043-05/rs

PERSISTENT SURVEILLANCE FORPIPELINE PROTECTION AND THREAT INTERDICTION

BPO Input to the ITER Design Review for In-vesselComponents and Discharge Evolution

prepared byR.D. Stambaugh

APS-DPP Town Meeting on ITER Design Review

November 14, 2007Orlando, Florida

043-05/rs

043-05/rs

BPO Input to the ITER Design Review onDischarge Evolution

• Leaders of Work in the US– Tom Casper– Dave Humphreys– Chuck Kessel

• Issues– High li versus low li startup– Vertical stability– Discharge Evolution and PF current limits

2WG1 presentations 17/09/07

The ITER PF systemThe ITER PF system

Differential currents in CS

Elongation andtriangularity

Elongation,Triangularity and X-point

u

VerticalField

Shaping

3WG1 presentations 17/09/07

PID Requirements for the PF coilsPID Requirements for the PF coils

• The four “design” and three “assessed” scenarios, ofsection 3.2 of the PID, have defined the requirements forthe PF coils. The number of turns in each and a maximumconductor current of 45kA lead to the following, maximumAmpere turns (Table 4.1-11 of the PID):

PF1 PF2 PF3 PF4 PF5 PF6Max. coil capacity(MA-t) 11.21 4.35 8.33 7.61 9.90 19.3

Scenario Evaluation Control - evaluate controller performanceduring scenario modeling

• Scenario modeling mode – used fixed boundary with prescribed reference shapes to studyperformance of scenarios (e.g. earlier performance shown)

» Robust operation and fast execution» Controller (shape and vertical position) not required

• Switch to free-boundary mode to re-solve equilibrium using fixed-boundary results» ne, Te, Ti from fixed boundary transport evolution» Find new separatrix with mean-square constraint from fixed-boundary solution» Evaluate external flux linkages – “back out” PF and CS coil current demand for shape evolution

Corsica Simulations: Nominal Control During Rampup;

JCT2001 Controller, Hi Density, Hi-Resistance Wall

• Scenario2 “small bore”, n/nG=0.8

• Maximum li(3)~1.2 [~1.0 at end of ramp]

• Good nominal control during rampup

• PF1 violates current limits briefly

• Growth rate near control limit !~12-13 rad/s

DIII–D Has Simulated the ITER Baseline Startup Scenario

NATIONAL FUSION FACILITYDIII–D

G. Jackson, APS07

Constant q95 Baseline Scenario

Diverted

q95

Vp(m3)

Gapout (cm)

Te0(KeV),

li

128317

Time (ms)

Disruptionn/ncrit= -0.91

t=100ms

0 400 800 1200

36

9

12

05

1015

1.0

1.4

1.8

0.4

1.0

1.6

t=800ms

ITER design limit

0

0.5

1.0Ip(MA)

DIII–D has Simulated the ITER Baseline Startup Scenarioand Helped Develop an Improved ITER Startup

NATIONAL FUSION FACILITYDIII–D

G. Jackson, APS07

Constant q95 Baseline Scenario

Diverted

q95

Vp(m3),Gapout (cm)

Te0(KeV)

li128317,129789

Time (ms)

Disruptionn/ncrit= -0.91

t=100ms

0 400 800 1200

36

9

12

05

1015

1.0

1.4

1.8

0.4

1.0

1.6

t=800ms

Large bore scenario

ITER design limit

0

0.5

1.0Ip(MA)

• li (large bore, red) is near or below ITER design limit

• Higher qmin (delayed saw- teeth) with large bore scenario

Full Bore, EarlyDiverting, and Late H-

modeAE = aperture expansionFB = full bore initial plasmaLow level heating is applied in the FB casesafter diverting

PF Coil limits notexceeded, except PF4,which is due to controlleraction

40 MW injectedH-mode onsetIp ramp ends at 100s

Example of How H-mode TransitionAffects li and PF Coil Currents

ITER Vertical Stability Analysis and Experience from USTokamaks for ITER Design Review

D. Humphreys, T. Casper, M.

Ferrara, D. Gates, I. Hutchinson, C.

Kessel, J. Leuer, L.D. Pearlstein, M.

Walker, A. Welander

Presented at

STAC Briefing

Video Conference

November 1, 2007

VS1

CircuitNew Coils inNew Coils in

VS2 CircuitVS2 Circuit

ITER System Geometry

Why ITER07 Has Overall Degraded Vertical Control

Margins Relative to ITER EDA

• Size was reduced to reduce cost

• Coils were consolidated (number

reduced) to reduce cost

• Plasma elongation was increased to

recover some performance

• Approximately fixed shielding depth

caused PF coils to move farther away

from plasma as fraction of minor

radius

• Power supply capabilities and coil

operating points were not increased

to compensate sufficiently

Nominal Operating Point Issue: Analytic Model and

C-MOD Data Show li(3)~1.2-1.3 Attainable at q95=3.0

• Analytic model relating li,

kappa, q95 matches

experimental experience

• Increasing q95 increases li

range

• If ITER operates at low

current (high q95) the

elongation must be

reduced in order to

maintain vertical control

0.5

1

1.5

2

2 3 4 5 6 7 8

li(3)high-li

li(3)sw

li(3)no-sw

li(3)low-lil i

(3)

1-5MA, 1.7-4T, 45-90m3ε=0.2-0.3, κ=1.4-1.8,δ=0-0.2, βp

ITER Vertical Control Analysis with TSCUsing JCT2001 controller, Kavin Dec. 2001

Using plasma parameters from worst case identifiedby Corsica simulation (t=80 s)

Turn controller voltages to zero, allow plasma to driftto a specified Z, then turn controller back on

Vmax(PF2,3,4,5) = 5.7 kV

ITER Simulations Show ~50% Improvement in "Zmax

with Each Upgrade from VS1 at 6 kV

• TokSys linear model calculations with power suppliesonly (no controller) for !Z ~ 11 rad/s

• Corsica nonlinear calculation of "Zmax is

reasonably close to TokSys value

9.06.0VS1+VS2

6.34.2

(Corsica: 3.5)

VS1

"Zmax

(cm)

for 9 kV

"Zmax

(cm)

for 6 kV

Control

Circuit

5WG1 presentations 17/09/07

Increased requirements for PF systemIncreased requirements for PF system

• It is required that the maximum current in PF2, PF5, PF6and, if possible, PF4 increase by 10%.

• Further measures to improve shape and position controlcould include an upgrade of the CS tie-plates and anintegrated design of the dome shape to optimize pumpingand shape control.

• The effects of these modifications are to allow headroomfor control variations and to increase the range of li at15MA.

• The best means of achieving this PF performanceincrease should be chosen by the IO.

043-05/rs

BPO Input to the ITER Design Review onIn-Vessel Components

• Leaders of Work in the US– Richard Nygren– Tom Rognlein– Peter Stangeby– Dennis Whyte

• Pulled together an assessment of 12 issues

1

BPO Input to ITER WG1 PFC Task Provided by U.S. BPO members (lead author listed first)

Issue 1 - Need for shaping the First Wall configuration, and requirements for additional

limiters for protection of the first wall. Stangeby (DIII-D), Nygren (Sandia)

Issue 2 - Characterization of heat loads to walls and limiters during startup. Stangeby (DIII-D), Rognlien (LLNL), Kessel (PPPL), Luce (GA)

Issue 3 - Impact of enhanced sputtering by RF-induced sheaths at antenna and B-field-line-connected divertor/wall surfaces. Wukitch (MIT), Myra (Lodestar), Lipschultz (MIT)

Issue 4 - Heating and melting of Be from distributed heat loads during disruption mitigation. Petrie (GA), Whyte (MIT)

Issue 5 - Document the tritium co-deposition, implantation and diffusion properties for Be, C, W, including the dependence on substrate temperature, based on both modeling and empirical extrapolation. Document the tritium removal options for the three candidate materials. Haasz (U. Toronto), Causey (LANL), Davis (U.Toronto), Skinner (PPPL), Whyte (MIT), Stangeby (DIII-D), Brooks (ANL), Doerner (UCSD)

Issue 6 - Examine the requirements on baking temperatures and tritium handling systems arising from periodic oxygen bake-out of the divertor and vessel and assess the changes needed in affected systems, including costs. Davis (U. Toronto), Willms (LANS), Doerner (UCSD), Haasz (U. Toronto)

Issue 7 - Examine the data on out-gassing molecular hydrogen and water from graphite walls with long pulse pumped divertor operation as a substitute for high baking temperature. Maingi (ORNL), Wade (GA), LaBombard (MIT), Whyte (MIT), Petrie (GA)

Issue 8 - Knowledge on dust production mechanisms and rates for tokamaks, and clean-up procedures. Sharpe (INEL), Skinner (PPPL), West (GA), Rudakov (UCSD)

Issue 9 - Properties (melting temperature, sputtering yield, hydrogen retention) of mixed materials expected in ITER, e.g., Be-W, Be-C, and C-W. Doerner (UCSD), Nygren (Sandia), Brooks (ANL)

Issue 10 - High heat-flux capability of Be for the divertor including possible localized RF-sheath acceleration of ions. Smith (GA), Petrie (GA)

Issue 11 - Damage to various types of W and Mo from transients and high energy Helium ions (e.g. from RF-sheaths). Lipschultz (MIT), Baldwin (UCSD), Causey (Sandia), Nygren (Sandia)

Issue 12 - Modification of thermo-mechanical properties of C, Be, W by neutron radiation. Nygren (Sandia), Wong (GA), Whyte (MIT), Causey (Sandia)

043-05/rs

Present ITER Plan Uses a Mixture of PlasmaFacing Materials

• Carbon at Divertor Strike Points– Can handle ELM, disruption heat pulses– Tokamak high performance basis on carbon– Tritium codeposition problem

• Be main chamber first wall– Plasma most vulnerable to main chamber

wall materials, low Z most tolerable.– May melt in a mitigated disruption

• Tungsten in lower heat flux divertor zones– Low erosion in reactor– Melting, cracking– Plasma has very low tolerance for W

• Mixed materials complicates situation– Little known about mixed materials T retention– Harder to understand what is going on– Coatings needed on metals (IT/1-4)but not

carbon (OV/1-4)

2

Issue 1: Need for shaping the First Wall configuration, and requirements for additional limiters for protection of the First Wall.

Stangeby (DIII-D), Nygren (Sandia) 1. Need to consider parallel heat flux to first wall. Edges.

Alignment of Blanket Modules. Shape modules or raise up some “protection limiters.”

2. Limit the number of BMs that have substantial damage risk.

Reduce remote handling requirements. 3. Need for improved experimental SOL measurements. 4. Protection limiters could also constitute attractive start-up

and ramp-down limiters and off-normal event protection.

3

Issue 2: Characterization of heat loads to walls and limiters during startup. Stangeby (DIII-D), Rognlien (LLNL), Kessel (PPPL), Luce (GA) Issue: Can the presently designed two movable port limiters (PL) accept the plasma heat load during start-up and ramp-down without excessive erosion or melting? Present design requirements are: “15 MW total power to the limiters. Protruding limiter may receive maximum power of 9 MW.” [ITER Project Integration Document, 26 Jan. 2007, p. 88]. Results: Design requirement of 9 MW to single PL will yield a peak heat-flux in excess of the limit of 8 MW/m2. Transport modeling of basic ohmic reference scenario indicates ~5 MW produced within the start-up discharge, for which the PL’s are probably adequate. However, for other standard scenarios (e.g., hybrid), substantially more input power is expected and coupled with possible asymmetries, the heat-load to PL heat-load capability may well be unacceptable.

IIC-STAC Meeting, Cadarache, 5-7 November 2007

10

Proposed FW Power Loads - Steady State

Parallel (_)

Perpendicular (_)

~1°

81515MW_Start-up & Ramp-down

0.61.41.3MWm-2_MARFE

Baffle Region

0.8 (17)0.77 (3.8)NoneMJm-2_Energy Conducted byELMs

333.8NoneMWm-2_Power Conducted byELMs

5NoneNoneMWm-2_Power Conductedbetween ELMs

0.50.50.5MWm-2_Radiation

Near second X-point

0.06 (1.7)0.19 (0.93)NoneMJm-2_Energy Conducted byELMs

2.5 (3.4)0.93NoneMWm-2_Power Conducted byELMs

3NoneNoneMWm-2_Power Conductedbetween ELMs

0.50.50.5MWm-2_Radiation

Outer Midplane

LatestProposal

2006/72001UnitsType of Interaction

IIC-STAC Meeting, Cadarache, 5-7 November 2007

9

In-Vessel Component Power LoadsIn-Vessel Component Power Loads- Assessment / Impact- Assessment / Impact

1a. Design to shield edges,holes, gaps to be finalized

1b. Polodal limiters or shapedFW proposed.

1c. ELM’s mitigation required.

Note – No VV design impactexpected.

1. See detailtable

1. See detailtable

1. Power load requirementsto the FW need to be revisedbecause of 1) recentexperience, 2) newunderstanding, 3)clarification.

(DCR-074 – Accepted forstudy)

2001BaselineRequirement

Design Assessment /Impact

ProposedChange /

Clarification

Issue

Shaped FW ConceptsShaped FW ConceptsShaped FW Design Developed in 2005Shaped FW Design Developed in 2005

(Note - 2001 FW was a flat design)(Note - 2001 FW was a flat design)

IIC-STAC Meeting, Cadarache, 5-7 November 2007

13

Plug Type Hydraulic Connection ConceptPlug Type Hydraulic Connection Concept

- Good access (+)

- Conventional technology forcutting and welding (+)

- Concept appears consistentwith FW replacement in VV (+)

- Large hole may not be consistentwith space available in current design

- Shielding must be confirmed

- Concept not demonstrated

FeaturesIssues

4

Issue 3: Impact of enhanced sputtering by RF-induced sheaths at antenna and B-field-line-connected divertor/wall surfaces

S.J. Wukitch and B. Lipschultz, MIT Plasma Science and Fusion Center D. D’Ippolito and J.R. Myra, Lodestar Research Corporation One of the leading models for ICRF induced impurity production is enhanced sputtering from RF rectified sheaths that have substantially higher sheath voltages than expected for thermal sheaths (~3T) and in excess of the threshold energy for deuterium (as well as higher-Z impurities) sputtering of tungsten. A number of experimental observations are consistent with this sheath rectification model. The sheath rectification model also has potentially significant implications for ITER. Self-sputtering can … lead…to “avalanche” for high-Z impurities where the self-sputtering yield can exceed unity.

5

Issue 4: Heating and melting of Be from distributed heat loads during disruption mitigation.

Petrie (GA), Whyte (MIT)

a) It is possible, even likely, that sectors of the Be wall on the outboard midplane side could experience some transient surface melting during disruption mitigation

b) The uncertainties in the quantity of Be melting lie in the toroidal and temporal uniformity of the radiation pulse following radiative termination for disruption mitigation.

c) Best estimates yield ~50-100 microns Be melt layers forming for areas up to ~ 25% (200 m2). The fate of these layers is unknown due to the highly transient EM effects occurring simultaneously with the rapid termination.

d) The operational consequences of the melting, liquid movement and re-solidification are unknown since non-disruptive heat and particle loads to these main-wall surfaces is unclear. Likely result is that direct plasma contact with previously melted/distorted surfaces will result in further melting.

6

Issue 5: Document the tritium co-deposition, implantation and diffusion properties for Be, C, W, including the dependence on substrate temperature, based on both modeling and empirical extrapolation. Document the tritium removal options for the three candidate materials

Tony Haasz/lead (U Toronto), Jeff Brooks (ANL), Rion Causey (LANL), Jim Davis (U Toronto), Russ Doerner (UCSD), Charles Skinner (PPPL), Peter Stangeby (DIII-D and U Toronto), Dennis Whyte (MIT) Summary of key findings (i) Tritium retention due to implantation/diffusion in W and Be --

including neutron effects The retention limit in Beryllium is mainly due to codeposition. DIFFUSE calculations for Tungsten show T retention in a neutron –irradiated W will remain below the 350 g-T limit for several thousand discharges. In a full W ITER, based on upper-limit estimates, 330 g-T would be reached after several hundreds of discharges.

(ii) Tritium retention due to implantation/diffusion in carbon and

CFC Although the structure of CFCs allows deep diffusion and trapping of tritium, a saturation effect is expected at very high fluences (after 10-100 ITER discharges), which will limit the trapping capacity of CFCs to well below 350 g-T. In ITER with a CFC divertor target, tritium retention will be dominated by codeposition; the 330 g-T limit could be reached after a few hundred discharges.

(iii) Tritium retention due to codeposition in Be, C, W and mixed

materials In general, the tritium content in codeposits will be highest in C, somewhat smaller (but not much) in Be and smallest in W. The level of tritium in mixed codeposits will probably be between the levels expected in the pure materials.

(iv) Modelling results comparing T retention due to codeposition in

Be, C, and W REDEP/WBC modelling shows (i) generally high T/C codeposition, (ii) possibly high T/Be codeposition, and (iii) essentially negligible T/W codeposition.

7

(v) Estimates of tritium codeposition based on empirical scaling from present tokamaks Based on results from various all-C tokamaks, it appears that the rate of codeposition increases as the linear dimension of the device (i.e., the total C/s scales linearly with R), rather than with the plasma or wall surface area.

(vi) Tritium removal from codeposits via oxidation

Based on oxidation results for JET codeposits with ~50% Be/(Be+C), we would expect that thermo-oxidation at 623 K and 20 kPa O2 pressure would remove > 90% within a day – independent of the codeposit thickness.

(vii) Tritium removal from codeposits via photons and other

techniques Laboratory experiments have shown that tritium can be removed from JET and TFTR C and C/Be codeposits as gaseous T2 by laser heating. Tile heating and tritium outgassing may also be achieved by radiative termination of discharges. Tritium retained in codeposits can be converted to DTO by thermal oxidation (above) or oxidation in conjunction with glow or ion cyclotron resonance discharges

Haasz et al

US-B

PO: ITER

Summ

ary/WG

-1/Task 5/Topic #5

Updated - 2007sep13

- 8 -

Fig 1:

Retained am

ount of tritium as a function of cum

ulative discharge time. B

asic parameters for the

different cases are given on page 9:

9

Issue 6: Examine the requirements on baking

temperatures and tritium handling systems arising from periodic oxygen bake-out of the divertor and vessel and assess the changes needed in affected systems, including costs.

Davis (U. Toronto), Willms (LANS), Doerner (UCSD), Haasz (U. Toronto) Overview: Thermo-oxidation has the ability to remove T/C films from all locations in the vacuum vessel; however, the current ITER design is not optimized for this procedure. A. In-Vessel Systems ITER baking temperature Temperatures required to remove codeposits O2 Pressure B. Tritium Handling Systems Chemical Products of Thermo-Oxidation Properties of Highly Tritiated Water Impact on ITER Exhaust Method of Processing Tritiated Water 1) Thermo-oxidation at the current ITER baking temperature of 240ºC would be marginal, while at the temperature achievable by removing the coolant, ~ 350ºC, would be quite satisfactory. Alternatively, local heating of selected regions of the divertor is recommended. 2) For thermo-oxidation to be viable on ITER, it is likely that a highly tritiated water processing system, separate from the Tokamak Exhaust Processing system would be needed. The capital cost associated with installing equipment to process this tritiated water might range from $5M - $10M.

10

Issue 7: Examine the data on out-gassing molecular hydrogen and water from graphite walls with long pulse pumped divertor operation as a substitute for high baking temperature.

W. Gardner, D. Hillis, B. LaBombard, R. Maingi, M. Wade, D. Whyte

1. For relatively long pulse (5-30 sec) graphite-PFC diverted

machines, the cryopump can actually be used to degas the

wall and reduce wall particle inventory (DIII-D, JT-60U).

2. Removal of chemically adsorbed water requires

temperatures > 500 oC

11

Issue 8: Knowledge of dust production mechanisms and rates for tokamaks, and clean-up procedures

Phil Sharpe (INL), Charles Skinner (PPPL), Phil West (GA), Dimity Rudakov (UCSD)

1. Plasma-material interaction mechanisms generating dust in tokamaks are poorly understood.

2. Plasma operation conditions, plasma-exposed materials, particle wall loadings, etc. are obviously quite different for ITER.

3. Dust detection and removal systems for ITER are in the infancy stage of development

4. On-going R&D is clearly required for practically all aspects of dust in ITER.

12

Issue 9: Properties (including melting temperature, sputtering yield, hydrogen retention) of mixed materials expected in ITER, e.g. Be-W, Be-C and C-W.

Russ Doerner, Jeff Brooks, Richard Nygren and Charles Skinner. Executive Summary

It appears that the mixing of Be and C will be beneficial for ITER from an erosion viewpoint, but the lower melting temperature of Be2C could be a concern during off-normal events. A similar conclusion of W and C mixing holds. Thin mixed-material layers of either C, or Be, on plasma exposed W surfaces inhibit their blistering, but Be/W alloy formation reduces the melting temperature of material and may result in higher erosion rates of the tungsten component of the alloy.

The CFC divertor target plates in ITER operate at high temperature where the possibility of beryllium carbide is maximized. If the carbide forms, it will mitigate carbon erosion from those surfaces during normal operation and thereby reduce the problem of tritium retention with the eroded carbon.

If the divertor baffles in ITER were made of graphite, where the surface temperature during operation is lower, then one would expect a slower mitigation of their chemical erosion and hence the problem of tritium retention with the eroded carbon would be much more worrisome.

The present ITER design uses tungsten as the divertor baffle material. The surface temperature in this region is only in the 400°C range during normal operation. Even if a beryllium layer forms on the tungsten surface at this temperature, one expects little beryllide growth and a very small diffusion rate for Be in W. If, however, ITER were to switch to tungsten divertor plates, the higher surface temperature expected near the strike points would lead to concerns about possible beryllide formation and growth in these surfaces. The choice of CFC divertor plates, tungsten baffles, and a beryllium first wall, may be an acceptable choice from the existing, but limited, understanding of mixed-material effects .

13

Issue 11: Damage to various types of W and Mo from transients and high energy Helium ions B. Lipschultz, MIT Plasma Science & Fusion Center M. Baldwin, UCSD Center For Energy Research R. Causey, Sandia National Laboratories In current tokamaks with metal PFCs it is common that tiles have melted surfaces during the run period. The primary regions are\ where high-energy electrons (runaways) impact them. The melting on divertor components (leading edges) is believed to principally occur during disruptions. ELMs are less of an issue in current tokamaks. When the surface reaches high temperatures, even without melting, surface cracks form which may reduce the heat load capability Studies of the sources of high-Z impurities in C-Mod point towards surfaces outside the divertor as the source of Mo/W affecting the core (as opposed to melting or melted surfaces). The sources are a mixture of thermal and accelerated ion sputtering of those surfaces, such as rectification of ICRF electric fields to form high-voltage sheaths. Non-thermalized alphas can also be a source of damage to W surfaces. Many studies have identified deleterious effects on hot tungsten surfaces caused by He ions. At the operational temperature for ITER W surfaces surface blistering due to bombardment by either He or hydrogen isotopes may occur. Should ITER attempt an all W divertor, He ion induced effects are likely to become more severe as the W strike points will likely operate around 1500 K.

14

Issue 12: Modification of thermo-mechanical properties of C, Be and W by neutron radiation

R.E. Nygren, Sandia National Laboratories, Lance Snead, ORNL; Russ Doerner, USCD; Dennis Whyte, MIT

Issue: Neutron irradiation will degrade the thermal and mechanical properties of C, Be or W armor used in ITER through the increase in lattice defects and transmutations that produce He, that promote the growth of voids in W and Be, and produce tritium in Be. At moderate fluences (few dpa), graphite swells, and Be and W begin to show damage, such as tangles of dislocations and voids, that affect mechanical properties and trapping of tritium. The effects of neutron irradiation on bond joints and on the near surface of armor modified by deposition of material and ion damage are other concerns. Bulk Damage Effects Carbon Be W Near Surface Effects: Trapping Dust Effects on Joints