Prsentation PowerPoint

1
Gas balance and Fuel retention
Th Loarer with contributions from D Borodin, C
Brosset, J Bucalossi, S Droste, G Esser, G Haas,
A Herrmann, A Kirscher, A Kreter, K Krieger, J
Likonen, A Litnovsky, M Mayer, V Mertens, Ph
Morgan, V Philipps, G Ramos, S Richter, V Rohde,
J Roth, M Rubel, A Sergienko, E Tsitrone, E
Vainonen-Ahlgren, P Wienhold, EU TF on PWI and
JET EFDA contributors
- Overview of Gas balance and fuel retention
results ? Tokamak experiments (JET, TS, AUG,
TEXTOR) ? Post mortem analysis (Laboratories) -
Summary and further plans
2
Introduction

• Evaluation of hydrogenic retention in present
  tokamaks is of high priority to establish a
  database for ITER (400 sec 7min10-20 sec
  today). T-retention constitutes an outstanding
  problem for ITER operation particularly for the
  choice of the materials (carbon ?)
• A retention rate of 10 of the T injected in
  ITER would lead to the in- vessel mobilisable
  T-limit (350 g) in 35 pulses.

- Retention rates of this order (10-20) or
higher are regularly found using gas balance in
C-wall tokamaks. - Retention rate 5 times lower
are obtained using post mortem analysis
- Are these two methods reliable to evaluate the
retention and is it possible to understand why
they lead to different results ? - SEWG to
clarify Gas Balance vs post mortem analysis
3
Mechanisms for fuel retention

• physics material erosion, migration fuel
  retention
• QMB measurements
• Spectroscopy
• Gas balance measurements
• Deposition probes
• 13C migration
• Post mortem tile analysis

4
Particle balance procedure on JET
Regenerate cryopumps before and after expt. ?
collect total pumped gas (accuracy1.2)
Repeat sets of identical discharges (no intershot
conditioning)
Plasma
Calibrated Particle Source (Gas, NBI)
Wall Retention Long Short Term
Divertor cryo-pumps
Injection Pumped Short Term Ret Long Term
Ret
Total recovered from cryo-regeneration Pumped
intershot outgassing over 800s (assumed equal to
Short Term Ret )
5
Particle fluxes H mode Type I
Ip2.0MA, BT2.4T 13MW NBIICRH ELM Energy100kJ
_at_16 sec, Ret5.2x1021Ds-1 LongRet ShortRet
From L mode to Type I ELM H-mode? Increase of
long term retention - with the recycling flux -
with ELMs Energy
Th Loarer et al
6
Integrated particle fluxes

• Integrated CIII and Ha horizontal light (L-mode,
  Type III and Type I ELMs)
• Slope for Type I ELMy H-mode shows both enhanced
  recycling and total carbon source.

Higher recycling and ELM ?Enhanced carbon erosion
and transport leading to stronger carbon
deposition and fuel codeposition
7
ELM induced C deposition
Non-linear dependence of carbon erosion on ELM
energy ? thermal decomposition of surface layers
and favourable geometry rapidly increases QMB
deposition
Can explains high deposition rates on
water-cooled louvres during 97-98 JET DT
experiments ? high T-retention
A Kreter, G Esser et al
8
Particle Balance summary on JET

• - Long term retention increases from L-mode to
  H-mode
• Increased C erosion and transport due to
  increased recycling and effect of ELMs ? enhanced
  C erosion ? enhanced co-deposition and retention.
• - Recovery between pulses (short term retention)
  always constant within a factor 2 in the range
  1-3?1022D
• Independent of discharge type, ELM energy,
  quantity of injected particles

9
Tore Supra the DITS project
(Deuterium Inventory in Tore supra)

• Objectives
• Clarify post mortem analysis vs Gas Balance
• Retention mechanisms (codeposition vs bulk
  migration)

• 3 phases
• dedicated experimental campaign ? Gas Balance
• dismantling of a sector of the limiter ? samples
  for post mortem analysis
• sample analysis (collaboration with european
  labs, EU PWI TF)

10
Scenario of the DITS campaign
E Tsitrone et al
Repetitive pulses every 20 mn ( 40 mn of plasma
each day) 5 h of plasma w/o conditionning
Scenario 1 (nominal 120 s)
Main issue UFOs (C metals D ? ) ?
detachment ? disruptions

• scenario at lower LH power (lt 1.8 MW) slow ramp
  up
• - No evolution for C
• - Fe and O level increasing to values before
  carbo/boronisation

Scenario 2 (lower power 80 s)
11
UFOs on CCD imaging of the TPL
E Tsitrone et al
12
No wall saturation observed after 5h00
Injected 5.8x1024D (19.5 g)
Trapped 3.3x1024D (11 g)
E Tsitrone et al
13
Inventory proportional to discharge duration
Disch. OK
Disruptions
Trapping
Outgassing
E Tsitrone et al
14
Long term recovery ltlt wall inventory
Total exhausted (610-5 Pa) (1.3106 s) 10
m3/s 700 Pa.m3/s

3.51023 D atoms to be compared to
DWI 3.31024 D atoms ( 10 ) (upper limit D2
concentration in pumped gas decreases rapidly)
E Tsitrone et al
15
Summary

• DITS experimental campaign successfully
  completed
• 13C carbonisation / 11B boronisation performed
• 5h of plasma w/o conditionning 1 year of
  operation in 2 weeks
• Reliable operation (LH, cooling loops, PFCs)
• Main limit UFOs ? disruptions ? operational
  limit ?
• 80 of the objective reached (WI 3.3 1024 D
  or 11g)
• ok for qualitative and quantitative analysis
• Particle balance
• No wall saturation, retention proportional to
  discharge duration.
• Exhausted gas dominated by D during the shots
• Disruptions at low Ip, long term recovery
  negligible in the balance
• DITS project on tracks TPL sector dismantled,
  selected fingers extracted ? samples available
  for analysis november 2007

16
Phases of discharges observed in C
V Rohde et al.
Typical discharge puff and pump steady phase
reached after 2sec
17
Full W configuration Carbon free machine, ?
How does it compare to C in terms of fuel
retention ?
V Rohde et al.
In typical discharge puff and pump steady phase
not reached
18
Gas balance with W wall
Wall loading observed, no steady state reached
V Rohde et al.
19
Gas Balance summary from AUG in W

• Gas Balance is needed to verify the benefit of
  full tungsten wall.
• Support from EU TF on PWI to investigate gas
  balance, but support more difficult from man
  power point of view.
• However, experiments performed and detailed
  analysis to start soon.
• Data set exits, but direct comparison with C is
  very difficult due to different plasma scenario.
• Accuracy is dominated by pumping of cryo pump.
• - Due to the high gas puffing rate (gt1022Ds-1),
  an accuracy of 1 is required in AUG.
  Improvement of the accuracy by adding a separated
  volume to store all the gas (as in AGHS in JET)

V Rohde et al.
20
Deuterium retention in CFC
Deuterium retained in the samples (by TDS)
Photograph of the test limiter with material
stripes exposed in TEXTOR
Comparison with PISCES-A data (J.Roth PSI 06)
Ts 500K
TEXTOR
EK98
DMS780 JET
NB31 ITER
Retention in both CFCs slightly higher than in
EK98 Good agreement with N11 exposed in
PISCES-A No saturation observed for obtained
fluences Fuel retention in TEXTOR is dominated by
co-deposition (Contribution of in-bulk retention
to total retention 10)
A.Kreter et al.
21
Exposure of W castellated limiter in the SOL of
TEXTOR
2 shapes of castellation studied
The shape of a castellation cells can be
optimized to reduce impurity and fuel transport
into gaps

• Experimental details
• ? Shaped and rectangular cells
• exposed under the same plasma
• conditions
• ? 16 repetitive discharges
• 112 sec, Te20eV, ne6x1018m-3
• Fluence averaged over plasmawetted area
• Rectangular cells 2.21020 D/cm2
• Shaped cells 4.21020 D/cm2
• ? Post-exposure analyses with
• SIMS, Dektak, NRA and EPMA
• on all sides of poloidal and
• toroidal gaps.

SOL Plasma
Poloidal direction
Toroidal gaps
Poloidal gaps
Toroidal direction
Rectangular cells 10x10x15 mm
Shaped cells 10x10x12(15)mm
20o
Gaps 0.5 mm
A. Litnovsky
22
Fuel accumulation in toroidal gaps

• Toroidal gaps exposed deep in plasma

Shaped geometry
Rectangular geometry
D/C ()
NC, ND
NC, ND
D/C ()
N?, 1016 at./cm2
N?, 1016 at./cm2
ND, 1015 at./cm2
D/C ()
D/C ()
ND, 1014 at./cm2
Distance from the top of a gap, mm
Distance from the top of a gap, mm
DS3.461015 at/cm2
Plasma-closest edge
DS1.461015 at/cm2
Plasma-closest edge
Less fuel in gaps of shaped cells
23
Different fuel retention in the poloidal and
toroidal gaps
Plasma flow
Ongoing research short summary

• Poloidal gaps

Plasma-
open side
shadowed side

• Toroidal gaps

A. Litnovsky et al., Phys. Scr. T 128 (2007) 45
24
Deposition and Fuel Inventory in Castellated
Beryllium Limiters from JET
Be toroidal belt limiter Operation 1989
1992 56 000 s of plasma (16 hours) 2000
castellated blocks.
Be
Be
Be

• Studies performed with Ion Beam Analysis on two
  tiles
• Castellated grooves both sides of 6 grooves
• Side surface between the tiles
• Top surfaces of tiles.

M. Rubel et al
25
Top and Side Surfaces of Cleaved Beryllium
Limiter Tiles
Cleaved limiter blocks mounted in the chamber for
IBA

• Bridging of some gaps by molten Be.
• Grooves are not filled with Be.

M. Rubel et al.
26
Deposition in the Castellated Grooves of the
Beryllium Limiter Tiles

• Messages
• Deuterium deposition in the castellation is
  always associated with Carbon.
• Short decay length of deposition in the
  castellation l 1.5 mm.
• D content in the castellated groove does not
  exceed 8 x 1017 cm-2.
• No deuterium detected in bulk beryllium.

M. Rubel et al.
27
Deposition at divertor (MkIISRP, 2001-2004)
- Carbon inner total 625 g (1.0 g/cm3) 3.1x1025
C-atoms 3.7x1020/sec, D/C from NRA ? 30g
D Injected D 1800g, retention fraction 1.7 -
Carbon outer 507 g 2.5x1025 C
3.1x1020/sec Deuterium D/C from NRA ? 13
g retention fraction 0.7 ? Total D retention
43 g 2.4 of injected No SRP included
72µm
67cm3
26µm
7µm
44cm3
10µm
10µm
38µm
464cm3
33µm
19cm3
99cm3
41µm
105cm3
24cm3
130µm
60g on louvre
233cm3
17cm3
18µm
300µm
32µm
200µm
22µm
Thicknesses surface analyses Volumes
integration over torus
J Likonen et al
28
Deposition at OPL and IWGL (MkIISRP, 2001-2004)
J Likonen et al
29
Conclusion for MkIISRP, 2001-2004

• Deposition at divertor very asymmetric (70
  inner divertor, 30 at the outer)
• Main D retention at divertor
• OPL limiters have minor contribution to D
  retention
• IWGL have most likely a small contribution
• D retention 10 (MkIIA), 4 (MkIIGB), 3
  (MkIISRP, SRP analysis under way)
• Long term fuel retention 13 (TFTR), 8
  (TEXTOR), 5 (DIII-D) and 4 (AUG with C)

J Likonen et al
30
upper divertor
AUG Wall areas and analysis methods

• Analysis methods
• NRA D(3He,p)a - 1000 keV D inventory in 2
  µm - 2500 keV D inventory in 10 µm
• Marker stripes for RBS - Deposition of B, C
  (talk on 9.5.2007)
• SIMS
• Data for 2002-2003 and 2004-2005 Campaigns ?
  Carbon dominated machine

upper PSL
inner heat shield
ICRH limiter
lower PSL
inner divertor
pump duct
outer divertor
roof baffle
31
Deuterium retention in 20022003
Retention
Fuelling
from (BC), assuming D/(BC)0.4
Long-term D retention 34 of fuelling Majority
on divertor tiles (50-60), followed by remote
areas (20)
Gas balance (V Mertens 2003) 1020Marginal
agreement, taking error bars into account
M Mayer et al, Nuc Fus 2007
32
Tungsten machine Preliminary results Analysed
tile for D inventory

• Exposed for 2 campaigns 2003 2005 about
  7000 plasma seconds
• Thin W-coating with 4 µm thickness using PVD

• Surface temperature close to RT, with maximum
  of 500 K
• D/W 20 30 at surface ? trapping with C
  241021 C/m2
• D/W 0.01 0.1 in W-layer

M Mayer et al
33
Evaluation of the total amount of D retained in W
D-inventory 1.51021 D/m2 AUG wall area 72 m2?
11023 D-atoms 0.3 g D-input in 2 campaigns
160 g ? Retention with W-walls lt 0.2 of
input (Retention with C-walls 4 of input)
M Mayer et al
34
Summary Comparison Gas balance-Post mortem

• - Post mortem analysis confirm that the long term
  retention in the PFCs is low.
• AUG (4 in C, less in W), JET (3-4), TEXTOR
  and TS (DITS) 8
• Post mortem analysis is representative of the
  averaged over a campaign of a small area
  (difficult for extrapolation flakes in JET
  during DTE ) cumulative effects of thermal
  release (plasma ops.), GDC, disruptions, .. (eg
  JET Averaged power with MkIIGB4MW, and averaged
  fuel rate 5x1021Ds-1 in 2007)
• Retention in PFCs, mainly in the divertor (30
  Outer leg/ 70 inner leg)
• - Retention in gaps always associated to carbon,
  typical length 4mm

• Gas balance Long term retention evaluated in the
  range 10-20 for carbon machine.
• Analysis generally carried out for plasma
  conditions different from averaged
• Retention increases with recycling (gas/NBI
  injection) and the ELMs (Type III to Type I)
• eg interesting pulse5 times the average JET
  15-20MW, and fuel rate 2.5x1022Ds-1
• ?Long term recovery between pulses is negligible
  in the overall balance
• Gas balance or Post mortem analysis Carbon leads
  to high retention

Further results and experiments (main) - AUG
analysis of the retention in a full W machine?
answer to the question of C - Tore Supra DITS
project? Where is the D trapped ? In the carbon
structure ? - JET Preparation of the ILW (no
carbon), reference pulses to be quantify in
Carbon - Complementary experiments of post
mortem analysis