Power and Particle Control in JT-60SA to Support and Supplement ITER and DEMO

Transcription

1 Power and Particle Control in JT-60SA to Support and Supplement ITER and DEMO Shinji Sakurai and JT-60SA design team Division of Advanced Plasma Research, Japan Atomic Energy Agency, Naka, Japan Abstract: JT-60 is planned to be modified as a fully superconducting coil tokamak (JT-60 Super Advanced, JT-60SA). Divertor targets are water-cooled to handle heat flux up to 15 MW/m 2. JT-60SA allows exploitation of high beta regime with stabilizing shell covered with ferritic plates and internal resistive wall mode (RWM) stabilizing coils. A remote handing system is equipped to maintain in-vessel components even for high dose rate due to a substantial annual neutron production. Divertor cassettes are introduced to be maintained by a remote handling. In the present design, a monoblock type carbon fibre composite (CFC) divertor target will be used to withstand a heat load of ~15 MW/m 2. CFC divertor targets and other bolted armor tiles will be mounted on the divertor cassette. All of the plasma facing components including the first wall armor are water-cooled to handle heat load during 100s or more. Divertor heat load and pumping efficiency for an ITER-like configuration has been evaluated, using 2D plasma fluid (SOLDOR) and neutral Monte-Carlo (NEUT2D) code. The pumping speed of 50 m 3 /s is specified at an albedo for neutrals in front of the in-vessel cryopanel. In the simulation for the divertor with a V-shaped corner like as that in ITER, the plasma detachment occurs near the outer-strike point within the V-shaped corner, as well as near the inner-strike point, which results in low peak heat flux density 5.8 MW/m 2 for the case with additional gas puff of 5x10 21 /s compared to 11.4 MW/m 2 for the case without V-shaped corner. Keywords: JT-60SA, Design, Divertor, Heat Flux, Particle Exhaust 1. INTRODUCTION JT-60 is planned to be modified as a fully superconducting coil tokamak (JT-60 Super Advanced, JT-60SA) with equivalent break-even class plasma confinement capability [1,2]. JT-60SA program is a combined program of Japan-EU satellite tokamak program under the Broader Approach (BA) Program and Japan Atomic Energy Agency (JAEA) s program for national use with equal operating opportunity between the two programs. Its mission is to contribute to the early realization of fusion energy by support and supplement of the International Thermal nuclear fusion Experiment Reactor (ITER) program, and by addressing key physics issues for ITER and DEMO. Figure 1 shows a bird s eye view of JT-60SA. Major Parameters of JT-60SA are shown in Table 1. Fig. 1 Bird s eye view of JT-60SA tokamak with heating and current drive system Table 1 JT-60SA Parameters Parameters Low A ITER like Plasma Current I p 5.5 MA 3.5 MA Toroidal Field B T 2.68 T 2.6 T Major Radius R p 3.06 m 3.15 m Minor Radius a p 1.15 m 1.02 m Aspect Ratio A Elongation κ Traiangularity δ Safety Factor q Flat Top 100 sec (8 hours in future option) H&CD Power 41 MW x 100 sec P-NB 24 MW (85 kev) N-NB 10 MW (500 kev) ECRF 7 MW (110,140 GHz) Divertor Heat Flux 15 MW/m 2 Annual Neutron 4 x JT-60SA allows exploration of configuration optimization for ITER and DEMO with a wide variety of plasma shape (elongation and triangularity) and aspect ratio (A=R p /a p down to 2.6) including that of ITER and double null configuration. ITER relevant plasma regimes far exceeding the H-mode power threshold can be studied with 41 MW heating, including dominant electron heating by 110&140 GHz Electron Cyclotron Radio Frequency (ECRF) and 500 kev Negative ion-sourced Neutral Beam Injection (N-NBI). High beta and fully non-inductive steady-state operation can be achieved with the stabilizing shell, internal Resistive Wall Mode (RWM) stabilizing coils, and 10 MW / 500 kev tangential N-NB current drive and 7 MW of EC current drive. Therefore, power and particle handling up to 41 MW x 100s with top and 305

2 bottom water-cooled divertors compatible with the maximum heat flux of 15MW/m 2 should be required [3]. A remote handing system is equipped to maintain in-vessel components [4] under high dose rate environment [5] due to a substantial annual neutron production of 4x This paper presents conceptual design of a divertor for JT-60SA to control power and particle exhaust. Latest engineering design of a divertor is briefly summarized. Heat and particle controllability in lower single null divertor for ITER-like plasma is estimated with 2D fluid (plasma) and Monte-Carlo (neutrals) codes [6-8]. 2. CONCEPTUAL DESIGN OF A DIVERTOR 2.1. Requirements of a divertor Required performances of a divertor for JT-60SA are as follows. (1) Flexibility in plasma configuration for high beta research and ITER-like plasma, (2) Handling of 41 MW plasma heating power during 100 seconds, (3) Capability for maintenance with remote handling system, (4) Baking at 473K and operation temperature as higher as possible, (5) Flexibility in plasma facing materials for plasma material interaction research toward DEMO For demonstration of day-long (8 hours) operation with non-inductive current drive in future option, all plasma facing components (PFCs) should be actively cooled against steady-state plasma heating power of 15 MW. Figure 2 shows tentative design plan of PFCs. Bottom divertor for ITER-like (medium triangularity) plasma and Top divertor for low A plasma top divertor for low A (high triangularity) plasma will be installed at bottom and top of vacuum vessel, respectively. Inboard surface of vacuum vessel is protected by an inner first wall. Stabilizing plate will be installed at outboard side to stabilize vertical positional instability and free boundary ideal Magneto-Hydro Dynamics (MHD) instability. Stabilizing plate is also protected by first wall armors and water-cooled brazed armors for tangentially injected N-NB. Fast position control coil and sector coils for RWM control will be installed behind a stabilizing plate. Cryopanels will be installed between divertor cassette and vacuum vessel as divertor pumping for particle control Divertor and target A divertor consists of inner and outer targets, where high heat flux is expected, private dome and inner and outer baffles for medium heat flux region and divertor cassette for remote handling. V-shaped corner is introduced to enhance particle recycling and reduce target heat flux similar to ITER divertor. To maximize flexibility in main plasma shape and aspect ratio, bottom and top divertor are optimized for medium triangularity (ITER-like) configuration and high triangularity configuration, respectively. Design example of bottom divertor is shown in Fig. 3. Inner and Outer Baffles Outer Target Private Dome Divertor cassette Stabilizing plate Inner Target V-shaped Corner Divertor Cassette Exhaust Hole 6m ITER-like Plasma Sector coils Cryopanel Cooling Water Pipes Fig. 3 Bottom divertor for ITER-like plasma First wall Cryopanel Bottom divertor for ITER-like plasma Fig. 2 Plasma facing components in JT-60SA Expected heat flux on outer divertor target exceeds 10MW/m 2 for high power heating. Outer divertor target should have maximum power handling capability within achieved performance in R&D for ITER divertor [9]. A monoblock type CFC target is most promising candidate to handle heat flux >10 MW/m 2. Monoblock type is also required to minimize possibility of exfoliation of CFC armor. A flat type brazed CFC target[10,11], which has advanced cooling channels such as screw tubes or hypervapertron, has also possibility to handle heat flux >10 MW/m 2 after the sufficient R&D. Basic structure of monoblock target is shown in Fig. 4. A target is fixed to a backplate with dove-tail sliding support for allowing 306

3 thermal expansion of target in axial direction. Coolant tubes to header should be long enough to allow this thermal expansion by its bending. An additional bolted armor is introduced to cover a coolant tube. It is also useful to avoid long overhang of baffle armor to cover between divertor target and baffle. Monoblock target Tube cover CuCrZr tube CFC brazed Cu alloy type CFC block type 30mm 30mm CFC monoblock Dove-tail sliding Thermal expansion Coolant header Backplate Outer target plate Fig. 4 Basic structure of an outer target plate and a monoblock target with a bolted armor tube cover 2-D thermal analyses were performed to determine target width and thickness and coolant tube type and diameter within a limited total coolant flow rate [3]. Maximum temperature of CuCrZr tube should be 673 K. In order to avoid water leakage due to burn out, wall heat flux inside of tube should be enough smaller than the critical heat flux at inside (CHF in ) even for unexpected short transient high heat flux onto target. Maximum coolant velocity is limited at m/s due to coolant inlet space into a Vacuum Vessel (VV). Incident critical heat flux (ICHF) at outside of M10 screw tube was measured experimentally.[12] Estimated CHF in might be ~60 MW/m 2 for coolant velocity of 10 m/s with assuming proportional to surface area {CHF in =ICHFx(OD/ID)}, where ID and OD are inner and outer diameter of a coolant tube. Thermal analyses of monoblock target show CHF in is 37 MW/m 2 with armor width of 30 mm and coolant velocity of m/s for 20 MW/m 2 at monoblock surface. Therefore, margin for burn out at 20 MW/m 2 is >1.6 and same as that for ITER outer divertor target. Expected heat flux is less than 15 MW/m 2 as shown in section 3. Therefore, maximum surface temperature of CFC armor with thickness of 5 mm is <1573 K, which can reduce erosion rate of CFC similar with those observed in existing tokamaks. Joining yield between a CFC monoblock to a CuCrZr tube is very important for mass production, because total yield of a target with 10 monoblocks is 0.9 even if yield for each monoblock reaches Therefore, we concentrate on R&D to improve brazing yield between CFC to CuCrZr. A bolted armor on a water-cooled heatsink, which can remove heat load of ~1MW/m 2 [3], will be adopted to a private dome, baffles and V-shaped corner to protect divertor cassette from radiation and neutral particle heat load Remote handling maintenance of divertor High power (41 MW) and long pulse (100 s) heating causes the large annual neutron fluence of neutrons/year. The vacuum vessel is made with low cobalt (0.05 wt%) stainless steel, Stainless Steel 316L, in order to reduce radio activation. However, because the expected dose rate at the VV may exceed 1 msv/h after 10 years operation and three month cooling[5], the human access inside the VV is restricted. This indicates a remote handling (RH) system is necessary for the maintenance and repair of in-vessel components[4]. Divertor cassettes are introduced to be maintained by a remote handling. Toroidal width of a divertor cassette is limited to 10 degree due to the size of a large horizontal port. Maximum handling weight (moment) of a manipulator is 500 kg in preliminary and conservative design, which strongly depends on the required function of a manipulator to trace required movement of a cassette. Therefore, maximum handling weight is expected to be increased by simplification of a cassette movement and manipulator function. After removing some bolted armors Rail Cassette Manipulator Cutting pipes Removing screws Carrying out from VV Lifting cassette Palette Handing to palette Fig. 5 Maintenance procedure for a divertor cassette Maintenance procedure of a divertor is shown in Fig. 5. Some of the bolted armor tiles are removed by the manipulator for bolted armors in order to grip the divertor cassette and to cut the cooling pipes. Cooling water pipes of a cassette are cut and welded by a special laser cutter/welder before and after exchange of the cassette, respectively. The manipulator for a divertor cassette lifts and carries a cassette to the palette. A palette carries out cassette through a large horizontal port to a cask for transportation. A removed cassette is maintained in a hot cell. A bolted armor and its water-cooled heatsink on a cassette can be replaced by hand. Replacement of mon- 307

4 oblock targets, which needs much time to cut and re-weld coolant tubes, is executed after removing a target plate from a cassette by cutting between coolant header and a cassette, in order to reduce exposure of worker from a radio-activated cassette. 3. DIVERTOR PERFORMANCE 3.1. Heat flux reduction in the ITER-like divertor Lower single-null divertor is designed for ITER-like plasma configuration as shown in Fig. 2. Divertor geometry is designed to study physics concept of the ITER divertor. Control of the plasma detachment inside the divertor region is the most important research for JT-60SA as well as ITER, which can reduce large power during long pulse discharges. Vertical target is used for both inner and outer divertors to increase the recycling and radiation loss efficiently along the divertor leg. A private dome is installed similar to ITER. Neutrals are pumped from the private region, and the inner and outer exhaust slots are connected under the dome. The outer exhaust slot at the dome side is tentatively designed between 8 and 18 cm above the strike point, and the plasma detachment near the strike-point, i.e. partially detached divertor can be produced in the V-shaped corner geometry more effectively compared to a conventional geometry, where exhaust slot is located at the bottom ( L-shaped corner ). Width of the exhaust slot will be optimized to control the divertor detachment and to obtain sufficient helium exhaust. Divertor heat load and pumping efficiency for an ITER-like configuration has been evaluated, using 2D plasma fluid (SOLDOR) and neutral Monte-Carlo (NEUT2D) code[6-8]. Radiation power from carbon impurity is calculated by a simple non-corona model, assuming residence parameter of n e τ res =4x10 15 s/m 3, where τ res is the impurity residence time in the plasma, and a uniform carbon fraction of 1 % of deuterium density in SOL and divertor plasma. Calculation mesh is shown in Fig. 6. Core plasma boundary ( edge ) is set on r/a =0.95, where power flux of 37 MW (assuming 4 MW loss in core) and ion flux of 5x10 21 D/s are exhaust. The pumping speed of 50 m 3 /s is specified at an albedo with including transparency of chevron and exhaust holes in divertor cassette in front of the cryopanel. Recycling coefficient of deuterium is set to unity at all first wall surfaces. Electron and ion thermal diffusivities are assumed to be 1 m 2 /s at whole edge plasma. Density profile of H-mode edge plasma is modeled with combination of particle diffusion coefficient and pinch velocity. Those in inside-separatrix and scrape off layer (SOL) are assumed to be 0.15 m 2 /s, -5 m/s and 0.3 m 2 /s, 0 m/s, respectively. Electron density and temperature distribution in the outer divertor region with additional gas puffing of 5x10 21 D/s are shown in Fig.7. Exhaust Backflow n e 2.5 V-shaped corner R (m) 3.0 T e V-shaped corner 2.5 R (m) 3.0 Fig. 7 Electron density and temperature distribution in the outer divertor region Fig. 6 Calculation mesh for SOLDOR and NEUT2D Electron density along the separatrix increases especially in V-shaped corner due to particle backflow from private region and recycling enhancement by V-shaped corner. Partially detachment reduces electron temperature along the separatrix in the V-shaped corner. Heat flux profiles at the outer divertor target are evaluated, which includes heat fluxes due to radiation and neutral particle flux. Heat flux profiles with and without additional gas puffing in above configuration are shown in Fig. 8 with thick lines. Outer divertor plasma is detached with medium gas puffing and peak heat flux is reduced to 5.8 MW/m 2. The peak heat flux remains 10.9 MW/m 2 even in attached case without gas puffing. Normal divertor geometry of existing tokamaks is so called L-shaped corner here. Thin lines in Fig. 8 show heat flux profiles 308

5 in L-shaped corner. Outer divertor plasma remains attached and the peak heat flux reaches 11.4 MW/m 2 even when the additional gas is puffed. The peak heat flux reaches 16.3 MW/m 2 and exceeds allowable level for divertor target without additional gas puffing. The peak heat flux is close to 15 MW/m 2 without gas puffing when the X-point is shifted upward by 11 cm and the outer strike point is elevated to the exhaust slot. The additional gas puff of 5x10 21 D/s makes outer divertor plasma partially detached and reduces peak heat flux to 7.5 MW/m 2. Therefore, V-shaped corner can reduce heat flux to the target effectively. Heat flux at outer target (MW/m 2 ) allowable level V-shaped corner with gas puff V-shaped corner without gas puff L-shaped corner with gas puff L-shaped corner without gas puff Distance from strike point (m) Fig. 8 Comparison of heat flux profile Peak heat flux to V-shaped corner and private dome due to the radiation loss in the outer divertor region increases from 0.6 MW/m 2 (attached case) to 0.9 MW/m 2 (detached case). Therefore, bolted armor tiles on a water-cooled heatsink can be adopted those regions, if hitting of strike point longer than several seconds due to operation failure can be avoided Control of particle exhaust and detachment Particle control under the partially detached divertor plasma is crucial for handling the large power flow into SOL during long pulse discharge. Divertor plasma configuration and geometry of exhaust slots are important to control the plasma detachment through the pumping efficiency. Therefore, influence of the divertor plasma configuration on the detachment and exhaust efficiency is investigated. Table 2 shows neutral particle flux (D 0 and D 2 ) balance in the divertor region in steady state. Pumping flux of 10x10 21 D/s is balanced to the total flux of ion flux from core and additional gas puffing. In the standard configuration, large neutral particle flux is exhausted from the inner divertor due to its short distance to exhaust slot, while large part of the flux is supplied to outer divertor though under the private dome. Such circulation from the inner divertor to the outer divertor can make the outer divertor detachment efficiently as well as the vertical target with V-shaped corner, which is seen in the simulation of ITER divertor. Exhaust efficiency decreases at the inner divertor and increases at the outer divertor when X-point is shifted upward by 11 cm. Therefore, exhausted flux from the inner divertor and backflow flux to the outer divertor is strongly reduced. Net exhaust from the outer divertor is observed. Consequently, the detached outer divertor plasma approaches attachment. The outer divertor is completely changed from detachment to attachment when the additional gas puffing is 2.5x10 21 D/s. Table 2 Comparison of neutral particle flux balance Flux from target (10 21 D/s) Net exhaustion (10 21 D/s) Pumped by cryopanel (10 21 D/s) Standard configuration Inner Outer divertor divertor X-point is shifted upward by 11 cm Inner Outer divertor divertor SUMMARY JT-60SA allows exploration of configuration optimization for ITER and DEMO with a wide variety of plasma shape and aspect ratio including that of ITER and double null configuration with high power long pulse heating. Power and particle handling up to 41 MW x 100s with top and bottom water-cooled divertors compatible with the maximum heat flux of 15MW/m 2 should be required. A remote handing system is equipped to maintain in-vessel components under high dose rate environment due to a substantial annual neutron production. A divertor consists of inner and outer targets, where high heat flux is expected, private dome and inner and outer baffles for medium heat flux region and divertor cassette for remote handling. V-shaped corner is introduced to enhance particle recycling and reduce target heat flux similar to ITER divertor. To maximize flexibility in main plasma shape and aspect ratio, bottom and top divertor are optimized for medium triangularity (ITER-like) configuration and high triangularity configuration, respectively. A monoblock type CFC target is most promising candidate to handle heat flux >10 MW/m 2. Armor width of 30 mm and coolant velocity m/s with M10 screw tube can obtain margin of 1.6 for burn out at unexpected short transient heat flux of 20 MW/m 2 under the limitation of total coolant flow rate. Maximum surface temperature of CFC armor with thickness of 5 mm is < 1573K. A divertor cassette is introduced to be maintained by a remote handling system. Conceptual design of the procedure and a manipulator for divertor cassette maintenance has been conducted. Lower single-null divertor is designed for ITER-like plasma configuration to study physics concept of the ITER divertor. Control of the plasma detachment inside the divertor region is the most important research for JT-60SA as well as ITER, which can reduce large power during long pulse discharges. Vertical target is used for both inner and outer divertors to increase the recycling and radiation loss efficiently along the divertor leg. A private dome is installed similar to ITER. Neutrals are pumped from the private region, and the inner and outer 309

6 exhaust slots are connected under the dome. Physic assessment on heat flux reduction and particle control in the ITER-like divertor has been performed with using 2D plasma fluid (SOLDOR) and neutral Monte-Carlo (NEUT2D) code. In the simulation for the divertor with a V-shaped corner, the plasma detachment occurs near the outer-strike point within the V-shaped corner, as well as near the inner-strike point, which results in low peak heat flux density 5.8 MW/m 2 for the case with additional gas puff of 5x10 21 /s compared to 11.4 MW/m 2 for the case without V-shaped corner. Exhaust efficiency can be changed by shifting X-point height. Results of simulation show outer divertor plasma changes between attached and detached by shifting X-point height. REFERENCES 1. M. Kikuchi, JA-EU satellite tokamak working group and JT-60SA design team, in Proc. of 21 st IAEA Fusion Energy Conference (Chengdu, China, 2006), IAEA-CN-149/FT/ M. Matsukawa, JA-EU satellite tokamak working group and JT-60SA design team, in Proc. of 21 st IAEA Fusion Energy Conference (Chengdu, China, 2006), FT/P S. Sakurai, K. Masaki, Y. K. Shibama, H. Tamai and M. Matsukawa, Fusion Eng. Des, in press. 4. T. Hayashi, S. Sakurai, K. Masaki, H. Tamai, K. Yoshida and M. Matsukawa, in Proc. of 15th Int. Conf. on Nuclear Engineering (Nagoya, Japan, 2007) 5. A. M. Sukegawa, S. Sakurai, K. Masaki, K. Kizu, K. Tsuchiya, Y. K. Shibama, et al., Fusion Eng. Des, in press. 6. K. Shimizu, T. Takizuka, S. Sakurai, H. Tamai, H. Takenaga, H. Kubo, et al., J. Nucl. Mater, (2003) H. Kawashima, K. Shimizu, T. Takizuka, S. Sakurai, T. Nakano, N. Asakura, et al., Plasma Fusion Res, 1 (2006) N. Asakura, H. Kawashima, K. Shimizu, S. Sakurai, T. Fujita, H. Takenaga, et al., in Proc. of 34th EPS conf. on Plasma Phys. and Cont. Fusion (Warsaw, Poland, 2007), P K. Ezato, M. Dairaku, M. Taniguchi, K. Sato, S. Suzuki, M. Akiba, et al., Fusion Sci. Technol, 46 (2004) P. Garin, Fusion Eng. Des, (2001) K. Masaki, M. Taniguchi, Y. Miyo, S. Sakurai, K. Sato, K. Ezato, et al., Fusion Eng. Des, (2002) J. Boscary, M. Araki, S. Suzuki, K. Ezato and M. Akiba, Fusion Technology, 35 (1999)