T. Nakano, N. Asakura, H. Takenaga, H. Kubo, Y. Miura, S. Konoshima, K. Masaki, S. Higashijima and the JT-60Team Japan Atomic Energy Research Institute, Ibaraki, Japan. Impact of nearly-saturated divertor plates on particle control in long and high-power-heated discharges in JT-60U 20th IAEA Fusion Energy Conference Vilamoura, Portugal, 1-6 November 2004

Background ISSUE: Experience and knowledge of operation with R recycling ~ 1 JT-60 Sec. Long pulse, steady-state operation JT-60NCT Min. ITER Min.~ Hour DEMO Day - Year Conditioned R recycling < 1 Unconditioned R recycling ~ 1 or >1 (Wall saturation)  wall 15s => 65s a

Introduction Particle control; For a constant density Fueled = Pumped aaaaaaaaaaaaaaaa Short pulse: Divertor-pumping + Wall-pumping Long pulse: Divertor-pumping + (Co-deposition?) Long pulse in JT-60 P heat ~12 MW, 30s Particle control at wall saturated First wall Gas-puff Pellet NB Core Plasma Baffle Divertor-Pumping Baffle Plates Divertor

Outline Density controllability w at wall saturation Identification of wall saturation Summary

No “carbon bloom” Energy Input : 350 MJ Div. surface temperature : ~ 1300 K

Identification of Wall Saturation

Wall Saturation identified in long pulse operation t = s, constant wall retention  Wall saturation  ELMy H-mode  Z eff ~3,H 89PL ~1.7 Then,  MARFE ( detach ) MARFE Saturation Quasi-steady-state, V p dn p /dt ~ 0, V n dnD 0 /dt ~ 0,  R wall = R P-NB + R N-NB  + R Gas - R pump

Saturation Area, Wider than Divertor plates Particle Balance between pulses Until 44018, Wall retention increases Wall saturation After 44019, 3x10 22 are released. 3x10 22 Saturation level of D + to C tile at 300eV = 1x10 21 m -2 Saturation area (Minimum) = 3x10 22 / 1x10 21 m -2 = 30 m 2 > Divertor plates ( 20 m 2 ) During 44020, 3x10 22 retained in walls.  Active wall-pumping capacity ~ 3x10 22 Particle Balance during 44020

T wall = 520 K Identical discharge condition; I p,B t,n e,P NB,H 89PL,T surf div,,, Release-Rate from walls Significant Particle Release at T wall = 520 K T wall = 420 K The only difference : first-wall-temperature  Suggests particle release from first wall / Baffle plates Gas-Puff-Rate

Source of particles, Baffle Plates T wall = 420 KT wall = 520 K Divertor similar First wall> Baffle< 1ch 8ch 11ch 20ch 19ch Viewing chord for D  Divertor Inner Baffle Outer Baffle Main Chamber First wall D  Brightness T wall = 520 K No gas-puff T wall = 420 K Gas-Puff

Density Controllability by Active Divertor-Pumping at Wall Saturation

Pumping GAP Density controllability of divertor-pumping GAP9.5 cm4.5 cm P0P0 +~ 0, - nene > Wall saturation Penning gage Suggests; Large GAP => Increase of P 0 => Increase of n e With Gas-puff No gas-puf

High pumping-rate suppresses increase of plasma particles Difficult to prevent undesirable density rise of high  plasmas ( Large GAP )  Limited period of high  N ; ex s for  N = 2.3 Higher pumping-rate is required even for low  plasmas ( Small GAP ) Condition : wall saturation Density : n e /n e GW = 62-66% Decreasing Gap

Controllability of detachment by divertor-puming at R = 1 SOLDOR simulation R Pump 6x10 21 /s R recycle =1 3x10 21 /s 10 MW  e =  i = 1 m 2 /s, D = 0.3 m 2 /s Indicates higher pumping speed by a factor of can avoid MARFE at the end of long pulse discharges R Gas 3x10 21 /s Pumping speed 25m 3 /s50m 3 /s P 0 OutDiv 2.0 Pa1.2 Pa n e OutDiv 4x10 20 m -3 2x10 20 m -3 T e OutDiv 0.8 eV9.7 eV

Summary Modification for Long Pulse Operation, 15 sec => 65 sec ELMy H-mode plasma ( ~30 s,~ 12 MW, 350 MJ, No “carbon bloom”) Wall saturation was identified (Minor role of co-deposition) Divertor plates Wall/baffle plates 1 No sudden changes of plasma ( Z eff, H 89PL ) Undesirable increase of plasma density => Confinement Degradation, MARFE Higher divertor-pumping efficiency ( x ) required to avoid MARFE NB 10 sec => 30 sec