Near term PWI, SOL and divertor physics issues for ITER

1
Near term PWI, SOL and divertor physics issues
for ITER

• R. A. Pitts
• on behalf of the ITER Organisation
• Fusion Science and Technology Department
• with thanks to A. Loarte, A. Kukushkin, M.
  Shimada, D. Campbell (IO) B. Lipschultz, E.
  Tsitrone (Chair and co-chair of the ITPA SOL
  Divertor Topical Group)

2
Outline

• Brief introduction
• Important near term RD PWI Issues
• Identified in preparation for ITER 5th Science
  Technical Advisory Committee (STAC) Meeting in
  May 2008
• Supported by STAC report
• to be addressed by IO contracts, RD in the
  Domestic Agencies and by ITPA
• Discuss the most important key areas identified
• Tritium retention and retention control
• Dust
• Transients and heat fluxes to PFCs

3
Upscale to ITER is a big step
Comparison with JET for illustration
Parameter JET MkIIGB(1999-2001) JET MkIIGB(1999-2001) ITER
Integral time in diverted phase 14 hours 0.1 hours 0.1 hours
Number of pulses 5748 1 1
Energy Input 220 GJ 60 GJ 60 GJ
Average power 4.5 MW 150 MW 150 MW
Divertor ion fluence 1.8x1027 6x1027 6x1027
Extracted from Matthews et al. EPS 1999
Code calculation
1 ITER pulse 0.5 JET years energy input
1 ITER pulse 6 JET years divertor fluence

• Stored energy goes ? R5 ? 35? higher on ITER
  than JET
• But deposition area for power in the divertor ?
  Rlp ? lp,ITER lp,JET ? 3.5 m2 ITER cf. 1.0 m2
  ? ITER must project 35? the energy into only 3
  times the area
• High stored energy means that unmitigated
  disruptions and ELMs far beyond anything
  tolerable by todays materials.

4
Materials choice big RD driver
DIVERTOR
In the current ITER baseline CFC at the strike
points, W on the baffles through the H and D
phases All-W from the start of DT operations
W
W
W
CFC
CFC

• Rationale
• Carbon easier to learn with
• Lack of melting makes it easier to test ELM and
  disruption mitigation strategies
• T-retention expected to be too high in DT phase
  with CFC targets
• Dust probably a big issue with CFC, even in
  steady state

W reflector plates

• But (limited) DT operations with CFC target still
  in place not excluded .

5
Materials choice big RD driver
FIRST WALL

• Rationale
• Low Z
• Good oxygen getter ? automatic wall
  conditioning
• Low(er) T-retention
• But
• Very little experience to-date in research
  tokamaks
• Low melting temperature
• Mixed material issues (alloying)
• T-retention could be more problematic than
  initially thought

Beryllium
Surface areas Be 700 m2 W 100 m2
(bafflesdome) CFC 50 m2 (targets, first
divertor)
6
T-retention

• A 400 s QDT 10 ITER discharge will require 50
  g of T fuellingfor 5050 DT mix (cf. 0.01-0.2 g
  of D in todays tokamaks)
• Maximum in-vessel mobilisable T in ITER limited
  to 1 kg (max on-site inventory 4 kg)
• This is a safety issue
• In practice, administrative limit of 700 g
• 120 g in cryopumps
• 180 g uncertainty
• Providing the best possible estimates of the
  expected retention in ITER before DT operation
  begins is an important part of the ITER physics
  programme
• Good progress is already being made but
  continuous refinement is required through
  improvements in physics models

7
T-inventory build-up estimates

• Considerable effort stimulated by ITER design
  Review
• New estimates made by both EU PWI Task Force and
  ITPA

ITPA Wall erosion simple assumption YBe YC
0.02, divertor erosion from ERO with 1 Be in
inner divertor flux, 0.1 in outer flux Main
chamber fluxes derived from experimental
scalings, divertor fluxes from B2-Eirene Co-deposi
tion from with T/C, T/be, T/W from exptl.
scalings No wall retention by co-deposition
EU PWI TF Wall erosion/redeposition from DIVIMP,
divertor from ERO Main chamber fluxes from
B2-Eirene simulations (with multiplying factor)
Divertor fluxes from B2-Eirene Co-deposition
with fixed T/C, T/Be, T/W Retention in W from lab
data extended to ITER fluxes by diffusion
codes No wall retention by co-deposition
8
T-inventory build-up estimates
ITPA
EU PWI TF
all-C materials
Be wall with CFC divertor
initial ITER mix
150W/mK
Be wall W divertor
150W/mK
50W/mK
Be wall W divertor
50W/mK
all-W materials
C vessel wall
high wall flux
all-W materials
low wall flux
J. Roth et al., PPCF 50 (2008) 1
J. Roth et al., IAEA 2008, IT/P16

• Reasonable agreement between the two approaches
• With Be wall CFC divertor ? few 100 discharges
  before limit is reached
• Might be allowed gt2500 full power pulses with Be
  wall W divertor

9
Refining T-inventory build-up estimates
The baseline ITER safety case strategy requires
that estimates of the expected T-retention for DT
phase wall material combinations be constantly
refined through construction. What outstanding
near term issues need to be addressed?
10
Refining T-inventory build-up estimates

• The baseline ITER safety case strategy requires
  that estimates of the expected T-retention for DT
  phase wall material combinations be constantly
  refined through construction.
• What outstanding near term issues need to be
  addressed?
• Main chamber T co-deposition with Be completely
  ignored so far

Blanket shield module PFCs (450!) will be shaped
to protect leading edges and misalignments
creates shadowed regions where net impurity
redeposition can occur ? associated T-retention?
11
Refining T-inventory build-up estimates
G. De Temmerman et al., NF 48 (2008) 075008

• The baseline ITER safety case strategy requires
  that estimates of the expected T-retention for DT
  phase wall material combinations be constantly
  refined through construction.
• What outstanding near term issues need to be
  addressed?
• Main chamber T co-deposition with Be completely
  ignored so far

(DT)/Be 5.82?10-5E1.17(G(DT)/GBe)-0.21
e2273/TT-conc. increases with increasing
deposition rate and incident energy, decreases
with surface T ? large first wall surface area ?
potential problem even if (DT)/Be low
12
Refining T-inventory build-up estimates

• The baseline ITER safety case strategy requires
  that estimates of the expected T-retention for DT
  phase wall material combinations be constantly
  refined through construction.
• What outstanding near term issues need to be
  addressed?
• Main chamber T co-deposition with Be completely
  ignored so far

TCV Central Column, March 2007 (photo R. A .Pitts)
Erosion/deposition patterns are a reality in
tokamaks with shaped walls! Little work has been
done in this area and more is required ? being
pursued at ITER and through ITPA
13
Refining T-inventory build-up estimates

• The baseline ITER safety case strategy requires
  that estimates of the expected T-retention for DT
  phase wall material combinations be constantly
  refined through construction.
• What outstanding near term issues need to be
  addressed?
• Main chamber T co-deposition with Be completely
  ignored so far

• T-retention in gaps (between mono-blocks and
  castellated structures)

Divertor 100,000 CFC monoblocks500,000 W
monoblocks and tiles (baffles, dome and reflector
plates)First wall as yet undefined number of Be
flat tiles ? huge number of gaps
See talk by M. Merola
14
Refining T-inventory build-up estimates

• The baseline ITER safety case strategy requires
  that estimates of the expected T-retention for DT
  phase wall material combinations be constantly
  refined through construction.
• What outstanding near term issues need to be
  addressed?
• Main chamber T co-deposition with Be completely
  ignored so far

• T-retention in gaps (between mono-blocks and
  castellated structures)
• Effect on T-retention (reduction?) when W
  surfaces are simultaneously bombarded with
  hydrogenic and He ions
• Permeability and retention of T in plasma ion and
  neutron damaged W
• T-retention associated with dust

15
T-inventory control/removal
Whatever we do, tritium will accumulate. Strategie
s must be in place to recover it Good
housekeeping approach can reduce inventory
16
T-inventory control/removal

• Whatever we do, tritium will accumulate.
• Strategies must be in place to recover it
• Good housekeeping approach can reduce
  inventory
• Use of ion (or electron) cyclotron wall
  conditioning (inter-shot), but many issues still
  to assess T-removal rate, discharge uniformity,
  optimum pressures, gas mixtures .

AUG
TEXTOR
A. Lyssoivan
17
T-inventory control/removal

• Whatever we do, tritium will accumulate.
• Strategies must be in place to recover it
• Good housekeeping approach can reduce
  inventory
• Use of ion (or electron) cyclotron wall
  conditioning (inter-shot), but many issues still
  to assess T-removal rate, discharge uniformity,
  optimum pressures, gas mixtures .

R. Doerner et al., PISCES-B, UCSD

• Divertor bake to 350?C now in the ITER baseline

High temperature bake choice based on expts. with
laboratory deposited Be films. Important to
continue validation of this approach including
more realistic mix of co-deposited materials,
effect of implantation temperature, influence of
gaps
18
T-inventory control/removal

• Whatever we do, tritium will accumulate.
• Strategies must be in place to recover it
• Good housekeeping approach can reduce
  inventory
• Use of ion (or electron) cyclotron wall
  conditioning (inter-shot), but many issues still
  to assess T-removal rate, discharge uniformity,
  optimum pressures, gas mixtures .

• Divertor bake to 350?C now in the ITER
  baseline
• T-removal by disruption flash heating of
  co-deposited layers and in gaps ? modeling for
  ITER, and more experimental work required (e.g.
  PISCES-B)
• Isotope tailoring ? careful attention to
  optimisation of T-fuelling efficiency, use of
  D-phases in discharge tail ? does this help
  (T-removed in previous shotduring ICWC must
  simply be added again in the next pulse)?
• Continued studies of O2 baking efficacy for
  removal of co-deposited C layers

19
Dust why worry?

• Expectation is that increase in duty cycle and
  erosion in ITER will lead to large scale-up in
  quantity of dust particles produced
• Like T-retention, dust is a safety issue
• dust particles radioactive (tritium activated
  metals)
• potentially toxic (Be)
• potentially responsible for a large fraction of
  in-VV mobilisable tritium
• chemically reactive with steam or air
• Radiological or toxic hazard depends on how well
  dust is contained in accident scenarios and
  whether it is small enough to remain airborne and
  be respirable
• depends on how dust is produced, e.g. crumbling
  of co-deposited layers or destruction (thermal
  overload) during off-normal events, surface
  melting during transients

20
Dust

• Safety prescribes limits of
• 1 tonne of in-vessel mobilisable dust
• On hot surfaces 6 kg each of C, Be, W (for
  C/Be/W mix), higher for Be/W mix
• A strategy for dust measurement and removal

• Physics needs to clarify
• Dust generation how, where, how much?
• Dust transport effect on confinement, rate of
  destruction, where does it go?

21
Dust formation

• Dust formation in ITER PFC mix from several
  possible sources
• Deposited layer disintegration under transient
  loads ? most likely in divertor were layers most
  likely to grow

R. A. Pitts et al., PPCF 47 (2005) B303
Layers expected to grow quickly at targets in
ITER ? likely to be more easily destroyed by
transient heat fluxes (ELMs/disruptions)Question
for physicsconversion factor from gross erosion
to dust?
e.g. erosion-deposition balance at JET
1999-2001 campaign (14 hrs plasma)
22
Material migration

• Dust formation in ITER PFC mix from several
  possible sources
• Deposited layer disintegration under transient
  loads ? most likely in divertor were layers most
  likely to grow
• Whole issue of material erosion, migration and
  transport (edge flows) still to be fully
  understood in tokamaks
• Quantify Be concentrations in divertor fluxes,
  magnitude of first wall erosion, inner to outer
  divertor transport
• Intensified modelling efforts including drift
  physics, realistic plasma-wall interactions in
  main chamber and divertor

23
Dust formation

• Dust formation in ITER PFC mix from several
  possible sources
• Deposited layer disintegration under transient
  loads ? most likely in divertor were layers most
  likely to grow

M. J. Baldwin et al., PSI 2008

• He-induced nano-morphology ? dust formation in
  steady state, enhanced non-atomistic erosion
  rates on W

Ts 1120 K, GHe 461022 m2s1, Eion 60 eV
24
Dust formation

• Dust formation in ITER PFC mix from several
  possible sources
• Deposited layer disintegration under transient
  loads ? most likely in divertor were layers most
  likely to grow

N. Klimov et al., PSI 2008

• He-induced nano-morphology ? dust formation in
  steady state, enhanced non-atomistic erosion
  rates
• Intense heat loads during disruptions ? brittle
  destruction, melt layer loss, droplet ejection

1.4 MJm-2 in QSPA (500 ms) corresponds to 2.5
MJm-2 (1.5 ms) average L-mode disruption2.2
MJm-2 in QSPA (500 ms) corresponds to 3.9
MJm-2 (1.5 ms) maximum L-mode disruption
W droplet ejection study in QSPA at plasma
pressures typical of ITER disruptions
25
Dust formation

• Dust formation in ITER PFC mix from several
  possible sources
• Deposited layer disintegration under transient
  loads ? most likely in divertor were layers most
  likely to grow

N. Klimov et al., PSI 2008

• He-induced nano-morphology ? dust formation in
  steady state, enhanced non-atomistic erosion
  rates
• Intense heat loads during disruptions ? brittle
  destruction, melt layer loss, droplet ejection.
• Melt layer dynamics still not well understood (RT
  and KH instabilities).
• Situation for Be largely unstudied (QSPA-Be
  expts. expected to begin soon)
• Conversion factor - how much dust created for
  given heat load incident above threshold?
• How much collects in tile gaps and castellations?

W droplet ejection study in QSPA at plasma
pressures typical of ITER disruptions
26
Heat loads
J. Linke et al., 13th ICRFM, Nice, France 2007

• Transient loads remain a major concern for ITER
• ELMs
• Must stay below 0.5 MJm-2 energy density on
  divertor targets to avoid CFC damage and
  macrobrush edge melting

DWELM qELM ? A?,in ? (1 Eout/Ein) 0.5
MJm-2 ?1.4 m2 ?1.5 1 MJ
Eout/Ein ELM divertor energy asymmetryA?,in
divertor energy deposition area, qELM ELM energy
flux density
27
Heat loads
J. Linke et al., 13th ICRFM, Nice, France 2007

• Predicted natural ELM size for Type I Baseline
  H-mode 20 MJ
• Must be mitigated by factor 20 (see talk by T.
  Evans for one method)
• Require more data for ELM statistics on targets
  and first wall mitigated and natural ELMs ELM
  footprints, radial filament velocities, heat load
  amplitude variation

• Divertor heat loads during ELMs suppressed with
  resonant magnetic perturbations
• Balance between penetration to core of W eroded
  by ELMs and ELM impurity flushing action

DWELM qELM ? A?,in ? (1 Eout/Ein) 0.5
MJ/m2 ?1.4 m2 ?1.5 1 MJ
Eout/Ein ELM divertor energy asymmetryA?,in
divertor energy deposition area, qELM ELM energy
flux density
28
Heat loads

• Transient loads remain a major concern for ITER
• Disruptions
• Thermal loads and runaway electrons must be
  mitigated by large factors

e.g. for reference QDT ELMing H-mode discharge
stored energy at thermal quench W (120-175)
MJ 21 inout divertor loading asymmetryEin
(80-117) MJ, Eout (60-88) MJ Expansion of
wetted area of divertor 5-10 Ain (6.5-13)
m2, Aout (8.5-17) m2 Energy deposition rise
time ? (1.5-3) ms
Most severe case (inner plate) W 117 MJ, Ain
6.5 m2, ? 1.5 ms ? ? fT-1Ein/(Ain?0.5) ? 388
MJm-2s-0.5 (fT 1.2-1.5 ? factor accounting for
energy pulse shape compared with square-wave
pulse) Critical value for melting of tungsten
?melt ? 48 MJm-2s-0.5 Energy flux due to
disruptions needs to be reduced by ? / ?melt ?
388 / 48 ? 8 ? Target value of mitigation ?10
29
Heat loads

• Transient loads remain a major concern for ITER
• Disruptions
• Thermal loads and runaway electrons must be
  mitigated by large factors

e.g. VDEs
28 MJm-2
Most severe case 270 MJ, ?p 3cm Preliminary
studies with shaped wall panels28 MJ/m2 ? ?
700 MJ/m2/s0.5 ? ? factor 30 mitigation
30
Heat loads

• Transient loads remain a major concern for ITER
• Disruptions
• Thermal loads and runaway electrons must be
  mitigated by large factors
• Key edge physics/PWI issues needing urgent
  attention
• Best technique or combinations of techniques for
  ITER (e.g. gas injection, killer pellets),
  influence on runaway electron suppression/generati
  on
• Number of injection locations required
• Degree of asymmetry of radiation flash ? if
  number of injection locations limited, how
  serious will local wall melting be?
• Quantification of likely heat loads by improved
  data from operating devices for normal
  disruptions and VDEs
• Consequences on lifetime/melting/dust production
  if mitigation fails

31
Heat loads

• Transient divertor reattachment recently
  identified as a concern
• Prompt loss of divertor radiation (loss of
  extrinsic seeding, gas puff or confinement
  transitions) can lead to extreme power loads

R. A. Pitts, ITER_D_2DMGEF
If radiation fraction in the divertor falls to
20 ? peak outer target power flux reaches 40
MWm-2 for an out-in power asymmetry of 21 ? this
is 4x the allowable steady state heat flux for
the actively cooled divertor ? cannot be
tolerated for long
qpk,outer PDIV,outersin(aouter)/2plqRouter
(Bq/B)sep qpk,inner PDIV,innersin(ainner)/2plqRi
nner (Bq/B)sep PDIV,outer PSOL(1
fRAD)Asym/(1 Asym) PDIV,inner PSOL(1
fRAD)/(1 Asym)
32
Heat loads

• Transient divertor reattachment recently
  identified as a concern
• Prompt loss of divertor radiation (loss of
  extrinsic seeding, gas puff or confinement
  transitions) can lead to extreme power loads

Target First CriterionDta,Temp (s) Second CriterionDta,CWHF (s)
CFC lt1.1 2.0
W lt1.5 2.0
If radiation fraction in the divertor falls to
20 ? peak outer target power flux reaches 40
MWm-2 for an out-in power asymmetry of 21 ? this
is 4x the allowable steady state heat flux for
the actively cooled divertor ? cannot be
tolerated for long
Dta,temp design allowable duration below which
no decrease in target lifetime Dta,CWHF design
allowable duration before which target
destruction can be excluded
33
Heat loads
A. Kukushkin, ITER B2-Eirene case 585

• Transient divertor reattachment recently
  identified as a concern
• Prompt loss of divertor radiation (loss of
  extrinsic seeding, gas puff or confinement
  transitions) can lead to extreme power loads
• Require experimental studies of validity of this
  concern ? data mining and new expts.
• Backed up by modelling ? already started for ITER
  in the IO

Power density (MWm-2), Te (eV)
Density (m-3), Flux (m2s-1)

• General issue of divertor detachment modelling
  still outstanding
• Not correctly reproduced in todays experiments
  by simulation codes ? baseline operation mode for
  ITER to maintain tolerable steady state heat
  loads
• For both carbon and impurity seeded (W target)
  scenarios

Distance along divertor target (m)
34
Summary

• Work plan for near term RD PWI and SOL/divertor
  physics has been established and agreed by STAC
• It is being addressed by efforts within the IO,
  RD tasks launched by the IO, by the ITPA SOL and
  Divertor Topical Group and other Task Forces in
  the parties
• Key areas identified
• Tritium retention and retention control
• Dust
• Heat fluxes to PFCs, transient and steady state
• Erosion, migration, transport of impurities