1. 1 Estimating the upper wall loading in ITER Peter Stangeby with help from J Boedo 1, D Rudikov 1, A Leonard 1 and W Fundamenski 2 DIII-D 1 JET 2 10 th ITPA SOL/divertor meeting Avila 7-10 January, 2008 U NIVERSITY OF T ORONTO Institute for Aerospace Studies 1 U NIVERSITY OF T ORONTO Institute for Aerospace Studies

2. 2 Motivation How to shape the front faces of the blanket modules – particularly at the top where a 2 nd X-point tends to form – in order to handle the power load due to t, the time-averaged (through ELMs), normal operation, plasma parallel power flux density?

3. 3 Questions What is the maximum t at: (a) the 2 nd separatrix? (b) the outside midplane? What fraction of the maximum t is due to ELMs? What is the uncertainty band on the estimated maximum t ? ____________________________________________ and what light can be shed on these questions by DIII-D experiments?

4. 4 ITER Magnetic configuration in normal operation The 2 nd separatrix can be as close as 3 cm from the 1 st separatrix, allowing for mechanical uncertainty (1 cm). The most challenging location is therefore at the top. At the bottom the 2 nd separatrix strike points are in the divertor. Chris Lowry

5. 5 Estimates by Lowry (presented at WG8 meeting 8/10/07) Type of InteractionUnits2001PIDLatest Proposal Outer Midplane Radiation┴MWm -2 0.5 Power Conducted between ELMs װ MWm -2 None 3 Power Conducted by ELMs װ MWm -2 None0.932.5 (3.4) Energy Conducted by ELMs װ MJm -2 None0.19 (0.93)0.06 (1.7) Near second X-point Radiation┴MWm -2 0.5 Power Conducted between ELMs װ MWm -2 None 5 MWm -2 Power Conducted by ELMs װ MWm -2 None3.8 33 MWm- 2 Energy Conducted by ELMs װ MJm -2 None0.77 (3.8)0.8 (17) Baffle Region MARFE┴MWm -2 1.31.40.6 Lowry estimates based on work by Loarte and for ELMS, work of Fundamenski, Pitts, et al.

6. 6 2 nd (upper) X-point region

7. 7 Recent relevant DIII-D papers “Edge-localized mode dynamics and transport in the scrape-off layer of the DIII-D tokamak”, Boedo, et al, PoP 12 (2005) 072516. “Far SOL transport and main wall plasma interaction in DIII-D”, Rudakov, et al, NF 45 (2005) 1589.

8. 8 A fast reciprocating probe at the DIII-D outer midplane. Rudakov

9. 9 Fraction of particle and power wall fluxes due to ELMs Rudakov particles power fraction n e /n GW Power assumedwith γ = 7 and T e = T i. n e /n GW

10. 10 Radial decay of peak burst ELM n e and T e Boedo n e /n GW = 0.8n e /n GW = 0.45 TeTe nene TeTe nene

11. 11 v perp of ELM density front measured by reflectometry in DIII-D. v perp decreases from ~ 500 m/s at separatrix to ~ 120 m/s near the wall.

12. 12 Comparing the DIII-D radial decay of ELM filaments with the Fundamenski model “A model of ELM filament energy evolution due to parallel losses”, Fundamenski, R A Pitts and JET EFDA contributors, PPCF 48 (2006) 109–156. This is a more sophisticated, numerical treatment of an analytic model in: “Cross-field blob transport in tokamak scrape-off-layer plasmas”, D’Ippolito, Myra and Krasheninnikov, PoP 9 (2002) 222.

13. 13 Simple analytic model: sheath-limited, T e = T i, no time delay, etc. particle equation: energy equation: gives:

14. 14 Between-ELM radial profiles. DIII-D. Shot 110494. Low density case: n e /n GW = 0.45. n e [10 20 m -3 ] (thomson) T e [keV] (thomson) T i [keV] (CER) Tony Leonard

15. 15 DIII-D. Shot 110494. Low density case: n e /n GW = 0.45. Radial variation of the ELM burst peak T e and n e. temperaturedensity

16. 16 DIII-D. Shot 110497. High density case: n e /n GW = 0.8. Radial variation of the ELM burst peak T e and n e. temperaturedensity

17. 17 Conclusions An initial and limited comparison of DIII-D ELM filament data with the Loarte-Fundamenski-Lowry method of estimating power loads on the ITER wall was broadly and roughly confirmatory. Comparisons for more DIII-D shots needed. The DIII-D comparisons use measured. How reliably can we predict for ITER? Can we reliably estimate the maximum t at the 2 nd separatrix strike point to better than a factor of 3? 5? 10?