46th Annual Meeting of Division of Plasma PhysicsAmerican Physical Society

November 15 – 19, 2004Savannah, Georgia

A solenoid-free current start-up scenario utilizing outer poloidal field

coils and center-post

Supported by

Columbia UComp-X

General AtomicsINEL

Johns Hopkins ULANLLLNL

LodestarMIT

Nova PhotonicsNYU

ORNLPPPL

PSISNL

UC DavisUC Irvine

UCLAUCSD

U MarylandU New Mexico

U RochesterU Washington

U WisconsinCulham Sci Ctr

Hiroshima UHIST

Kyushu Tokai UNiigata U

Tsukuba UU Tokyo

JAERIIoffe Inst

TRINITIKBSI

KAISTENEA, Frascati

CEA, CadaracheIPP, Jülich

IPP, GarchingU Quebec

Wonho Choe, Jayhyun Kim Korea Advanced Institute of Science and Technology

J. Menard, Masayuki Ono, and the NSTX team

Princeton Plasma Physics Laboratory

Yuichi TakaseTokyo University

Outer PF coil-only inductive plasma start-up

• Ohmic solenoid has been the work horse of fusion research for decades.

• Attractive fusion CTF and power plant design requires OH elimination– Compact CTF requires elimination of OH regardless of R/a.

– ARIES-AT and ARIES-ST design assumes no OH.

• PF coils have been used to start-up the plasma– MAST / START: Merging/compression using in-vessel PF coils

– JT-60U / TST-2: ~150 kA / 10 kA obtained using strong

(1 MW / 100 kW) ECH pre-ionization with ~0.02 kV/m

• Plasma start-up using appropriate combination

of out-board and outer PF coils, and/or using

a conducting center-post.

Plasma

Conventional ohmic solenoid

Ra

Out-boardregion

In-boardregion

Null field generation using out-board induction coils

Plasma

#1

R

Z

#2

#3

Null field regionInitial plasma

• Small radius PF coil #1 produces a ‘peaked’ BV profile.

• Large radius PF coil #2 produces a flat BV profile.

• Near-midplane ‘trim’ PF coil #3 produces a follow BV profile.

22,ZZBBRR∂∂∂∂

matched

22,,0ZZZBBBRR∂∂=∂∂Field null

region

a

Plasma

Midplane R a

Plasma

#1

#2

#3

Major axis

Available flux

22,ZZBBRR∂∂∂∂

matched

Mid-plane vertical field profiles

Coil #1 (+16 MA-t)Coil #2 (-10 MA-t)

Coil #3_1 (-3.7 MA-t)

Coil #3_2 (+6.7 MA-t)

22,,0ZZZBBBRR∂∂=∂∂Field null region

Net vertical field profile

Coil # R (m) Z (m) I (kA-turn)

1 2.00 2.63 +16,000

2 3.80 1.90 -10,006

3_1 3.10 1.20 -3,689

3_2 3.40 1.00 +6,670

Sufficient mid-plane space for blanket access (NSST)

Flux contours Mod-B contours (Gauss)

• Significant amount of volt-sec available for current ramp-up:~4.5 V-s at R0 = 1.75 m

• Generation of good quality multi-pole field null• Excellent out-board access (~1.8 m vertical spacing) Suitable for the interchangeable blanket modules for CTF.

Radial profile of flux

Plasma axis

Case 5 cont’dOuter PF Start-Up could provide multi-MA ‘seed’ current for future STs NSST

592 ms 596 ms 600 ms

604 ms 608 ms 612 ms

• Successful start-up with ET·BT/BP ≥ 0.02 kV/m

(Takase et al., EX/P4-34)

• Large 0.02 kV/m contour (~1.2 m diameter)

Breakdown contours (0.1 kV/m and 0.02 kV/m) (NSST)

0.02 kV/m

0.1 kV/m

~1.2 m

Time-dependent calculationNSST

Simulation for NSTX

Flux contours+20 kA/turn

- 20 kA/t

+2.8 kA/t

0.02 kV/m

0.1 kV/m

Breakdown contours

• Lloyd’s condition, with strong pre-

ionization, ET·BT/BP ≥ 0.1 kV/m

satisfied in a significant volume.

*J. Kim, W. Choe, M. Ono, Plasma Phys. Control. Fusion 46 (2004) 1647

26 ms 27 ms 28 ms

Time-dependent calculation with vacuum vessel eddy currents considered*

Mod-B contours (Gauss)23 ms 24 ms 25 ms

contours (kV/m)

TTPBEB

~ 35 cm

23 ms 24 ms 25 ms

27 ms26 ms 28 ms

• Successful start-up with

ET·BT/BP ≥ 0.02 kV/m

(Takase et al., IAEA EX/P4-34)

• Large 0.02 kV/m contour

Loop voltage and poloidal flux

Coil current Flux vs time (at 1.4 m)

• Significant V-s is available for current ramp-up.

• Full ramp-up scenario will require bi-polar PF5. But, initial breakdown

experiment to ~100 kA should be possible with the existing power supplies.

Loop voltage

20 22 24 26 28 30-202468

10

Loop Voltage (V)

Time (ms)

Null

20 22 24 26 28 300.090.100.110.120.130.140.150.16

Flux (Wb)

Time (ms)

Null

0 5 10 15 20 25 30-20

-10

0

10

20

PF current (kA/turn)

Time (ms)

Null

PF2

PF3

PF4

PF5

23 24 25 26 27 28 29 30

-40

-30

-20

-10

0

10

Bz (Gauss)

Time (ms)

Calculated Bz Required Bz for Ip

23 24 25 26 27 28 29 300.0

0.4

0.8

1.2

1.6

2.0

0.0

0.4

0.8

1.2

1.6

2.0

Field Index R (m)

Time (ms)

Plasma Position Field Index

• Force balance in the initial start-up phase needs to be checked carefully.

• Self-consistent code including evolution of plasma current and parameters

Force balance and stability issues

Vertical field for force balance Stable field structure against

axis-symmetric mode

23 0 <

∂∂

−<RB

BR z

zfor stability

Preliminary PF-only Start-up Experiments in NSTX

• Pre-ionize plasma near RF antenna with ECH + 400 kW HHFW, nD2

= (1 - 2)10-5 Torr

• Create high-quality field-null with (5 - 15) loop-volts at antenna

So far, require EfBf /BP > 0.1 kV/m over substantial plasma volume

• Have created 20 kA plasmas that terminate near center-stack

Shot 114405

J. Menard

Null Size Evolution During PF-Only Start-Up

0

0.5

1

1.5

2

2.5

3

0 2 4 6 8 10Time (msec)

Null Size (m^2)

(Null size: ETBT/BP > 0.1kV/m)OH

XP433-I

XP448-IXP431

XP433-II

XP448-II

Successful initiation:OH:112152, 4.5 kGXP433-I: 113612, 3.5 kGXP433-II:114405, 3 kG

Unsuccessful initiation:XP431: H:11293, 4.5 kGXP448-I: 113609, 3.5 kGXP448-II:114484, 3 kG

Successful initiation thus far requires a large null region

Coil current waveforms used in XP-443

Similar to OH startup waveforms

114405

PF2

PF4

Field and loop voltage

114405

At plasma breakdown

Near maximum plasma current

Camera images and reconstructions show plasmas are born on LFS and have an inward radial trajectory

• LRDFIT code used for reconstructions– IVessel 10 IP

• Careful control of BZ after breakdown helped raise IP from 10kA to 20kA

• More BZ evolution optimization possible

8ms

9ms

10ms

114405

Thomson measurements consistent with plasma motion and peaked pe profiles

Thomson Te < 35eV and camera images consistent with lack of burn-through

need more plasma heating power:• More HHFW power during breakdown• Higher VLOOP – keep plasma outboard• EBW power could be very helpful

114405 t=10ms

B. LeBlanc

C. Bush, ORNL

Solenoid-free current-start-up scenario including PF4

• Try to store more poloidal flux at null region for IP ramp

• Start PF2 & 3 coils with large positive bias– Balanced by negative PF4– Store 50-100mWb at null– Null size 1/3 of XP448

• Null formation very sensitive to coil current time-history and vessel current model

113609

LRDIAG simulations predict vertical merging of X-points

Shot 113608 - C. Bush, ORNL

3 ms 4 ms

6 ms 7 ms

Camera Gain = 95

s113609 (N. Nishino - Hiroshima fast camera)

2.442 ms 2.664 ms 2.886 ms 3.108 ms 3.330 ms 3.552 ms

3.774 ms 3.996 ms 4.218 ms 4.440 ms 4.662 ms 4.884 ms

5.106 ms 5.328 ms 5.550 ms 5.772 ms 5.994 ms 6.216 ms

6.438 ms 6.660 ms 6.882 ms 7.104 ms 7.326 ms 7.548 ms

Magnetics bound IP to < 15kA (probably only few kA)

B_L1DMPPPGU7

B_L1DMPPPGL7

B_L1DMPPPGU8

B_L1DMPPPGL8

IPF4

Rogowski

Bright emission (Hiroshima camera) in 2.5 – 8 ms

~50 G

- RF noise- IPF4 diff.

Ip = ~ 15 kA

113611 (vac. shot) B-dot loop signal (near PF4)

Results

• HHFW pre-ionization necessary– Need sufficient neutral density, 0-0 phasing

– Increase from 0.5 MW to > 1 MW w/ more straps

– EBW could be very helpful (was on TST-2)

• Large null required for IP initiation thus far

– Need more work on finding optimal balance of stored flux vs. null size vs. initial plasma shape

• Good plasma position evolution following breakdown crucial

to high IP

– DINA modeling should be helpful here

Inductive plasma start-up scheme using a conducting center-post

Single turn TF leads to an attractive ST CTF

Wall Loading at Test Modules (MW/m2) 1.0 3.0

HH (ITER98pby2) 1.4 1.8

Applied toroidal field (T) 2.4 2.2

Plasma current (MA) 12.6 11.4

Normalized beta (N) 4.1 7.0

Toroidal beta (T, %) 26.8 45.1

n/nGW (%) 17 52

Q (using NBI H&CD) 2.4 5.8

Fusion power (MW) 72 214

Number of radial access ports 7 7

Radial access test area (m2) 12.8 12.8

PHeat/R (MW/m) 37 67

Tritium burn rate (kg/full-power-year) 4 12

Total facility electrical power (MW) 286 272

Fraction of neutron capture (%) 81.6 81.6

Local T.B.R. for self-sufficiency 1.23 1.23

Toroidal field coil current (MA) 14.6 13.2

Center post weight (ton) 89 89

Capital cost ($B) with 40% contingency 1.47. R = 1.2 m, a = 0.8 m

Plasma

PF Coils

Cen

ter-

post

• Possible as long as the poloidal flux is available at the time of the plasma start-up.

• Assuming the center-post radius is 50 cm for NSST (or CTF-like device), Bz ~ 0.7 T would mean the center-post may provide ~0.55 Wb.

• However, if we use the same coils energized in the same current direction, Bz ~ 8 T can be obtained in the center-post.

~ 6 Wb is stored in the center-post.

• Then, can one reverse PF #2 and #3_1 before the center-post eddy current decays away?

Solid center-post envisioned for ST reactor could provide additional poloidal flux

TF column charging sequence

Field null formation and plasma initiation

Charging up

TF column fully charged

Plasma current ramp up

t

PF3_2 current

t

PF2 and 3_1 current Center-post resistive skin time

t=T1 T2

t

PF1 currentVacuum vessel resistive skin time

Copper TF center-post charging sequence

At t = T1 , All PF currents in the same polarity to maximize the charging

At t = T2 , PF2 current reverses to create field null before center-post flux decays

Cop

per T

F C

ente

r-po

st

0.0 0.5 1.0 1.5 2.0 2.5 3.0012345678

Bz (T)

R (m)C

oppe

r TF

Cen

ter-

post

0.0 0.5 1.0 1.5 2.0 2.5 3.0012345678

Bz (T)

R (m)

-0.5 0.0 0.5 1.0 1.5 2.0-100

0

100

200

300

400

Induced CP currents (kA)

Time (s)

CPshell at 0.08 m CPshell at 0.16 m CPshell at 0.24 m CPshell at 0.32 m CPshell at 0.40 m CPshell at 0.48 m CPshell at 0.56 m

-0.5 0.0 0.5 1.0 1.5 2.0-202468

101214

Induced eddy currents (kA)

Time (s)

Innermost VV Outermost VV

PF1,2Outermost shell

At 2 sec

-0.5 0.0 0.5 1.0 1.5 2.0

0

5

10

15

20

Flux (Wb)

Time (s)

Net PF1 PF2 VV CP

Flux from CP

Loopvoltage

Net

PF2PF1

VV

Plasma initiation time

At (1.0 m, 0)

0.0 0.5 1.0 1.5 2.0 2.5 3.00.00.51.01.52.02.53.0

Flux (Wb)

R (m)

Center-post

Induced current waveforms and poloidal flux

Mid-plane

-20 -15 -10 -5 0 50

50

100

150

200

PF coil current (kA/turn)

Time (s)

PF1 PF2

Evolution of Bz profile

Driving current waveform

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0

10

20

30

40

Bz (kG)

R (m)

-10.0 s -5.0 s +0.00 s +0.25 s +0.50 s +1.00 s +2.00 s

Total Bz profile

0.0 0.5 1.0 1.5 2.0 2.5 3.0

-20-10

010203040

Bz (kG)

R (m)

-10.0 s -5.0 s +0.00 s +0.25 s +0.50 s +1.00 s +2.00 s

Bz by center-post

0.0 0.5 1.0 1.5 2.0 2.5 3.00

5

10

15

20

Bz (kG)

R (m)

Bz profile of a typical solenoid

PF1,2

Due to vessel

Solenoid

Center-post

Centerpost Central solenoid

Null region

Stray field due to the charged center-post

• Like OH, the stray field is relatively small and uniform in the breakdown region.

• The stray field can be easily nulled out by compensation coils (on-going).

• The present method is quite compatible with the PF-only start-up concept utilizing the same hardware.

Some thoughts

• The center-post flux storage concept can essentially double the available flux by using the same hardware of the outer-PF-only start-up concept.

• ~ 10 Wb could potentially solve the CTF start-up and ramp-up issue. NSTX gets 1 MA with 0.35 Wb. So, it should be possible to reach 15 MA with twice the major radius (R = 1.7 m) with 10 Wb ( R2).

• The concept works better with high R/a. If we go with R/a = 2 CTF, then one can envision R = 75 cm core, which can essentially double the flux compared to R = 50 cm core.

• The induced loop voltage is determined by the stored flux and the flux decay rate. It can be adjusted by controlling the center-post resistivity (material, temperature) or resistance (thickness).

• If the inner torus radius is 4 m for example for R = 6 m advanced tokamak reactor, one can conceive a metal ring structure (a combination of vacuum vessel, shield, and support structure) of mean-radius of 3.5 m. In that case, one can store up to 230 Wb of flux assuming 7 T charging field ( x 3.52 x 7)! Even a fluctuation of that would be sufficient for start-up.

• Since pure metal is much more radiation resistance and simple compared to highly stressed OH solenoid, it could also be a more attractive reactor option for tokamak as well.

Summary

• Two complementary inductively-based concepts to aid the solenoid-free start-up for future ST and tokamak reactors are presented.

• A combination of out-board PF coils placed outside the vacuum vessel is shown to create a good quality field null region while retaining significant volt-second capability for current ramp-up.

- For NSST, 4 - 5 Wb possible for ramping the current to a few MAs.- For NSTX, ~0.12 Wb possible for Ip ~ a few hundred kA.

• Inboard-side conducting material to store the magnetic flux which is initially charged up by the outboard-side outer PF coils. For ST, it is conceivable to utilize the central TF conducting post as the flux storage.

- The NSST (CTF) size device can provide additional 2 - 4 Wb with this method.

• Like the OH solenoid, the stray field generated by the center-post flux is relatively small which would make it suitable for the plasma start-up utilization.