HFS PELLET REFUELING FOR HIGH DENSITY TOKAMAK OPERATION

Transcription

1 ASDEX Upgrade Session: "Issues and prospects of effcient fueling for magnetic confinement" HFS ELLET REFUELING FOR HIGH DENSITY TOKAMAK OERATION.T. Lang for the ASDEX Upgrade and JET teams Cubic mm size pellet injected with 4 m/s from the inboard (magnetic HFS) at ASDEX Upgrade

2 Outline of the talk Motivation: Why high density operation? Why inboard (HFS) launch in tokamaks? Figures of merit : what means "high performance refueling"? τ > τ E ellet fueling cycle Optimization of pellet fueling in "conventional" ELMy H-modes High density operation with maintained confinement : an example from the JET Strategies for further improvements

3 High density operation enhances operational headroom 3 Operation at high density yields improved flexibility of operation Gas puffing: Decreasing efficiency with plasma temperature and size; Reduction of energy confinement when approaching "Greenwald" density Deep particle deposition e.g. by pellet injection (cryogenic mm size km/s velocity) bears the potential to improve this situation

4 Improvements of HFS injection (empirical) Injection from the LFS: loss of "efficiency" and outward shift of deposition profile with increasing T Can be improved by HFS launch FUELLING EFFICIENCY ε F HEATING OWER (MW) Cubic pellets of side length mm at 17 Hz rate and velocity 13 m/s: AUG.8 MA 1.9 T 7.5MW NBI n (1 m ) e 19-3 LFS HFS TIME (s) Now, looping setup for high speed launch Total length of injection path : L 17 m. 7 Centrifuge

5 Investigate underlying physics causing improvements of HFS injection 5 Formation of a localized high-β-plasmoid B Density gradient layer Drift direction Separatrix Diamagnetic outwarddrift Inboard launch favoured Inboard launch Outboard launch R Strong indications that precooling of ambient plasma plays a significant role But why does strongly "tilted" injection work as well? Comparison of stellarator and tokamak can help to improve our physics understanding Optimize operational scenario Understanding of underlying physics

6 Valuating refueling performance: figures of merit 6 N ABLATION & DRIFT DOMINATED 1ms n e ε τ fast τ slow TRANSORT DOMINATED ELMs 1 ms LN (TIME) TIME Slowing down of τ due to peaking of deposition profile R (m) 3.8 x. Example shown is a simulation of typical JET discharge assuming D = 1 m /s. 1 dn (119 e m -3 ) 3 1 Analysis of ASDEX Upgrade data shows D transiently enhanced after the injection, gradually returning to initial value. 7 1 τ = ms τ = 6 ms 3 τ = 11 ms n (119 e m -3 ) N εn 6 < n e > τ = 16 ms 1 time (ms)

7 Convective losses causes boundary of operational area 7 articles lost at diffusive time scale τ are already thermalized Γ Φ loss loss.8 MA. T n e,gw n e,gw 6.5 MA.4 T.6 τ W (MJ).4. Gas puffing ASDEX Upgrade W = n (1 m ) e 1 + W c Φ loss W heat JET Enhance τ Reduce Φ loss Reduce W for same increase of particle inventory ( n ) Extend operational area e

8 8 Deeper penetration better refueling performance "Sprint" looping setup (max. launch speed 56 m/s): increased launch speed beneficial from HFS as well? Deposited atoms (1 m ) Reflectometer Thomson scattering Interferometer deconvolution #143 (.393s) v = 4 m/s Increasing the pellet launch velocity from 4 m/s to 56 m/s increases from 18.±1.7 to 4.±1.1 cm; increasing the deposition depth as well 1 5 #144 (.36s) v = 56 m/s..3.4 FIT τ = 58 ms v = 56 m/s n e -3 (1 m ) r (m).65 #144 (.36s) v = 56 m/s.65.6 τ = 48 ms v = 4 m/s 1 ms Same exponential decay for both v.6 Deeper deposition reduces energy loss rate from 1.5 ±.18 MW to.84 ±.14 MW. Indeed advantages by higher HFS speed ellet deposition elletlike Gaspufflike r / a

9 Transient enhancement by τ > τ pellet fueling cycle E n (1 e m -3 ) n egw 1 FIT τ fast = 1 ms τ slow = 1 ms Enhanced density without confinement loss 3 AUG.8 MA. T 5MW NBI.5.45 W MHD (MJ) W 5 kj FIT τ = 5 ms TIME (s)..5 LASMA ENERGY (MJ).4 5 n e,gw 1 1 Gas puffing n e (1 m ) 15 Energy recovery takes place faster than the density enhancement gets lost. Transient phase with full initial energy and enhanced density. For a pellet repetition time τ > t rep. > τ ersistent density build up without consecutive energy (confinement) losses E

10 Optimization of the pellet refueling scenario 1 Apply optimized injection setup (HFS launch at high speed and according repetition rate) Get rid of "parasitic gas puff" imposed by particle losses by sufficient pumping capability AUG.8 MA. T 7.5MW NBI n e,gw ellets Gaspuff n (1 e m -3 ) n (1 esep m -3 ) W MHD (MJ) TIME (s) Avoid onset conditions for NTM Avoid drawback to type III ELM (proximity to H-mode threshold) Sufficient heating power roper B t, I Note from pellet penetration scalings: n -1/9 e T -5/9 e Improved pellet performance at higher density (at a certain energy level) Gently prepeare high density "starting" level

11 High density operation with good confinement: an example from JET 11 6 W Dia. (MJ) JN531. JET.5 MA.4 T 18MW NI&IC 4 β N 1.5 heat = 18 MW n e Gw n e n (1 e m -3 ).5 ellet monitor (V) TIME (s) H MA;.4 T <δ> W = const Gas ellet Gas ellet n/ngw

12 High density operation with good confinement: an example from JET 1 1. final n e Gw JET.5 MA.4 T 18MW NI&IC n (1 e m -3 ) T (kev) e initial ellets can yield adiabatic density enhancement (constant pressure profile while density profile gets more peaked and temperature profile is reduced accordingly). 3 1 p (ka) e No significant change of pedestal values Z eff Dilution of impurities by the pellet.4. NBI power (MW/m 3) R (m) keeps NBI power deposition profile unchanged.

13 Strategies for further improvements 13 JN5316 (5314) tot ellet can benefit from the developement of advanced szenarios (e.g. high δ configurations) W Dia. (MJ) NI rad (MW) Benefits from high δ shape and pellets can be combined. article flux reduce to 1/5 in pellet discharge still "more than sufficient". JET.5 MA.7 T δ u =.53 Improvement of injection system AUG: reduce max. impact angles in looping JET: revised setup for fast central HFS launch ITER: new machine, big size more relaxed construction constraints n e Gw n e ellet monitor (V) n (1 e m -3 ) Γ (1 Gas s -1) 59 TIME (s) 6 Improvement of pellets (compound pellets) Compose pellet from doped ice for ellet hardening (higher given setup) Additional radiation shield (deeper given speed) Ar T (Xe) D (Kr)

14 Improvements: Latest results from ASDEX Upgrade 14 Transfer efficiency through guiding tube system for mm pellets launched at 88 m/s Improvement of the looping setup: Looping I undoped Looping I doped Looping II doped ellet hardening by nitrogen doping: smashed pellet good pellet Relative transfer efficiency,5, 1,5 1,,5 Undoped,5% Nitrogen,% Nitrogen ellet speed [m/s] Nitrogen doped pellets: enhacement of the radiative shielding enetration from experiment [m],4, a b c d = C T e n e m d v a = -,8 b =,7 c =,1 d =,43 Undoped Doped (.5%),,,,4 enetration from scaling [m]

15 Conclusions 15 ellets do have the potential (beyond others) to extend a tokamaks operational area and improve the experimental headroom in a given machine Inboard (HFS) launch towards the plasma center at maximum speed yields maximum deposition depths and hence the most favourable performance Advantage of HFS launch caused by curvature drift and precooling of ambient plasma "Optimised coupling" of pellet tool required: sufficient heating and pumping, adapted repetition rate Clear improvements demonstrated when applying the pellet tool (shown here: AUG and JET) O U T L O O K Still headroom for further enhancements: comes "naturally" in bigger sized machine can be imposed e.g. by optimized pellet guiding system and/or compound pellet approach