Microwave Spherical Torus Experiment and Prospect for Compact Fusion Reactor

Microwave Spherical Torus Experiment and Prospect for Compact Fusion Reactor Takashi Maekawa 1,*, Hitoshi Tanaka 1, Masaki Uchida 1, Tomokazu Yoshinaga 1, Satoshi Nishio 2 and Masayasu Sato 2 1 Graduate School of Energy Science, Kyoto University, Kyoto 606-8501, Japan 2 Naka Fusion Institute, Japan Atomic Energy Agency, Ibaraki, 311-0193, Japan Abstract: Fusion power may be a major energy resource free from CO2 emission in future. The International Thermonuclear Experimental Reactor (ITER) project has been approved and the construction phase has just started for experiments after 2016 to prove DT fusion burning in tokamak magnetic confinement system. There is, however, a concern in fusion reactors of conventional tokamak type. They are large and their construction cost may not meet the economical requirement. In these circumstances, Spherical Tokamak (ST) concept has been developed to overcome the concern on the reactor size. Furthermore, a reactor design study on Very Compact Tokamak Reactor (VECTOR) has shown that removal of the central solenoid (CS) from tokamaks makes the reactors drastically compact and simple in structure, thus the economical requirement may be fulfilled. Without CS, another method is required to start plasma discharge and drive the plasma current toward full plasma current for confinement of burning plasma. Recent experiments in the Low Aspect ratio Torus Experimet (LATE) device in Kyoto University has demonstrated that ST plasma can be formed by injection of microwave power alone without CS. The plasma has been initiated and started-up by electron cyclotron heating (ECH) by injected microwave. A simulation study is also conducted to explore the microwave method for ramp-up of the plasma current toward fusion burning. Keywords: Fusion Energy, Fusion Reactors, Tokamak, Spherical Tokamak, Microwave, Electron Cyclotron Heating 1. INTRODUCTION In future fusion energy may be one of major energy sources. The fuel is a mixture of deuterium and tritium. Deuterium is plenty in water. Tritium is radioactive with a short half-life and therefore does not exist in nature. However, it can be bred from lithium by the irradiation of fusion generated neutron in the lithium blanket surrounding the fusion plasma. Lithium itself is also sufficiently contained in seawater for electricity generation over one million years. Therefore, fusion energy can be considered practically inexhaustible and sustainable. The inherent safety of a reactor (no chain reaction), no emission of gas causing greenhouse effect and the possibility to clear or to recycle the activated components of the reactor after about 100 years are other positive key features of fusion, which would favor the social acceptance of this form of energy. Thus, fusion energy is potentially very attractive. Worldwide efforts over half a century toward realization of magnetic fusion reactor were put together by international collaboration as the International Thermonuclear Experimental Reactor (ITER) project. Recently this project was officially approved and ten-years-long construction phase has just started. ITER is due to achieve and study burning fusion plasmas under conditions very close to those expected in a fusion power plant. The objectives of ITER are to demonstrate the scientific and technological feasibility of fusion energy and the safety and environmental acceptability of fusion. The ITER project is very important since accomplishing its objectives is essential to step farther toward fusion power plant. A lot of issues including both scientific and technical ones must be resolved in the course of ITER project to attain its objectives. While the ITER project is, as mentioned above, a key step toward realization of fusion power plant, there are further broad issues to be investigated and resolved to realize truly useful fusion power plant that meets simultaneously economical, environmental and safety criterions. Especially, there is a concern in size of fusion reactors by conventional tokamak type. They are large and their construction cost may not meet the economical requirement. In these circumstances, Spherical Tokamak (ST) concept has been developed to overcome the concern on the reactor size [1]. Furthermore, a reactor design study on Very Compact Tokamak Reactor (VECTOR) has shown that removal of the central solenoid (CS) from tokamaks makes the reactors drastically compact and simple in structure, thus the economical requirement may be fulfilled [2, 3]. In addition to the final reactors based on the CS-free ST concept, Component Test Facility (CTF) also based on the CS-free ST concept for testing various reactor materials under strong fusion neutron irradiation is now seriously considered and designed [4]. Construction cost of such a CTF is quite low compared ITER, which would allow multi facilities for various testing and accelerate fusion energy realization. Without CS, another method is required to start plasma discharge and drive the plasma current toward full plasma current for confinement of burning plasma. Recent experiments in the Low Aspect ratio Torus Experiment (LATE) device in Kyoto University has demonstrated that ST plasma can be formed by electron cyclotron heating (ECH) and current drive (ECCD) by injection of microwave power without CS [5, 6]. The plasma has been initiated and started-up by electron cyclotron heating by injected microwave. A simulation study is also being conducted to explore the microwave method for ramp-up of the plasma current toward fusion burning. 2. MICROWAVE SPHERICAL TORUS EXPERIMENT 2.1 Experimental apparatus LATE is a small device as shown figure 1. The vacuum chamber is a cylinder made of stainless steel with an inner diameter of 1.0 m and a height of 1.0 m each. The center post is also a cylinder made of stainless steel with an outer diameter of 11.4 cm, enclosing 60 turns of conductors of a coil for toroidal magnetic field. Its return conductors are grouped into 6 limbs and located far from the vacuum vessel as shown in the figure, which allows good accessibility to the vacuum chamber and suppresses toroidal field ripple at a low level (the level is 1.5% at the outer wall of the vessel (R=50 cm) and 0.07% at R=30 cm). There are four sets of poloidal field coils. The second set from the inboard side is for feedback control of the vertical position of the plasma loop and the other three sets Corresponding author: maekawa@energy.kyoto-u.ac.jp 1

are for the external vertical fields for the radial equilibrium of the plasma loop and their currents are preprogrammed. Note that there is no central solenoid (CS) for inductive current drive. Three 2.45 GHz magnetrons including two 5 kw CW magnetrons and one 20 kw 2 s magnetron and a 5 GHz klystron (160 kw, 100ms) are used for electron cyclotron heating and current drive (ECH/ECCD) of plasma. In all cases, microwaves are injected from the radial ports, usually with the linearly polarized electric fields on the equatorial plane and with injection angles slightly deviate (~15 degree) from the normal direction to the toroidal field. The diagnostics of LATE comprises two channels of 70 GHz interferometer including the vertical and the oblique ones denoted as chords A and B in the figure 1(b), respectively for the plasma density measurement, a spectrometer, a video camera in visible range, single channel NaI scintillator for hard X-ray detection, Langmuir probes and four soft X-ray cameras for the computerized tomography imaging of the soft x-ray emission profile. Magnetic data from seventeen flux loops are used for the analysis of the plasma current profile and the poloidal flux surfaces. Fig. 1 The Low Apect ratio Torus Experiment (LATE) device ((a) side view and (b) top view) 2.2 Spontaneous formation of microwave spherical torus It is interesting that a closed flux surface can be spontaneously produced by ECH under a steady Bv field [5, 6]. A characteristic in the present LATE case is the appearance of a clear current jump, where plasma current increases rapidly in the time scale of a few milliseconds even at a low decay index of n < 0.1. After the jump a closed flux surface is formed. Figure 2 shows a case by a 5 GHz microwave pulse of 130 kw under Bv=85 Gauss. Time evolution of the plasma images on the video camera shows that the breakdown takes place at the fundamental ECR layer at R=10 cm and the plasma expands quickly to the lower field side. In accordance with the plasma expansion, a plasma-current starts to flow and grows slowly up to 2 ka, and then suddenly it rises rapidly and reaches 7 ka, after which the current is maintained to the end of the microwave pulse as shown in figure 2(a). Time evolution of the plasma current distribution is analyzed by using the magnetic data from thirteen flux loops and displayed in figures 2.(d)-(i). The distribution before the first jump is stretched vertically near the second harmonic resonance layer ((d) and (e)). After the first jump, it expands to the stronger field side and a small closed flux surface touched to the centre post appears ((f)). After the second jump, the current distribution as well as the closed flux surface expands to the lower field side ((g)). At the final stage the current distribution is detached from the centre post and a broad current profile expanded to the outboard wall of the vessel is formed ((h)). The location of the plasma current centre (+) in the final stage is between the second and third harmonic resonance layers. The line averaged electron density exceeds significantly the plasma cutoff density. These results suggest that electron Bernstein waves (EBW) supports the plasma. It is remarkable that the open field configuration of external fields spontaneously changes into a closed field configuration by ECH alone ((p) and (q)). Plasma images reflect these changes of field topology ((j)-(o)). Formation of closed field equilibria via rapid current rise has been observed for a wide range of Bv in both 2.45 and 5 GHz experiments as shown in figure 3. All the characteristic currents including the current just before current rise, the current at which initial closed surface appears and the final current after rapid current rise are roughly proportional to Bv, suggesting that the current begins to rise when the deformation of poloidal field structure due to the self field from plasma current reaches a certain level. 2

Fig. 2 Typical discharge of spontaneous formation under steady B v field Fig. 3 Various characteristic currents versus external vertical field Bv 2.3 Equilibrium, field topology and current generation during rapid current rise For analytical simplicity we consider equilibria of plasma loops by using the Shafranov formula; µ 0I p 8R l = + i 3 Bv ln + β p 4π R a 2 2 where µ p p µ I β 2 0 0, and B p p = a = 2 2 B I 2π a a p The formula can be cast into the dimensionless form as shown in figure 4. Here, the first term is proportional to the plasma current and responsible for the current-hoop-force and the second term is inversely proportional to the plasma current and responsible for the pressure-hoop-force. Both terms become the same at the pressure-current (PC) turning point in this figure and the lower part from this point is the pressure-hoop-force dominant regime and the upper part is the current-hoop-force dominant regime. The ratio, B a /B v, 3

is a measure of deformation of poloidal field from the vacuum external field. When I p increases and B a becomes comparable to B v, closed flux surfaces appear. This point is near the PC turning point when aspect ratio is low as shown in figure 4. This result and the results in figure 3 suggest that the plasmas before rapid current rise are in the pressure-hoop-force dominating regime while those after rapid current rise are in the current-hoop-force dominating regime. The rapid current rise bridges both regimes through the PC turning point. Fig. 4 Equilibrium characteristics for various aspect ratios 2.4 Plasma current ramp-up with a ramp of Bv for equilibrium at larger plasma current Once a large closed flux surface is formed via the rapid current rise under a steady B v field, the plasma current is further ramped up by increasing the microwave power with a slow ramp of B v for the equilibrium of the plasma loop at larger currents. In the case of a 5GHz experiment, the plasma current has ramped-up and reached 12 ka at B v =100 Gauss by 130 kw of injection power after spontaneous formation of I p =6 ka under a steady field of B v =60 Gauss as shown in figure 6. The current profile is vertically elongated along the second and third EC resonance layers. The plasma image at the final stage of discharge reflects strongly elongated shape of last closed flux surface with an elongation of κ=2.2 and an aspect ratio of R/a=20.0cm/14.3cm=1.4. Fig. 5 Plasma current ramp-up with a ramp of Bv field 4

Fig. 6 Microwave spherical torus produced in the LATE device (5 GHz experiment). In the case by the 2.45 GHz microwaves, the final current reaches 8.1 ka at B v =80 Gauss by 31 kw of total injection power from the three magnetrons. Present long pulse discharge (2 s) has conveniently allowed pulse height analysis for photon energy spectra in X-ray range and the X-rays up to the photon energy range of ~100keV have been observed, showing that high energy tail of fast electrons appears and develops as I p increases. The maximum plasma currents so far amount to 13.5 % of the total toroidal coil currents flowing through the centre post for both cases of 2.45 GHz and 5 GHz experiments. The field line pitches are still moderate as shown in figure 6, which are comparable to those in tokamaks with conventional aspect ratios. In order to approach full ST equilibrium we need more microwave power with sufficient pulse length. The maximum current ramp up rate is so far 1kA/10ms for both cases of 2.45 GHz and 5 GHz experiments. However, this seems to be limited at the moment by the present ability of power supply for B v ramp up. Therefore, we need also more powerful power supply for B v field. 3. PROSPECT FOR COMPACT FUSION REACTOR 3.1 Benefit from removal of central solenoid As to a tokamak concept, which is most promising among all of the magnetic fusion concepts, the dependence of plasma performance and engineering capability on the cost of electricity (COE) are reported [7, 8]. The COE of tokamak power plants improves drastically as the plasma beta (β) increases. Here, β is the ratio of plasma pressure to the pressure of toroidal magnetic field that confines the fusion plasma. In the cases of the conventional power reactor designs such as SSTR [9] and A-SSTR2 [10], for example, the construction cost of toroidal field (TF) coils was estimated to be more than half of the whole tokamak construction cost. This means that the required fusion power should be achieved by a TF coil system as low cost as possible. The β value expected in the conventional tokamaks such as ITER should be improved more than twice to have tokamak power plant ready for competition in future energy market. As a general approach, higher β plasmas have been pursued by higher grade control on plasma current profile, plasma density profile and plasma cross-sectional shape. But reactor plasmas are in almost the self-ignited state with high bootstrap current fraction, a self-organized state. Therefore it is risky to depend on the high grade control for such reactor plasmas. On the other hand, M. Peng pointed out that high β plasmas were easily obtainable by lowering the aspect ratio (plasma major radius/minor radius) [1, 11]. In this approach of low aspect ratio torus, all non-essential components such as inboard blanket and inboard poloidal coil (PF) systems like a center solenoid (CS) coil system should be discarded from the inner side of tokamaks. In other words, not only attainable plasma β can be optimized but also the central structure of tokamak reactor can be greatly simplified in the case of CS free tokamaks. However, non-inductive current generation and sustaining methods such as RF wave or NBI must be realized in this case. 3. 2 Key conditions for reactor design and the VECTOR The plasma elongation (κ) and normalized beta (β N ) must be formulated by the function of the aspect ratio (A) for systematic parameter survey. The formulations are proposed by Lin-Liu and Stambaugh [12]. In the CS free tokamak reactor, the inner legs of TF coils can be bundled so as to form a center post structure. From engineering point of view, the maximum achievable magnetic field (B MAX ) was formulated by the function of the center post radius [2]. In addition to β and B MAX, a scaling law for plasma energy confinement time, plasma density limit, neutron shield requirement, design stress of structure material and etc. are also considered as the design condition.based on above mentioned design conditions, the design optimization was carried out with an objective function of a weight power density. The optimum design concept is VECTOR [3] as shown in figure 7. The VECTOR s lightweight and high power density is shown in figure 8. 3. 2 Prospect of microwave spherical torus for application to fusion reactor In order that microwave spherical torus is truly useful in fusion reactor operations with no CS, required microwave power should be modest compared with the fusion output power. This is the primal concern since the microwave method has essentially no other concern. Once fusion burning starts, only small power would be required. Most serious phase is just before the ignition, where largest power would be required to ramp-up the plasma current. Whether is this maximum power allowable or not is still not clear. 5

Fig. 7 The Very Compact TOkamak Reactor (VECTOR) Fig. 8 VECTOR improves significantly Cost of Electricity 5. ACKNOWLEDGMENTS The authors gratefully acknowledge continuous encouragements from Dr. Martin Peng. 6. REFERENCES [1] Peng, Y-K. M. and Strickler, D.J.,(1986) Nuclear Fusion 26, 769. [2] Nishio, S. et al., (2002) Proc. 19 th Int. Conf. on Fusion Energy, CD-ROM file FT/P1-21 and http://www-pub.iaea.org/mtcd/pu blications/pdf/csp_019c/pdf/ftp1_21.pdf [3] Nishio, S. et al., (2004) Proc. 20 th Int. Conf. on Fusion Energy, CD-ROM file FT/P7-35 and http://wwwnaweb.iaea.org/napc/physics/fec/fec2004/papers/ft_p7-35.pdf. [4] Peng, Y-K. M. et al., (2004) Proc. 20 th Int. Conf. on Fusion Energy, CD-ROM file FT/3-1rb and http://wwwnaweb.iaea.org/napc/physics/fec/fec2004 /papers/ft_3-1rb.pdf. [5] Maekawa, T. et al. (2005) Nuclear Fusion 45, 117. [6] Yoshinaga, T. Uchida, M., Tanaka, H. and Maekawa, T. (2006) Phys. Rev. Letters 96, 125005. [7] Galambos, J.D. et al.(1995) Nuclear Fusion 35, 551. [8] Okano, K. et al., (1996) J. Plasma and Fusion Res. 72, 365. [9] Seki, Y. et al., (1990) 13th IAEA Fusion Energy Conf., IAEA-CN-53/G-1-2. [10] Nishio S., et al., (2002) J. Plasma Fusio Res. 78, 1218. [11] Peng, Y-K. M. et al., (1985) Fusion Tech. 8, 338. [12] Lin-Liu, Y.R. and Stambaugh, R.D. (2004) Nucl. Fusion 44, 548. 6