Lithium and Liquid Metal Studies at PPPL
High Power Devices LTX R. Maingi, M. Jaworski, R. Kaita, R. Majeski, J. Menard, M. Ono PPPL NSTX-U Surface Analysis EAST IAEA TM on Divertor Concepts Vienna, Austria 1-Oct-2015 Test stands Options for liquid metal uses, and R&D needs
Goal: cohesive program for liquid metals to be considered as PFC candidates for fusion devices Outline Motivation: high steady power exhaust and high confinement scenarios with liquid metal PFCs Options for liquid metal uses, and R&D needs Excerpt of results from existing PPPL collaborations Summary 2 Motivation 1: power exhaust challenge harder than thought
Development of liquid metal PFCs for fusion devices: transformative area, deserving of effort Motivation 1: power exhaust challenge harder than thought Heat flux footprint decreases with Ip; no increase with R Both steady and transient loads can exceed solid PFC limits Motivation 2: Confinement difficult with bare high-Z PFCs Good confinement is challenging with high-Z walls in e.g. JET Evidence that liquid metal PFCs can exhaust higher power than solids, and lithium PFCs (liquid or solid) provide access to high confinement 3 Lithium Capillary Porous System (CPS) targets
Lithium loaded targets withstood high steady and transient heat loads in plasma gun experiments Steady operation with heat loads from 1-11 MW/m2 withstood for 3 hours Heat loads < 25 MW/m2 withstood with Li targets (5-10 minutes, limited by Li inventory) Transient loads < 50 MW/m2 withstood with Li targets without cooling (up to 15 s) Lithium Capillary Porous System (CPS) targets IAEA Technical Meeting on Assessment of Atomic and Molecular Data Priorities Vienna, Austria, December 2006 4 Lithium (solid and liquid) PFCs increase confinement
NSTX LTX H98~1 2-4x improvement over ITER98P(y,2) (H-mode scaling) H98y2 increased from 0.8 -> 1.4 (H-mode scaling) J.C. Schmitt, Phys. Plasmas 22 (2015) D.P. Boyle, J. Nucl. Mater. 438 (2013) S979 5 Options for liquid metal uses, and R&D needs
Goal: cohesive program for liquid metals to be considered as PFC candidates for fusion devices Outline Motivation: high steady power exhaust and high confinement scenarios with liquid metal PFCs Options for liquid metal uses, and R&D needs Excerpt of results from existing PPPL collaborations Summary 6 Provide particle exhaust, access to high H-factor for FNSF
Liquid metals can be used in several ways to address part or all of the particle and power exhaust challenge Provide particle exhaust, access to high H-factor for FNSF Slow flow; may have to use advanced divertors for power exhaust Open issue on liquid metal substrate material Provide power exhaust: high PFC temperature solutions Vapor-box configuration (Goldston, this conf.), compatible with W Low recycling may be achievable in low flux areas; will that allow access to high confinement? Provide particle and power exhaust, access to high confinement: low PFC temperature solutions Fast flowing liquid metal option (Majeski, this conf.) Use ferritic steel as option to W as substrate, which reduces needed R&D to develop neutron resistant W with low DBTT 7 Disadvantages and R&D needs
Liquid metal PFCs are alternatives to solid PFCs, but have substantial R&D needs to assess viability Advantages Erosion tolerable from PFC view: self-healing surface No dust; main chamber material and tritium transported to divertor could be removed via flow outside of tokamak Liquid metal is neutron tolerant; protects substrate from PMI Liquid (and solid) lithium offer access to low recycling, high confinement regimes under proper conditions Very high steady, and transient heat exhaust, in principle (50 MW/m2 from electron beam exhausted; also 60 MJ/m2 in 1 msec) Disadvantages and R&D needs Need stable surface & flows to maintain a fresh surface Liquid metal chemistry needs to be controlled Temperature windows need optimization Most of experience in fusion is with Li, but Sn and maybe Ga offer some promise in terms of broader temperature windows 8 R&D need: PFC erosion and temperature limit studies initiated in divertor simulators (MAGNUM-PSI)
Issue: what are PFC erosion rates for lithium targets? Temperature-enhanced erosion seen with Li at low flux not well understood High flux data (Abrams) showed substantial reduction in Li erosion, and agreement with adatom evaporation model (Doerner) Issue: what is lithium PFC operating temperature range with Li? Li melts at 180o C, high evaporation rates > 400o C Hot target (> 700o C) showed lithium emission localized near surface Pre-loaded Li target lasted longer than gross evaporation time Abrams JNM 463 (2015) 1169 Target T>700C Neutral D emission Neutral Li emission Plasma Li Trapping 10 mm MAGNUM-PSI: M. Jaworski,ISLA 2013/2015 R&D need: LM stability and hydrogen retention/release evaluated in confinement and surface science devices Deuterium release in lithium Theme: Horizontal layer of dense fluid over less dense fluid is unstable (drips): Rayleigh-Taylor instability Rayleigh-Taylor analysis of NSTX Liquid Lithium Divertor showed droplet stability DIII-D DiMES expt only marginally stable Temperature programmed desorption (TPD) shows oxygen inhibits formation of LiD and reduces thermal stability of D in Li films Jaworski et al., Nucl. Fusion 53 (2013) Capece, J. Nucl. Mater. 463 (2015) 10 Options for liquid metal uses, and R&D needs
Goal: cohesive program for liquid metals to be considered as PFC candidates for fusion devices Outline Motivation: high steady power exhaust and high confinement scenarios with liquid metal PFCs Options for liquid metal uses, and R&D needs Excerpt of results from existing PPPL collaborations Summary 11 PPPL Lithium Program Elements: present, future
Liquid Li walls as a primary PFC (LTX) Liquid Li as divertor PFC in long pulse tokamak (EAST) Fundamental Li surface science studies (PU Li deposition on divertor PFCs in high b ST (NSTX-U) Li aerosol injection to probe pedestal physics and stability in high b advanced tokamaks (DIII-D) Li granule injection for ELM control and possible conditioning (NSTX-U, DIII-D, EAST) Plasma slow-flow test stand & fast-flow test facility (PPPL) Li lifetime & trapping in divertor-like plasma (MAGNUM-PSI) Liquid Li limiter (EAST) and divertor modules (NSTX-U) 12 Main result: good H-factor obtained with LLD
NSTX: Liquid lithium divertor operation was compatible with reliable H-mode operation Liquid lithium divertor on NSTX: Cu heat sink, SS permeation barrier, Mo mesh, with overhead Li ovens filling the LLD Main result: good H-factor obtained with LLD Jaworski et al., Nucl. Fusion 53 (2013) 13 Bakeable high-Z PFCs for liquid Li
NSTX-U goal: Establish high tE and b + 100% non-inductive,then assess compatibility with metallic PFCs Base: year plan steps for implementation of cryo-pump + high-Z PFCs + LLD 2022 Flowing Li module All high-Z FW + divertors + flowing LLD module or Cryo + full lower outboard high-Z divertor High-Z tile row Cryo + high-Z FW and OBD + liquid Li divertor (LLD) C BN Li High-Z Lower OBD high-Z row of tiles Downward Li evaporator + Li granule injector High-Z tile row Bakeable high-Z PFCs for liquid Li Up + downward Li evaporator Cryo-pump EAST: Long-pulse ELM-free H-modes with constant radiated power enabled with lithium dropper (on loan from PPPL) Li dropper J.S. Hu et al, Phys. Rev. Lett. 114 (2015) 15 Designed and fabricated at PPPL
EAST: Liquid lithium limiter developed at PPPL and inserted via midplane port Designed and fabricated at PPPL Used a DC EM pump, with BT of EAST, for steady-state recirculation Implemented in EAST in Oct. 2014 J. Ren et al., Rev. Sci. Instrum. 86 (2015) 16 EAST: Liquid lithium limiter compatible with EAST H-mode discharges
Lithium light increased when current driven in limiter system Performance improved in ohmic discharges Da and impurities decreased in both divertors J.S. Hu et al., Nucl. Fusion in preparation 17 Closing thought - development of LM PFCs is a transformative area
Motivations: power exhaust very challenging, and access to good confinement difficult with bare high-Z PFCs Evidence that liquid metal PFCs can exhaust higher power than solids, and provide access to high confinement This is an area ripe for advancement 18 Thank you for your attention!
19 Limiter emission uniform later in pulses
H-mode discharges maintained in EAST when liquid lithium limiter inserted past separatrix Plasma-limiter interaction resulted in filamentary emission during ramp-up Limiter emission uniform later in pulses After exposure, limiter showed damage from plasma-wall interactions J.S. Hu et al., Nucl. Fusion in preparation 20 W chosen for divertor PFCs & Be for wall PFCs of ITER
W advantages Low physical sputtering yield & high energy threshold No chemical sputtering with hydrogen Low in-vessel tritium retention in most scenarios Reparable by plasma spray; good joining technology W PFCs can exhaust 5-10 MW/m2 steady heat flux W disadvantages Low allowable core concentration Melts under large transient loads High ductile-brittle transition temperature Recrystallizes & becomes brittles at temperatures >1500 K High activation Blisters and generates fuzz under He bombardment Confinement reduced in tokamaks as compared with low-Z PFCs G. Federici, et. al., Nucl. Fusion 41 (2001) 1967 21 Several thrusts to address knowledge gaps in LM program
LM PFC technology and science in flowing, self-cooled and externally cooled test systems Flow rates from 1 mm/sec 10 m/sec Use capillary or j x B forces to overcome MHD forces that could cause mass ejection Determine operating temperature windows Assess hydrogenic species control and He entrainment Fundamental LM surface science studies Keep LM surface clean for reliable flow; understand PMI Predict flow of LM, including wetting and de-wetting Compatibility with attractive core/edge plasma Plasma power and momentum exhaust; particle control Applicability of low recycling regimes with excellent confinement: target H98 > 2, enabled by LM resilience to transients and high peak heat flux exhaust 22 PPPL Lithium Program Elements
Liquid lithium walls as a primary PFC: effects on electron transport, fueling requirements (LTX) Lithium as a divertor PFC in high b ST: power/particle exhaust and confinement improvement (NSTX-U) Lithium lifetime and trapping in divertor-like plasma stream (MAGNUM-PSI) Liquid lithium as primary PFC in long pulse tokamak (EAST) Liquid lithium flowing loop R&D (PPPL test stand) Fundamental lithium surface science studies (PU Lithium injection in high b advanced tokamaks: confinement and stability changes (DIII-D, AUG) Lithium granule injection for ELM control and possible conditioning: (DIII-D, EAST NSTX-U) 23 Pedestal performance and core confinement in JET scenarios was reduced with installation of ITER-like wall Beurskens PPCF 2013 Partial performance recovery with N2 seeding, which cannot be used in D-T campaign; less favorable results with Ne Projected performance less than in 1990s D-T experiment Degradation and partial recovery seen in AUG, C-Mod (metallic PFCs) 24 In NBI heated H-modes, Li injection via dropper reduced ELM frequency but did not eliminate ELMs
Milestone: Compare access to ELM-free operation with active Lithium aerosol injection between lower hybrid heated and neutral beam heated H-modes NBI H-mode Stored energy decays slowly Li emission saturates ELM frequency drops ~25-30% Zeff unaffected Li dropper Courtesy of J.S. Hu & Z. Sun 25 Active discharge modification with Lithium injection
Additional issues for NSTX-U LM deployment addressed by collaborations on international & domestic devices Active discharge modification with Lithium injection Li droppers for 18 sec H-modes in EAST (Maingi IAEA 14, Hu, PRL 15) Li dropper, Li granule injector in DIII-D (Jackson, IAEA 14, Bortolon EPS 15) Li pellet injector (and dropper) studies in AUG with W-PFCs (2015) Flowing LM in EAST H-modes Very slow flowing, midplane liquid lithium limiter tested in 10/2014: compatible with H-mode! Next module being prepared for test in FY2016 Research will inform design of flowing LM modules for NSTX-U Figure 1: Schematic of heated copper plate and small liquid lithium reservoir, mounted on an insertable probe for testing in EASTin 2014 26 Scenarios in tokamak discharges with High-Z PFCs can affect pedestal performance and core confinement C-Mod Beurskens PPCF 2013 Lipshultz, PoP 2006 C-Mod Beurskens PPCF 2013 Kallenbach NF 2011 27 Impact of low lqmid studied for ITER
28 Cross-field transport can be reduced to get lower lqmid, but higher divertor neutral pressure Pn reduces qpeak 29 Heat flux profile width lq measured in divertor
Update on gaps since ReNeW: Steady heat flux exhaust is more challenging than projected at ReNeW Heat flux profile width lq measured in divertor lq projected to outer midplane with flux expansion International effort found that lq varies inversely with Bpol,MP No increase with R, PSOL Low gas puff attached plasmas; some broadening and heat flux dissipation with detachment Projected width in ITER ~ 1/5 previous value; operating window narrows Much more challenging for reactors, due to higher Pfusion ITER Kukushkin, JNM 2013 Eich, NF 2013 30 Operating window in ITER reduced with lower lqmid
Adjust cross-field transport to get desired lqmid (SOLPS) Vary the divertor gas puffing to reduce qpeak < 10 MW/m2 Assess ratio of divertor neutral pressure Pn to critical value for detachment Pncrit; ratio m = Pn/Pncrit Use boundary simulation as input to core simulation for Q, Pa, PLH, for operating window m < 0.8 Q > 5 P > PLH 20 Q P > PLH 5 Q > 5 Pa m < 0.8 31 Operating window gets reduced with lower lqmid
ITPA Projection ~ 0.9 mm 20 15 7 Q Q Q P > PLH 5 5 5 Q > 5 Pa Pa m < 0.8 Pa Technology improvements (e.g. liquid metals) or complete elimination of ELM transients could increase qpeakmax by 50%, and partly restore operational window 32 Development of LM science and technology via dedicated test stands
Dedicated linear device with integrated liquid lithium loop can address physics and technology goals Arc-source proposed to provide divertor-relevant heat fluxes Material transport, recapture requires integrated lithium loop Extensive water cooling to be avoided with lithium PFCs Dedicated toroidal devices can demonstration basic stability Similarity experiments with GaInSn could be restarted quickly Dedicated lithium facilities will address low-density fluid and hydrogen cycle aspects directly UCLA MTOR GaInSn Experiment NSTX: H-factor increased by 50% with lithium coatings
Lithium injection improves confinement on many machines, generally (but not always) via recycling reduction NSTX: H-factor increased by 50% with lithium coatings Reduction in recycling reduced the ne profile gradient and shifted it inward; very little Li penetrated inside the separatrix Te gradient clamped by stronger driver for ETG Pressure profile followed ne profile -> stabilizing to ideal MHD peeling/ballooning modes, so discharge went ELM-free DIII-D: H-factor increased transiently by 60% with lithium injection of microscopic dust particles No change in recycling, but change in ne, Te, pressure profiles nearly identical to NSTX when a newly observed instability occurs Substantial Li penetration and ion dilution in edge Profile inward shifts and ion dilution both stabilizing to ideal MHD New ITPA multi-machine experiment to understand the effect of low-Z impurities on pedestal 34 Profile changes in DIII-D in ELM-free H-mode qualitatively similar to NSTX ELM-free H-mode with inter-shot Li evaporation Shifting gradients away from separatrix improved edgestability in both DIII-D and NSTX DIII-D NSTX No Li With Li NSTX: Maingi, NF 2012 DIII-D: Maingi, PRL 2014 submitted Osborne, NF 2015 submitted Flow chart of how lithium results in ELM elimination similar for highly and weakly shaped discharges
N from (recycling region) Li coatings reduce recycling and core fueling (SOLPS) ne and ne reduced Te fixed ne, Pe gradient reduced Jbs, J|| reduced - stabilizing (ELITE) e increased ETG (GS2) Pi gradient unchanged N from ne and ne Te and Te increased Jbs, J|| increased but far from separatrix; improves stability to kink/peeling modes (ELITE) Edge Pe follows ne and Te; peak P shifts farther from separatrix Deff reduced, most in ELM-free (SOLPS) e reduced strongly mT stable (GS2) 36 36 Lithium injection induces a bifurcation to higher pedestal pressure and width in DIII-D
PNBI [MW] ne [m-3] Da [au] HH98y2 Pradtot [MW] Teped [keV] Peped [kPa] Pewidth [cm]#159643 Lithium injection ELM-free bifurcated state canbe seen in Da emission H98y2