- Email: [email protected]
Blanket R&D activities in Japan towards fusion power reactors
Fusion Engineering and Design 51 – 52 (2000) 299 – 307 www.elsevier.com/locate/fusengdes
Blanket R&D activities in Japan towards fusion power reactors Satoru Tanaka a,*, Yoshihiro Ohara b, Hiroshi Kawamura c a
Department of Quantum Engineering and Systems Science, The Uni6ersity of Tokyo, 7 -3 -1, Hongo, Bunkyo-ku, Tokyo 113 -8656, Japan b Naka Research Establishment, Japan Atomic Energy Research Institute, Naka, Ibaraki, Japan c Oarai Research Establishment, Japan Atomic Energy Research Institute, Oarai, Ibaraki, Japan
Abstract Research and development on both the solid breeder blanket and the liquid breeder blanket have been conducted in Japan towards the fusion power reactors. The solid breeder blanket, one of the most realistic systems based on the present technical data base, is under development within Japan Atomic Energy Research Institute in collaboration with the Japanese universities. On the other hand, the liquid breeder blanket has been studied mainly within the universities as a future alternative blanket system considering its merits like a simple structure and a high accommodation to radiation damage. The present paper overviews the long-term development program and the current status of R&D on the breeding blankets towards the DEMO reactor. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Breeding blankets; Japan; Fusion power reactors
1. Introduction Recent progresses on the fusion plasma researches have made it possible to proceed intensively a development plan for a breeding blanket system aiming at the electric power generation in the fusion reactors. As a most realistic blanket system to produce high heat for the power generation, a solid breeder blanket cooled by high temperature pressurized water has been proposed and
* Corresponding author. Tel.: +81-3-38122111; fax: + 813-58006860. E-mail address: [email protected] (S. Tanaka).
applied for the design of DEMO reactor SSTR [1]. It utilizes layered small pebbles of tritium breeder and neutron multiplier installed in a reduced activation ferritic steel box structure. The electric power generation efficiency is expected to be comparable with that of the fission reactors due to its cooling conditions. In order to realize a higher efficiency, an advanced solid breeder blanket cooled by high temperature pressurized helium gas has been proposed for the design of the power reactor DREAM [2]. In the design, SiC/ SiC composite is utilized instead of the ferritic steel as a structural material. Presently, our major efforts are concentrated on the R&D on the water-cooled solid breeder blanket for the DEMO
0920-3796/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 0 - 3 7 9 6 ( 0 0 ) 0 0 2 9 5 - 7
300
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307
reactor considering the ferritic steel as a most realistic structural material. As an alternative, one also pursues an advanced blanket system based on liquid breeder cooling considering its merits like a simple structure and a high accommodation to radiation damage. Force free helical-like reactor (FFHR) is a design study for DEMO reactor which utilizes Flibe (Li2BeF4) as breeding material and coolant conducted by National Institute for Fusion Science (NIFS) and researchers in Japanese universities. Development of the solid breeder blanket is primarily conducted by Japan Atomic Energy Research Institute in collaboration with the Japanese universities, while the liquid breeder blanket has been studied mainly within the universities and NIFS. The present paper overviews the long-term development program and the current status of R&D on the breeding blankets towards the DEMO reactor.
2. Development program of breeding blanket Major purposes of the development plan on the solid breeder blanket are to develop fabrication technologies of the blanket module, to establish an engineering database for mock-up design and fabrication, and then to validate the design concepts by in-pile small-scaled mock-up tests and also by out-of-pile prototypical mock-up tests [3]. This program has been carried out so that the DEMO blanket test module to be tested in the
International Thermonuclear Experimental Reactor (ITER) will be in time for the ITER operation, according to the following phases (See Fig. 1). In Phase 1 where the authors are presently involved, elemental development on the blanket has been carried out. It covers fabrication technology development of a blanket structure, breeder and multiplier pebbles and tritium permeation barriers, as well as material-oriented engineering tests on the breeder and multiplier pebbles and a variety of engineering tests on packed pebble beds, including safety-related issues. In Phase 2, in-pile irradiation tests and out-of-pile thermomechanical performance tests will be carried out on the basis of the elemental development. Within the in-pile tests, small-scale mock-ups, typically 10 cm in diameter and 1 m long will be irradiated in the Japan Material Testing Reactor (JMTR) to examine thermal responses and tritium release characteristics. In the out-of-pile performance tests, prototype blanket mock-ups will be manufactured and their thermo-mechanical performances will be demonstrated. In Phase 3, one plans to construct a DEMO blanket test module based on the elemental development and the inpile and out-of-pile tests, and to install the test module in the ITER test port. The test module is to be operated under the fusion reactor environment in order to confirm high heat extraction, efficient tritium recovery, and reliability of the system. It would also be essential to conduct an electric power generation test using the test module prior to the DEMO reactor. In parallel to this
Fig. 1. Development plan for the solid breeder blanket.
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307 Table 1 Stages for liquid breeder R&D Stage 1
Stage 2
Stage 3
– establishment of material database – establishment of material system database – element test a few candidates selected – blanket module test – other engineering R&D tasks 1–2 candidates selected – engineering demonstration the most promising DEMO blanket concept determined after the comparison with solid blanket concepts
R&D program, many supporting and fundamental studies by experiments and modern calculation methods will be continued in the universities and JAERI. These include irradiation effects on tritium release from solid breeders, interaction of defects with tritium release and trapping behavior, and elucidation of tritium adsorption/desorption reactions on the surface of solid breeders. R&D status on liquid breeders is much behind that on solid breeders. In particular in Japan, not enough data not only on Flibe but also on the liquid metal breeders have been obtained to complete a liquid blanket design. Thus, the accomplishment of well-organized basic researches followed by a long-time extensive R&D is strongly required. Table 1 shows the R&D stages for liquid breeder research. Stage 1 is for the establishment of material database including interaction with structural materials. The database should be completed in the blanket conditions such as high temperature irradiation. This stage also contains ‘element test’ for tritium release behavior, and irradiation effects by using research reactors. At the end of this stage, a few candidates are selected for further investigation. In the 2nd stage, ‘blanket module test’ and other engineering R&D tasks should be done and the selection of the most promising one or two candidates is carried out. The final stage is the engineering demonstration, in which the technological establishment for construction, operation, maintenance and decommissioning of a blanket, and general evaluation of the feasibility from safety and economy point of view. Finally the most promising
301
DEMO blanket concept will be determined after the comparison with solid blanket concepts. These R&D tasks should be carried out in a strong relationship to blanket design studies. Though liquid blanket R&D is slower in progress than solid blanket R&D, it has a great potential for DEMO class reactors since it can operate in severe conditions such as high heat flux at high temperature. Thus, sufficient level of R&D should be carried out to realize the comparison in the feasibility with solid blanket concept by the time for the blanket concept selection. Because the R&D tasks of liquid blanket, in particular, selfcooled liquid Li concept is overlapped a lot with those of high-intensity fast neutron source, both R&Ds are expected to proceed in good consistency.
3. R&D status on solid breeder blanket
3.1. Materials de6elopment 3.1.1. Tritium breeder pebbles The breeder materials in the form of small pebbles have been applied to the recent designs of the solid breeder blanket. The characteristics like the pebble shape, size, density, purity, yield strength, and tritium release rate depend on the fabrication method. Three methods, i.e. a melting granulation method [4], a rotating granulation method [5] and a wet granulation method [6], have been studied. From the preliminary tests, the wet granulation method has turned out to be advantageous in that mass fabrication of small pebbles is relatively easy through the reprocessed lithium-bearing solution. The flow chart of the wet granulation method is shown in Fig. 2. Up to now, Li2O, Li2TiO3 and Li2ZrO3 pebbles with a target density of 80–85% T.D. have been manufactured. Particularly, Li2TiO3 pebbles with a diameter from 0.3 to 1.5 mm were manufactured successfully. Density dependence and thermal hysteresis of the thermal conductivity, the specific heat and the thermal expansion of Li2TiO3 pebbles have been investigated [7] and an in-situ irradiation test has been carried out in JMTR. According to the material data obtained so far,
302
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307
Li2TiO3 turned out to be promising as a breeder material.
3.1.2. Beryllium multiplier pebbles A rotating electrode method (REM), a magnesium reduction method (MRM) and a gas atomizing method (GAM) have been proposed so far for the beryllium pebble production. Among them, the REM has been studied in detail because this method was considered to be superior from the view point of impurity control and mass production. Concept of the REM is shown in Fig. 3. From the recent trial production tests, pebbles with a diameter of 0.3 – 2 mm has been fabricated at the production rate of about 120 kg/year using
Fig. 2. Flow chart of wet granulation method for Li2TiO3 pebble fabrication.
a small-scale machine (w 1.5× l 1 × h 1.5m) [8]. The tritium release and mechanical properties of the neutron irradiated REM pebble were measured. The tritium diffusion coefficient was almost similar to that of the hot-pressed beryllium and the tritium release properties turned out to be dependent on the growth of surface oxide layer [9]. The compressive strength at the room temperature did not change by the neutron irradiation up to 1000 appm He at 600°C [10]. In order to achieve a lower swelling and lower reactivity with the structural materials at a higher temperature, a possibility of beryllide intermetallic compounds is studied as a future alternative.
3.1.3. In-pile tritium release test Lithium titanate (Li2TiO3) is a promising tritium breeding material for its low chemical reactivity. An in-pile mockup with 1 mm diameter Li2TiO3 pebbles has been irradiated in JMTR in order to evaluate the effects of the irradiation temperature, the hydrogen content in the sweep gas and the sweep gas flow rate on the tritium release [11]. The irradiation facility and vertical cross section of the in-pile mockup are shown in Fig. 4. The tritium generated in Li2TiO3 pebbles and released from them was recovered by the helium sweep gas under the on-line analysis of tritium content. Tritium release from Li2TiO3 pebbles started when the center temperature of Li2TiO3 pebble bed was about 140°C and tritium release rate increased with increasing temperature. The saturated tritium release from Li2TiO3 pebble bed did not depend on the sweep gas flow rate, while it is dependent on the flow rate of hydrogen in the sweep gas [12]. 3.2. Engineering R&D in JAERI
Fig. 3. Concept of rotating electrode method for Be pebble fabrication.
3.2.1. Thermal fatigue test of a first wall model made of reduced acti6ation ferritic steel F-82H Reduced activation ferritic steel F-82H is one of the candidate blanket structural materials for DEMO reactor. A first wall panel made of F-82H [13] based on the design of ITER test blanket and SSTR [14]was fabricated by hot isostatic pressing (HIP). Major sizes of the panel are 200 mm long, 113 mm wide and 18 mm thick. The panel has ten
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307
303
Fig. 4. Block diagram of irradiation facility and vertical cross-section of in-pile mockup.
rectangular tubes and cooling manifolds are attached on both sides. A high heat flux test was performed to evaluate thermo-mechanical and fatigue performance of the panel and validity of HIP bonding method. The surface of the panel was heated by electron beam to investigate thermal fatigue performance. In order to simulate the maximum F-82H temperature of 450°C based on the design of water-cooled blanket even with the water temperature of about 30°C, the test was conducted at the surface heat flux of 2.7 MW/m2. The panel withstood more than 5000 thermal cycles. No particular external damage of the panel was observed after the test. Additionally, any change of the temperature response, which agreed well with a finite element method (FEM) analysis, was not observed. This also suggests no internal damage of panel after 5000 cycles. From the results of the high heat flux test, the HIPed F-82H first wall structure could be applied for the DEMO blanket.
3.2.2. Fabrication of a small blanket box model made of reduced acti6ation ferritic steel F-82H Based on the optimization study of HIP conditions [15] and the fabrication of a first wall panel
[13], a small blanket box model made of F-82H was fabricated. The model is composed of the first/side walls with built-in eight rectangular cooling tubes (8 mm× 8 mm in inner dimensions, 1.5 mm in thickness) and a back wall which includes inlet and outlet coolant manifolds. Top and bottom cover plates of the box are not included for this model. The rectangular tubes made of F-82H were manufactured by roll-forming or cold-drawing of circular tubes. The rectangular tubes, front and rear plates of the first/side walls sandwiching the tubes, and the back wall were joined by HIP method under condition of 1040°C temperature, 150 MPa pressure and 2-h holding time. In addition, heat treatments such as for releasing residual stresses or tempering after machining and welding were appropriately performed. An appearance of the model fabricated is shown in Fig. 5, from which a piece for destructive examination were cut away. The macroscopic image shows no excessive deformation of rectangular tubes. In a typical scanning electron microscopy (SEM) image and electron probe X-ray microanalysis (EPMA) results, any harmful voids and peculiar segregation and precipitation of material elements were not observed at the HIP bonded interface. Thus the
304
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307
fabricability of the F-82H blanket box structure by the HIP method has been demonstrated.
3.2.3. Effecti6e thermal conducti6ity measurements of the ceramic breeder pebble beds The equivalent thermal conductivity, packing characteristics and purge gas flow characteristics through the pebble bed have been investigated as important engineering data for designing a breeder/multiplier pebble bed. Recently, the hot wire method, a standardized method for measuring the thermal conductivity of less thermally conductive materials, has been applied to measure the equivalent thermal conductivity of Be, Li2O
and Li2TiO3 pebble beds. Also, a mockup thermal test by the cylindrical system was performed to demonstrate the integrated thermal performance of the packed pebble bed layers. With respect to the effective thermal conductivity measurement by the hot wire method, consistent data was obtained using Be, Alumina, Li2O single and binary beds [16]. Fig. 6 shows the measured results of Li2O 1 mm-diameter single packing bed as the dependency of the bed temperature, along with the data of Li2TiO3 sent from CEA under an IEA co-operative programme on Nuclear Technology of Fusion Reactors. The observed results were consistent with the correlations by Schluender, Zehner and Bauer (SZB) modified model and Hall and Martin (HM) modified model, assuming the contact area fraction and accommodation factor are the same values as those for Li2O pebble beds. The estimation of Li2TiO3 bed thermal conductivity with no mechanical stress has become possible by the result obtained in this work. The mock-up tests have been conducted to confirm the thermal design by using a cylindrical model with layered configuration of Be and Li2O or Li2TiO3. The preliminary results of the mock-up test [17] showed relatively good consistency with the thermal conductivity of Li2O pebble bed.
3.3. Fundamental researches in uni6ersities Fig. 5. Ferritic steel first wall box structure mock-up manufactured by hot isostatic pressing (HIPing).
Fig. 6. Measurement of thermal conductivity of Li2O and Li2TiO3 pebble beds.
Tritium recovery and inventory in solid blanket are influenced by tritium migration in the solid breeding material and tritium desorption on the surface. These processes are affected by irradiation defects and surface nature which is influenced by the chemical composition of the sweep gas. Fundamental studies to elucidate these have been intensively conducted in Japanese universities. Taniguchi et al. showed by in-situ FTIR observation of surface –OD on Li2O that surface is not homogeneous for D2O or D2 adsorption and that surface nature is influenced by the chemical composition in the sweep gas, suggesting the possibility of non-stoichiometry near the surface [18]. In order to elucidate surface nature, ab-initio quantum chemical calculation and XPS, UPS surface analysis were found to be helpful. [19,20]. The vapor pressures and work function changes of
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307
candidate Li ceramics at elevated temperatures were measured to investigate their thermochemical behavior [21,22]. Application of two independent experimental techniques made it possible to understand such behavior from both macroscopic and microscopic standpoints. The experimentally obtained results of vapor pressures and work function changes were closely related to each other, indicating some relevance between macroscopic and microscopic behaviors of hydrogen at the surface of ceramic breeders [23]. Behavior of irradiation defects and interaction of them with tritium are important subjects. Defect behaviors in Li2O have been studied by luminescence observation under reactor, ion- or g-irradiation by Grishmanov et al. They elucidated the behavior of self-trapped exciton, F+ center, F-center aggregate and lithium colloid. Interaction of colloid with tritium release was pointed out by them [24]. Tanigawa et al. conducted experimental studies on interaction of hydrogen isotopes with defects in single crystal Li2O by using in-situ FTIR observation of O–D at high temperature, combining with quantum chemical studies [25]. In order to evaluate precisely tritium inventory in the blanket, all migration processes should be systematically integrated. Nishikawa et al. constructed this system model which includes tritium diffusion in the grain, tritium percolation along grain boundaries, tritium desorption and exchange reaction on the surface. Calculation results were able to reproduce the tritium inventories in in-situ tritium release experiments [26].
4. R&D status on liquid breeder blanket In order for Flibe to be utilized in fusion reactor, several key issues should be solved. They are tritium release, compatibility of structural materials and molten salt technology. On compatibility of Flibe to structural materials, analysis of corrosion mechanism and corrosion protection is very important from the point of reliability and safety. Flibe produces HF– H2O –O2 –F2 gas mixture by the nuclear reaction with neutrons, and these species result in a serious
305
corrosion problem. Be is added as the neutron multiplier in FFHR blanket design, and it is expected to act as a reducing agent and a scavenger for the corrosive species. Thus, corrosion behavior of Flibe should be investigated under controlled Redox potential. A compatibility study with Flibe has initiated in 1997 in the University of Tokyo in the collaboration with NIFS. Up to the present moment, thermodynamic analysis and preliminary dipping tests in a pot have been done [27,28]. In the thermodynamic analysis, it was concluded that JLF-1 and V-based alloy (V–5Cr– 5Ti) would have sufficient corrosion resistance, if oxide film works as a protective scale, and that Be addition is very effective for keeping the blanket in a reducing condition. However, a kinetic study is necessary whether or not they actually work well. From these points of view, preliminary dipping experiments were performed in an inert atmosphere with a trace of HF as the first step. No severe corrosion was observed in case of SUS430 specimens, and Cr2O3 and FeCr2O4 were formed on the surfaces. In case of vanadium specimens, on the other hand, severe corrosion was observed. In order to improve the compatibility of vanadium, addition of corrosion-resistant elements, Be utilization as a Redox buffer, and corrosion resistant coating should be investigated in the future. Tritium recovery and release should be elucidated for the design of tritium recovery systems and the estimation of tritium inventory and tritium leakage rate. In-pile tritium released experiments from Flibe named INTREX-FLIBE have been carried out at the University of Tokyo [29]. Temporal change of tritium release rate was measured with different H2 and HF concentrations in He purge gas, and the overall mass transfer coefficient of HT from molten Flibe to purge gas was obtained. Tritium release rate is different between HT and TF, and they change to each other depending on the Redox potential. In order to explain these two results, a primary model on tritium release mechanisms from molten Flibe including an isotopic exchange reaction, permeation through structural walls contacting molten Flibe was constructed. One of the key issues of the liquid metal blanket is the development of ceramic coating acting
306
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307
as a barrier for insulating, tritium permeation and corrosion. Compatibility test and manufacturing studies have been conducted. In order to select the material which has good compatibility with liquid breeder, bulk specimens for coating materials, e.g. Y2O3, MgO, 3Al2O3 – MgO, Al2O3, BN, AlN–BN and AlN were examined in compatibility with liquid lithium and Li17 – Pb83 at 773 K for 2000 h [30]. For Li17 – Pb83, all of the compounds examined were intact, while only Y2O3 and the nitrides gave good results for liquid lithium. These results were in good agreement with thermodynamic predictions. However, in case of Y2O3, a mixed compound LiYO2 and a non-stoichiometric compound Y2O3 − x were formed, and the latter was electrically conductive. In case of the nitrides, Al2O3 as an impurity on grain boundaries was dissolved in Li, which results in the degradation. Al2O3 coatings were prepared on austenitic stainless steel (SUS316) by the hot-dipping and oxidation method. It has a multi-layered structure consisting of Al2O3, AlFe, AlNi3 and Fe, and has higher electrical resistivity than 10 MV, high compatibility with Li17 – Pb83 at 873 K. Trial fabrication of Y2O3 coating on the surface of SUS316 was also carried out by the plasma-spraying method. This coating was severely attacked by liquid lithium because lithium penetrated into the coating layer through small pores or cracks and reacted on Y2O3 to form LiYO2. Several microns of Al2O3 and AlN coatings were successfully prepared on SUS430 substrate by RF sputtering method in several hours. Both coating have enough high resistance \1012 V. However, the coating was not intact with liquid lithium [31].
5. Summary Research and development on both the solid breeder blanket and the liquid breeder blanket have been conducted in Japan towards the fusion power reactors. R&Ds on the solid breeder blanket, the main option for the design of the DEMO reactor, have made progress in recent years, particularly in the area of manufacturing technologies for the blanket structure made of the reduced activation ferritic steel F-82H and also for the
Li2TiO3 breeder pebbles and the Be multiplier pebbles. Engineering database on the solid breeder blanket has been accumulated steadily aiming at the module test in ITER. On the other hand, the liquid breeder blanket has been studied mainly by the universities and NIFS as a future alternative blanket system. Though the researches on the liquid breeder blanket, including researches on ceramic coating, are still in a fundamental level or an experiment-room scale, many useful results have been obtained so far. Additionally, a systematic study on flibe (Li2BeF4) has started in order to clarify the applicability or limit of this material to the fusion blanket from the view points of hydraulics, heat transfer, compatibility and tritium control. Intensive research and development on both the solid breeder blanket and the liquid breeder blanket will be continued within Japanese universities, NIFS and JAERI collaboratively to realize an attractive DEMO blanket system. Acknowledgements The authors would like to acknowledge Professor T. Terai and Drs. M. Enoeda, K. Furuya, T. Hatano, E. Ishitsuka, T. Kuroda and K. Tsuchiya for providing information to this review paper. References [1] Y. Seki et al., Proc. 13th Int. Conf. on Plasma Phys. and Contr. Nucl. Fusion Research, Washington, DC, IAEACN-53/G-I-2, Vienna, 1991. [2] S. Nishio et al., Maintenance oriented tokamak reactor with low activation material and high aspect ratio configuration, Proc. 16th Int. Conf. on Fusion Energy, Montreal, IAEA-CN-64/GP-27, Vienna, 1997. [3] H. Takatsu, H. Kawamura, S. Tanaka, Development of ceramic breeder blankets in Japan, Fus. Eng. Des. 39–40 (1998) 645 – 650. [4] N. Asami, K. Nagashima, H. Akiyama, M. Nagakura, N. Suemari, A. Ohya, Development of fabrication methods for lithium ceramics, Ceram. Trans. 27 (1989) 17 – 35. [5] T. Suzuki, O. Murata, S. Hirata, Development of small spherical lithium ceramics for fusion reactor blanket, Ceram. Trans. 27 (1989) 37 – 56. [6] K. Tsuchiya, H. Kawamura, K. Fuchinoue, H. Sawada, K. Watarumi, Fabrication development and preliminary characterization of Li2TiO3 pebbles by wet process, J. Nucl. Mater. 258 – 263 (1998) 1985 – 1990.
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307 [7] S. Saito, K. Tsuchiya, H. Kawamura, T. Terai, S. Tanaka, Density dependence on thermal properties of Li2TiO3 pellets, J. Nucl. Mater. 253 (1998) 213 – 218. [8] T. Iwadachi, N. Sakamoto, K. Nishida, H. Kawamura, Production of various size and some properties of beryllium pebbles by the rotating electrode method, JAERIConf. 98-001 (1998) 33–38. [9] E. Ishitsuka, H. Kawamura, T. Terai and S. Tanaka, Effect of surface oxide layer on tritium release from beryllium pebbles, In: K. Herschbach, W. Maurer, J.E. Vetter (Eds.), Fus. Technol. 1994, 2 (1995) 1345 – 1348. [10] E. Ishitsuka, H. Kawamura, T. Terai, S. Tanaka, Microstructure and mechanical properties of neutron irradiated beryllium, J. Nucl. Mater. 258–263 (1998) 566 – 570. [11] H. Kawamura, K. Tsuchiya, M. Nakamichi, J. Fujita, H. Sagawa, Y. Nagao, Y. Gohar, Y. Ikejima, T. Saito, S. Sakurai, I. Hisa, H. Kumahara, N. Nemoto, Tritium release behavior from lithium titanate pebbles at low irradiation temperature, Proc. of the 20th Symposium on Fusion Technology, France, vol. 2, 1998, pp. 1289 – 1292. [12] K. Tsuchiya, M. Nakamichi, Y. Nagao, J. Fujita, H. Sagawa, S. Tanaka and H. Kawamura, Integrated experiment of blanket in-pile mockup with Li2TiO3 pebbles, Fusion Eng. Des. (in press). [13] K. Furuya, M. Enoeda, T. Hatano, et al., Application of HIP bonding to first wall panel fabrication made from reduced activation ferritic steel F82H, J. Nucl. Mater. 258 – 263 (1998) 2023–2029. [14] S. Mori, S. Yamazaki, J. Adachi, et al., Blanket and divertor design for the Steady State Tokamak Reactor (SSTR), Fus. Eng. Des. 18 (1991) 249–258. [15] M. Oda et al., Development of HIP bonding procedure and mechanical properties of HIP bonded joints for reduced activation ferritic steel F82H, JAERI-Tech 97-013 (1997). [16] M. Enoeda, et al., Fus. Technol. 34 (1998) 877 – 881. [17] S. Sato et al., Proceedings of 20th SOFT, Marseille, France, 1998. [18] M. Taniguchi, et al., In-situ observation of surface hydroxyl group on lithium ceramics at high temperature by infrared absorption spectroscopy, Fus. Technol. 28 (1995)
.
307
1284 – 1289. [19] M. Taniguchi, S. Tanaka, Ab-initio Hartree – Fock study on surface desorption process in tritium release, J. Nucl. Mater. 258 – 263 (1998) 531 – 536. [20] S. Tanaka, M. Taniguchi, H. Tanigawa XPS and UPS studies on electronic structure of Li2O, J. Nucl. Mater. (in press). [21] A. Suzuki, M. Tonegawa, M. Yasumoto, M. Yamawaki, Sweep gas chemistry effect on vaporization property of Li2ZrO3, J. Nucl. Mater. 258 – 263 (1998) 562 – 565. [22] A. Suzuki, K. Yamaguchi, M. Yamawaki, Study on the surface nonstoichiometry of solid breeder blanket materials by means of work function measurement, Fus. Eng. Des. 39 – 40 (1998) 699 – 705. [23] A. Suzuki, K. Yamaguchi, T. Terai, M. Yamawaki, Application of high temperature and spectrometry and work function measurement to evaluation of thermochemical performance of ceramic breeders, Fus. Eng. Des. (in press). [24] V. Grishmanov, et al., Influence of radiation defects on tritium release parameters from Li2O, Fus. Eng. Des. 39 – 40 (1998) 685 – 691. [25] H. Tanigawa et al., FT-IR study on interaction of hydrogen isotopes with defects in lithium oxide, Fusion Eng. Des. (in press). [26] M. Nishsikawa, A. Baba, S. Odoi, Y. Kawamura, Tritium inventory estimation in solid blanket system, Fus. Eng. Des. 39 – 40 (1998) 615 – 625. [27] T. Terai, et al., Compatibility of structural materials with Li2BeF4 molten salt breeder, J. Nucl. Mater. 258–263 (1998) 513. [28] T. Terai et al., Corrosion behavior of molten LiF –BeF2 mixture, Fusion Eng. Des. [29] A. Suzuki, et al., In-situ HT release behavior from molten Li2BeF4 salt, Fus. Eng. Des. 39 – 40 (1998) 781 – 785. [30] T. Mitsuyama, et al., Compatibility of insulating ceramic materials with liquid breeders, Fus. Eng. Des. 39–40 (1998) 811 – 817. [31] T. Terai et al., Fabrication and property of ceramic coating for CTR liquid blanket by sputtering method, Fusion Eng. Des. (in press).
.
Blanket R&D activities in Japan towards fusion power reactors Satoru Tanaka a,*, Yoshihiro Ohara b, Hiroshi Kawamura c a
Department of Quantum Engineering and Systems Science, The Uni6ersity of Tokyo, 7 -3 -1, Hongo, Bunkyo-ku, Tokyo 113 -8656, Japan b Naka Research Establishment, Japan Atomic Energy Research Institute, Naka, Ibaraki, Japan c Oarai Research Establishment, Japan Atomic Energy Research Institute, Oarai, Ibaraki, Japan
Abstract Research and development on both the solid breeder blanket and the liquid breeder blanket have been conducted in Japan towards the fusion power reactors. The solid breeder blanket, one of the most realistic systems based on the present technical data base, is under development within Japan Atomic Energy Research Institute in collaboration with the Japanese universities. On the other hand, the liquid breeder blanket has been studied mainly within the universities as a future alternative blanket system considering its merits like a simple structure and a high accommodation to radiation damage. The present paper overviews the long-term development program and the current status of R&D on the breeding blankets towards the DEMO reactor. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Breeding blankets; Japan; Fusion power reactors
1. Introduction Recent progresses on the fusion plasma researches have made it possible to proceed intensively a development plan for a breeding blanket system aiming at the electric power generation in the fusion reactors. As a most realistic blanket system to produce high heat for the power generation, a solid breeder blanket cooled by high temperature pressurized water has been proposed and
* Corresponding author. Tel.: +81-3-38122111; fax: + 813-58006860. E-mail address: [email protected] (S. Tanaka).
applied for the design of DEMO reactor SSTR [1]. It utilizes layered small pebbles of tritium breeder and neutron multiplier installed in a reduced activation ferritic steel box structure. The electric power generation efficiency is expected to be comparable with that of the fission reactors due to its cooling conditions. In order to realize a higher efficiency, an advanced solid breeder blanket cooled by high temperature pressurized helium gas has been proposed for the design of the power reactor DREAM [2]. In the design, SiC/ SiC composite is utilized instead of the ferritic steel as a structural material. Presently, our major efforts are concentrated on the R&D on the water-cooled solid breeder blanket for the DEMO
0920-3796/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 0 - 3 7 9 6 ( 0 0 ) 0 0 2 9 5 - 7
300
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307
reactor considering the ferritic steel as a most realistic structural material. As an alternative, one also pursues an advanced blanket system based on liquid breeder cooling considering its merits like a simple structure and a high accommodation to radiation damage. Force free helical-like reactor (FFHR) is a design study for DEMO reactor which utilizes Flibe (Li2BeF4) as breeding material and coolant conducted by National Institute for Fusion Science (NIFS) and researchers in Japanese universities. Development of the solid breeder blanket is primarily conducted by Japan Atomic Energy Research Institute in collaboration with the Japanese universities, while the liquid breeder blanket has been studied mainly within the universities and NIFS. The present paper overviews the long-term development program and the current status of R&D on the breeding blankets towards the DEMO reactor.
2. Development program of breeding blanket Major purposes of the development plan on the solid breeder blanket are to develop fabrication technologies of the blanket module, to establish an engineering database for mock-up design and fabrication, and then to validate the design concepts by in-pile small-scaled mock-up tests and also by out-of-pile prototypical mock-up tests [3]. This program has been carried out so that the DEMO blanket test module to be tested in the
International Thermonuclear Experimental Reactor (ITER) will be in time for the ITER operation, according to the following phases (See Fig. 1). In Phase 1 where the authors are presently involved, elemental development on the blanket has been carried out. It covers fabrication technology development of a blanket structure, breeder and multiplier pebbles and tritium permeation barriers, as well as material-oriented engineering tests on the breeder and multiplier pebbles and a variety of engineering tests on packed pebble beds, including safety-related issues. In Phase 2, in-pile irradiation tests and out-of-pile thermomechanical performance tests will be carried out on the basis of the elemental development. Within the in-pile tests, small-scale mock-ups, typically 10 cm in diameter and 1 m long will be irradiated in the Japan Material Testing Reactor (JMTR) to examine thermal responses and tritium release characteristics. In the out-of-pile performance tests, prototype blanket mock-ups will be manufactured and their thermo-mechanical performances will be demonstrated. In Phase 3, one plans to construct a DEMO blanket test module based on the elemental development and the inpile and out-of-pile tests, and to install the test module in the ITER test port. The test module is to be operated under the fusion reactor environment in order to confirm high heat extraction, efficient tritium recovery, and reliability of the system. It would also be essential to conduct an electric power generation test using the test module prior to the DEMO reactor. In parallel to this
Fig. 1. Development plan for the solid breeder blanket.
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307 Table 1 Stages for liquid breeder R&D Stage 1
Stage 2
Stage 3
– establishment of material database – establishment of material system database – element test a few candidates selected – blanket module test – other engineering R&D tasks 1–2 candidates selected – engineering demonstration the most promising DEMO blanket concept determined after the comparison with solid blanket concepts
R&D program, many supporting and fundamental studies by experiments and modern calculation methods will be continued in the universities and JAERI. These include irradiation effects on tritium release from solid breeders, interaction of defects with tritium release and trapping behavior, and elucidation of tritium adsorption/desorption reactions on the surface of solid breeders. R&D status on liquid breeders is much behind that on solid breeders. In particular in Japan, not enough data not only on Flibe but also on the liquid metal breeders have been obtained to complete a liquid blanket design. Thus, the accomplishment of well-organized basic researches followed by a long-time extensive R&D is strongly required. Table 1 shows the R&D stages for liquid breeder research. Stage 1 is for the establishment of material database including interaction with structural materials. The database should be completed in the blanket conditions such as high temperature irradiation. This stage also contains ‘element test’ for tritium release behavior, and irradiation effects by using research reactors. At the end of this stage, a few candidates are selected for further investigation. In the 2nd stage, ‘blanket module test’ and other engineering R&D tasks should be done and the selection of the most promising one or two candidates is carried out. The final stage is the engineering demonstration, in which the technological establishment for construction, operation, maintenance and decommissioning of a blanket, and general evaluation of the feasibility from safety and economy point of view. Finally the most promising
301
DEMO blanket concept will be determined after the comparison with solid blanket concepts. These R&D tasks should be carried out in a strong relationship to blanket design studies. Though liquid blanket R&D is slower in progress than solid blanket R&D, it has a great potential for DEMO class reactors since it can operate in severe conditions such as high heat flux at high temperature. Thus, sufficient level of R&D should be carried out to realize the comparison in the feasibility with solid blanket concept by the time for the blanket concept selection. Because the R&D tasks of liquid blanket, in particular, selfcooled liquid Li concept is overlapped a lot with those of high-intensity fast neutron source, both R&Ds are expected to proceed in good consistency.
3. R&D status on solid breeder blanket
3.1. Materials de6elopment 3.1.1. Tritium breeder pebbles The breeder materials in the form of small pebbles have been applied to the recent designs of the solid breeder blanket. The characteristics like the pebble shape, size, density, purity, yield strength, and tritium release rate depend on the fabrication method. Three methods, i.e. a melting granulation method [4], a rotating granulation method [5] and a wet granulation method [6], have been studied. From the preliminary tests, the wet granulation method has turned out to be advantageous in that mass fabrication of small pebbles is relatively easy through the reprocessed lithium-bearing solution. The flow chart of the wet granulation method is shown in Fig. 2. Up to now, Li2O, Li2TiO3 and Li2ZrO3 pebbles with a target density of 80–85% T.D. have been manufactured. Particularly, Li2TiO3 pebbles with a diameter from 0.3 to 1.5 mm were manufactured successfully. Density dependence and thermal hysteresis of the thermal conductivity, the specific heat and the thermal expansion of Li2TiO3 pebbles have been investigated [7] and an in-situ irradiation test has been carried out in JMTR. According to the material data obtained so far,
302
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307
Li2TiO3 turned out to be promising as a breeder material.
3.1.2. Beryllium multiplier pebbles A rotating electrode method (REM), a magnesium reduction method (MRM) and a gas atomizing method (GAM) have been proposed so far for the beryllium pebble production. Among them, the REM has been studied in detail because this method was considered to be superior from the view point of impurity control and mass production. Concept of the REM is shown in Fig. 3. From the recent trial production tests, pebbles with a diameter of 0.3 – 2 mm has been fabricated at the production rate of about 120 kg/year using
Fig. 2. Flow chart of wet granulation method for Li2TiO3 pebble fabrication.
a small-scale machine (w 1.5× l 1 × h 1.5m) [8]. The tritium release and mechanical properties of the neutron irradiated REM pebble were measured. The tritium diffusion coefficient was almost similar to that of the hot-pressed beryllium and the tritium release properties turned out to be dependent on the growth of surface oxide layer [9]. The compressive strength at the room temperature did not change by the neutron irradiation up to 1000 appm He at 600°C [10]. In order to achieve a lower swelling and lower reactivity with the structural materials at a higher temperature, a possibility of beryllide intermetallic compounds is studied as a future alternative.
3.1.3. In-pile tritium release test Lithium titanate (Li2TiO3) is a promising tritium breeding material for its low chemical reactivity. An in-pile mockup with 1 mm diameter Li2TiO3 pebbles has been irradiated in JMTR in order to evaluate the effects of the irradiation temperature, the hydrogen content in the sweep gas and the sweep gas flow rate on the tritium release [11]. The irradiation facility and vertical cross section of the in-pile mockup are shown in Fig. 4. The tritium generated in Li2TiO3 pebbles and released from them was recovered by the helium sweep gas under the on-line analysis of tritium content. Tritium release from Li2TiO3 pebbles started when the center temperature of Li2TiO3 pebble bed was about 140°C and tritium release rate increased with increasing temperature. The saturated tritium release from Li2TiO3 pebble bed did not depend on the sweep gas flow rate, while it is dependent on the flow rate of hydrogen in the sweep gas [12]. 3.2. Engineering R&D in JAERI
Fig. 3. Concept of rotating electrode method for Be pebble fabrication.
3.2.1. Thermal fatigue test of a first wall model made of reduced acti6ation ferritic steel F-82H Reduced activation ferritic steel F-82H is one of the candidate blanket structural materials for DEMO reactor. A first wall panel made of F-82H [13] based on the design of ITER test blanket and SSTR [14]was fabricated by hot isostatic pressing (HIP). Major sizes of the panel are 200 mm long, 113 mm wide and 18 mm thick. The panel has ten
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307
303
Fig. 4. Block diagram of irradiation facility and vertical cross-section of in-pile mockup.
rectangular tubes and cooling manifolds are attached on both sides. A high heat flux test was performed to evaluate thermo-mechanical and fatigue performance of the panel and validity of HIP bonding method. The surface of the panel was heated by electron beam to investigate thermal fatigue performance. In order to simulate the maximum F-82H temperature of 450°C based on the design of water-cooled blanket even with the water temperature of about 30°C, the test was conducted at the surface heat flux of 2.7 MW/m2. The panel withstood more than 5000 thermal cycles. No particular external damage of the panel was observed after the test. Additionally, any change of the temperature response, which agreed well with a finite element method (FEM) analysis, was not observed. This also suggests no internal damage of panel after 5000 cycles. From the results of the high heat flux test, the HIPed F-82H first wall structure could be applied for the DEMO blanket.
3.2.2. Fabrication of a small blanket box model made of reduced acti6ation ferritic steel F-82H Based on the optimization study of HIP conditions [15] and the fabrication of a first wall panel
[13], a small blanket box model made of F-82H was fabricated. The model is composed of the first/side walls with built-in eight rectangular cooling tubes (8 mm× 8 mm in inner dimensions, 1.5 mm in thickness) and a back wall which includes inlet and outlet coolant manifolds. Top and bottom cover plates of the box are not included for this model. The rectangular tubes made of F-82H were manufactured by roll-forming or cold-drawing of circular tubes. The rectangular tubes, front and rear plates of the first/side walls sandwiching the tubes, and the back wall were joined by HIP method under condition of 1040°C temperature, 150 MPa pressure and 2-h holding time. In addition, heat treatments such as for releasing residual stresses or tempering after machining and welding were appropriately performed. An appearance of the model fabricated is shown in Fig. 5, from which a piece for destructive examination were cut away. The macroscopic image shows no excessive deformation of rectangular tubes. In a typical scanning electron microscopy (SEM) image and electron probe X-ray microanalysis (EPMA) results, any harmful voids and peculiar segregation and precipitation of material elements were not observed at the HIP bonded interface. Thus the
304
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307
fabricability of the F-82H blanket box structure by the HIP method has been demonstrated.
3.2.3. Effecti6e thermal conducti6ity measurements of the ceramic breeder pebble beds The equivalent thermal conductivity, packing characteristics and purge gas flow characteristics through the pebble bed have been investigated as important engineering data for designing a breeder/multiplier pebble bed. Recently, the hot wire method, a standardized method for measuring the thermal conductivity of less thermally conductive materials, has been applied to measure the equivalent thermal conductivity of Be, Li2O
and Li2TiO3 pebble beds. Also, a mockup thermal test by the cylindrical system was performed to demonstrate the integrated thermal performance of the packed pebble bed layers. With respect to the effective thermal conductivity measurement by the hot wire method, consistent data was obtained using Be, Alumina, Li2O single and binary beds [16]. Fig. 6 shows the measured results of Li2O 1 mm-diameter single packing bed as the dependency of the bed temperature, along with the data of Li2TiO3 sent from CEA under an IEA co-operative programme on Nuclear Technology of Fusion Reactors. The observed results were consistent with the correlations by Schluender, Zehner and Bauer (SZB) modified model and Hall and Martin (HM) modified model, assuming the contact area fraction and accommodation factor are the same values as those for Li2O pebble beds. The estimation of Li2TiO3 bed thermal conductivity with no mechanical stress has become possible by the result obtained in this work. The mock-up tests have been conducted to confirm the thermal design by using a cylindrical model with layered configuration of Be and Li2O or Li2TiO3. The preliminary results of the mock-up test [17] showed relatively good consistency with the thermal conductivity of Li2O pebble bed.
3.3. Fundamental researches in uni6ersities Fig. 5. Ferritic steel first wall box structure mock-up manufactured by hot isostatic pressing (HIPing).
Fig. 6. Measurement of thermal conductivity of Li2O and Li2TiO3 pebble beds.
Tritium recovery and inventory in solid blanket are influenced by tritium migration in the solid breeding material and tritium desorption on the surface. These processes are affected by irradiation defects and surface nature which is influenced by the chemical composition of the sweep gas. Fundamental studies to elucidate these have been intensively conducted in Japanese universities. Taniguchi et al. showed by in-situ FTIR observation of surface –OD on Li2O that surface is not homogeneous for D2O or D2 adsorption and that surface nature is influenced by the chemical composition in the sweep gas, suggesting the possibility of non-stoichiometry near the surface [18]. In order to elucidate surface nature, ab-initio quantum chemical calculation and XPS, UPS surface analysis were found to be helpful. [19,20]. The vapor pressures and work function changes of
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307
candidate Li ceramics at elevated temperatures were measured to investigate their thermochemical behavior [21,22]. Application of two independent experimental techniques made it possible to understand such behavior from both macroscopic and microscopic standpoints. The experimentally obtained results of vapor pressures and work function changes were closely related to each other, indicating some relevance between macroscopic and microscopic behaviors of hydrogen at the surface of ceramic breeders [23]. Behavior of irradiation defects and interaction of them with tritium are important subjects. Defect behaviors in Li2O have been studied by luminescence observation under reactor, ion- or g-irradiation by Grishmanov et al. They elucidated the behavior of self-trapped exciton, F+ center, F-center aggregate and lithium colloid. Interaction of colloid with tritium release was pointed out by them [24]. Tanigawa et al. conducted experimental studies on interaction of hydrogen isotopes with defects in single crystal Li2O by using in-situ FTIR observation of O–D at high temperature, combining with quantum chemical studies [25]. In order to evaluate precisely tritium inventory in the blanket, all migration processes should be systematically integrated. Nishikawa et al. constructed this system model which includes tritium diffusion in the grain, tritium percolation along grain boundaries, tritium desorption and exchange reaction on the surface. Calculation results were able to reproduce the tritium inventories in in-situ tritium release experiments [26].
4. R&D status on liquid breeder blanket In order for Flibe to be utilized in fusion reactor, several key issues should be solved. They are tritium release, compatibility of structural materials and molten salt technology. On compatibility of Flibe to structural materials, analysis of corrosion mechanism and corrosion protection is very important from the point of reliability and safety. Flibe produces HF– H2O –O2 –F2 gas mixture by the nuclear reaction with neutrons, and these species result in a serious
305
corrosion problem. Be is added as the neutron multiplier in FFHR blanket design, and it is expected to act as a reducing agent and a scavenger for the corrosive species. Thus, corrosion behavior of Flibe should be investigated under controlled Redox potential. A compatibility study with Flibe has initiated in 1997 in the University of Tokyo in the collaboration with NIFS. Up to the present moment, thermodynamic analysis and preliminary dipping tests in a pot have been done [27,28]. In the thermodynamic analysis, it was concluded that JLF-1 and V-based alloy (V–5Cr– 5Ti) would have sufficient corrosion resistance, if oxide film works as a protective scale, and that Be addition is very effective for keeping the blanket in a reducing condition. However, a kinetic study is necessary whether or not they actually work well. From these points of view, preliminary dipping experiments were performed in an inert atmosphere with a trace of HF as the first step. No severe corrosion was observed in case of SUS430 specimens, and Cr2O3 and FeCr2O4 were formed on the surfaces. In case of vanadium specimens, on the other hand, severe corrosion was observed. In order to improve the compatibility of vanadium, addition of corrosion-resistant elements, Be utilization as a Redox buffer, and corrosion resistant coating should be investigated in the future. Tritium recovery and release should be elucidated for the design of tritium recovery systems and the estimation of tritium inventory and tritium leakage rate. In-pile tritium released experiments from Flibe named INTREX-FLIBE have been carried out at the University of Tokyo [29]. Temporal change of tritium release rate was measured with different H2 and HF concentrations in He purge gas, and the overall mass transfer coefficient of HT from molten Flibe to purge gas was obtained. Tritium release rate is different between HT and TF, and they change to each other depending on the Redox potential. In order to explain these two results, a primary model on tritium release mechanisms from molten Flibe including an isotopic exchange reaction, permeation through structural walls contacting molten Flibe was constructed. One of the key issues of the liquid metal blanket is the development of ceramic coating acting
306
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307
as a barrier for insulating, tritium permeation and corrosion. Compatibility test and manufacturing studies have been conducted. In order to select the material which has good compatibility with liquid breeder, bulk specimens for coating materials, e.g. Y2O3, MgO, 3Al2O3 – MgO, Al2O3, BN, AlN–BN and AlN were examined in compatibility with liquid lithium and Li17 – Pb83 at 773 K for 2000 h [30]. For Li17 – Pb83, all of the compounds examined were intact, while only Y2O3 and the nitrides gave good results for liquid lithium. These results were in good agreement with thermodynamic predictions. However, in case of Y2O3, a mixed compound LiYO2 and a non-stoichiometric compound Y2O3 − x were formed, and the latter was electrically conductive. In case of the nitrides, Al2O3 as an impurity on grain boundaries was dissolved in Li, which results in the degradation. Al2O3 coatings were prepared on austenitic stainless steel (SUS316) by the hot-dipping and oxidation method. It has a multi-layered structure consisting of Al2O3, AlFe, AlNi3 and Fe, and has higher electrical resistivity than 10 MV, high compatibility with Li17 – Pb83 at 873 K. Trial fabrication of Y2O3 coating on the surface of SUS316 was also carried out by the plasma-spraying method. This coating was severely attacked by liquid lithium because lithium penetrated into the coating layer through small pores or cracks and reacted on Y2O3 to form LiYO2. Several microns of Al2O3 and AlN coatings were successfully prepared on SUS430 substrate by RF sputtering method in several hours. Both coating have enough high resistance \1012 V. However, the coating was not intact with liquid lithium [31].
5. Summary Research and development on both the solid breeder blanket and the liquid breeder blanket have been conducted in Japan towards the fusion power reactors. R&Ds on the solid breeder blanket, the main option for the design of the DEMO reactor, have made progress in recent years, particularly in the area of manufacturing technologies for the blanket structure made of the reduced activation ferritic steel F-82H and also for the
Li2TiO3 breeder pebbles and the Be multiplier pebbles. Engineering database on the solid breeder blanket has been accumulated steadily aiming at the module test in ITER. On the other hand, the liquid breeder blanket has been studied mainly by the universities and NIFS as a future alternative blanket system. Though the researches on the liquid breeder blanket, including researches on ceramic coating, are still in a fundamental level or an experiment-room scale, many useful results have been obtained so far. Additionally, a systematic study on flibe (Li2BeF4) has started in order to clarify the applicability or limit of this material to the fusion blanket from the view points of hydraulics, heat transfer, compatibility and tritium control. Intensive research and development on both the solid breeder blanket and the liquid breeder blanket will be continued within Japanese universities, NIFS and JAERI collaboratively to realize an attractive DEMO blanket system. Acknowledgements The authors would like to acknowledge Professor T. Terai and Drs. M. Enoeda, K. Furuya, T. Hatano, E. Ishitsuka, T. Kuroda and K. Tsuchiya for providing information to this review paper. References [1] Y. Seki et al., Proc. 13th Int. Conf. on Plasma Phys. and Contr. Nucl. Fusion Research, Washington, DC, IAEACN-53/G-I-2, Vienna, 1991. [2] S. Nishio et al., Maintenance oriented tokamak reactor with low activation material and high aspect ratio configuration, Proc. 16th Int. Conf. on Fusion Energy, Montreal, IAEA-CN-64/GP-27, Vienna, 1997. [3] H. Takatsu, H. Kawamura, S. Tanaka, Development of ceramic breeder blankets in Japan, Fus. Eng. Des. 39–40 (1998) 645 – 650. [4] N. Asami, K. Nagashima, H. Akiyama, M. Nagakura, N. Suemari, A. Ohya, Development of fabrication methods for lithium ceramics, Ceram. Trans. 27 (1989) 17 – 35. [5] T. Suzuki, O. Murata, S. Hirata, Development of small spherical lithium ceramics for fusion reactor blanket, Ceram. Trans. 27 (1989) 37 – 56. [6] K. Tsuchiya, H. Kawamura, K. Fuchinoue, H. Sawada, K. Watarumi, Fabrication development and preliminary characterization of Li2TiO3 pebbles by wet process, J. Nucl. Mater. 258 – 263 (1998) 1985 – 1990.
S. Tanaka et al. / Fusion Engineering and Design 51–52 (2000) 299–307 [7] S. Saito, K. Tsuchiya, H. Kawamura, T. Terai, S. Tanaka, Density dependence on thermal properties of Li2TiO3 pellets, J. Nucl. Mater. 253 (1998) 213 – 218. [8] T. Iwadachi, N. Sakamoto, K. Nishida, H. Kawamura, Production of various size and some properties of beryllium pebbles by the rotating electrode method, JAERIConf. 98-001 (1998) 33–38. [9] E. Ishitsuka, H. Kawamura, T. Terai and S. Tanaka, Effect of surface oxide layer on tritium release from beryllium pebbles, In: K. Herschbach, W. Maurer, J.E. Vetter (Eds.), Fus. Technol. 1994, 2 (1995) 1345 – 1348. [10] E. Ishitsuka, H. Kawamura, T. Terai, S. Tanaka, Microstructure and mechanical properties of neutron irradiated beryllium, J. Nucl. Mater. 258–263 (1998) 566 – 570. [11] H. Kawamura, K. Tsuchiya, M. Nakamichi, J. Fujita, H. Sagawa, Y. Nagao, Y. Gohar, Y. Ikejima, T. Saito, S. Sakurai, I. Hisa, H. Kumahara, N. Nemoto, Tritium release behavior from lithium titanate pebbles at low irradiation temperature, Proc. of the 20th Symposium on Fusion Technology, France, vol. 2, 1998, pp. 1289 – 1292. [12] K. Tsuchiya, M. Nakamichi, Y. Nagao, J. Fujita, H. Sagawa, S. Tanaka and H. Kawamura, Integrated experiment of blanket in-pile mockup with Li2TiO3 pebbles, Fusion Eng. Des. (in press). [13] K. Furuya, M. Enoeda, T. Hatano, et al., Application of HIP bonding to first wall panel fabrication made from reduced activation ferritic steel F82H, J. Nucl. Mater. 258 – 263 (1998) 2023–2029. [14] S. Mori, S. Yamazaki, J. Adachi, et al., Blanket and divertor design for the Steady State Tokamak Reactor (SSTR), Fus. Eng. Des. 18 (1991) 249–258. [15] M. Oda et al., Development of HIP bonding procedure and mechanical properties of HIP bonded joints for reduced activation ferritic steel F82H, JAERI-Tech 97-013 (1997). [16] M. Enoeda, et al., Fus. Technol. 34 (1998) 877 – 881. [17] S. Sato et al., Proceedings of 20th SOFT, Marseille, France, 1998. [18] M. Taniguchi, et al., In-situ observation of surface hydroxyl group on lithium ceramics at high temperature by infrared absorption spectroscopy, Fus. Technol. 28 (1995)
.
307
1284 – 1289. [19] M. Taniguchi, S. Tanaka, Ab-initio Hartree – Fock study on surface desorption process in tritium release, J. Nucl. Mater. 258 – 263 (1998) 531 – 536. [20] S. Tanaka, M. Taniguchi, H. Tanigawa XPS and UPS studies on electronic structure of Li2O, J. Nucl. Mater. (in press). [21] A. Suzuki, M. Tonegawa, M. Yasumoto, M. Yamawaki, Sweep gas chemistry effect on vaporization property of Li2ZrO3, J. Nucl. Mater. 258 – 263 (1998) 562 – 565. [22] A. Suzuki, K. Yamaguchi, M. Yamawaki, Study on the surface nonstoichiometry of solid breeder blanket materials by means of work function measurement, Fus. Eng. Des. 39 – 40 (1998) 699 – 705. [23] A. Suzuki, K. Yamaguchi, T. Terai, M. Yamawaki, Application of high temperature and spectrometry and work function measurement to evaluation of thermochemical performance of ceramic breeders, Fus. Eng. Des. (in press). [24] V. Grishmanov, et al., Influence of radiation defects on tritium release parameters from Li2O, Fus. Eng. Des. 39 – 40 (1998) 685 – 691. [25] H. Tanigawa et al., FT-IR study on interaction of hydrogen isotopes with defects in lithium oxide, Fusion Eng. Des. (in press). [26] M. Nishsikawa, A. Baba, S. Odoi, Y. Kawamura, Tritium inventory estimation in solid blanket system, Fus. Eng. Des. 39 – 40 (1998) 615 – 625. [27] T. Terai, et al., Compatibility of structural materials with Li2BeF4 molten salt breeder, J. Nucl. Mater. 258–263 (1998) 513. [28] T. Terai et al., Corrosion behavior of molten LiF –BeF2 mixture, Fusion Eng. Des. [29] A. Suzuki, et al., In-situ HT release behavior from molten Li2BeF4 salt, Fus. Eng. Des. 39 – 40 (1998) 781 – 785. [30] T. Mitsuyama, et al., Compatibility of insulating ceramic materials with liquid breeders, Fus. Eng. Des. 39–40 (1998) 811 – 817. [31] T. Terai et al., Fabrication and property of ceramic coating for CTR liquid blanket by sputtering method, Fusion Eng. Des. (in press).
.