Divertor Requirements and Performance in ITER M. Sugihara ITER International Team 1 th International Toki Conference Dec. 11-14, 001 Contents Overview of requirement and prediction for divertor performance taking ITER case as an example (1) Required divertor performance () Predicted divertor performance (3) Further model development needed and remaining uncertainty (4) effects and mitigation (5) Summary With contributions from : A. Kukushkin, A. Loarte and ITER IT Physics Unit _
1. Required divertor performance (i) Heat removal (ii) Fuel density control (iii) Exhaust of helium ash and other impurities (iv) Providing proper magnetic configuration for enhanced confinement (H-mode) Requirements 1. Peak power load on the target plates ( q pk ). Helium concentration in the core plasma ( C He ) qpk 10 C He 006. MW/ m 3. Z eff in the core plasma Z eff 16. 4. Upstream plasma density ( n s ) ns ne/3 5. D-T particle throughput Γ DT 00 Pa m 3 / s ( Γ DT )* 6. Core fuelling ( Γ core core DT )* 0 Γ DT 100 Pa m s 6 requirements must be simultaneously satisfied * are also control actuators Specific features for divertor control (i) Control actuators; not so many - Divertor geometry - Gas-puffing (Throughput ; Γ DT ) * - Core fuelling ( Γ core DT )* - Pumping speed - Impurity seeding (Ne, Ar) 3 1
(ii) Divertor and Core Performance are closely linked ; SOL/Divertor Pedestal Core Core performance Core o Pedestal separatrix SOL Core <T>, <n> (Stiffness) Pedestal T n Separatrix n sep Transport o Divertor performance - Development of Modelling to include Pedestal continues, but not yet complete. - Presently CEI (5cm inside separatrix) is calculation boundary and transport barrier is not yet properly modelled.
. Predicted divertor performance Prediction by B/Eirene divertor code Basic models - D = 03. m / s, χ=1 m / s w/o parameter and spatial dependence - effect is not included (time averaged) - Carbon sputtering (physical + chemical), but they are absorbed at every surface encounterd - Partial detachment (only near separatrix is detached) Optimization of divertor geometry Key strategy to reduce peak power load: Enhance neutral accumulation, in particular, near the separatrix region for outer target plate Vertical Target Plate + Divertor Dome V-shape Target geometry ; effective in accumulating neutrals near separatrix (JET) 30% reduction of q pk in ITER V-shape 0 straight q pk [MW/m ] 16 1 8 4 0.4 0.6 0.8 0.3 0.3 0.34 0.36 n s [10 0 m -3 ] 86MW straight 86MW old V 86MW new V 100MW straight 100MW old V 100MW new V (Kukushkin, EPS 001)
Gas flow between inner and outer divertor; increase neutral recirculation in the outer divertor target region (higher power flux) reduce peak heat load (JET, JT-60U) Γ in Γ out Γ pump - 0% reduction of q pk by gas flow between inner and outer divertor Separatrix density Dominant effect on divertor performance and can be controlled by gas puffing (throughput; Γ DT ) to some extent gas Γ = Γ + Γ DT DT core DT Saturation corresponds to detach. of inner divertor (increased neutral density) Higher n s for higher power to detach
Inductive operation Peak power load and helium concentration X Reference operation P SOL =86MW(P f =410MW, P total =13MW, Q=10, f rad =0.3) High fusion power with high Q P SOL =100MW (P f =600MW, P total =145MW, Q=4, f rad =0.4) High fusion power with low Q P SOL =130MW (P f =600MW, P total =187MW, Q=9, f rad =0.3) Peak power load and helium concentration for reference operation mode is well within the requirement. Fusion power (helium source) and throughput dominate helium concentration, while pumping speed is less important Reasonably wide operation window is available for the reference inductive operation mode, while density window is not so wide ( n s between q max and complete detachment)
Steady state operation Steady state P SOL =100MW P f =340MW, Q=5.7, P total =18MW, f rad =0. Longer connection length with q95=4.5 => same q pk with lower n s cf. Inductive operation P SOL =100MW q95=3.0 (Kukushkin) n s (at q pk =10MW/m )=0.6 => somewhat higher than ns ne => Impurity seeding will be needed / 3 0. 3 Initial calculations with neon seeding (0.4%) ; 30% reduction of q pk (radiation region is getting far from target plate compared with carbon) q pk =10MW/m at n s 03. 04. Z eff 0.4 (total Z eff 1.6)
3. Further model development needed and remaining uncertainty (1) Transport in SOL region () Separatrix density under good H-mode confinement (3) Consistent estal model is not yet develo; - D = 03. m / s, χ=1 m / s are too large in the estal (transport barrier) region => low estal density ( n ( 35. 45. ) 10 19 m 3 ) => e.g., neoclassical level D 006. m / s and proper width model for estal must be implemented => Consistent boundary condition for core plasma transport (to be develo) => By proper estal model, core fuelling requirements can be properly specified, which is consistent with the expected density estal in ITER Core fuelling is needed because; Gas-puffing is very inefficient due to thick SOL in ITER Only small fraction of gas-puffed neutrals can penetrate across separatrix
Specification of required core fuelling for expected density estal in ITER Particle balance across separatrix and estal C Core fuelling Γ core Fuelling inside estal ; Core fueling Pedestal fueling n(r) Γ C core estal n Γ P core separatrix SOL Γ sep i ~Γ sep 0 +Γ core P Pedestal fuelling Γ core Fuelling between separatrix-estal ; r Γ 0 sep ~10cm a n sep Γ 0 puff With proper transport model in the estal and core or estal fuelling can achieve the expected estal density Fuelling in the estal region is also possible but factor of two larger fuelling is needed Pedestal Density n (10 1 9 m -3 ) 10 8 6 4 Expected Pedestal density in ITER D=0.06 m /s =10cm Γ C core =8Pam3 /s (56) Γ C core =14Pam3 /s (8) Γ C core =14Pam3 /s (8) D=0.3 m /s ( ):Pedestal fuelling 3 4 5 Separatrix density n sep (10 1 9 m - 3 ) High field side pellet is prepared for ITER required core fuelling is possible 50-100 Pa m 3 / s 500 m/s (deposition depth 015. a; inside estal)
4. effects and mitigation High estal pressure required for good confinement can result in large divertor erosion due to Type-I s Limit for divertor erosion due to s W /( S τ ) ; surface temperature rise (Federici; SOFE, 00) Specification for τ 00 µ s τ in JET for various density and triangularity τ 00µ s (JET, Becoulet)
Specification for S S ss λ q ; Power deposition width map on midplane has large uncertainty Experimental data for S ss are mostly taken from attached condition => λ q 5mm => S 6 m From power load profile in ITER ; λ q =(10-13) mm due to detachment => S 15 m Power to the target - total, with plasma, and radiation - map to the mid-plane [MW/m ] 10 7 10 6 total plasma radiation λ q =5.6 mm 0 0.01 0.0 x MP [m] ITER-FE P in = 86 n s = 3 10 S p = 40 A. S. Kuk Jun 000 Criteria of W / W for surface temperature rise up to critical one τ = 00 µ s S 6 15 m CFC W Allowable W / W (%) for 10 6 events with deposition area S SSS 6 15 m W 100 MJ 3.4-8.3 4.4-11 * This is also necessary to maintain plasma purity ( 10 carbon/ event is produced)
Proposed models for experimental data summary Collisionality ( ν * ) (Loarte, IAEA 000) 0.5 0. DIII-D JET ASDEX-U W * ν * P ( ) ( ν ) W P -0.33 (15-0) % for ITER W / W 0.15 0.1 0.05 ν* ITER 0 0.01 0.1 1 10 ν* (collisionality) Parallel transport ( τ // ) (Janeschitz, PSI 000) 0.5 0. DIII-D JET ASDEX-U W tau // W ( ) = ( ) W W τ // 1 0 1 + τ // τ * = ( 1+ 3/ ν ) L Cs τ 00 µ s (1-15) % for ITER W / W 0.15 0.1 0.05 τ // ITER 0 0 00 400 600 800 1000 100 τ // (µs) Sheath model (Shimada, 001) W = γkγt ( B / B ) 4πR t p T om om om Upper limit ( 5) % for ITER W / W 0.3 0. 0.1 DIII-D_exp DIII-D_model JET_exp JET_model AUG_exp AUG_model 0 0 500 1000 1500 T (ev) om=a/100 t=0.ms All models still need much more work for ITER extrapolation
ITER Prediction W / W 0.3 0. 0.1 Allowable W / W Model prediction W (inclined) (inclined) (Further inclined) ν* model τ / / model CFC 0 100 00 300 400 500 Fusion power (MW) sheath model Very severe predictions based on ν * and τ // models. Range of uncertainty and difference between models are significantly large.
Possible mitigation methods (1) Further inclination of divertor plate Poloidal projected angle.8 (.1 real) 11.4 Possible disadvantage can be acceptable - Particle recycling on the upper part of divertor plate ; not significantly increased (B/Eirene) - Separatrix line position control; may be acceptable once operation mode is fixed for engineering testing (life time becomes a more important issue for this phase). use of W divertor plate may also be possible during this phase due to low disruption probability.
() Discharge regime of high estal pressure with small s (Type II) Most of the present machines show that high q 95 ( 3.5-4) high δ X ( 0.4-0.5) are needed to obtain this alternative regime. δ X for ITER (=0.5) satisfies the required condition. Q=10 and P fusion 50MW operation with q 95 =3.5 (Ip=13MA) is possible, though window is narrow. Further increase of HH-factor - with lower density (many machines) - with higher q 95 (HH=1.3 with q 95 =3.6, n/ n G 1 and very small s in AUG) window becomes much wider. This small regime will be accessible for Hybrid and steady-state scenario ( q 95 >3.5). Further R&D is needed to extend this small regime to the reference high Q inductive operation mode. - Type II s in-between Type I ( q 95 =3, δ X =0.5; JET) could be a clue for R&D
5. Summary Divertor requirements for ITER are summarised. B/Eirene code calculations show that these requirements will be satisfied for inductive operation mode. For non-inductive operation mode, impurity seeding will reduce the peak heat load to meet the requirement. Further model development is necessary for B/Eirene to include proper estal model. It is indicated that gas-puffing cannot fuel across the separatrix to form proper density estal. High field side pellet fuelling is prepared in ITER to fuel inside the estal. Effect of Type-I s on divertor plate could be severe for high estal pressure required for good confinement, while present prediction by proposed models are still primitive, and thus further development/improvement of the models as well as the database are necessary. Possible mitigation methods for Type-I effect are summarised; inclination of target plate and Type-II s. Hybrid and steady-state scenarios can be operated with Type-II regime. Further exploration to extend this regime to high Q inductive operation mode should be promoted.