LH Systems International Status and Prospects

LH Systems International Status and Prospects Liang Liu, Jiangang Li, Fukun Liu, Jiafang Shan, Bojiang Ding, Hua Jia, Miaohui Li, Mao Wang, Lianmin Zhao Institute of Plasma Physics, Chinese Academy of Sciences 15 th -18 th, Nov. 2016, KIT, Karlsruhe, Germany

Outline Introduction Current status and future plan of LH systems Some consideration for Demo LHCD Summary 2

Function of LHCD for DEMO LHCD is a powerful tool for DEMO To drive far off-axis current and to save the volt-second during the current ramp-up phase To suppress or mitigate ELMs and to optimize the heat load distribution on the divertor plates Demonstration of the effect of LHCD on ELM behavior by modulating LH power in a series of target H-mode plasmas produced by ICRH with different q 95 IR image of the outer lower divertor plate in a toroidal range of φ=1.3π-1.5π rad during the application of LHW. J. Li et al., Nature Physics 9, 817-821 (2013) Y. Liang et al., PRL 110, 235002 (2013) 3

Defects of LHCD for DEMO Poor accessibility Deficiency of LHCD for DEMO CD efficiency could be degraded sharply with density increasing, due to PI, CA, and SDF. CD efficiency would be affected by high temperature. Including the effects of collisional damping Without ne~0.9 x10 20 m -3 (black), ne~1.1x10 20 m -3 (Red) PI measured by Langmuir Probes in C-mod PI occurs at higher density PI can occur in either HFS or LFS S G Baek, et al., PPCF 55 (2013) 052001 CA in SOL may decrease CD efficiency (GENRY/C3QLD) G M Wallace et al., Phys. of Plasmas 17 (2010) 082508 4

Outline Introduction Current status and future plan of LH systems Tore Supra LH system FTU LH system C-Mod LH system KSTAR LH system EAST LH system ITER LH system Some consideration for Demo LHCD Summary 5

Tore Supra LHCD system Designed for long pulse operation 1996: 2 minutes long pulses 2003: World record 6 min,1 GJ injected 2011: 6.3 MW/150s, 1 GJ injected Generator: 16 CW klystrons, f = 3.7GHz PAM n // = 1.7-2.0 TH2103C klystron FAM n // = 1.7-2.3 9.2 MW available / 1000 s Klystron efficiency: 37.4% Launchers: Fully Active Multi-junction (FAM) First operation in 1999 4MW - 25MW/m 2 1000s Passive Active Multi-junction (PAM) First operation in 2009, ITER relevant design 2.7MW - 25MW/m 2 1000s 6

Tore Supra current experimental results ITER-relevant power density on PAM Long pulse operation #47979 2011 E LHCD = 220MJ 2.75MW (25MW/m 2 ) coupled with PAM for 78 seconds Low RC (< 2%) at large plasmalauncher gap (> 10cm) E Total = 960MJ 5.3MW LHCD combined with 1MW ICRH Ip=0.7MA; >150s ne=3 10 19 m -3 Higher LHCD power widens operational space (n e, I P ) for long pulse operation. 7

Tore Supra LH research plan Development of oversized components for transmission line of a reactor-relevant LH system CEA ENEA IPR Collaboration F=3.7GHz Rectangular / Circular Convertor Over-sized waveguides low RF losses Corrugated waveguides to avoid spurious modes in bends R&D for the slotted waveguide antenna Low power prototype has been manufactured First test will be carried out on the Compass tokamak 5GHz / low field side Limiter fixing structure Corrugated Bend Rectangular / Circular Convertor Transition to WR187 SWA 8

FTU LHCD system 8 GHz, 6 MW, 1 s LHCD system, aiming at studying high density plasmas B Tmax =8T Density max =3 10 20 m -3 8GHz gyrotron assembly, TE 0 51 output mode Oversized C18 WG transmitting TE 0 01 mode 2 classical grill Power splitter - Mode convertor Using gyrotron as RF sources Circular WG => Rec WG Classical grill Gyrotron 8GHz 1MW 1s TE 01 => TE 10 LHCD Module 10

FTU LHCD PAM launcher Demonstrate coupling characteristics of the launcher No cooling (1s pulse length) PAM (R.C 2%) Vs CG (R.C~15%) PAM (Dp~65%) Vs CG (Dp~80%) The very first prototype of a PAM launcher has been tested To demonstrate the optimum coupling characteristics in presence of the expected really harsh ignited plasma s characteristics of the future fusion machines PAM and CG performances PAM performance FTU PAM layout PAM experiment on FTU (EX/5-5) - 20 th FEC.Villamoura 1-6 Nov, 2004 - V. Pericoli Ridolfini et al FTU PAM view 11

Alcator C-Mod LHCD system Investigate LH physics with ITER-like parameters Source: 250kW 4.6GHz klystrons Power: 4MW Antenna: 16 four-way splitters Mid-plane launcher LH2 CPI VKC-7849 klystron Above mid-plane launcher LH3 design Parasitic edge power losses are considered to cause unexpectedly large degradation of LH efficiency at high density LH3 aims to minimize parasitic edge power losses by enhancing strong single pass power absorption LH3 is designed to maximize synergistic effects between LH2, by placing the launcher off-midplane and selecting N// higher than LH2 S. Shiraiwa et al, Plasma and Fusion Research, Vol 8, 2402143 (2013) 13

Current status of KSTAR LH system Control q profile aiming for high beta plasmas Generator: E3762 klystron - Toshiba Electron Tubes and Devices 5GHz 0.5MW CW with matched load Transmission line: Oversized rectangular WR-284 (80m, ~34% loss) Launcher: Grill type with 4-way splitter 8 N // ~ 2.0 P LH ~ 100 kw (15% reflection) [#11145] Grill type antenna R=1.78m R=1.35m ECE P FWD [kw] P REV [kw] <ne> [x10 19 /m 3 ] Observation of plasma response by 100 kw LH power Changes of ECE radiation spectrum were observed Measurement of deposition profile using HXR camera will be tried 15

KSTAR LH research plan Minor upgrade of current system before 2021 Prototype of Mid-plane PAM launcher is under development Basic RF design was completed Mock-up construction is planned KSTAR new mid-plane PAM launcher # P2.32, J. Kim, Development of 4-MW KSTAR LHCD system, SOFT 2016 # P2.33, T. Seong, Mode Converters for Low-loss Transmission-line of KSTAR LHCD system, SOFT 2016 Test of low loss transmission line system propagating circular TE 01 mode Cold model developed for TE 10 to TE 01 mode converter + 3.5 MW - Antenna & system 2016 2017 2018 2019 2020 2021~ Study on launching position and N // Antenna Design & Development System Development Experiment LHCD upgrade is planned after 2021 Mid-plane, off mid-plane, or high field side launching are under consideration 5GHz 4MW with low loss transmission line 16

Overview of EAST LHCD system EAST aims at high performance long pulse operations 4.6GHz 6MW LHCD Consists of two LH systems 2.45GHz LH 4.6GHz LH EAST Hall RF Source 200kW 20 250kW 24 Transmission WR430 WR229 Line ~15% loss ~25% loss Active WG = Active WG = 2.45GHz 4MW LHCD Launcher 5 4 8 Passive WG = 12 6 8 Passive WG = Location of EAST LHCD system Top view 5 5 12 7 Play a key role concerning the controllability of shaping plasma current profile rather than plasma heating 2.45GHz 2MW LH: Start of operation in 2008 Upgraded 2.45GHz 4MW LH: Start of operation in 2012 4.6GHz 6MW LH: Start of operation in 2014 4.6GHz and 2.45GHz LH antenna spectrum (ALOHA code) 18

Two LH systems on EAST Transmission line Antenna Klystron EAST Output power: 6MW Frequency: 4600±5MHz Pulse length: CW Power spectrum: 1.79~2.23 VKC-7849A klystron 4.6GHz Multi-junction 576A+84P The framework of 2.45GHz LHCD system for EAST KU-2.45B klystron parameters Parameters Value Units Frequency 2.45 GHz Output power 200 kw Input power 0.3~1.0 W Gain 53 db Input HV power Max.420 kw Beam voltage -40~-46 kv Beam current 10 A Efficiency 50 % Klystrons and Transmission lines 19

LHCD experiments on EAST 4.6GHz LHCD Long pulse discharge with LH (1.8MW) + EC (0.4MW), 0.4MA for 102s, ne~1.5 10 19 m -3, Te 0 ~5keV 2.45GHz LHCD Minute-scale H-mode operation with LH (2.5MW) + IC (0.8MW) + EC (0.3MW), 0.45MA for 61s, ne~3 10 19 m -3 20

ITER LHCD programme ITER is an important step towards DEMO 20MW 5GHz system N // =1.8-2.2 for all scenarios ITER main parameters Parameters ITER To drive far off-axis current, at ρ=0.6-0.8 current profile control in advanced regimes To save volt-second during the current ramp-up phase Report of the EFDA Task, WP09-HCD-03-01 LH4IT Summary of the main results from LHCD calculations for ITER Toroidal field (plasma center) 5.3 T Major radius 6.2 m Minor radius 2.0 m Elongation (95%) 1.7 Plasma current (MA) 15.0 <n e > [10 19 ] 6.7 m -3 β N 3.0 Scenario Steady-state (4) Inductive (2) Hybird Ramp-up Optimum n // 1.85 2.10 1.85 <1.75 Corresponding I LH 0.70 MA 0.45 MA 0.60 MA >3.0 MA Corresponding r/a 0.65 0.90 0.80 <0.50 I LH for n // =2 0.65 MA 0.40 MA 0.60 MA 2.40 MA A scan in the injected N// was systematically performed also accounting for variations in the temperature and density profiles in order to characterize the performance and robustness of the scheme 22

ITER PAM launcher Nuclear heating on PAM components Nuclear Heating (due to neutron and secondary gamma) Component Plasma facing components (PFC) MW 6.28 10-2 PAM 4 WG 1.31 PAM 2 WG 6.77 10-3 HFSS model for ITER PAM-B launcher (DDD-2001) 20MW/5GHz/CW 1.5 times the cut-off density 3.1x10 17 m -3 PAM 1 WG 1.84 10-4 Mode converter + Tapers 3.89 10-4 Shield 1.62 10-5 Secondary Window 3.96 10-7 Frame 1.71 Stainless steel CuCrZr Overall nuclear load ~ 3MW Mostly deposited on the PAM 4 WG region Various materials for PAM front copper or copper alloy, for thermal and electrical conductivity Be PAM front thermal analysis stainless-steel, for mechanical strength and reduce Foucault currents beryllium (Be), facing the plasma, to prevent pollution of the plasma with high z materials 23

ITER LH installation 2 x 48 RF windows (500kW/CW) 48 rectangular T-Lines (or 24 circular T-Lines) 48 Klystrons 500kW / CW To minimize the interaction with the ICRH antenna and to improve the power coupling in ELMy plasmas Foreseen distance between the klystrons and the launcher, could be reduced at about 40m for minimum RF losses Report of the EFDA Task, WP09-HCD-03-01 LH4IT Side view of LH system near Port # 11 Port # 11 24

Outline Introduction Current status and future plan of LH systems Some consideration for Demo LHCD Summary 25

The choice of LH frequency is determined by complicated trade-off between manifold counteracting elements Technological viability of RF sources Alpha particle absorption High power RF source for DEMO Accessibility requirements and multipactor effect Several tens of MW coupled power is required by DEMO Comparison between different operation frequency 3.7GHz 4.6GHz 5GHz Klystron TH2103C VKC-7849A E3762 Output power ~700kW/CW 250kW/CW 500kW/CW Manufacturer Thales Electron Devices Communications & Power Industries Conclusion Alpha particle can absorb a significant part of the coupled power Limited output power, requiring a higher number of klystron Toshiba Electron Tubes & Devices Maybe best choice 26

High field side launcher HFS launching offers significant advantages over LFS launching ADX Main parameters Major/Minor Radius Magnetic Field Plasma Current P AUX (net) 0.73 / 0.2 m 6.5 Tesla (8 Tesla) 1.5 MA 8 MW ICRF 2 MW LHCD J LH [A/cm 2 ] LFS launch 0.5 MA / 10 MW HFS launch 0.7 MA / 10 MW Advanced Divertor and RF Experiment Divertor and firstwall material Pulse Length Tungsten/Molybdenum 3s, with 1s flat-top Lower N // of the waves yields a higher CD efficiency Higher B opens the window between wave accessibility and the condition for strong ELD, allowing LH waves from the HFS to penetrate into the core of burning plasma, while waves launched from the LFS are restricted to the periphery of the plasma G.M. Wallace, SOFE2015 27

High field side launcher High field side launch is highly favorable for LHCD Transport in tokamak sends heat and particles to LFS SOL Quiescent SOL eliminates the need to couple LH waves across a long distance to the separatrix Measurements show impurities penetration is 10 times smaller on HFS HFS mid-plane has 25% lower neutron wall loading than LFS HFS location gives space at LFS for tritium breeding < 10 cm ADX HFS LH launcher The engineering of HFS launcher seems more difficult simply due to the smaller area on the inside of the torus Transportation in LFS SOL Simulated ITER neutron flux VULCAN Podpaly, et al., FED 87 (2012) 215. G.M. Wallace, SOFE2015 28

More effective CD HFS LH + Top EC LFS I LH =1.43MA f=4.6ghz P=30MW N//=1.8 I LH =1.83MA f=4.6ghz P=30MW N//=1.8 HFS CFETR (Chinese Fusion Engineering Test Reactor) Simulation Results HFS power injection has a higher CD efficiency than LFS injection The synergy of both HFS LH and top EC has a significant higher CD efficiency than LFS LH & LFS EC HFS LHCD TOP ECCD I = 0.3MA I CD =4.0MA LFS LHCD LFS ECCD I = 0.34MA I CD =2.49MA 29

Passive Active Multi-junction PAM launcher is a good candidate for Demo LHCD Conventional grill Multi-junction PAM C-Mod KSTAR EAST JET JT-60 Tore Supra FTU ITER Heavy heat load and different plasma configurations In the H mode plasma, the edge density will decrease rapidly enough Long distance coupling (>10cm) Powerful water cooling capacity Good mechanical stiffness Less radiating surface PAM Structure 30

Local gas puffing improving coupling Density and density gradient determining wave-plasma coupling Operation with large plasma-launcher gap H-mode edge 4.6GHz LHCD Local gas puffing Gas puffing is utilized to improve coupling(e.g., CD 4, D 2 ) JT-60U, ASDEX, JET, Tore-supra, EAST Local gas puffing 2.45GHz LHCD Top view P LHCD (MW) RC (%) Good LH coupling obtained at plasma-launcher distance up to 15cm, using local gas injection in H-mode plasmas. A. Ekedahl et al., PPCF, 51 (2009) JET H-mode plasmas 12-13cm -2cm Da Gas puffing is efficient to improve LHW-plasma coupling, especially from electron-side R.C EAST plasmas with / without local gas puffing 31

RAMI Analysis RAMI is an acronym of Reliability, Availability, Maintainability, Inspectability Indispensable for Demo tokamak and its auxiliary systems Assuring continuous operations in specified conditions Reliability: Assuring lowest failure rate during system operation. Availability: Determined by reliability and maintainability of the whole system. Functional relation between RAMI elements Maintainability: Prompting restoration of the full performance of the system Inspectability: Directly affecting the maintainability for inspection and replacement of faulty items. If λ i is the failure rate of the i th subsystem, the failure rate λ S of the whole LHCD module is: LHCD Module Block Schematic S i i 32

Definition of RAMI elements Reliability, survival probability R t e t Maintainability, probability to be repaired with λ = failure rate defined as: t 1 N dn dt [no. of failures 10 6 hours] Availability, probability to be in operation The inherent availability is defined as: A MTTF MTTF MTTR M t 1 e t with μ = repair rate of the item. Inspectability, not time dependant It depends only on the experience of the designers. Take ITER for example to analyze Klystron Sub-assembly Unit Failure Rate [h -1 ] Klystron (500 kw CW) 75 10-6 Anode HVPS 140 10-6 Focusing Coils Sub-assembly 55 10-6 HV Thank 20 10-6 Total 294 10-6 The 48 klystrons must be considered a series structure from reliability point of view To have a high availability of the LHCD system, at least 10 spare klystrons must be stored at the system site 33

R&D needed for Demo LHCD Key components of Transmission line: RF window (500kW/CW), Mode filters, Mode converter Be adapted to Demo, in terms of minimization of the number of the TL Antenna front: R&D a new front end to LH launcher exposed to Demo harsh environment RF source development: Higher frequency (>=5GHz) & higher power (>=500kW/CW) klystron 34

Outline Introduction Current status and future plan of LH systems Some consideration for Demo LHCD Summary 35

Summary Corrugated oversized waveguides should be used for transmission lines to reduce RF loss and to avoid spurious modes, and try to shorten the distance between the RF source and the launcher. 5GHz maybe the best choice for DEMO LH operation frequency. The mechanisms affecting CD efficiency show that edge parameters play an important role, how to effectively control the edge parameters is a challenge. LHW-plasma coupling can be improved by local gas puffing, and grill density feedback-control is necessary. Complex systems must be carefully split into self consistency elementary functional units, for each unit a reasonable number of spare units must be adequately stored. Real-time protection of antenna during long pulses and survey of the erosion of the front face are key points in view of application of LHCD on a reactor. Some components such as high power klystron, new front to launcher, key transmission line devices need urgently R&D. 36

Acknowledge Specially thank: F Mirizzi (FTU), M Goniche (Tore Supra), A Ekedahl (Tore Supra), S Shiraiwa (C-Mod), J Kim (KSTAR) Thank: Tore Supra LH group, FTU team and LHCD staff, Alcator C-Mod LHCD group, KSTAR LH group We acknowledge: J Hillairet, T Hoang, J-M Bernard, L Delpech, W Helou, M Prou, S Ceccuzzi, P K Sharma, G Wallace, R Parker, O Meneghini, Y.S. Bae, S. Wang 37