Thrust 02 ChitID 02.001 Bill Dorland 6/08/2009 16:02 Thrust 02 The suggested ties to 17 are real, but greatly overshadowed by the rest of the story. Namely, the whole point of thrust 17 is to avoid having to do all the hard work of thrust 2. This is by far the dominant interaction between these thrusts, I think. ChitID 02.002 Richard Callis 6/08/2009 16:04 Thrust 02 Owing to the sensitivity of insulating materials to the neutrons, interior coils are problematic in reactors like DEMO. If 3D coils are needed an element of the research plan should be to evaluate how close to the plasma the 3D coils need to be and what response times are required. ChitID 02.003 Neil Morley 6/08/2009 16:10 Thrust 02 You mention thrusts 9-12, but thrust 13 which includes the tritium breeding blanket and integrated first wall is one of the components that must take heat and EM load. A linkage here is probably important -- thin walled ferritic steel blanket full of liquid metal must be shown to take these loads and may help set your goal posts. They may also limit the response time of control systems by interposing large conducting structures between your sensor, coils, and the plasma. ChitID 02.004 Rob Goldston 6/08/2009 16:13 Thrust 02 I believe that the capability to predict disruptions is a big problem. My understanding is that JT-60U saw no signals whatsoever in advance of their high-beta disruptions. I think it is important to describe how difficult this problem is, and the status of the work to date, not just include it in a to-do list. ChitID 02.005 Rob Goldston 6/08/2009 16:16 Thrust 02 You can include the Thrust 12 device as a place to test disruption avoidance at Demo-relevant heat flux, with Demo-relevant first-wall components, and long-pulse / high duty factor. Probably much higher duty factor than planned for the Asian devices. Note that CTF/FDF is not a place to test this technology, since it cannot survive lots of disruptions. ChitID 02.006 Allan Reiman 6/08/2009 16:17 Thrust 02 It would be helpful to have a set of benchmarks to assess where we are in terms of disruptions, and to measure progress in dealing with disruptions. For this purpose, a few reactor-relevant equilibria should be targeted. Given the limited pulse lengths available on present day tokamaks, the equilibria should be run repeatedly to get a measure of the disruption frequency. It would not be unreasonable to devote a day of run time to repeatedly running some of these cases for 20 or 30 shots. Methods of avoiding and mitigating disruptions could be tested in this context as they are developed. ChitID 02.007 Allen Boozer 6/08/2009 16:17 Thrust 02 On viewgraph 13: Inventions and theoretical investigations are required to before disruption prediction and avoidance concepts can be tested. Said differently, concepts must be developed before they can be tested. At present there are few concrete suggestions especially for methods of ways of steering the plasma away from disruption prone regimes while maintaining high performance. The basic talk was very good. ChitID 02.008 Larry Baylor 6/08/2009 16:20 Thrust 02 No mention is made of the required reliability of the tools used to avoid and/or mitigate off normal events. Some mention of this seems warranted to make clear the engineering difficulty of implementing such systems. ChitID 02.009 Vincent Chan 6/08/2009 16:25 Thrust 02 I am surprised to hear the statement that we cannot "postdict" disruptions. There has been extensive work in understanding disruptions caused by approaching stability boundaries e.g. due to lock modes, NTMs and RWMs. Significant work has been done in validating the physics against experiments. We should be able to predict hence avoid or mitigate such instabilities. The more relevant question is what fraction of the disruptions in tokamaks is caused by "predictable" instabilities and what fraction is caused by accidental "junks" falling off the wall? And whether disruptions caused by "junks" can be mitigated? It seems to me this question can be answered with some dedicated effort within reasonable time. ChitID 02.010 Allan Reiman 6/08/2009 16:27 Thrust 02 I have been told that because disruptions do not cause problems on DIII-D, there is little incentive for operators to make any effort to avoid them. This greatly decreases the value of the disruption database on DIII-D. If the operators were asked to attempt to avoid disruptions on all shots where it is possible to do so without compromising the other goals of those shots, that would immediately render the DIII-D disruption database more valuable. Statistics on disruptivity in various regimes would become meaningful. The remaining disruptions could be analyzed as to causes, with a view to further reducing the incidence of disruptions. One could look at the data to determine whether precursors were present, or whether the imminence of the disruptions could have been detected in some other way. This would allow us to assess whether it is reasonable to expect that we will be able to avoid or mitigate disruptions. ChitID 02.011 Allan Reiman 6/08/2009 16:38 Thrust 02 There is a great deal of overlap between thrusts 2, 5 and 8. It would appear that they should logically be merged to form two thrusts. ChitID 02.012 John Wesley 6/08/2009 16:41 Thrust 02 The issue that some disruptions may not be 'predictable' is raised. It is arguable that all disruptions owed to internal causes should be predictable in the sense that some advance warning of onset is possible; questions do exist whether prediction lookahead time for certain internal causes will be sufficient for some avoidance or slow-acting mitigation methods. With regard to external causes, eg falling micro or macro objects, velocities are low enough that there may be sufficient 'initial interaction time' to be able to actuate 'fast mitigation' schemes. Such considerations put a premium on minimal detection + actuation delay strategies / technologies. ChitID 02.013 John Wesley 6/08/2009 16:52 Thrust 02 ITER current = 9 - 15 MA is a 'game modifier' with regard to achieving soft landing disruption avoidance or rapid shutdown disruption mitigation without inducing high levels of runaway electron.current. DM tests in present low-current and moderate-current tokamaks (eg JET) will not necessarily be definitive for ITER. ITER will be a 'first-of-kind' test of ability to avoid/mitigate disruptions while limiting or avoiding excess RE. Successful resolution of this issue in ITER is essential for DEMO with less-robust PFC designs. Ability to test various DM/RE-Avoidance methods in ITER plus PFC robustness is critical. ChitID 02.014 richard nygren 6/08/2009 17:02 Thrust 02 The technology requirements in regard to gas injection, pellets, etc. are embedded in the control of transients but do not seem to be explicitly called out at the level of the one pagers. ChitID 02.015 Jim Callen 6/08/2009 17:42 Thrust 02 Your viewgraphs are better than the 6 pager. The description on pages 5 to 6 of the 6 pager give a good description of what to do and how the U.S. is well-positioned for research in this area. But your one-pager does not include these issues. Shouldn't the one-pager end with a strong "punch-line" about what should be done and the possible U.S. role? ChitID 02.016 John Wesley 6/08/2009 17:51 Thrust 02 Also applies for Thrust 5. Even 'normal' fusion power and current shutdown in ITER is acknowledged to be a delicate coordinated action and control sequence. Faster, but not rapid shutdown for disruption avoidance without itself causing disruption looks even more difficult, and successful disruption avoidance or retreat strategies that work in present medium-scale tokamaks may or may not be similarity successful for ITER (or DEMO). Research for T2 and T5 need to assess extrapolation of technologies and strategies to ITER and DEMO. ChitID 02.017 richard nygren 6/08/2009 17:53 Thrust 02 I suggest that we all use the term "off-normal" only for events that are truly outside the expected operation, such as unexpected failures of components, as would be of interest for a safety analysis. ChitID 02.018 richard nygren 6/08/2009 17:55 Thrust 02 I would suggest that the implied technology in terms of RF antennae and equipment such as launchers for killer pellets and massived gas injection be clearly identified as enabling technology. ChitID 02.019 John Wesley 6/09/2009 10:49 Thrust 02 Need/intent to develop and test a 'broad portfolio' of disruption prediction means, plasma and hardware in present expts should be more explicitly stated. One or two prediction means is unlikely to be effective for all types and causes of disruption, and 'voting' or redundant predictions may be important to minimize false positives. Final complement and deployment for ITER TBD. ChitID 02.020 John Wesley 6/09/2009 11:05 Thrust 02 Need a range of ITER qualified mitigation strategies and/or hardware, plus effective mitigation capabilities for 'target' plasmas ranging from low-current to full-current OH, heated and burning full-current plasmas (9-15-17 MA) and during-shutdown lower-energy plasmas. Also after-thermal-quench 'already-disrupted' plasmas. RE threat may be common to all; methods for mass delivery or other RE mitigation may differ. Action/delay times may differ. Need some statement on this 'range of mitigation strategies' and likely need for multiple means 'on-ready' for ITER. ChitID 02.021 Tony Taylor 6/09/2009 14:05 Thrust 02 After listening to thrusts 9, 10 11, & 12, It is clear that ELM suppression techniques are assumed to have been developed elsewhere in thrust 2. Thrust 2 seems to imply heavy reliance on non-axisymmetric fields as a viable ELM suppression technique ---- Yet there have been some comments that ELM suppression by RMP coils may not be credible for DEMO. It is my opinion that removing the impulsive heat and particle fluxes from repetitive ELMs is the hihgest priority task for "taming the plasma material interface." Therefore, I think thrust 2 should add to the ELM suppression effort "exploration of innovative techniques" for ELM control with high pedestal pressure.