Plasma-surface interactions, SOL and · Plasma-surface interactions, SOL and divertor physics: Implications for ITER Bruce Lipschultz For the ITPA SOL/divertor committee Transport

ITPA SOL& Divertor

Plasma-surface interactions, SOL anddivertor physics: Implications for ITER

Bruce Lipschultz

For the

ITPA SOL/divertor committee

Transport

ELMs & disruptions

D/T retention & removal

Materials

Summary

ITPA SOL/divertor presentation, 2006, Chengdu

2ITPA SOL& Divertor

It is a challenge to use our currentexperience to predict ITER performance

No boronization planned (but Be

may serve that purpose)

Surfaces coated with low-Z material

(e.g. boronization)

High transient heat loads - limits

PFC lifetime

Low ELM/disruption transient

loadings

T retention should be ~ 0.1% to

maximize operational availability

D/T retention ~ 3-30% of injected gas

Primarily Be with lesser amounts

of carbon and tungsten

Operational experience primarily

carbon Plasma Facing Components

(PFCs)

ITERCurrent tokamaks


3ITPA SOL& Divertor

We are building a basic understanding of radialtransport in the SOL

• => ITER parallel power flow width similar

(normalized to R) to current tokamaks.

0.0 0.2 0.4 0.6 0.80.000

0.001

0.002

0.003

0.004

0.005

AUGC-ModDIII-DJETJT-60U

λ Te/R

0

ne,sep/nGreenwald

Kallenbach J. Nuc. Mat. 337–339 (2005) 381

=> Q|| ∝ P/R2 NOT P/R


4ITPA SOL& Divertor

We are building a basic understanding of radialtransport in the SOL

• => ITER parallel power flow width similar

(normalized to R) to current tokamaks.

• Pressure gradients just outside the separatrix

are well-organized by Electromagnetic Fluid Drift

Turbulence parameters => direct connection

between gradients and underlying turbulence.

αd0.0 0.2 0.4 0.6 0.8

0.0

0.4

0.8

1.20.80.51.0

IP (MA)

αMHD

LaBombard, Nucl. Fusion 45 (2005) 1658

Inaccessib

le

increasing ν*

∇P

IP2

0.0 0.2 0.4 0.6 0.80.000

0.001

0.002

0.003

0.004

0.005

AUGC-ModDIII-DJETJT-60U

λ Te/R

0

ne,sep/nGreenwald

Kallenbach J. Nuc. Mat. 337–339 (2005) 381

=> Potential to predict plasma profiles

from first principles.

=> Q|| ∝ P/R2 NOT P/R


5ITPA SOL& Divertor

Much better understanding of flows in theedge

SOL flows are a controlling process in impurity

transport as well as tritium co-deposition

• Standard models predict ~ stagnant flows in the SOL

opposite the divertor

0

0.5

1

S

0

0

0.5

-0.5

1

0.5 1Flux-tube coordinate, S

Inner SOL

Pa

rall

el

Mach

Nu

mb

er

Outer

SOL


6ITPA SOL& Divertor




• Standard models can’t match measured flows

• New inner wall probe measurements provide clues:

Pressure drop from from low- to high-field SOL

M~1 flows at high field side*

* LaBombard, Phys Plasmas 12 (2005)

0

0.5

1

S

0

0

0.5

-0.5

1


Inner SOL

Pa

rall

el

Ma

ch

Nu

mb

er

Outer

SOL

C-ModJT-60UJETTCV


7ITPA SOL& Divertor




• Standard models can’t match measured flows

• New inner wall probe measurements provide clues:

Pressure drop from from low- to high-field SOL

Pressure imbalance => driving parallel flows

Pressure imbalance driven by low-field side

ballooning transport out of core, across separatrix*

• Evidence of transport-driven flows setting toroidal

rotation boundary condition for confined plasma

0

0.5

1

S

0

0

0.5

-0.5

1


Inner SOL

Pa

rall

el

Ma

ch

Nu

mb

er

Outer

SOL

C-ModJT-60UJETTCV

Allows better understanding of

impurity migration and T retention

* Gunn, EX/P4-9, LaBombard, Phys Plasmas 12 (2005)


8ITPA SOL& DivertorELM filaments travel far through the SOL to the wall

Loarte, IT/P1-14, Boedo, EX/P4-2, *Kirk , EX/9-1

• Filamentary nature of ELMs (n ~ 7-15)

rotating toroidally and poloidally*

Type I ELMs reduce ITER divertor and main chamber PFC lifetime

MAST/Kirk


9ITPA SOL& DivertorELM filaments travel far through the SOL to the wall

Loarte, IT/P1-14, Boedo, EX/P4-2, *Kirk , EX/9-1

• ELMs travel far into the SOL having

a substantial effect on the density

and temperature at the limiter

• Filamentary nature of ELMs (n ~ 7-15)

rotating toroidally and poloidally*

Type I ELMs reduce ITER divertor and main chamber PFC lifetime

MAST/Kirk

3186 3188 3190 3192 31940.0

0.4

0.8

Time ( ms )

DIII-D/Zeng

Dα

( a.u

.)

wall

R (m)

ne (

10

19 m

-3) separatrix

2.2 2.25 2.30 2.35 2.40

1

2

3

4

1 2 3 4 5

1 234

5

lim

iter


10ITPA SOL& Divertor

Type 1 ELM filaments lead to variable heat

loads on first wall surfaces

•Visible image of ASDEX-

Upgrade limiter*

10 c

m

*Herrmann et al J. Nucl. Mater. 337-339 (2005) 697



0

5

10

15

20

25

30

heat

flu

x M

W/m

2

Type 1 ELM filaments lead to local hot spots on

limiters

• Individual ELM filaments lead

to hot spots on limiter surface*

10 c

m




0

5

10

15

20

25

30

heat

flu

x M

W/m

2


limiters

• Individual ELM filaments lead

to hot spots on limiter surface*

10 c

m




0

5

10

15

20

25

30

heat

flu

x M

W/m

2


limiters

•The instantaneous heat load

is high*.

10 c

m




0

1

2

3

4

5

6

heat

flu

xM

W/m

2

Heat loads are less localized when averaged over

ELMs

•When averaged over ELMs

the heat load is more uniform*

10 c

m**Loarte et al, Paper IT/P1-14, ***Moyer et al, Paper EX/9-3

*Pitts et al, Paper EX/3-1, Herrmann et al J. Nucl. Mater. 337-339 (2005) 697

•ELMs need to be small enough such that the

divertor survives ( WELM/ WPED < 5%) - then

main chamber surfaces should be ok too**.

•But, a few strong ELMs can reduce the tile

resistance to thermal shock

•The community is pursuing small ELM regimes as

well as ELM mitigation***.



WThermal Quench/WPlasma(max)

Re

lati

ve

pro

ba

bil

ity

ASDEX Upgrade

JET

0.0

0.1

0.2 0.4 0.6 0.8 1.0

0.2

0.3

0.4

Disruption statistics reveal details of energy

balance during a disruption

• A significant fraction of the stored energy is

often lost before the thermal quench

Energy lost through L-H transitions…..

• Advanced scenario (ITB and high- )

disruptions are the most dangerous: All the

stored energy comes out rapidly

Loarte et al, Paper

IT/P1-14

• => specify fewer ITER high power

disruptions for ITER reference scenario



0.00 0.04 0.08 0.120.0

0.5

1.0

1.5

2.0JET

DIII-D

ASDEX-Upgrade

WThermal Quench/VPlasma(MJ/m3)

WD

ive

rto

r/W

Th

erm

al

Qu

en

ch

WThermal Quench/WPlasma(max)

Re

lati

ve

pro

ba

bil

ity

ASDEX Upgrade

JET

0.0

0.1

0.2 0.4 0.6 0.8 1.0

0.2

0.3

0.4

Disruption statistics reveal details of energy

balance during a disruption

• A significant fraction of the stored energy is

often lost before the thermal quench

Energy lost through L-H transitions…..

• Advanced scenario (ITB and high- )

disruptions are the most dangerous: All the

stored energy comes out rapidly

• The divertor receives less of the disruption

energy as the stored energy increases

Loarte et al, Paper

IT/P1-14

• Surfaces outside the divertor become more

of a concern

• Disruption mitigation is being pursued with

success*

• => specify fewer ITER high power

disruptions for ITER reference scenario

*Granetz, EX/4-3, Pautasso, EX/P8-7, Izzo, TH/P3-15



Tritium retention is a central emphasis ofSOL/divertor work

• Estimates of T retention in ITER are uncertain

All-carbon PFC tokamaks have D retention per discharge ~3-50% of that injected

ITER retention of 0.1% needed for continuous operation

ITER will have much less carbon, replaced with Be and tungsten (W).

Be does co-deposit with tritium but releases it at a lower temperature than C

Be will not migrate to remote cooled locations as easily as carbon => less likely to

accumulate thick co-deposited layers

Predicted to lead to lower T retention than

current tokamaks

TF

TR

/JE

T30%

/pu

lse

Bro

oks

4-10

%/p

ulse

Roth

1%/pulse

00

50 100 150 200

200

400

600

Number of 400 s pulses

Tri

tiu

m r

ete

nti

on

(g

ram

s)

ITER site limit

1 w

eek

1 d

ay

=>Modelling estimates give a range of

1-3 weeks operation before T site limit

reached



T retention on tile sides could be more

important in ITER

• 20% of the total D retention is on the

sides of tiles

Co-deposition with C ions and molecules

• ITER design increases the ratio of tile

side to front surface areas over current

carbon PFC tokamaks

1

10

100

1000

0 1 2 3 4 5 6

Poloidal running gap (0.5mm)

ASDEX-Upgrade

C-Mod

Distance down tile gaps (mm)

D t

hic

kn

es

s (

10

15 a

tom

s/c

m2

~ A

ng

str

om

s)



T retention on tile sides could be more

important in ITER

• 20% of the total D retention is on the

sides of tiles

Co-deposition with C ions and molecules

• ITER design increases the ratio of tile

side to front surface areas over current

carbon PFC tokamaks

1 10 100 1000

0.1

1.0

10.0

100.0

0.01

Tile

sid

e D

-in

ve

nto

ry

gro

wth

ra

te [

10

20 D

/m2/s

]

D+-flux to front surface [1020 D/m2/s]

10%

1%0.1%

DIII-D C-ModTEXTOR ASDEX-U• Cross-tokamak studies indicate tile side

D retention

proportional to surface ion fluence

Lowest in fully high-Z tokamak

Reduced by elevated tile temperatures

• More difficult to remove

1

10

100

1000

0 1 2 3 4 5 6

Poloidal running gap (0.5mm)

ASDEX-Upgrade

C-Mod

Distance down tile gaps (mm)

D t

hic

kn

es

s (

10

15 a

tom

s/c

m2

~ A

ng

str

om

s)



Studies have revealed another process besides

co-deposition that leads to T retention

• A number of tokamaks have reported that co-deposition on tile surfaces cannot

explain the level of D retention measured (e.g. Tore Supra*, C-Mod**, JT-60U)

• New laboratory studies have found that D can be stored deep below surface

True for carbon AND molybdenum

• Deep retention in tiles will add to ITER T

retention levels

Potentially dominate over co-deposition

in high flux regions

Potentially more difficult to remove

through surface T removal techniques

Exploring for Be, W as well

0.01

0.1

1

10

D a

tom

ic c

on

ce

ntr

ati

on

(%

)

Depth into surface (µm)0 2 4 6 8 10

D/CFC

PISCES-A, UCSD

D/Mo

DIONISOS, U. Wisc.

100 eV D+, Φ~1-15 x 1024 D+/m2, T ~ 500 K

*Loarer, EX/3-6. **Whyte, EX/P4-29


21ITPA SOL& DivertorMixed materials in ITER are a mixed blessing

•Carbon tiles could even be doped with

metals before installation such that

the chemical erosion is reduced

•A number of alloys form

Beryllides (e.g. Be2W) lowers tungsten melting temperature

Carbides can increase T retention (WC)

Alloys could form barriers to the out-diffusion of T

•Be or W on carbon surfaces reduces carbon chemical sputtering

Ch

em

ica

l E

ros

ion

Yie

ld (

C/D

)

Temperature (K)

Pure C

NW/NC =.03

200 eV D+ -> Carbon layers

0.00300 500 700 900 1100

0.04

0.08

0.12

0.16M. Balden, PSI Hefei (2006)



Tritium removal techniques are being

developed

• Tritium removal techniques include

Heating the surface to increase T diffusion (e.g. laser, disruptions)

Chemical removal of carbon & T (Oxygen exposure, discharge cleaning…)

Ablation of the carbon - freeing the T (e.g. flash-lamps, lasers)

•All techniques must

Remove T from wherever it is stored (tile front, sides, bulk)

Be compatible with ITER toroidal field

Not cause dust

Be able to remove T from mixed material surfaces such as

Be, W, C, BeC, WC, Be2W

Not cause problems for subsequent operation

Impurities or damage to vessel


23ITPA SOL& DivertorITER not ready to go to fully high-Z

ITER operation with high-Z PFCs is a goal

in support of DEMO

• Tokamaks with primarily high-Z PFCs

ASDEX-Upgrade (85% W-coated carbon)

C-Mod (100% solid Mo tiles)

• Core high-Z content rises quickly after

boronization*

• ICRF erodes B layer (and Mo/W underneath)

more quickly than NBI or Ohmic heating

• Erosion localized to small fraction of PFCs

• Questions remain

Core high-Z concentration (and radiation)

Boronization needed?

Melting

10-5

10-6

10-4

Imp

urity

co

nce

ntr

atio

n (

CZ =

nZ/n

e)

CMoC-Mod (~ 3MW ICRF)

ne ~ 3.8-4.3x1020 m-3

CWASDEX Upgrade

(5-7.5 MW NBI, > 1MW ICRF)

ne ~ 6-8x1019 m-3

CWASDEX Upgrade

(5-7.5 MW NBI, < 1MW ICRF)

ne ~ 6-8x1019 m-3

0 5 10 15discharges after boronization

*Dux et al, EX/3-3, Marmar et al, EX/3-4



Better understanding but uncertainties are still

a concern

• We are making progress towards first-principles prediction of transport

Much better capability of predicting parallel transport (and impurity transport)

Connection of radial transport to underlying turbulence

• Tritium retention rate estimated to be lower than before but still uncertain

Combined Be/W/C reduces T retention over pure carbon

A number of T removal techniques are being explored with success

• Transient loading on PFC surfaces is very complicated

Much of the stored energy can be lost before disruption thermal quench

The loading of first-wall surfaces by ELMs and disruptions is uncertain

• Material characteristics and their interactions strongly affect ITER operation

A variety of alloys are created whose behavior is difficult to include in predictions

High-Z operational experience for ITER is being developed



The Interaction with the first-wall is central to

the success of ITER

We cannot afford to ignore problems

Nor can we say: ‘the sky is falling’

Documents

Plasma-surface interactions, SOL and · Plasma-surface interactions, SOL and divertor physics: Implications for ITER Bruce Lipschultz For the ITPA SOL/divertor committee Transport