1. L.R. Baylor 1, N. Commaux 1, T.C. Jernigan 1, S.J. Meitner 1, N. H. Brooks 2, S. K. Combs 1, T.E. Evans 2, M. E. Fenstermacher 3, R. C. Isler 1, C. J. Lasnier 3, R.A. Moyer 4, T. H. Osborne 2, P.B. Parks 2, P.B. Snyder 2, E.J. Strait 2, E.A. Unterberg 1, A. Loarte 5 1 Oak Ridge National Laboratory, Oak Ridge, TN, USA 2 General Atomics, San Diego, CA, USA 3 Lawrence Livermore National Laboratory, Livermore, CA, USA 4 University of California San Diego, San Diego, CA, USA 5 ITER Organization, Cadarache, France Demonstration of ELM Pacing by Pellet Injection on DIII-D and Implications for ITER Presented at the 24 th IAEA Fusion Energy Conference San Diego, CA 11-Oct-2012 LB/IAEA/Oct12 1

2. D 2 pellets injected in ITER configuration 12x increase in f ELM with 12x reduction in ELM  W H factor maintained Impurities greatly reduced DIII-D has Demonstrated Viability of Pellet ELM Pacing Mitigation for ITER Talk to cover motivation, pellet technology, and physics results. Pellets 0 2 LB/IAEA/Oct12

3. ITER ELM size needs to be < 0.7 MJ for low erosion Implying ELM triggering is needed > 30Hz if f ELM *  W = Const (>30x increase in natural ELM frequency) ELM Erosion of the Divertor is an Important Issue for ITER [A. Loarte, et al., IAEA 2010 and ITR/1-2 this conference ] No Erosion Some Erosion Significant Erosion 50 Shots Significant Erosion 10 Shots  Can pellets be used to trigger frequent ELMs on-demand to reduce  W? [Zhitlukhin et al., JNM, 2007] 0.5 1.0 1.5 3 LB/IAEA/Oct12

4. New ITER like LFS X-point injection line installed on DIII-D. All 1.3mm pellets from X- point and Midplane reliably trigger ELMs – no fueling. HFS 1.3mm pellets trigger ELMs, but provide fueling. DIII-D Pellet ELM Pacing Experiment Performed with D 2 Pellets Injected from Low Field Side Midplane and X-point D3DPI 1.3mm pellets ~100-150 m/s (2 mbar-L, 1x10 20 atoms per pellet) ORNL 3 barrel injector 20 Hz per gun X-pt HFS Mid 4 New ITER like X-point Injection line added in 2011 LFS Pellet Guide Tube x3 ITER [S. Maruyama, ITR/P5-24] LB/IAEA/Oct12

5. ELMs Triggered On-demand by 60 Hz Pellets at 12x the Natural ELM Frequency 147690 5 LB/IAEA/Oct12 ITER shape,  N ~1.8, q 95 =3.4

6. ELMs Triggered On-demand by 60 Hz Pellets at 12x the Natural ELM Frequency Pellets 147690 6 ITER shape,  N ~1.8, q 95 =3.4 ELM Pacing demonstrated 12x the natural ELM rate. 60 Hz 1.3mm pellets injected from LFS mid (20 Hz), X-pt (40Hz) at 100-150 m/s. Much lower energy loss ELMs observed with the pellets. Negligible fueling and reduction in H-factor observed. Reduced Ni in the plasma core with high rep rate pellets, f ELM *  E > 15. LB/IAEA/Oct12

7. X-pt Ave Mid Ave Divertor Heat Load from Pellet Triggered ELMs is Significantly Smaller than Natural ELMs 12x Increase in ELM Frequency Results in >12x Decrease in Energy Deposited in the Divertor x12 Natural Ave 7 The energy deposited (log scale) in the divertor from the IR camera data is significantly lower than the 5 Hz natural ELM case. LB/IAEA/Oct12

8. 0 5 10 15 20 12345 Mid 20Hz X-pt 20Hz The energy deposited (log scale) in the divertor from the IR camera data is significantly lower than the 5 Hz natural ELM case. The energy distribution spread is somewhat larger for the X-pt pellets.  Wtot_IR (kJ)  of ELMs X-pt Ave Mid Ave x12 Natural Ave 8 Divertor Heat Load from Pellet Triggered ELMs is Significantly Smaller than Natural ELMs LB/IAEA/Oct12

9. 0 5 10 15 20 25 30 35 40 45 010203040506070  W_IR (kJ) Pellet Frequency (Hz) Natural Pellet Divertor Heat Deposition per ELM Decreases as the Pellet Frequency Increases 9 f ELM *  W= Const The average energy deposited in the divertor per ELM decreases with increased pellet frequency. LB/IAEA/Oct12

10. 0 5 10 15 20 25 30 35 40 45 010203040506070  W_IR (kJ) Pellet Frequency (Hz) Natural Pellet 0 100 200 300 400 500 600 700 010203040506070 Outer Inner Peak_IR (W/cm 2 ) Pellet Natural Pellet Frequency (Hz) f ELM *  W= Const The average energy deposited in the divertor per ELM decreases with increased pellet frequency. The peak heat flux in the outer divertor is greater than the inner divertor. − The IR camera data is integrated assuming axisymmetric deposition Divertor Heat Deposition per ELM Decreases as the Pellet Frequency Increases f ELM * q peak = Const 10 LB/IAEA/Oct12

11. Reduction of Impurities Observed with Pellet ELM Pacing – A Function of the Pellet Frequency Reduced high-Z and lower Z impurities during pellet ELM pacing. STRAHL calculations indicate a reduced Ni density explains reduced emission. Higher frequency pellets reduce impurity levels. Impurities and natural ELMs return within  E. Ni 26 Intensity Plasma Center O VIII Intensity Plasma Edge Divertor D  Inner Strike Point No Pellets 60 Hz Pellets 20 Hz Pellets 40 Hz Pellets Time (s) 1.8 2.0 2.2 2.4 2.6 2.8 3.0 0 1 2 0 2 4 6 0 4 8 12 11 LB/IAEA/Oct12

12. All Impurities are Reduced with Pellet ELM Pacing as a Function of the Pellet Frequency SPRED and CX show a reduced impurities with pellet ELM pacing. (Data shown from 0.8s after the start of pellets) Consistent with earlier gas puffing results where impurities were entrained in the divertor with enhanced SOL flow. [Wade, et al., JNM 1999] Atomic No. Z Intensity (a.u.) B5 C6 08 Fe23 Ni26 f ELM 12 LB/IAEA/Oct12

13. Pellet ELM Pacing Produces a Narrower Pedestal Width Total pressure profile from the pellet ELM pacing has a narrower pedestal width and lower height. Note: Pellet case averaged over 200ms. − Lower Z eff and higher rotation shear help to compensate for lower pedestal pressure.  P tot (kPa) No-pellet before ELM Pellet Paced (average over ELMs) No-pellet after ELM  T e (keV) n e (10 20 m -3 ) No-pellet after ELM No-pellet before ELM Pellet Paced No-pellet before ELM No-pellet after ELM 13 LB/IAEA/Oct12

14. Natural ELMs Occur When Pedestal Width Becomes Unstable to Peeling Ballooning Modes Natural ELMing Case Pre-ELM Post-ELM Unstable Stable 14 The boundary for peeling and ballooning stability in this plasma configuration is calculated by ELITE (P. Snyder, NF 2007). LB/IAEA/Oct12

15. Pellet ELM Pacing Produces a Narrower Pedestal Width - Stable to Peeling Ballooning Modes The boundary for peeling and ballooning stability in this plasma configuration is calculated by ELITE (P. Snyder, NF 2007). Pellet case is far in stable region, ELMs likely triggered by local effect. Natural ELMing Case Pellet ELMing Case Pre-ELM Post-ELM Unstable Stable Pellet 15 LB/IAEA/Oct12

16. Fast Camera Images of Pellets Triggering ELMs Show Individual Filaments Being Perturbed X-point pellet exciting a filament as it enters the plasma. ELM is subsequently triggered. Midplane pellet triggers ELM while X-point pellet fragment enters plasma. 147691 - 2142 147691 - 2678 D  Filter Images consistent with hypothesis of a local ballooning mode triggered by pressure gradient in pellet cloud. Mid X-pt 16 LB/IAEA/Oct12

17. Minimum Pellet Size Needed for ELM Triggering is Under Investigation What are the minimum pellet size and speed requirements to reliably trigger ELMs ? Small fragments and 1mm slow pellets do not trigger ELMs. ≥ 1.3mm do trigger ELMs. New smaller 0.9x1.3mm pellets (~60% of 1.3mm size) also trigger ELMs. Non-linear MHD modeling is consistent with observations of local ballooning instability and minimum pellet size. [Futatani ITR/P1-22] ELM triggered No ELM triggered ? ELM triggered No ELM triggered 1.3, 150 17 LB/IAEA/Oct12

18. NN 1.8  * ped B  (T) 5.3 n e _ped (10 19 m -3 ) 8.0 LL x f ELM 30 10 5 2 q 95 * ped  ped (cm) 6 4 1.3 n e/ n GW 6.5 0.6 0.85 1.0 ITER parameters P/P H 1. 1 0.7 3.4-3.7 0.58% 0.48% 0.05% 0.05 0.16 1.9 3.0 DIII-D Pellet ELM Pacing Operating Space Compared to ITER Except for magnetic field and  * the operating space on DIII-D is comparable to that expected for ITER. DIII-D parameters 18 LB/IAEA/Oct12

19. Summary – DIII-D has Tested Key Principles of Pellet ELM Pacing for ITER Pellets trigger ELMs on demand with lower X-point injection. ELM frequency increase of 12x achieved with > 12x reduction in divertor ELM energy deposition. No confinement degradation Impurities screened out ELMs triggered locally before pellet reaches pedestal top – enables smaller pellets and lower fueling Future optimization and extension to higher frequencies with smaller pellets is planned. 19 LB/IAEA/Oct12

20. Backup Slides 20 LB/IAEA/Oct12

21. Pellet Ablation of X-pt Pellets Indicates ELMs are Triggered Within 1cm of Entering the Plasma ELM is observed by magnet loops ~50  sec after pellet starts to ablate – implies ELM starts at 1cm penetration.  =0.95 R-2 LFS Trajectory Distance to  =.95 location is 4cm along R-2 trajectory. DIII-D 149301 21 Time (ms) 2965.0 2965.2 2965.4 2965.6 2965.8 Pellet Ablation (a.u.) dB/dt (a.u.) n e L (10 14 ) 0 5 10 2.50 2.52 2.54 2.56 0 2 4 0.9x1.3mm D 2 Pellet LB/IAEA/Oct12