MHD Issues and Control in FIRE

1
MHD Issues and Control in FIRE

• C. Kessel
• Princeton Plasma Physics Laboratory
• Workshop on Active Control of MHD Stability
• Austin, TX 11/3-5/2003

2
Layout of FIRE Device
R2.14 m a0.595 m ?x2.0 ?x0.7 Pfus150 MW
PF4
PF1,2,3
H-mode Ip7.7 MA BT10 T ?N1.85 li(3)0.65 ?flat
20 s AT-mode Ip4.5 MA BT6.5 T ?N4.2 li(3)0.40
?flat31 s
TF Coil
CS3
Cu stabilizers
PF5
CS2
CS1
Cu cladding
VV
3
FIRE Description
R 2.14 m, a 0.595 m, ?x 2.0, ?x 0.7, Pfus
150 MW

• H-mode
• IP 7.7 MA
• BT 10 T
• ?N 1.80
• 2.4
• ?P 0.85
• ???? 0.075
• q(0) lt 1.0
• q95 3.1
• li(1,3) 0.85,0.66
• Te,i(0) 15 keV
• n20(0) 5.3
• n(0)/?n? 1.15
• p(0)/?p? 2.4

• AT-Mode
• IP 4.5 MA
• BT 6.5 T
• ?N 4.2
• 4.7
• ?P 2.35
• ???? 0.21
• q(0) 4.0
• q95, qmin 4.0,2.7
• li(1,3) 0.52,0.45
• Te,i(0) 15 keV
• n20(0) 4.4
• n(0)/?n? 1.4
• p(0)/?p? 2.5

Cu passive plates
plasma
Port
Cu cladding
4
FIRE H-mode Parameters and Profiles
total
bootstrap
5
FIRE H-mode Parameters and Profiles
6
FIRE H-mode m1 Stability

• Sawteeth
• Unstable, r/a(q1) 0.33, Porcelli sawtooth
  model in TSC indicates weak influence on plasma
  burn due to pedestal/bootstrap broadening current
  profile, and rapid reheat of sawtooth volume
• Requires 1 MA of off-axis current to remove q1
  surface
• RF stabilization/destabilization of sawteeth? To
  remove or weaken drive for low order NTMs

7
FIRE H-mode Neo-Classical Tearing Modes

• Neo-Classical Tearing Modes
• Unstable or Stable?
• Flattop time (20 s) is 2 current diffusion times,
  j(?) and p(?) are relaxed
• Sawteeth and ELMs as drivers are expected to be
  present
• Operating points are at low ?N and ?P, can they
  be lowered further and still provide burning
  plasmas ----gt yes, lowering Q
• EC methods are difficult in FIRE H-mode due to
  high field and high density (280 GHz to access
  Ro)
• LH method of bulk current profile modification
  can probably work, but will involve significant
  power, affecting achievable Q ----gt is there
  another LH method such as pulsing that needs less
  current?

8
FIRE H-mode Neo-Classical Tearing Modes
TSC-LSC simulation
POPCON shows access to lower ?N operating points
(3,2) surface P(LH)12.5 MW I(LH) 0.65
MA n/nGr 0.4
9
FIRE H-mode Ideal MHD Stability

• n1 external kink and n8 ballooning modes
• Stable without a wall/feedback
• Under various conditions sawtooth flattened/not
  flattened current profiles, strong/weak
  pedestals, etc. ?N3
• EXCEPT in pedestal region, ballooning unstable
  depending on pedestal width and magnitude
• Intermediate n peeling/ballooning modes
• Unstable, primary candidate for ELMs
• Type I ELMs are divertor lifetime limiting, must
  access Type II, III, or other lower energy/higher
  frequency regimes
• FIRE has high triangularity (?x 0.7) and high
  density (n/nGr lt 0.8), what active methods should
  be considered?

10
FIRE H-mode Ideal MHD Stability
Self consistent bootstrap/ohmic equilibria
No wall ?N(n1) 3.25 ?N(n8) ? 4.5 Other cases
with different edge and profile conditions yield
various results -----gt ?N 3
11
FIRE AT-mode Operating Space

• Database of operating points by scanning q95,
  n(0)/?n?, T(0)/?T?, n/nGr, ?N, fBe, fAr
• Constrain results with
• installed auxiliary powers
• CD efficiencies from RF calcs
• pulse length limitations from TF or VV nuclear
  heating
• FW and divertor power handling limitations
• identify operating points to pursue with more
  detailed analysis

12
FIRE AT-mode Parameters and Profiles
13
FIRE AT-mode Parameters and Profiles
14
FIRE AT-mode Neoclassical Tearing Modes

• Neoclassical Tearing Modes
• Stable or Unstable?
• q(?) gt 2 everywhere, are the (3,1), (5,2), (7,3),
  (7,2).going to destabilize? If they do will
  they significantly degrade confinement?
• Examining EC stabilization at the lower toroidal
  fields of AT
• LFS launch, O-mode, 170 GHz, fundamental
• 170 GHz accesses Ra/4, however, ?p e ?ce
  cutting off EC inside r/a 0.67
• LFS deposition implies trapping degradation of CD
  efficiency, however, Ohkawa current drive can
  compensate
• Current required, based on (3,2) stabilization in
  ASDEX-U and DIII-D, and scaling with IP?N2, is
  about 200 kA ----gt 100 MW of EC power! Early
  detection is required
• Launch two spectra with LHCD system, to get
  regular bulk CD (that defines qmin) and another
  contribution in the vicinity of rational surfaces
  outside qmin to modify current profile and resist
  NTMs ----gt this requires splitting available
  power

15
FIRE AT-mode Neoclassical Tearing Modes
J. Decker, MIT
145?155 GHz -30o?L-10o midplane launch 10 kA
of current for 5 MW of injected power
?149 GHz ?L-20o
Ro
Roa
Bt6.5 T
?
fce182
fce142
170 GHz
Ro
Roa
Bt7.5 T
?
?
fce210
fce164
200 GHz
Ro
Roa
Bt8.5 T
?
fce190
fce238
16
FIRE AT-mode Neoclassical Tearing Modes
??ce170 GHz
r/a(qmin) 0.8 r/a(3,1) 0.87-0.93 Does (3,1)
require less current than (3,2)? Local ?, ?,
Rem effects? 200 GHz is better fit for FIRE
parameters
Rays are launched with toroidal directionality
for CD
?pe?ce
Short pulse, MIT
Rays are bent as they approach ??pe
17
FIRE AT-mode Ideal MHD Stability

• n 1, 2, and 3external kink and n 8 ballooning
  modes
• n 1 stable without a wall/feedback for ?N lt
  2.5-2.8
• n 2 and 3 have higher limits without a
  wall/feedback
• Ballooning stable up to ?N lt 6.0, EXCEPT in
  pedestal region ballooning instability associated
  with ELMs
• Specifics depend on po/?p?, H-mode or L-mode
  edge, pedestal characteristics, level of LH
  versus bootstrap current, and Ip (q)
• FIREs RWM stabilization with feedback coils
  located in ports very close to the plasma, VALEN
  analysis indicates 80-90 of ideal with wall
  limit for n1
• n 1 stable with wall/feedback to ?Ns around
  5.0-6.0 depending on edge conditions, wall
  location, etc.
• n 2 and 3 appear to have lower ?N limits in
  presence of wall, possibly blocking access to n
  1 limits ----gt how are these modes manifesting
  themselves in the plasma when they are predicted
  to be linear ideal unstable?
• Intermediate n peeling/ballooning modes
• Unstable under H-mode edge conditions

18
FIRE AT-mode Ideal MHD Stability

• H-mode edge
• Ip 4.8 MA
• BT 6.5 T
• ?N 4.5
• ? 5.5
• ?p 2.15
• li(1) 0.44
• li(3) 0.34
• qmin 2.75
• p(0)/?p? 1.9
• n(0)/?n? 1.2
• ?N(n1) 5.4
• ?N(n2) 4.7
• ?N(n3) 4.0
• ?N(bal) gt 6.0

19
FIRE AT-mode Ideal MHD Stability

• L-mode edge
• Ip 4.5 MA
• BT 6.5 T
• ?N 4.5
• ? 5.4
• ?p 2.33
• li(1) 0.54
• li(3) 0.41
• qmin 2.61
• p(0)/?p? 2.18
• n(0)/?n? 1.39
• ?N(n1) 6.2
• ?N(n2) 5.2
• ?N(n3) 5.0
• ?N(bal) gt 6.0

20
AT Equilibrium from TSC-LSC Dynamic Simulations
TSC-LSC equilibrium Ip4.5 MA Bt6.5 T q(0)3.5,
qmin2.8 ?N4.2, ?4.9, ?p2.3 li(1)0.55,
li(3)0.42 p(0)/?p?2.45 n(0)/?n?1.4 Stable
n? Stable n1,2,3 with no wall
L-mode edge
vV/Vo
21
FIRE AT-mode Ideal MHD Stability
Growth Rate, /s
?N4.2
?N
RWM Feedback Coil
ICRF Port Plug
22
FIRE H-mode and AT-Mode Other

• Alfven eigenmodes and energetic particle modes
• Error fields from coil misalignments, etc. ----gt
  install Cu window coils outside TF coil,
  stationary to slow response
• Disruptions ----gt
• Pellet and gas injectors will be all over the
  device, resulting radiative heat load is high
• Up-down symmetry implies plasma is at or near the
  neutral point, not clear if this can be used to
  mitigate or avoid VDEs
• Vertical position control
• Cu passive stabilizers providing growth time of
  30 ms, vertical feedback coils located outside
  inner VV on outboard side
• Fast radial position control, antenna coupling,
  provided by same coils as vertical control
• Shape control provided by PF coils

23
FIRE H-mode and AT-mode Other
PF4
PF1,2,3
Error correction coils

TF Coil
CS3
PF5
CS2
CS1
Fast vertical and radial position control coil
24
FIRE H-mode and AT-mode Other