sensitivity to perturbation field
  will have to be assessed
• gt re-ionised particles in duct!!
• Interaction with diagnostics will be considered
• gt in particular with magnetic probes
• Position control will have to be modified.
• Forces on coils will have to be calculated, both
  for normal and off-normal operation.
• Electrical pick-up during VDEs will have to be
  estimated and insulation chosen appropriately.

22
The Future

• Edge ergodisation holds considerable promise for
  ELM suppression in ITER.
• It is the only proposed method for suppression
  and has been proven on DIIID.
• It does not reduce energy confinement.
• However, practical implementations for ITER and
  JET are proving to be less straightforward than
  had been hoped, with respect to currents required
  and core perturbations.
• Nevertheless, a worthwhile proof-of-principle
  ITER Prototype can be mounted on JET in the
  2008/9 shutdown at a cost 2Meuro.
• A detailed design will be ready by the end of
  2005.