ITER Plasma Vertical Stabilization

Transcription

1 ITER Plasma Vertical Stabilization A. Portone, R. Albanese, G. Ambrosino, M. Ariola, A. Brooks, D.J. Campbell, T. A. Casper, M. Cavinato, V. Chuyanov, G. De Tommasi, M. Ferrara, R. Fresa, H. Fujieda, D. Gates, Y. Gribov, R. Hawryluk, I. H. Hutchinson, D. Humphreys, A. Kavin, G. D. Loesser, M. Mattei, C. Neumeyer, A. Pironti, G. Rubinacci, G. Saibene, F. Sartori, F. Villone Experimental Vertical Stability Studies for ITER Performance and Design Guidance D.A. Humphreys, T.A. Casper, N. Eidietis, M. Ferrara, D. Gates, B. Hudson, I. Hutchinson, G.L. Jackson, E. Kolemen, J. Leuer, J. Lister, L.L. Lo Destro, T.C. Luce, W.H. Meyer, L.D. Pearlstein, F. Sartori, A. Portone, M.L. Walker, A.S. Welander, S.M. Wolfe 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page

2 OVERVIEW Old (2) baseline VS (Vertical Stabilization) system: VS circuit made of outboard coils (PF2-5) with max i V =6 kv to stabilize l (3)< plasmas ITER-like plasmas ramp-up/down experiments at JET, DIIID, AUG indicate li(3) ~.2; CMOD and DIII analyses suggest that to avoid VDE and operate robustly the VS system must stabilize offsets of Z/a~5-%; ITER VS stabilizes Z/a~% position offsets; stability li(3)>. also shows lack of robustness of ITER VS (phase margin ) Enhancements to ITER VS system proposed in 27-8 (Design Review/ STAC Issues ). Increase VS voltage to 9 kv 2. Add VS2 loop (CS2 coils 6 kv) 3. Add in vessel passive stabilizers. Add in-vessel VS coils 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 2

3 JET data on li(3) range Ohmic ramp-up li(3) ITER CONTROL RANGE βp PNBI PICRH Ip 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 3

4 Alcator C-Mod and DIII-D data on li(3)-q95 range Analytic formulae li=f(kx, q95) matches li,max C-Mod shots DIII-D C-Mod New ITER VS design criterion κx=.85, li(3)=.2, βp=. Ip=5 MA ( q95~3) G Sips et Al, IT/2-2, G Jackson et Al, IT/P7-2 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page

5 Disturbance metric: max. stabilizable VDE ( Zmax) Let plasma drift up to Z apply fully saturated step voltage to all power supplies in vertical control circuit Vary Z to find Zmax= maximum displacement beyond which VDE cannot be reversed. This is NOT a true control demonstration but reflects best possible Zmax/a (dimensionless) metric provides guidance to ITER from present devices li(3)=., γz= rad/s Key scaling min D Humphreys et Al, IT/2-Rb V max Z 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 p z I γ Page 5

6 Disturbances and Zmax/a capability DIII-D Noise sets typical control limits in DIII-D, Alcator C-Mod, JET, NSTX: RMS Noise-driven <Z> /a ~.5-% Noise reduction procedure demonstrated: RMS <Z> /a <.2% Plasma disturbances may set control limit in ITER DIII-D ITER demonstration discharges Experimental data suggests: max produce ELM-driven plasma perturbation ELM Robust operation requires Z /a ~5-% Z(DIII-D, /a ~ 3% C-Mod) ITER may require greater margin relative to ideal no-noise calculated value Even in absence of noise, disturbances Vertical Position Shift Due to ELM ~2 cm max IAEA Fusioncapability Energy Conference,>Geneva, will likely require 22nd ( Z/a) 5% 3-8 October 28 Page 6

7 Alcator C-Mod and DIII-D data on VDE onset Experiments in DIII-D and C-Mod tokamaks that exceed controllable growth rates show the following thresholds for operation: C-Mod τg (ms) ms Zmax Zmax/a (cm) (%) % % % % Marginal Control max Z /a ~ 2% Voltage Saturates VDE onset Marginal operating point max Marginal operation Z /a ~ % Reliable operation Zmax/a ~ 5 % 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Robust operation Zmax/a >~ % Page 7

8 Options to improve ITER VS: asymptotic analysis Cu thickness=.2 mm Toroidal straps In-vessel coils Diagnostics? Integration? Effectiveness? Zmax(mm) τg(ms): Zmax(mm) VS=6 kv 33 VS=9 kv VS=9kV 6 VS2=6kV7-7 Zmax (mm) 22 3 VS=6 kv 33 5 VS=9kV 6 68 VS=9kV VS2=6kV min V ~265 V/t max Remark: larger Zmax may ~3 kat)with FW! lead (Ito contact τg(ms): τg(ms): nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 9 Page 8

9 Trade-off Current-Voltage (żp, zp ) Curves refer to stabilization NOT to control plasma zp. This function can be: performed by PF coils on VS-time scale (.5 s) OR performed by in-vv coils BUT higher I are necessary (controller dependent!) Z=.2 m Minimum voltage design Tstop~.5 τg~3 ms Reduced Istop Promptness CREATE NL model PET model VS equivalent performances Z=. m Zstop=.22 m Vstop~ 5 kv, Istop~2 ka Pmax~ GW, Estop~5MJ Tstop~.3 τg~9 ms Braking energy ~ MJ 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 9

10 Assessment. Design choice: In-vessel coils provide best performances (stability domain, promptness...). For the same max(z) VS needs x 2 Pstop and x 5 Estop (feasibility!) Passive stabilizers may provide Zmax/a>5% but magnetic reconstruction of żp is an issue (on-line eddy currents measurement?) Cu cladding not effective (marginal increase of τg) 2. Design solution: max To stabilize offsets up to Z ~ 2 mm (=% a) AND reduce power supply cost 3-turn VS coils adopted, anti-series connected to power supply with Imax> 5 (~8 ka), Vmax> kv (~.5 kv). Coils clamped (grounded) on inside vv; 3. Key design/operation issues related to internal coils: Integration in present design (ELMs coils, blanket manifold, feeders routing ) Joule heating due to plasma control, magnetic noise <żrms>~.6 m/s (scaling from JET H/L shots) <Irms > ~< ka Disruption V/I/IxB: disruption/vdes lead to peak terminal voltages ~.3 kv (open-circuit configuration) and peak currents ~ 7 ka (short circuit) Page clamping/support22nd coils (Fz~.5 MN/m!) IAEAadequately Fusion Energy Conference, Geneva, 3-8 October 28

11 Summary and Conclusions Alcator C-Mod, DIII-D, JET, NSTX, TCV provide guidance to ITER for VS requirements: li > ~.2 can occur in ramp-up/down, Zmax/a > 5% is required for minimally robust control; in-vessel coils being designed to provide Zmax/a > ~5%; Zmax/a = 2% is capability of old ITER baseline system To provide the ITER VS system with the suggested margin the most effective solution is the use of in-vessel coils. This may be the only way to reliably control high li plasmas at full κ; 2 sets of 3-turn coils (anti-series) connected to power supply are now part of ITER baseline design; For back-up purposes all necessary provisions should be made to assure VS operation@9 kv & enable installation of VS2; The proposed coil set should become a fully integrated, maintainable component of the new ITER baseline design. 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page

12 κx-li trade-off in ITER High q95 widens li range at low Ip κx has to be reduced to keep vertical control S c e n a r io 2 : C S w ith m a x im u m fie ld 3 T 5 IT E R EO B 5 EO C ( 2.5 M A ).5 M A M A PF coils control? Z, m kx~.8 - q95~ τg(ms) R, m 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 2 9

13 Alcator C-Mod & DIII-D data on Zmax uncertainties Change κ to alter growth rate γz vertical control disabled for different intervals of time to vary initial displacements Z max DIII-D: Computed Z agrees reasonably well with data Computed different κ is ~ 2 x experiments max C-Mod: DIII-D C-Mod 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 3

14 ITER control loop scheme: VS and shape control 8 PF PF2 VS C S 3U 6 g5 PF3 C S 2U g g3 g6.. C S L Z, m C S U 2 C S 2L -2 PF g g2 C S3L - VS -6 PF5 PF R, m 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 3

15 ITER control loop stability analysis In V e s s e l c o il v o lt a g e P la s m a ve rt ic a l p o s it io n Z Z [m ] V [k V ] T im e [ s ] In V e s s e l c o il c u r r e n t 5.5 T im e [ s ] 2 I( t ) d t 3 [k A 2 s ] I [k A ] T im e [ s ].5 T im e [ s ] 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 5

16 Trade-off Current-Voltage (żp, zp ) Z=.2 m 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 6

17 Noise and Ieff In V e s s e l c o i l v o l t a g e P la s m a ve rt ic a l p o s it io n Z. Z [m ] V [k V ] T im e [ s ] In V e s s e l c o i l c u r r e n t [k A 2 s ] -2.5 T im e [ s ] 5 IV S [k A ] V S [k V ].5 T im e [ s ].5 T im e [ s ] V S c u rre n t Ieff~8 ka V S vo lt a g e -.5 T im e [ s ] 2 I( t ) d t 2 I [k A ] <żrms>~.6 m/s - 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28.5 T im e [ s ] Page 7

18 ELM & VS coils layout Upper VS coil Upper ELM coil mid leg Upper feeder toroidal leg sector Lower ELM coil Lower VS coil Multi purpose in vessel coils: - ELM control - Vertical Stability control - Resistive Wall Mode (RWM) control 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 8

19 Joule heating of VS and ELMs coil T in VS conductor center Coolant at C T in steel Welds zone on VV kept at ~ C 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 9

20 PF Coil Proximity JET D3D AUG ITER double-wall vessel is very efficient in screening the field from the VS coils 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 2

21 KSTAR Has Double-Wall Vessel and Eventually will Have In-Vessel Coils and Possibly Passive Stabilizers

22 EAST Has Double-Wall Vessel and In-Vessel Coils with No Passive Stabilizers

23 Open-circuit voltages 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 23

24 Short-circuit currents and induced forces 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 2

25 Short-circuit currents 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 25

26 Effects of plasma radial position offsets at high li Modified shape shifted inward by 9 cm Original shape li~., βp ~. Ip~5 MA, κxp~.8 Inward shift = 9cm Reference VS=6kV Growth rate γ s- 27 s- plasma vertically unstable! 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 26

27 Passive stabilizers nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 27

28 Passive stabilizers effects on magnetic diagn. & control Partial Loops Continuous Loops. Dedicated Saddle Loops Ex Vessel normal coils 2. Z (m) Ex Vessel tangential coils. Divertor Coils Normal Coils -2. Tangential Coils First Wall -. VV outer -6. VV inner VV triangle R (m) 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 28

29 Passive stabilizers effects on magnetic diagn. & control Time behaviour of the actual position and of the reconstructed position when the VS controller is driven by the reconstructed velocity δzp(t) VS controller driven by the actual velocity Z c [m ] T im e [ s ] δzp(t) VS controller driven by reconstructed velocity 2 6 T im e [ s ] 8 2 NB: VS and Shape controllers are always re-designed and re-optimized if the actual or reconstructed velocity is used in the feedback loop! 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 29

30 Passive stabilizers effects on magnetic diagnostics & control Time behaviour of the actual velocity and of the reconstructed velocity when the VS controller is driven by the actual/reconstructed velocity.2 5. A c t u a l ve lo c it y R e c o n s t ru c t e d ve lo c it y A c t u a l v e lo c it y R e c o n s t ru c t e d v e lo c it y.2. 5 δzp/dt Feed-back with actual velocity. 5 V e lo c it y [ m / s ] V e lo c it y [ m / s ] δzp/dt Feed-back with reconstructed velocity T im e [ s ] nd IAEA Fusion Energy Conference, Geneva, 3-8 October T im e [ s ] Page 3 5

31 Noise characterization in JET 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 3

32 Noise characterization in JET 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 32

33 Ex-vessel and In-vessel coils performance Ip=5 MA, βp=., li(3)=.2. CREATE and PET (italics) CODES τg (ms) ms Zmax (mm) VS 6 kv Zmax (mm) VS 9 kv Zmax (mm) VS 6 kv VS2 6kV Zmax (mm) VS 9 kv VS2 6 kv PF coils (PF2-5) n.a n.a..2 mm cu cladding n.a n.a. Toroidal straps n.a n.a. In-vessel coils 8 79 n.a.37 Zmax (mm) VS3 265 V/t 3 kat 2 In-vessel coils (single turn values) Zmax (mm) 2 2 Vmax (V/t) 33 5 Imax (kat) 5 7 Trade-off Vmax-Imax! 22nd IAEA Fusion Energy Conference, Geneva, 3-8 October 28 Page 33