Published on June 7, 2019
1. 1MIT-ANS Apr 07 The Challenges of Plasma-Surface Interactions in Magnetic Fusion Dennis Whyte, MIT MIT ANS Seminar April 9, 2007
2. 2MIT-ANS Apr 07 Why don’t we have fusion reactors yet? x Worst Jobs in Science = Popular Science Oct. 2003
3. 3MIT-ANS Apr 07 Materials impose one of our biggest challenges in developing fusion energy • Magnetic fusion produces a thermonuclear volumetric energy source, where all energy must be extracted through a single, surrounding material surface. • The nuclear fusion fuel cycle and the plasma physics of heat and particle exhaust simultaneously place severe demands on material surfaces. • The strongly coupled interaction between fusion plasmas and the material walls challenges our ability to control and predict plasma-surface interactions (PSI). • Significant worldwide research progress and effort continues on the PSI requirements of the ITER prototype fusion reactor.
4. 4MIT-ANS Apr 07 Deuterium-Tritium fusion represents a nearly inexhaustible energy source. Fuels: Deuterium: abundant in sea water Tritium: Half-life~12 years…must be produced? Tbred / Tburned > 1 “Real” fusion fuel cycle: 6Li + D = 24He + 22.4 MeV 30 Million years of world energy demand in oceans 6Li + n --> T + 4He + 4.8 MeV + others
5. 5MIT-ANS Apr 07 D-T fusion requires a confined thermonuclear plasma • E > 10,000 eV ® ionization separates electrons and ions. • Coulomb scattering >> fusion reaction ® thermalized plasma T ~ 10,000 eV ~ 100,000,000 K • Far out of thermodynamic equilibrium in terrestrial environment ® plasma exhausts power at te ® need confinement.
6. 6MIT-ANS Apr 07 Fusion energy by magnetic confinement • Exploits Lorentz force Ø No confinement // to B. • Fusion energy balance: Ø 4/5 in neutron leaves ® energy Ø 1/5 in He++ confined by B to heat & sustains plasma. • Every Joule of fusion and exhaust power must be extracted through a single surrounding surface. B
7. 7MIT-ANS Apr 07 The toroidal “tokamak” is the most prevalent magnetic confinement geometry • Self-closing helical magnetic fields. • Stabilizing toroidal field produced by external coils. Ø Bf µ 1 / R • Plasma current provides confining poloidal field against Ñ-B drifts. I R
8. 8MIT-ANS Apr 07 Plasma pressure, confinement and stability set the requirements for a fusion reactor. Pfusion µ n2 T2 Power Balance Power Density Confinement Kink Stability Triple Product Energy Gain Self- sustaining Size& Field $$$ Q=5 “burning”
9. 9MIT-ANS Apr 07 Translation of Fusion Strategy • Build it big, because bigger things hold their heat longer! • Make the magnetic field as big as possible to hold onto those particles! • Heat the heck out of it to get 100,000,000 K and light the fire!
10. 10MIT-ANS Apr 07 et voila, ITER • Mission: Ø “Burning plasma” Q=10 • Reactor level fusion power Ø Pfusion ~ 400 MW for 500 s Ø Heating = 40 MW Ø 20% duty cycle • Size R~6 m + Field B~6 T = 5 Billion dollars Ø 1000 m3 plasma Ø 1000 m2 plasma-facing wall
11. 11MIT-ANS Apr 07 Like a car, there are several things you have to do to keep the fusion “engine” running. Power removal Coolant in block, no melting Fuel control. Gas tank full, no flooding Helium ash exhaust Tailpipe Plasma purity No dirt in cylinders Magnetic topology, plasma physics & the fusion fuel cycle set unique requirements for wall materials in fusion Fusion Car
12. 12MIT-ANS Apr 07 Magnetic topology • Magnetic field line directly terminated on a solid has no confinement due to sink action of wall. • This separatrix flux surface is the primary interface between thermonuclear plasma and outside world. • (SOL) Scrape-Off Layer outside separatrix is the “exhaust pipe” of a fusion plasma. • Divertor targets.
13. 13MIT-ANS Apr 07 Divertor magnetic topology is presumed for most fusion reactors • SOL “fed” by cross-field heat and particle losses from “core” • Advantage Ø Concentrates power and particle exhaust (q, G) in one location. • Disadvantages: Ø Concentrates power and particle exhaust in one location. Ø Wastes magnetic volume. Core plasma SOL q// G// qtarget, Gtarget surface Ploss ~ 100-400 MW Bf
14. 14MIT-ANS Apr 07 Divertor magnetic topology is used in ITER (and probably a reactor) • ITER Magnetic Field- Line Geometry in SOL: Ø Bpoloidal / Bf ~ 1/10 Ø SOL: L// ~ 100 m Ø 2 p R ~ 50 m ITER Divertor Cross-section ~ 1m
15. 15MIT-ANS Apr 07 Extraordinary plasma heat conduction along B primarily set SOL properties • // e- Heat conductivity: • ITER energy exhaust: PLoss ~ 150 MW
16. 16MIT-ANS Apr 07 Conforming divertor surface Constant heat removal at divetor surfaces is daunting: #1 priority in edge “design” Field lines Melts 10 cm of tungsten in ~20 seconds ! Distorted surface “proud” to the field line receives q// ~ 1 GW/m2 and is immediately melted/ablated. Distorted divertor surface Field lines 2 p R
17. 17MIT-ANS Apr 07 The core fusion plasma has little tolerance to impurities • Plasma quasi-neutrality sets strict limits for impurities Ø Fuel dilution. Ø Radiation energy losses. • Dilute plasma (n~1020 m-3) extinguished by small particulate injection. ITER example: 10 mm3 “drop” of W ~ Ne,plasma W Mo C
18. 18MIT-ANS Apr 07 Heat exhaust, Tmelt, material stress and heat conductivity set armour thickness q dplasma coolantCoolant substrate • Limits material choices to refractory metals (W, Mo) or graphite. Tungsten CFC carbon k (W/m/K) 150 300 Tmelt (K) 3700 3700 dtile ~ 1 cm ~ 2 cm
19. 19MIT-ANS Apr 07 W & C bonding technology capable of exhausting ~ 25 MW/m2 ~ 1m ITER prototype divertor module castellations
20. 20MIT-ANS Apr 07 Plasma instability leads to large transient heating: Requires high Tmelt materials • MHD instability = Destruction of nested flux surfaces. • High Te flux surfaces connect to wall. • tMHD ~ 0.0001 seconds Ø Edge Localized Mode: ELM Ø Global instability: Disruption • All energy is “held up” near surface Kruger,APS04
21. 21MIT-ANS Apr 07 Materials pushed past their thermal limits even in present fusion devices. • Mo tile “limiter” positioned outside hot core plasma in MIT tokamak (C-Mod) • Plasma heat exhaust and “non-thermal” electron populations increase past Tmelt ~ 2900 K in < 2 seconds exposure. • Reactor must run 24/7.
22. 22MIT-ANS Apr 07 Movie Clips of Alcator C-Mod
23. 23MIT-ANS Apr 07 Understanding competition between “density- driven” radiation and “T-driven” conduction leads to safe shutdown technique Uniform radiation Prad µ n2 (1/T)a Localized heat conduction Q// µ k// µ T5/2 Solution: “Force” high density by massive impurity injection
24. 24MIT-ANS Apr 07 Competition between “density-driven” radiation and “T-driven” conduction critical to benign energy dissipation
25. 25MIT-ANS Apr 07 Even ideal radiative energy dissipation can cause material melting in ITER. Whyte 2004 PSI Prad ~ 3 TW ~ electricity output of US Bemelt ~ 100 kG
26. 26MIT-ANS Apr 07 Particle control #2 priority: Fusion reactors do not burn their fuel efficiently, forcing very large recycling of the tritium fuel. Global breeding ratio in Blanket Maximum allowed core helium fraction Measured helium (de)enrichment from core to divertor Allowed rate of Tritium Deposition in wall
27. 27MIT-ANS Apr 07 World tritium inventory impacts ability to “start” a fusion energy economy • A 1000 MWe will burn / produce ~ 0.5 kg ~ 1 pound of Tritium per day. • The time window for non-T-breeding burners (e.g. ITER, CTF) is short. • Tritium Breeding Ratio of starting reactors must be > 1 or whole system will grind to halt from lack of fuel. J. Schmidt, IEA 2005
28. 28MIT-ANS Apr 07 Developing neutron tolerant materials will probably be last problem solved for fusion • Uniform material bombardment by 14 MeV neutrons Ø ~ 1m thick blanket to thermalize, shield neutrons & breed Tritium • Displacements per atom in wall ~10-20 per year for 1 GW Ø Leads to serious thermal degradation of materials. Ø Internal p, He production by nuclear reactions
29. 29MIT-ANS Apr 07 Developing neutron tolerant materials will probably be last problem solved for fusion • Uniform material bombardment by 14 MeV neutrons Ø ~ 1m thick blanket to thermalize, shield neutrons & breed Tritium • Displacements per atom in wall ~10-20 per year for 1 GWth Ø Leads to serious thermal degradation of materials. Ø Internal p, He production by nuclear reactions ® not an issue in fission reactors due E • Solution will require dedicated experimental & modeling, probably exploiting self- annealing at high material temperatures
30. 30MIT-ANS Apr 07 Neutron irradiation tests of wall materials are needed for fusion reactors • International Fusion Materials Irradiation Facility will probably be part of broader fusion program. • Deuteron beam (0.25 MA, ~40 MeV) on flowing lithium target produces fast neutron spectrum Ø But only produces 0.5 L volume that will receive reactor-like neutron damage (~20 dpa/year)! IFMIF
31. 31MIT-ANS Apr 07 The material targets must also resist erosion caused by massive energetic particle throughput in wall • Vsheath ~ 5 x Te ~ 100-500 eV • Plasma ions can rapidly sputter target material away.
32. 32MIT-ANS Apr 07 Material migration is set by sputtering / recycling asymmetry. D/T saturated deposits
33. 33MIT-ANS Apr 07 Divertor “detachment”: Critical to easing power exhaust and sputtering ne increasing Alcator C-Mod
34. 34MIT-ANS Apr 07 Detachment solved erosion? DIII-D: Map of divertor Erosion / deposition Whyte,IAEA2000
35. 35MIT-ANS Apr 07 Detachment solved erosion? Yes! Yes, but another problem appeared DIII-D: Map of divertor Erosion / deposition Whyte,IAEA2000 Tritium trapped in plasma deposited films from other wall locations Rate ~ 1 in 10 fuelled T lost But requirement: < 1 in 1000 ! Solution: > 1000 K walls to deplete H/D/T ?
36. 36MIT-ANS Apr 07 Turbulent cross-field particle transport ® Erosion sources outside divertor ® Long-range transport to divertor S. Zweben, J. Terry, C-Mod A. Mclean, et al. DIII-D
37. 37MIT-ANS Apr 07 C-Mod now shows a strong link between ballooning transport, rotation, long-range SOL transport, T retention and H-mode! ALCATOR C-Mod, M.I.T. LaBombard, Greenwald APS 04 Ballooning transport SOL flow Core Rotation
38. 38MIT-ANS Apr 07 Besides groundbreaking research on tokamaks like C-Mod, laboratory experiments are coming online to enhance our understanding of PSI MAGNUM-PSI High power, low Te H plasmas DIONISOS Dynamics of PSI FOM, Netherlands MIT
39. 39MIT-ANS Apr 07 Besides groundbreaking research on tokamaks like C-Mod, laboratory experiments are coming online to enhance our understanding of PSI MAGNUM-PSI High power, low Te H plasmas DIONISOS Dynamics of PSI FOM, Netherlands MIT
40. 40MIT-ANS Apr 07 x = A Grand Scientific Challenge for Ours & Future Generations Worst Most Exciting Job in Science Whyte Fusion Wall equation
41. 41MIT-ANS Apr 07 The suppression of edge MHD ELMs is one of the most critical research issues for ITER & beyond Eich 2004 PSI Size& Field $$$
42. 42MIT-ANS Apr 07 Net redeposition or erosion: Deposition rate – erosion rate Impurity release via sputtering at PFC surface Impurity ionization & transport near surface > 90% impurity re- deposited at surface Local Net erosion & deposition arises from ~1-10% local flux imbalance Edge plasma modification by impurity Core plasma modification from impurities SOL transport Release of impurities to SOL Global
43. 43MIT-ANS Apr 07 Divertors concentrate particle flux and recycling, making practical He and H pumping exhaust possible ® particle control • This is not a trivial feat; fuel particle inventory in a fusion device (ITER) is dominated by the wall. Nwall = f Awall ~ 5•1021 m-2 103 m2 ~ 5• 1024 m Nplasma = n Vplasma ~ 1020 m-3 103 m3 ~ 1023 m
44. 44MIT-ANS Apr 07 Parallel heat conduction sustains particle ionization / recycling loop in divertor. • Consider simple pressure and particle conservation /w Bohm sheath criterion (vtarget = cs) • Two-point model (Stangeby) Ø Upstream: SOL Ø Target: Divertor Gion,e GH ionization Core plasma SOL q// G// qtarget, Gtarget surface Ploss Bf
45. 45MIT-ANS Apr 07 Competition between “density-driven” radiation and “T-driven” conduction critical to benign energy dissipation