Disruption mitigation with high-pressure noble gas injection

Transcription

1 Disruption mitigation with high-pressure noble gas injection D. G. Whyte*, University of California San Diego T.C. Jernigan, Oak Ridge National Laboratory D.A. Humphreys, A.W. Hyatt, T.E. Evans, P.B. Parks, A. Kellman, M.N. Rosenbluth General Atomics C.J. Lasnier, Lawrence Livermore National Laboratory, D.S. Gray, E.M. Hollman, University of California San Diego 9 th IAEA Fusion Energy Conference 4-9 October, Lyon, France * Current address: University of Wisconsin - Madison # D.G. Whyte, IAEA Oct

2 Burning plasma experiments must develop a thorough strategy to deal with disruptions Plasma operations Obtain needed performance away from known stability limits. Disruption avoidance Control of plasma pressure / current profiles (e.g. NTM suppression) Disruption detection Reliably determine onset of triggering instability in real-time. Disruption mitigation Provide a rapid and safe emergency shutdown technique in order to alleviate damage to costly internal components. DIII-D Experimental Results # D.G. Whyte, IAEA Oct

3 High-pressure gas injection of noble gases simultaneously satisfies the three requirements for mitigating the damage caused by disruptions. Surface thermal loading: Focused heat loss ablates/melts divertor material Solution: Deliver large quantities of impurity into core plasma to dissipate ~% plasma energy by relatively benign, isotropic radiation.. Poloidal halo currents: Large mechanical JxB stresses on vessel Solution: Rapid thermal quench, uniform resistive plasma and a plasma that remains centered in vessel during current quench substantially reduce vessel halo currents. 3. Runaway electrons: Relativistic MeV electrons (~I p ) from avalanche amplification during current quench in large-scale tokamaks (e.g. ITER) Solution: Runaway electrons can be suppressed by the large density of bound electrons in neutral gas in plasma volume. #3 D.G. Whyte, IAEA Oct

4 High-pressure gas injection on DIII-D delivers large impurity density to the plasma volume Gas jet parameters! 7 atmosphere reservoir! ~ ms response fast-valve! N inject ~ 3x ~ 3 N e,target! jet port/nozzle diameter ~.5 m! Gases: D, Helium, Neon, Argon gas jet.6 m reservoir & valve At entry to plasma:! n jet ~ 4 m -3 : P jet ρ v ~ kpa! v jet ~ c s ~ 3-7 m/s #4 D.G. Whyte, IAEA Oct

5 All gas jet species penetrate through to central plasma at approximately sonic velocity cold front location (r/a) ne (m-3) Central electron density ( < r/a <.3) D valve Gas jet propagates D through vacuum Gas jet propagates through plasma He Ne He Ne Ar 4 6 Elapsed time from valve opening (ms) velocity (m/s) 8 4 In vacuum Through plasma sound speed D He Ne Ar 3 4 Gas atomic weight #5 D.G. Whyte, IAEA Oct

6 All gas jet species penetrate through to central plasma at approximately sonic velocity cold front location (r/a) ne (m-3) Central electron density ( < r/a <.3) D valve Gas jet propagates D through vacuum Gas jet propagates through plasma He Ne He Ne Ar 4 6 Elapsed time from valve opening (ms) Empirical observations suggest neutral gas penetration: Hydrodynamics: P jet ~ P recoil > P e,plasma Electron dynamics: stopping distance of kev electrons << jet diameter. No dependence on radiative properties of gas species. Gas jet dynamics / penetration currently not well diagnosed or understood. B plasma electrons stopping distance for e - gas (P ~ kpa) r #6 D.G. Whyte, IAEA Oct

7 An example pre-emptive neon gas jet impurity injection into a stable plasma demonstrates a rapid, radiative plasma termination with no runaway electrons. gas jet Te (kev) central plasma Te current quench δb/bt (%) Ip (MA).5-5 nozzle pressure edge ECE central soft X-ray emission (a.u.) current quench loss of closed flux surfaces 5 5 time after gas jet trigger (ms) Plasma remains well-centered in the vessel ne ( m-3) Prad (GW) 4-5 midplane ne central plasma P rad 5 5 time after gas jet trigger (ms) Injection is very benign to tokamak " No pump system damage " Injected gas absent in breakdown of subsequent discharge. #7 D.G. Whyte, IAEA Oct

8 DIII-D has demonstrated real-time disruption detection, which is used to trigger gas jet injection for disruption mitigation. VDE detection algorithm tracks vertical stability in real-time. Triggered neon gas jet reproducibly terminates plasma in ~ 5 ms, before plasma vertically displaces into vessel wall. Other disruption detection algorithms developed & tested. Radiative/density limit Growing tearing modes leading to disruption. - - VDE detection and gas jet mitigation Plasma Vertical Position (cm) Vertical Instability Trigger from Control System to Neon Gas Jet Plasma Current (MA) no mitigation equilibrium position trigger levels 4 6 time after removal of vertical stabilization (ms) #8 D.G. Whyte, IAEA Oct

9 Radiative dissipation of plasma energy by impurities provides optimal mitigation of thermal loading and vessel forces Divertor thermal loading reduced to level similar to type-i ELM Divertor JxB forces caused by VDE reduced: Rapid, centered current quench.5 Econducted to divertor (Ethermal+Emagnetic) Vertical Instability Beta Limit ~ E thermal / (E thermal +E magnetic ) Radiative Limit Argon (Pre-emptive) Neon Triggered by PCS for Z<5cm Gas Jet Mitigation Peak poloidal halo current (MA).. No mitigation.5. VDE algorithm trigger threshold for vertical displacement (m).5 Toroidal peaking factor #9 D.G. Whyte, IAEA Oct

10 Physical models of disruption mitigation have been developed and validated Ionization / energy balance of plasma in presence of large density of injected impurity: KPRAD. Full charge-state dependent atomic data (non-coronal). Self-consistent time evolution of Z eff, <Z>, T e, T i,p rad,p ohmic. Injected impurity species, n imp and j // imposed from experiment. Performed on individual flux surfaces or volume averaged. Poloidal halo currents Analytic circuit equation for core/halo/wall coupling (GA halo). Runaway electrons Parallel electric field from KPRAD and Ohm s law: E=η j Rosenbluth, et al. formulation for RE avalanche amplification. # D.G. Whyte, IAEA Oct

11 Ionization/energy balance model (KPRAD) matches key features of gas jet mitigation experiments: Initial burnthrough # P rad # T e collapse # n e clamped Model result: Neon gas jet into DIII-D Te (kev) ne ( m-3) Prad (GW) 4-5 central plasma Te midplane ne central plasma P rad model data 5 5 time after gas jet trigger (ms) # D.G. Whyte, IAEA Oct

12 KPRAD and halo current modeling show effective mitigation using gas jets in a burning plasma device. Neon gas jet into example burning plasma device: ITER-EDA (R~8m) n ( m -3 ) Prad (TW) 8 n e T e ( 4 ev) W th (GJ) P rad (TW) nneon time after jet trigger (ms) Pressure from Halo Currents (MPa) ITER FEAT Simulation: Disruption Stress on Divertor Baffle Dome MITIGATED t [sec] UNMITIGATED # D.G. Whyte, IAEA Oct

13 KPRAD and halo current modeling show effective mitigation using gas jets in a burning plasma device. Neon gas jet into example burning plasma device: ITER-EDA (R~8m) n ( m -3 ) n e T e ( 4 ev) nneon Prad (TW) W th (GJ) 8 First wall heating: P rad (TW) ITER-EDA # 33 MJ m - s -/ ITER-FEAT # MJ m - s -/ time after jet trigger (ms) Melt/ablate limits: Beryllium ~ 5 MJ m - s -/ Tungsten ~ 45 MJ m - s -/ Carbon ~ 4 MJ m - s -/ #3 D.G. Whyte, IAEA Oct

14 Runaway electrons are controlled on DIII-D due to high total density of injected impurities Ip (MA) Soft X-ray emission (a.u.) Ne ( ) #7844 #9574 confined runaway electron tail emission from confined and lost runaway e- Argon gas jet N Ar ~ 4x Argon pellet N Ar ~ x (a) (b) (c) 5 time into current quench (ms) #4 D.G. Whyte, IAEA Oct

15 KPRAD model describes key features for runaways: E par and <Z> for gas jet and impurity pellets τ L/R (ms) Z imp Epar / Ec 4. pellets DIII-D runaway threshold gas jet 9 Injected n impurity (m-3) Experiment He Ne Ar Model He Ne Ar " Constant E par / τ L/R only sensitive to injected species. " Average impurity charge state falls below unity. " Dreicer evaporation criterion is broken at high n imp suppressing runaways: E par > E c (= - n e,total ) [V/m] e - acceleration > frictional drag. #5 D.G. Whyte, IAEA Oct

16 KPRAD model predicts runaway suppression in ITER as n impurity increases, E parallel ~ constant, Z imp decreases τ L/R (ms) Z imp Epar / Ec G Z imp 4. 4 pellets runaway threshold DIII-D runaway threshold RE gain ~ exp (G) gas jet ITER. 9 Injected n impurity (m-3) Experiment He Ne Ar Model He Ne Ar Runaway amplification growth rate: I RE exp(g) G (E par /E c ) τ L/R Reasonable scaling to ITER " < liter reservoir at atmospheres needed for ITER size device. " Conservative calculation since runaway transport losses are ignored. " Relatively simple technology. #6 D.G. Whyte, IAEA Oct

17 Issues regarding application of gas jet to burning plasmas Optimizing mitigation scenarios by choices of impurity Neon and argon: high radiation rates & optimal RE suppression Helium: low radiation rate, RE suppression at higher n imp, slower current quench. Development of reliable triggers for gas jet. Physics-based parallel disruption algorithms now being test on DIII-D Gas jet development Ram pressure > several atmosphere needed to penetrate burning plasma Create and benchmark detailed model for jet penetration. Experiments on larger, higher T e (JET) and high B (C-Mod) tokamaks. #7 D.G. Whyte, IAEA Oct

18 A simple and robust method of mitigating the damaging effects of disruptions has been developed High pressure jet penetrates to center of core plasma. Centrally deposited radiating impurity provides optimal thermal and halo current mitigation. A sufficient quantity of injected gas suppresses runaway electrons by collision damping on neutrals. Physical models of mitigation have been developed and validated on DIII-D, giving confidence in our extrapolation of this technique to burning plasma experiments. #8 D.G. Whyte, IAEA Oct