July 2005P Lomas Plasma Ops II1 PLASMA FORMATION CURRENT RISE (FLUX CONSUMPTION) STANDARD CONFIGURATIONS TIME EVOLVED SCENARIOS MAXFEA/PROTEUS PULSE SCHEDULE.

July 2005 P Lomas Plasma Ops II 1

•PLASMA FORMATION

•CURRENT RISE

•(FLUX CONSUMPTION)

•STANDARD CONFIGURATIONS

•TIME EVOLVED SCENARIOS

•MAXFEA/PROTEUS

•PULSE SCHEDULE EDITOR

•NEW CONFIGURATION FOR 2005/6

•DISRUPTIONS

PLASMA OPERATIONS II

44 slides


AVALANCHE PHASE

Breakdown in a torus Townsend discharge between electrodes

inter electrode distance connection length along field lines to parts of vessel

1. BREAKDOWN AND PLASMA FORMATION

Vloop = 17V, E~0.91 V/m i has a minimum 78m at P ~ 6 x10-5 Torr

8 x 10-5 bar

Need to go 4 times round the torus to create an ionisation!

P pressure torr (750 Torr = 1000mBar).

The ionisation length in meters

mean free path

0

50

100

150

200

250

300

350

400

0.00E+00 5.00E-05 1.00E-04 1.50E-04

pressure mbar

lam

bd

a m

1.82 V/m

1.35

0.91 V/m

0.67 V/m

0.35 V/M

E electric field Volts/m


The connection length

<Bz> is the average stray field over the

null

size of the field minimum toroidal

field

Number of ionisations per transit=LC/i

Avalanche successful if LC/i >>1

July 2005 P Lomas Plasma Ops II 4 = 0.0073 Wb

HEXAPOLAR NULLAt 20kA premag, 2turns P3m

and static P4 bias 425Amp(including vessel eddy currents).

Vloop = Ipre• R=20kA*0.6/710

=17volts<Bz> ~5 x 10-4 over anull 0.8m

Connection length Lc ~ 1100m at 2.8T.

many (~14) ionisations per lifetime

avalanche should occur. Need to catch Ip with net Bv and allow this to penetrate the vessel, so is better to ramp P4 from ~230amp to ~750A over the first 25ms. The null actually lasts longer because field due to inboard and outboard eddy currents cancel. Need also to compensate current in MKII divertor baseplate with radial field.


Townsend expression with LC=i min electric field for avalanche

Townsend avalanche.xls 16/07/06

0.01

0.10

1.00

10.00

1.00E-07 1.00E-06 1.00E-05 1.00E-04 1.00E-03pressure mbar

E v

olt/

mLc=5000 Mode B 2.6g in 2.6T a=2Lc=1100 Mode D 5g in 2.8T a=0.8Lc=560 Mode D 10g in 2.8T a=0.8Lc=340 40kA D 10g in 3.4T a=0.40.35 V/M Mode B0.67 V/M 15kA Mode D0.91 V/M 20kA Mode D1.35 V/M 30kA Mode D1.82 V/M 40kA Mode DExperimentsExperiments (error bars)Experiments (error bars)

Curves for different

connection lengths

Lines for various electric

fields in JET

A vertical field error of 0.5mT 38Amp in P4halve connection lengthstill margin!


COULOMB PHASEIonisation energy will keep the plasma cold (few eV) until all the neutrals are ionised (or driven into the wall). When fully ionised the temperature will rise to 10eV or so whereupon coulomb collisions dominate.

Between 10 and 20eV we reach the peak radiation from Carbon and Oxygen. Clearly must have enough power to burn through. Having burnt through the Poynting flux can then raise the plasma magnetic energy i.e. increase the plasma current.

Transition from atomic and molecular to Coulomb collisions when:

1.4% 2 eV= 3.8% at 4 eV

10% 8 eVThe current density at the transition is typically:

In JET, this current is 20 to 40 kA

/(1- ) 5 x 10-3 Te3/2

Where = ionization fraction:

= noeVde

~5 to 20 kA/m2


TYPICAL MODE D BREAKDOWN

BURN THRU’

The first 100ms of JET pulse 5485. The division between avalanche and Coulomb phases is based on equation 5.


CAN WE CATCH PLASMA?

Whilst the plasma current is low <100kA the fields due to the plasma current are comparable to stray fields and magnetic errors. This would be OK if the initial plasma position were stable. Unfortunately it is not.

The growth rate of the initial plasma in the stray fields at initiation is typically

Fortunately raising the plasma current is stabilising, so we must

• Get through to 100kA quickly• Keep the plasma current increasing• Until vertical and radial control take over.

~ 50 s-1


OH Mode B, C or D (or even A)Premagnetisation current (P1 voltage control wave in Mode B)Prefill pressureBias P4 (null formation)Pre-programmed P4 rise during avalancheCatch radial plasma position with flux control

(in Mode C or D requires vertical field booster amps)Compensate induced currents in MKII structure and enable

vertical stabilisationActivate imbalance for vertical location of plasmaActivate P1 voltage reduction switches (modes C D)

Sounds complicated?

BREAKDOWN CHECKLIST


Simply select from PlasmaOpsGroup/PremagLibrary

Need to change breakdown?


Ip and P1 controller

P1 voltage feedforward

waveform

Proportional controller

Poloidal flywheel generator excitation

P1 coil

PF/SC-VP1<CS

PF/SC-VFGC<CS

subtract add

PF/FGC-EXCTR<CHK:100

Poloidal flywheel generator convertor

output

P1 voltage measurement

PF/FGC-FIELD<AMP:100 PF/FGC-FIELD<VLT:100

PF/FGC-CONV<CUR:100

PF/SC-VP1<MS

Standard Mode B uses VP1=–5kV to give maximum generator excitation then ramps to +5kV around 40.4s to give maximum de-excitation ( note waveform sign convention after reconnection, generator currents and voltages are reversed and are negative to produce an increase in plasma current. The PPCC signals themselves adopt the standard R--Z convention.). Thus VP1 saturates the excitation voltage and it is not until VP1 goes to zero at 40.6s do we have any plasma current feedback.

The proportional controller has unity gain but clips the output SC-VFGC<CS to 1.5kV which is the limit of the excitation voltage FGC-FIELD<VLT:100. The term 255/NPFGC in Lennholm`s lecture notes scales with flywheel speed. The PPCC output is converted to 10V in the digital to analogue convertor which sends the signal to the generator where it is read back as FGC_EXCTR<CHK:100.

New Mode B scheme uses VP1 as a true feed-forward for the P1 voltage. The generator is connected to the P1 coil and R3/R4 in parallel. Thus, when the P1 current is low the P1 voltage is given by the generator current times the resistance of R3/R4. To get maximum generator excitation and the maximum ramp of P1 voltage VP1 should be 1.5kV more than the measured voltage and to get maximum de-excitation VP1 should be 1.5kV less than the measured voltage (when the same sign convention is used!). When the difference between VP1 and the measured voltage is less than the saturation level of 1.5kV then plasma current feedback can be active. In PPCC the plasma current feedback contribution to SC-VP1<CS is Ip-Ip

ref*0.0145. Using this scheme the vigour of the initial Ip rise can be varied between 4 and 8MA/s.

P1 voltage controlWhen VP1 is equal to the measured P1 voltage the generator excitation remains constant, the generator current remains constant and therefore the P1 voltage remains constant


Ip

Vp1

H

Gen Exc

li

“Old” BColours new B

Wrong gas timing mhd

Slow rise P1 volts,vary max.

Limit bounce

Sticks at zero

overshoot

Can vary initial li#52366, 60, 56, 59, 73


Mode B new v old

Suppress premature avalanche


CURRENT RISE INSTABILITIES

• During the current rise phase, j(r) may be rather flat with steep edge gradients and this can generate instabilities as integer values of q pass through plasma edge.

• At q (a)>6, these mhd modes are observed as bursts of rotating n=1 mhd activity which decay.

• For q (a)<6, the modes may lock and ultimately cause a disruption.

– The disruption can be delayed for many seconds and may not occur until the current decay begins.

• Although the mode is initiated with m= q (a), it is found that the mode which causes the disruption has m=2, n=1, indicating that the dominant poloidal mode number changes after locking.


MHD ACTIVITY IN THE CURRENT RISE

Bursts of n=1 activity are observed at q(a)~5,4,3 and the mode locks at q(a)~3


The li-q diagramThe behaviour of surface kinks with m~q,n=1, during the current rise phase can be understood in terms of the li-q diagram

Near the upper boundary of the diagram, disruptions (e.g. density limit) can occur as a result of the destabilization of modes on the q=2 surface.

See:JA Wesson et al, Nuclear Fusion 18 (1978) 87C Z Cheng et al, Plasma Phys and Contr Fus 29 (1987) 351P J Lomas et al, JET-IR(87)19 (Proc 1987 APS)J A Snipes et al, Nuclear Fusion 28 (1988) 1085))


CONTROL OF CURRENT RISE INSTABILITIES‘Knobs’ to modify the evolution of the current profile:

Breakdown premagnetization:

reducing premag reduces the rate of increase in current but wastes flux

Thyristor switched resistors: TMS

These reduce the loop voltage after breakdown and reduce the rate of current rise

Changing prefill pressure or requested density waveform:increased fuelling favours a more peaked current profile, but has to be balanced against the possibility of hitting the density limit in the current rise.

Ramping of toroidal field with current: break down at lower toroidal field and ramp, flux compression leads to more peaked current profile.

Aperture expansion:promotes current penetration and allows, in particular, q(a)=4 to be crossed at low

current; following this current can be increased at approximately constant q(a).Reduce current ramp rate in slow rise phase:

Best ramp rates achievable with all of the above are ~1MAs-1.


FLUX CONSUMPTION

The absolute flux at the plasma surface is given by

Pre is the flux at the end of premagnetisation and is positive for positive Ip and negative

for negative Ip The second term on the RHS is the resistive flux and the third the internal inductive flux. The normalised flux inductance hi is defined from the Faraday internal flux according to

If we have a computed full equilibrium (Proteus or IDENTD) we have the resistive consumption immediately from

Flux inductance hi versus normalised internal inductance li

How to compute possible pulse length.

ALL ON WEB

Good for the soul, but all computerised nowadays!


INDEX OF CONFIGURATIONS AND STANDARD SCENARIOS12/12/02 Issue 3.

Page 1 General purpose and Configs for ELMy H mode

Name Version Description Example pulse/time Additional RestrictionsLIMITER 1 Standard limiter config in proportional control #50000,45.3-50s NoneXFORMATION 1 X-point just formed on vertical target. For slow ramp from limiter to X-point.. Don`t forget to tune RIG/ROG. #50000, 50-54.5s NoneXFORM_HORIZ 1 X-point just formed on horizontal target. For slow ramp from limiter to X-point.. Don`t forget to tune RIG/ROG. 60870, 46.9s NoneStandardFat 1 Standard FAT in proportional control #49935. 63-65s NoneTERMINATION 1 Robust X-point scenario for current ramp down phase, note preferable to use Z0P control. #50000, 65-71s NoneC_SFE_LT 1 Corner, Standard Flux Expansion, Low Triangularity #50000, 63-65s NoneHC_SFE_LT 1 Horizontal to Corner, Standard Flux Expansion, Low Triangularity #50040, 54.5-60s NoneV_LFE_LT 1 Vertical target, Low Flux Expansion, Low Triangularity #50039, 54.5-64s NoneV_SFE_LT 1 Vertical target, Standard Flux Expansion, Low Triangularity. =0.25 #50000, 54.5-63s NoneVLO_LC_LT 1 Vertical target Low strike, Large wall Clearance Low Triangularity config.. #53318, 60s 2.5MA see approval documentation

VHI_LC_LT 1 Vertical target High strike, Large wall Clearance, Low Triangularity config. #53320, 60s 2.5MA see approval documentation

VIR_LC_LT 1 Intermediate strke position between VLO and VHI for infra-red camera. #60568,62s 2.5MA see approval documentation

ASDEX_LT 1 ASDEX like shape at Low Triangularity using standard turns. #53857, 65-75s 2.5MAV_SFE_HT 1 Vertical target, Standard Flux Expansion, High Triangularity. =0.32 #50054, 54.5-64s 3.0MAC_SFE_HT 1 Corner, Standard Flux Expansion, High Triangularity, sometimes called “deep throat”. #52312, 56-65s 2.7MAV_SFE_VHT 1 Vertical target, Standard Flux Expansion, Very High Triangularity. =0.38 #50056, 54.5-64s 2.7MAV_SFE_EHT 1 Vertical inner corner outer, Standard Flux Expansion, Extreme High Triangularity. =0.42-0.46. CMOD conf #50075, 54.5-64s 2.7MAV_S_EHTM 1 Vertical inner and outer, Standard flux expansion, Extreme High Triangularity. =0.42-0.46 #52520, 58-64s, 2.7MAVC_ EDA 1 Vertical inner Corner outer, standard flux expansion, max lower triangularity. =0.40. For Enhanced D Alpha #50663/7, 58.5-64s 2.0MAITER_EXPERT 2 Original ITER-LIKE config in RX/ZX control, Beware D3 limits in long pulses. =0.48 #50687, 57-64s 2.3MA, no mods allowedITER_ZSIRSO 2 ITER-LIKE with preferred control scheme ZSI/RSO. =0.48 #52988, 57-64s 2.3MA, Z0P/TOG mods only >2MA

HUD_LLD 1 High upper delta, low lower delta for ELMY H. This is an ITER-LIKE variant 53047, 57-64s Expert OnlyITER_OSVT 1 ITER-LIKE variant with Outer Strike just on Vertical Target 53046, 59-64s Expert OnlySEPTUM 2 Standard septum scenario, low triangularity #50040, 60-65s NoneSEP_HT 1 Septum, High Triangularity as devised by Saibene, like V/SFE/HT moved down rigidly #53045, 56-65s NoneHIXR_GB 1 High X point Reverse D1/D4 developed on MKIIA for Gas Box =0.35 #50078, 54.5-64s NoneH_4M5_LT 1 MKIIA 4.7MA pulse #42983 retuned for MKIIGB_SRP 60870, 52-58s 3.5MA see approval documentation

VH_3M5_HT3 1 Delta=0.42 potential 3.5MA configuration 60685 57-63s 2.5MA see approval documentation

DOC_L, DOC_L_probe 1 Diagnostic Optimised config and variant for target probes. 55933, 58s 2.5MA see approval documentation

Note, “None” in the additional restrictions column indicates that normal JOI and SLLicence limits apply. Operation above the plasma current limits indicated in therestrictions column requires either an Expert SL or explicit approval. Session leaders withrestricted licences should scale these plasma current limits by 0.75 for a 3MA licence andby 0.5 for a 2MA licence.

For the configuration ITER_ZSI/RSO above 2MA, SLs (other

than Expert) may only modify Z0P or TOG. The reference pulse in the “example”column illustrates an unmodified Standard Scenario except where marked . For theexamples marked , the standard scenario was created from this pulse. Version numbermarked indicates earlier versions automatically updated.

Approved POG ______________________, date___/___/___. ApprovedEAG _____________________, date___/___/___

Link to:-http://w3.jet.efda.org/pages/torus-ops/pages/op-instructions/index.html

http://w3.jet.efda.org/pages/torus-ops/pages/op-instructions/index.html


INDEX OF CONFIGURATIONS AND STANDARD SCENARIOS 12/12/02 Issue 3page 2, Configs for Advanced tokamak.

Name Version Description Example pulse/time Additional RestrictionsVC_OS_LT 3 Vertical/Corner strikes, standard Optimised Shear configuration, Low Triangularity, #48994, to 47.8s 3.8MAH_OS_LT 1 Horizontal target, standard Optimised Shear configuration, Low Triangularity, #54730, 45s 3.6MA See approval ducumentation

S_OS_LT 1 Sweepable version of VC_OS_LT #54624 3.0MA See approval ducumentation

VC_LC_LT 1 Vertical/Corner, Large wall Clearance, Low Triangularity for post Optimised Shear phase. #50940, 48-50s NoneVC_ELC_MT 2 Extra Large wall Clearance for post Optimised Shear phase with high PFX favoured by G.Sips. 2.5MAVC_ELC_LT 2 Extra large wall clearance config for post optimised shear phase as presently used. #50940, 50-55s NoneVC_LHCD_LT 3 Vertical/Corner strikes, config for LHCD coupling for use post optimised shear phase. #50950, 49-55s NoneVC_LHCD_LT 4 Best configuration for LHCD coupling post optimised shear phase #52371, 50-54s NoneLD_LK 2 Low delta, Low kappa for optimised shear #52659, 46s NoneLD_HK 1 Low delta, High kappa for optimised shear #52656, 46s 2.5MAHD_LK 1 High delta, Low kappa for optimised shear #52622, 47s 2.2MAHD_HK 1 High delta, High kappa for optimised shear 52642, 48s 2.2MA

Other Configs.Name Version Description Example pulse/time Additional RestrictionsZASPI, ZASPM, ZASPO 1 Inner, Middle and Outer position of aspect ratio scan. 52945 Expert onlyYTEST and ZTEST Various Commissioning only Plasma Ops SLs onlyDNX or Top_SNX - Any double null or near double null or Top Single Null X-point config 53702-15, 59-65s Explicit Approval RequiredNear circular limiter Any outboard limiter plasma with elongation<1.2 or inner wall plasma with minor radius<0.6 Explicit Approval Required

Note, “None” in the additional restrictions column indicates that normal JOI and SL Licence limits apply. Operation above the plasma current limitsindicated in the restrictions column requires either an Expert SL or explicit approval. Session leaders with restricted licences should scale these plasmacurrent limits by 0.75 for a 3MA licence and by 0.5 for a 2MA licence. The reference pulse in the “example” column illustrates an unmodifiedStandard Scenario except where marked . For the examples marked , the standard scenario was created from this pulse. Version number markedindicates earlier versions automatically updated. For H_OS_LT VDE tests suggest that 3.8MA should be OK but the approval documentation has notyet been revised, so please give advance warning if you want to avail yourself of this.

Approved POG ______________________, date___/___/___. Approved EAG _____________________, date___/___/___

Please note for configurations where an approval document has been completed this can be found in the red file. Thisdocumentation provides further information and advice about the use of these configurations. It also provides the mechanism bywhich approval can be granted for operation above the normal limits on plasma current. The Configuration ApprovalDocumentation provides information about configurations under development, for which the plasma current limit is 2.5MA.


Fully developed configsUnder development 2003/4New for 2005/6

Link to:- http://w3.jet.efda.org/index.shtml

http://w3.jet.efda.org/index.shtml





HC_SFE_LT:1

Low Triangularity0.24

SOL surfaces 10 mid-plane mm

Standard Flux Expansion

Horizontal Corner

PFXSHIMBD1D2D3D4

kA/MA65-0.1-257.2-2


CONFIG. V/SFE/LT 2.5MA/2.4T V/SFE/VHT


INTER-MACHINE SIMILARITY/IDENTITY PULSES

Also EDA regime studies


SCENARIO FOR ELMY H MODES

Slow X point formation from inner limiter, slow ramp to final config.


Scenario for low q95, High Ip H Modes

Small bore start upConst q limiter rise phase

expanding apertureramped TFearly sawteeth ~45s

q>3 limits Ip to ~3.6 MA.X Point must await P1 reversal (PFX)Continue ramp during X

Fully penetrated current at start of flat top. (used up to 6 MA q95 ~2)


X point formation for low density H modes

RAPID FORMATION OF HIGHLY TRIANGULAR PULSE FROM OUTBOARD LIMITERAt this high Ip limiter phase q~3.2. Also error fields a problem near q ~3.

MINIMUM DENSITYMINIMUM GASMINIMUM RECYCLING

High T


Scenario for Advanced Tokamak

Breakdown without premagnetisation

Large bore minimise current penetration

Early x with shape + divertors

PFX from 41.4

Replace shape with PFX for modest triangularity.

OPTIONS• Can retain large shaping for high triangularity • Can also hold limiter plasma through rise

Continue current ramp beware MHD at q95~4.

Start heating when desired q(r) has evolved.

Timing q(o) (Typ 1-4)

Ip ramp shear

Can also fine tune q(o) with off axis current drive.


MAXFEA, fast, interactive, user friendly

li

Circuits Currents

Nudge buttons Options


PROTEUSFiner mesh, more options, more outputs, more detail. Portable. < 1min/eq on PC.Less user friendly.


Single Control Window

All done with waveforms

Simplest pulse, limiter, using same control scheme throughout

Start time


Need outer gap control before X point

Transition FLUX to ROG Make X by ramping

and holding waveforms


Up to here user written waveforms

Choose Standard scenario V/SFE/LT

Button computes waveforms

Result

Thinking of changing to a new configuration?


Can run almost the entire pulse using standard scenarios.

Forward Transition Times

Dynamic phase using waveforms

LimiterX

FormationHC/SFE/LT

SeptumTermination

Still have to write ROG or RIG waveform!


ITER-LIKE equilibrium on divertor MKIIGB

Limited by D2 to about 2.7MA.

400Tonne disruption at 2.5MA!

Without septum cannot increase because inner SOL spills over.

However can lower X which reduces coil currents and allows higher Ip.

475 Tonne at 2.7MA!

10mm SOL

surfaces


3.5MA high triangularity without Septum

Forward D1~0.48

Zero D1

Zero D1


Family 3MA5_ITERPJL Run

New High-Field GapClosure Tile (HFGC)

New “Load Bearing” SeptumReplacement Plate (LBSRP)

Existing Vertical Targetsand inner/outer horizontaltargets unchanged.

P1EPFXShape-P4IMBD2D3D1D4

Most new configshave forward D1(same sign as Ip).Most old configshave reversed D1

Gap notcompletelyclosed forpol<80°>1.6 °


3.5MA HT3 up=0.44,low=0.37 D3D1 up=0.53,low=0.48

4MA HDLX up=0.43,low=0.45 3.5MA ITER up=0.45,low=0.54 Low X point


Disruptions, large forcessignificant displacements

See also Engineering Limits Lecture by I.Nunes

VS SWITCH OFF AND KICKS. Test_VDE.xls database 23/07/04. P.Lomas

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

0 50 100 150 200 250 300 350 400 450 500

Fswing Tonnes

Rol

l mm

Switch offKick upKick downELMn=2VS set upVDEH_4M5_LTVH_3M5_HT1VH_3M5_HT2VH_3M5_HT3V_3M5_HT4Z_IT_AT:001VC_OS_LT:003ITERDNXH_OS_LTLinear (ITER)


Levels and windows Matrix of inputs and responses

Shape Control Fast Stop initiates Fast

ramp-down

Set up using Plasma Protection System


SUMMARY

Shown how to

•Form plasma and ramp the plasma current avoiding instabilities

•Choose configuration and optimise the scenario

•An introduction equilibrium codes and the Pulse Schedule

•Some equilibria for the new divertor

•Mentioned some aspects of disruptions.

Tomorrow you will learn about

• plasma control and new in-vessel hardware

Then get your first hands on experience with the tools.

July 2005P Lomas Plasma Ops II1 PLASMA FORMATION CURRENT RISE (FLUX CONSUMPTION) STANDARD CONFIGURATIONS TIME EVOLVED SCENARIOS MAXFEA/PROTEUS PULSE SCHEDULE.

Documents

plasma formationvloop

p pressure torr

ka mode d2

ka mode d3

vm mode b1

ionisation length

field lines

torus townsend discharge