- Email: [email protected]
Chemical form of released tritium from solid breeder materials under the various purge gas conditions
Fusion Engineering and Design 82 (2007) 2147–2151
Chemical form of released tritium from solid breeder materials under the various purge gas conditions T. Kinjyo a,∗ , M. Nishikawa a , N. Yamashita a , T. Koyama a , T. Tanifuji b , M. Enoeda c a
Graduate School of Engineering Science, Kyushu University, Fukuoka 812-8581, Japan b Tokai Establishment, JAEA, Ibaraki 319-1195, Japan c Naka Establishment, JAEA, Ibaraki 319-1195, Japan Received 31 July 2006; received in revised form 2 July 2007; accepted 2 July 2007 Available online 30 August 2007
Abstract Understanding of the release behavior of bred tritium from solid breeder materials is necessary to design tritium recovery system from blanket of a fusion reactor because permeation loss of bred tritium in the piping system or type of tritium recovery system depends on the tritium release behavior. It has been reported by the present authors that behavior of tritium release from solid breeder grain is consisted of diffusion in grain, tritium transfer at surface layer and surface reactions on grain surface such as adsorption or isotope exchange reactions. Chemical form of released tritium from Li4 SiO4 (from FzK), LiAlO2 (from JAERI), Li2 TiO3 (from CEA) and Li2 ZrO3 (from MAPI) under various purge gas condition is discussed in this study by using the data obtained from the out-of-pile tritium release experiment in JAEA and fitting results estimated by the tritium release model formed by the present authors. And then the tritium release behavior and chemical form of tritium in the test blanket module with solid breeder under the ITER condition is also discussed based on the estimation obtained using the tritium release model. © 2007 Elsevier B.V. All rights reserved. Keywords: Tritium release model; Chemical form of released tritium; Surface reactions
1. Introduction It has been pointed out by the present authors that not only diffusion of tritium in bulk of grain but also tritium transfer at surface layer and surface reactions ∗
Corresponding author. E-mail address: [email protected] (T. Kinjyo).
0920-3796/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.fusengdes.2007.07.002
on grain surface give profound effects to the release behavior of bred tritium [1]. It is assumed in the tritium release model reported [1] that the bred tritium in crystal grain arrives at surface layer through diffusion and that tritium is transferred through the interfacial layer to the surface water on grain which consists of physical adsorbed water, chemical adsorbed water and structural water. Tritium on grain surface is liberated to
2148
T. Kinjyo et al. / Fusion Engineering and Design 82 (2007) 2147–2151
the purge gas from the surface water through such surface reactions as desorption, isotope exchange reaction between molecular hydrogen in purge gas and tritiated water on grain surface, and isotope exchange reaction between water molecular in purge gas and tritiated water on grain surface. Tritium release curves estimated by the tritium release model gave good agreements with experimental curves from Li4 SiO4 , Li2 TiO3 , Li2 ZrO3 and LiAlO2 under humid purge gas condition [2]. To analyze effects given by the surface reactions on tritium release behavior and chemical form of released tritium from solid breeder materials, tritium release experiments were carried out under different purge gas compositions (dry N2 , N2 with H2 , N2 with H2 O), and the release behavior was analyzed using the out-of-pile temperature programmed desorption techniques in the Japan Atomic Energy Agency (JAEA) by the present authors [3–6].
2. Experiment
experimental methods were explained in the previous paper [6,7]. 3. Results and discussion Fig. 1 shows comparison of experimental tritium release curves obtained from Li4 SiO4 (FzK, grain diameter is 1 m) under the dry purge gas condition, dry purge gas with hydrogen condition and humid purge gas condition. It is confirmed that tritium release behavior from solid breeder grains is composed with not only diffusion in bulk of grain but also surface reactions because tritium release behavior are different with purge gas compositions. The effect of surface reactions is the smallest under the humid purge gas condition because the tritium release under the humid purge gas condition is much faster than those obtained under any other purge gas conditions. Fig. 2 shows comparison of estimated tritium release curves with experimental curves obtained under
2.1. Drying operation before neutron irradiation The Li4 SiO4 (Forschungszentrum Karlsruhe, FzK), Li2 TiO3 (France Atomic Energy Commission, CEA and Kawasaki Heavy Industry, KHI), Li2 ZrO3 (Mitsubishi Atomic Power Industry, MAPI (now Mitsubishi Heavy Industry, MHI)) or LiAlO2 (Japan Atomic Energy Agency, JAEA) pebbles located in quartz tube were heated from room temperature to 800 ◦ C and the temperature was kept at 800 ◦ C for 8 h in dry He atmosphere. 2.2. Neutron irradiation
Fig. 1. Tritium release from Li4 SiO4 under various purge gas conditions.
Pebbles of Li4 SiO4 , Li2 TiO3 , Li2 ZrO3 or LiAlO2 were irradiated by the thermal neutron at the Japan Research Reactor 4 (JRR-4) in JAEA under the conditions of He atmosphere. The neutron flux was 4.0E13 cm−2 s−1 and irradiation time was 100 min. 2.3. Tritium release experiment Release curves of bred tritium from pebbles of breeder material pebbles were obtained applying the out-of-pile temperature programmed desorption (TPD) techniques. Details of experimental apparatus and
Fig. 2. Tritium release from Li4 SiO4 under humid purge gas condition.
T. Kinjyo et al. / Fusion Engineering and Design 82 (2007) 2147–2151
1000 Pa H2 O/N2 , 100 Pa H2 O/N2 or 10 Pa H2 O/N2 purge gas conditions, respectively. The tritium transfer coefficients of surface reactions obtained experimentally by the present authors previously [8–17], effective tritium diffusivity in bulk of grain [6] and tritium transfer coefficient at interfacial layer [1,2] are used in this estimation. The estimated curves under 1000 Pa H2 O/N2 or 100 Pa H2 O/N2 purge gas conditions give good agreement with the experimental curve, but the estimated curve under 10 Pa H2 O/N2 purge gas differs from the experiment curve. It is difficult to handle and measure a few amount of water vapor, especially lower than 10 Pa, in the purge gas, therefore water vapor lager than 10 Pa might exist in the 10 Pa H2 O/N2 purge gas. By the same reason, it is considered that a few amount of water vapor might exist in the purge gas used as “Dry purge gas”. In this study the partial pressure of water vapor in “Dry purge gas” is estimated by the curve fitting method. Estimated tritium release curves with experimental curves from various solid breeder materials under dry purge gas condition are shown in Fig. 3. In this estimation, the partial pressure of water vapor in purge gas is regarded as the adjustable parameter and the estimated curves give good agreement with experimental curves if several to several Pa of water vapor is assumed in the dry purge gas. Fig. 4 shows tritium release curve from Li4 SiO4 when 1 Pa of water vapor is considered in the dry purge gas with 100 Pa of hydrogen condition. Even though hydrogen is added to the dry purge gas to recover tritium as HT form, considerable amount of tritium is
Fig. 3. Tritium release from solid breeder materials under dry purge gas conditions.
2149
Fig. 4. Tritium release from Li4 SiO4 purged by dry gas containing 1000 Pa of H2 and 1 Pa of H2 O.
released as HTO form. Tritium on the grain surface is released to the gas phase by competition of surface reactions such as adsorption/desorption, isotope exchange reaction between tritium on grain surface and hydrogen in gas phase (isotope exchange reaction 1) or isotope exchange reaction between tritium on grain surface and water vapor in gas phase (isotope exchange reaction 2). HTO desorption from grain surface and isotope exchange reaction 2 with small amount of water vapor existing initially in the purge gas increase tritium release as HTO form. Plots of triangle, , or circle, , in Fig. 4 represent estimated tritium release as hydrogen form (HT) or whole amount of released tritium (HT + HTO), respectively. In this estimation, partial pressure of water vapor in the purge gas is assumed to be 1 Pa. The estimated curves assuming purge gas with 1 Pa of water vapor condition do not give good agreement with the experimental curve. The estimation assuming the purge gas with 10 Pa water vapor condition is shown in Fig. 5.
Fig. 5. Tritium release from Li4 SiO4 purged by dry gas containing 1000 Pa of H2 and 10 Pa of H2 O.
2150
T. Kinjyo et al. / Fusion Engineering and Design 82 (2007) 2147–2151
Fig. 6. Tritium release from Li2 TiO3 purged by dry gas containing 1000 Pa of H2 and 5 Pa of H2 O.
The estimated curves give much better agreement with the experimental curves when 10 Pa of water vapor is assumed to exist in the purge gas. Fig. 6 shows comparison of estimated tritium release curves with experimental curves from Li2 TiO3 under dry purge gas with 1000 Pa of hydrogen condition. The estimated curves give good agreement with experimental curves when 5 Pa of water vapor is assumed to exist in the purge gas. The tritium release curves estimated by the tritium release model gave good agreement with experimental curves from various solid breeder materials under humid purge gas conditions [2], and give good agreement with experimental curve under the dry purge gas condition when existence of some water vapor is considered in the “dry” purge gas. It is confirmed that the tritium release model developed by the present authors can estimate the tritium release behavior from solid breeder materials under dry purge gas, dry purge gas with hydrogen or humid purge gas condition. For application, tritium release behavior from Li2 TiO3 (CEA, grain diameter 1 m) under the ITER test blanket module (ITER-TBM) condition advanced by JAEA is simulated as shown in Fig. 7. Detail of simplified TBM model and conditions such as flow rate of purge gas, amount of packed Li2 TiO3 , temperature changing of solid breeder materials or neutron shots was explained in the previous paper [18]. Curves (1), (2) and (3) show tritium inventory in bulk, interfacial layer or surface water, respectively. Curve (4) shows tritium released to purge gas. First neutron shot starts from 600 to 1000 s and second one starts from 2400 to 2800 s.
Fig. 7. Estimation of tritium inventory in ITER test blanket module.
The tritium inventory in bulk shown by curve (1) is much smaller than inventory in interfacial layer, curve (2), and inventory in surface water, curve (3) under the ITER-TBM condition. Tritium diffusivity in bulk of grain does not affect so much to the inventory in solid breeder materials but tritium transfer at interfacial layer and surface reactions on surface water give large effect to the tritium inventory in solid breeder materials. Decrease of surface area of solid breeder materials by increasing of grain diameter might reduce tritium inventory. Optimisation of grain diameter is necessary to reduce and evaluate tritium inventory in solid breeder materials under various conditions. Fig. 8 shows simulation for chemical form of released tritium under ITER-TBM condition. Even though 100 Pa of hydrogen is added to the purge gas, more than half of tritium might be released as HTO form even when no water vapor in the purge gas is considered in this estimation at the inlet of TBM. A about
Fig. 8. Estimation of tritium release behavior from ITER test blanket module.
T. Kinjyo et al. / Fusion Engineering and Design 82 (2007) 2147–2151
80 Pa of water vapor is generated by water formation reaction [14] with decreasing the partial pressure of hydrogen to 20 Pa at the end of TBM. This is because tritium on the grain surface is released to the purge gas through desorption or isotope exchange reaction with water vapor generated by the water formation reaction in the purge gas. Under the ITER-TBM condition 0.18 kg of water vapor is generated by the water formation reaction from 19.9 kg of Li2 TiO3 made by CEA and 80 Pa of water vapor generated might continue to add to purge gas for 987 h (1974 times of neutron shots). The maximum tritium concentration in the purge gas under the ITER-TBM condition is set to be 1 Pa and it is much smaller than the concentration of water vapor in the purge gas. If it is preferred to recover tritium as the hydrogen form the suitable recovery system must be considered which can deal with HT and HTO. 4. Conclusion It is confirmed that not only diffusion of tritium in bulk of grain but also tritium transfer at interfacial layer and surface reactions on grain surface give profound effects to the release behavior of bred tritium. The tritium release behavior under various purge gas compositions is estimated in this study using the model considered by the present authors and the estimation results give good agreement with experiment data. Water vapor in purge gas affects the chemical form of released tritium if a small amount of tritium is handled. Even though hydrogen is added to the purge gas to recover tritium as hydrogen form, some of tritium might be released as water form affected by the water vapor that initially exists in purge gas, and by the water generated by the water formation reaction. References [1] T. Kinjyo, M. Nishikawa, Release behavior of bred tritium from irradiated Li4 SiO4 , Fusion Sci. Technol. 48 (2005) 646–649.
2151
[2] T. Kinjyo, M. Nishikawa, T. Tanifuji, M. Enoeda, Introduction of tritium release step at surface layer of breeder grain for modeling of tritium release behavior from solid breeder materials, Fusion Eng. Des. 81 (2006) 573–577. [3] M. Nishikawa, T. Kinjyo, Y. Nishida, Chemical form of tritium released from solid breeder materials, J. Nucl. Mater. 325 (2004) 87–93. [4] M. Nishikawa, T. Kinjyo, T. Ishizaka, S. Beloglazov, T. Takeishi, M. Enoeda, T. Tanifuji, Release behavior of bred tritium from LiAlO2 , J. Nucl. Mater. 335 (2004) 70–76. [5] S. Beloglazov, M. Nishikawa, T. Tanifuji, Modelling of tritium release from irradiated Li2 ZrO3 , Fusion Sci. Technol. 41 (2002) 1049–1053. [6] T. Kinjyo, M. Nishikawa, Tritium release behavior from Li4 SiO4 , Fusion Sci. Technol. 46 (2004) 561–570. [7] M. Nishikawa, T. Takeishi, Y. Matsumoto, I. Kumabe, Ionization chamber system to eliminate the memory effect of tritium, Nuclear Inst. Meth. Phys. Res. A278 (1989) 525–531. [8] M. Nishikawa, A. Baba, Y. Kawamura, Tritium inventory in LiAlO2 blanket, J. Nucl. Mater. 246 (1997) 1–8. [9] M. Nishikawa, A. Baba, Tritium inventory in Li2 ZrO3 blanket, J. Nucl. Mater. 257 (1998) 162–171. [10] Y. Kawamura, M. Nishikawa, K. Tanaka, H. Matsumoto, Adsorption characteristics of water vapor on GammaAluminate, J. Nucl. Sci. Technol. 29 (1992) 436–444. [11] Y. Kawamura, M. Nishikawa, Adsorption characteristics of water vapor on Li2 ZrO3 , J. Nucl. Mater. 218 (1996) 57–65. [12] Y. Kawamura, M. Nishikawa, K. Tanaka, Adsorption characteristics of water vapor on Li4 SiO4 , J. Nucl. Mater. 230 (1996) 308–312. [13] Y. Kawamura, M. Nishikawa, Adsorption and desorption rate of water on the various ceramic breeder materials, Fusion Technol. 27 (1995) 25–35. [14] K. Hashimoto, M. Nishikawa, N. Nakashima, S. Beloglazov, M. Enoeda, Tritium inventory in Li2 TiO3 blanket, Fusion. Eng. Des. 61/62 (2002) 375–381. [15] A. Baba, M. Nishikawa, T. Eguchi, T. Kawagoe, Isotope exchange reaction on solid breeder materials, Fusion. Eng. Des. 49/50 (2000) 483–489. [16] T. Kawagoe, M. Nishikawa, A. Baba, S. Beloglazov, Surface inventory of tritium on Li2 TiO3 , J. Nucl. Mater. 297 (2001) 27–34. [17] M. Nishikawa, N. Nakashima, K. Hashimoto, S. Beloglazov, Isotope exchange capacity on Li4 SiO4 and comparison of tritium inventory in various solid breeder blankets, J. Nucl. Sci. Technol. 38 (2001) 944–951. [18] T. Kinjyo, M. Nishikawa, M. Enoeda, Estimation of tritium release behavior from solid breeder materials under the condition of ITER test blanket module, J. Nucl. Mater. 367-370 (2007) 1361–1365.
Chemical form of released tritium from solid breeder materials under the various purge gas conditions T. Kinjyo a,∗ , M. Nishikawa a , N. Yamashita a , T. Koyama a , T. Tanifuji b , M. Enoeda c a
Graduate School of Engineering Science, Kyushu University, Fukuoka 812-8581, Japan b Tokai Establishment, JAEA, Ibaraki 319-1195, Japan c Naka Establishment, JAEA, Ibaraki 319-1195, Japan Received 31 July 2006; received in revised form 2 July 2007; accepted 2 July 2007 Available online 30 August 2007
Abstract Understanding of the release behavior of bred tritium from solid breeder materials is necessary to design tritium recovery system from blanket of a fusion reactor because permeation loss of bred tritium in the piping system or type of tritium recovery system depends on the tritium release behavior. It has been reported by the present authors that behavior of tritium release from solid breeder grain is consisted of diffusion in grain, tritium transfer at surface layer and surface reactions on grain surface such as adsorption or isotope exchange reactions. Chemical form of released tritium from Li4 SiO4 (from FzK), LiAlO2 (from JAERI), Li2 TiO3 (from CEA) and Li2 ZrO3 (from MAPI) under various purge gas condition is discussed in this study by using the data obtained from the out-of-pile tritium release experiment in JAEA and fitting results estimated by the tritium release model formed by the present authors. And then the tritium release behavior and chemical form of tritium in the test blanket module with solid breeder under the ITER condition is also discussed based on the estimation obtained using the tritium release model. © 2007 Elsevier B.V. All rights reserved. Keywords: Tritium release model; Chemical form of released tritium; Surface reactions
1. Introduction It has been pointed out by the present authors that not only diffusion of tritium in bulk of grain but also tritium transfer at surface layer and surface reactions ∗
Corresponding author. E-mail address: [email protected] (T. Kinjyo).
0920-3796/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.fusengdes.2007.07.002
on grain surface give profound effects to the release behavior of bred tritium [1]. It is assumed in the tritium release model reported [1] that the bred tritium in crystal grain arrives at surface layer through diffusion and that tritium is transferred through the interfacial layer to the surface water on grain which consists of physical adsorbed water, chemical adsorbed water and structural water. Tritium on grain surface is liberated to
2148
T. Kinjyo et al. / Fusion Engineering and Design 82 (2007) 2147–2151
the purge gas from the surface water through such surface reactions as desorption, isotope exchange reaction between molecular hydrogen in purge gas and tritiated water on grain surface, and isotope exchange reaction between water molecular in purge gas and tritiated water on grain surface. Tritium release curves estimated by the tritium release model gave good agreements with experimental curves from Li4 SiO4 , Li2 TiO3 , Li2 ZrO3 and LiAlO2 under humid purge gas condition [2]. To analyze effects given by the surface reactions on tritium release behavior and chemical form of released tritium from solid breeder materials, tritium release experiments were carried out under different purge gas compositions (dry N2 , N2 with H2 , N2 with H2 O), and the release behavior was analyzed using the out-of-pile temperature programmed desorption techniques in the Japan Atomic Energy Agency (JAEA) by the present authors [3–6].
2. Experiment
experimental methods were explained in the previous paper [6,7]. 3. Results and discussion Fig. 1 shows comparison of experimental tritium release curves obtained from Li4 SiO4 (FzK, grain diameter is 1 m) under the dry purge gas condition, dry purge gas with hydrogen condition and humid purge gas condition. It is confirmed that tritium release behavior from solid breeder grains is composed with not only diffusion in bulk of grain but also surface reactions because tritium release behavior are different with purge gas compositions. The effect of surface reactions is the smallest under the humid purge gas condition because the tritium release under the humid purge gas condition is much faster than those obtained under any other purge gas conditions. Fig. 2 shows comparison of estimated tritium release curves with experimental curves obtained under
2.1. Drying operation before neutron irradiation The Li4 SiO4 (Forschungszentrum Karlsruhe, FzK), Li2 TiO3 (France Atomic Energy Commission, CEA and Kawasaki Heavy Industry, KHI), Li2 ZrO3 (Mitsubishi Atomic Power Industry, MAPI (now Mitsubishi Heavy Industry, MHI)) or LiAlO2 (Japan Atomic Energy Agency, JAEA) pebbles located in quartz tube were heated from room temperature to 800 ◦ C and the temperature was kept at 800 ◦ C for 8 h in dry He atmosphere. 2.2. Neutron irradiation
Fig. 1. Tritium release from Li4 SiO4 under various purge gas conditions.
Pebbles of Li4 SiO4 , Li2 TiO3 , Li2 ZrO3 or LiAlO2 were irradiated by the thermal neutron at the Japan Research Reactor 4 (JRR-4) in JAEA under the conditions of He atmosphere. The neutron flux was 4.0E13 cm−2 s−1 and irradiation time was 100 min. 2.3. Tritium release experiment Release curves of bred tritium from pebbles of breeder material pebbles were obtained applying the out-of-pile temperature programmed desorption (TPD) techniques. Details of experimental apparatus and
Fig. 2. Tritium release from Li4 SiO4 under humid purge gas condition.
T. Kinjyo et al. / Fusion Engineering and Design 82 (2007) 2147–2151
1000 Pa H2 O/N2 , 100 Pa H2 O/N2 or 10 Pa H2 O/N2 purge gas conditions, respectively. The tritium transfer coefficients of surface reactions obtained experimentally by the present authors previously [8–17], effective tritium diffusivity in bulk of grain [6] and tritium transfer coefficient at interfacial layer [1,2] are used in this estimation. The estimated curves under 1000 Pa H2 O/N2 or 100 Pa H2 O/N2 purge gas conditions give good agreement with the experimental curve, but the estimated curve under 10 Pa H2 O/N2 purge gas differs from the experiment curve. It is difficult to handle and measure a few amount of water vapor, especially lower than 10 Pa, in the purge gas, therefore water vapor lager than 10 Pa might exist in the 10 Pa H2 O/N2 purge gas. By the same reason, it is considered that a few amount of water vapor might exist in the purge gas used as “Dry purge gas”. In this study the partial pressure of water vapor in “Dry purge gas” is estimated by the curve fitting method. Estimated tritium release curves with experimental curves from various solid breeder materials under dry purge gas condition are shown in Fig. 3. In this estimation, the partial pressure of water vapor in purge gas is regarded as the adjustable parameter and the estimated curves give good agreement with experimental curves if several to several Pa of water vapor is assumed in the dry purge gas. Fig. 4 shows tritium release curve from Li4 SiO4 when 1 Pa of water vapor is considered in the dry purge gas with 100 Pa of hydrogen condition. Even though hydrogen is added to the dry purge gas to recover tritium as HT form, considerable amount of tritium is
Fig. 3. Tritium release from solid breeder materials under dry purge gas conditions.
2149
Fig. 4. Tritium release from Li4 SiO4 purged by dry gas containing 1000 Pa of H2 and 1 Pa of H2 O.
released as HTO form. Tritium on the grain surface is released to the gas phase by competition of surface reactions such as adsorption/desorption, isotope exchange reaction between tritium on grain surface and hydrogen in gas phase (isotope exchange reaction 1) or isotope exchange reaction between tritium on grain surface and water vapor in gas phase (isotope exchange reaction 2). HTO desorption from grain surface and isotope exchange reaction 2 with small amount of water vapor existing initially in the purge gas increase tritium release as HTO form. Plots of triangle, , or circle, , in Fig. 4 represent estimated tritium release as hydrogen form (HT) or whole amount of released tritium (HT + HTO), respectively. In this estimation, partial pressure of water vapor in the purge gas is assumed to be 1 Pa. The estimated curves assuming purge gas with 1 Pa of water vapor condition do not give good agreement with the experimental curve. The estimation assuming the purge gas with 10 Pa water vapor condition is shown in Fig. 5.
Fig. 5. Tritium release from Li4 SiO4 purged by dry gas containing 1000 Pa of H2 and 10 Pa of H2 O.
2150
T. Kinjyo et al. / Fusion Engineering and Design 82 (2007) 2147–2151
Fig. 6. Tritium release from Li2 TiO3 purged by dry gas containing 1000 Pa of H2 and 5 Pa of H2 O.
The estimated curves give much better agreement with the experimental curves when 10 Pa of water vapor is assumed to exist in the purge gas. Fig. 6 shows comparison of estimated tritium release curves with experimental curves from Li2 TiO3 under dry purge gas with 1000 Pa of hydrogen condition. The estimated curves give good agreement with experimental curves when 5 Pa of water vapor is assumed to exist in the purge gas. The tritium release curves estimated by the tritium release model gave good agreement with experimental curves from various solid breeder materials under humid purge gas conditions [2], and give good agreement with experimental curve under the dry purge gas condition when existence of some water vapor is considered in the “dry” purge gas. It is confirmed that the tritium release model developed by the present authors can estimate the tritium release behavior from solid breeder materials under dry purge gas, dry purge gas with hydrogen or humid purge gas condition. For application, tritium release behavior from Li2 TiO3 (CEA, grain diameter 1 m) under the ITER test blanket module (ITER-TBM) condition advanced by JAEA is simulated as shown in Fig. 7. Detail of simplified TBM model and conditions such as flow rate of purge gas, amount of packed Li2 TiO3 , temperature changing of solid breeder materials or neutron shots was explained in the previous paper [18]. Curves (1), (2) and (3) show tritium inventory in bulk, interfacial layer or surface water, respectively. Curve (4) shows tritium released to purge gas. First neutron shot starts from 600 to 1000 s and second one starts from 2400 to 2800 s.
Fig. 7. Estimation of tritium inventory in ITER test blanket module.
The tritium inventory in bulk shown by curve (1) is much smaller than inventory in interfacial layer, curve (2), and inventory in surface water, curve (3) under the ITER-TBM condition. Tritium diffusivity in bulk of grain does not affect so much to the inventory in solid breeder materials but tritium transfer at interfacial layer and surface reactions on surface water give large effect to the tritium inventory in solid breeder materials. Decrease of surface area of solid breeder materials by increasing of grain diameter might reduce tritium inventory. Optimisation of grain diameter is necessary to reduce and evaluate tritium inventory in solid breeder materials under various conditions. Fig. 8 shows simulation for chemical form of released tritium under ITER-TBM condition. Even though 100 Pa of hydrogen is added to the purge gas, more than half of tritium might be released as HTO form even when no water vapor in the purge gas is considered in this estimation at the inlet of TBM. A about
Fig. 8. Estimation of tritium release behavior from ITER test blanket module.
T. Kinjyo et al. / Fusion Engineering and Design 82 (2007) 2147–2151
80 Pa of water vapor is generated by water formation reaction [14] with decreasing the partial pressure of hydrogen to 20 Pa at the end of TBM. This is because tritium on the grain surface is released to the purge gas through desorption or isotope exchange reaction with water vapor generated by the water formation reaction in the purge gas. Under the ITER-TBM condition 0.18 kg of water vapor is generated by the water formation reaction from 19.9 kg of Li2 TiO3 made by CEA and 80 Pa of water vapor generated might continue to add to purge gas for 987 h (1974 times of neutron shots). The maximum tritium concentration in the purge gas under the ITER-TBM condition is set to be 1 Pa and it is much smaller than the concentration of water vapor in the purge gas. If it is preferred to recover tritium as the hydrogen form the suitable recovery system must be considered which can deal with HT and HTO. 4. Conclusion It is confirmed that not only diffusion of tritium in bulk of grain but also tritium transfer at interfacial layer and surface reactions on grain surface give profound effects to the release behavior of bred tritium. The tritium release behavior under various purge gas compositions is estimated in this study using the model considered by the present authors and the estimation results give good agreement with experiment data. Water vapor in purge gas affects the chemical form of released tritium if a small amount of tritium is handled. Even though hydrogen is added to the purge gas to recover tritium as hydrogen form, some of tritium might be released as water form affected by the water vapor that initially exists in purge gas, and by the water generated by the water formation reaction. References [1] T. Kinjyo, M. Nishikawa, Release behavior of bred tritium from irradiated Li4 SiO4 , Fusion Sci. Technol. 48 (2005) 646–649.
2151
[2] T. Kinjyo, M. Nishikawa, T. Tanifuji, M. Enoeda, Introduction of tritium release step at surface layer of breeder grain for modeling of tritium release behavior from solid breeder materials, Fusion Eng. Des. 81 (2006) 573–577. [3] M. Nishikawa, T. Kinjyo, Y. Nishida, Chemical form of tritium released from solid breeder materials, J. Nucl. Mater. 325 (2004) 87–93. [4] M. Nishikawa, T. Kinjyo, T. Ishizaka, S. Beloglazov, T. Takeishi, M. Enoeda, T. Tanifuji, Release behavior of bred tritium from LiAlO2 , J. Nucl. Mater. 335 (2004) 70–76. [5] S. Beloglazov, M. Nishikawa, T. Tanifuji, Modelling of tritium release from irradiated Li2 ZrO3 , Fusion Sci. Technol. 41 (2002) 1049–1053. [6] T. Kinjyo, M. Nishikawa, Tritium release behavior from Li4 SiO4 , Fusion Sci. Technol. 46 (2004) 561–570. [7] M. Nishikawa, T. Takeishi, Y. Matsumoto, I. Kumabe, Ionization chamber system to eliminate the memory effect of tritium, Nuclear Inst. Meth. Phys. Res. A278 (1989) 525–531. [8] M. Nishikawa, A. Baba, Y. Kawamura, Tritium inventory in LiAlO2 blanket, J. Nucl. Mater. 246 (1997) 1–8. [9] M. Nishikawa, A. Baba, Tritium inventory in Li2 ZrO3 blanket, J. Nucl. Mater. 257 (1998) 162–171. [10] Y. Kawamura, M. Nishikawa, K. Tanaka, H. Matsumoto, Adsorption characteristics of water vapor on GammaAluminate, J. Nucl. Sci. Technol. 29 (1992) 436–444. [11] Y. Kawamura, M. Nishikawa, Adsorption characteristics of water vapor on Li2 ZrO3 , J. Nucl. Mater. 218 (1996) 57–65. [12] Y. Kawamura, M. Nishikawa, K. Tanaka, Adsorption characteristics of water vapor on Li4 SiO4 , J. Nucl. Mater. 230 (1996) 308–312. [13] Y. Kawamura, M. Nishikawa, Adsorption and desorption rate of water on the various ceramic breeder materials, Fusion Technol. 27 (1995) 25–35. [14] K. Hashimoto, M. Nishikawa, N. Nakashima, S. Beloglazov, M. Enoeda, Tritium inventory in Li2 TiO3 blanket, Fusion. Eng. Des. 61/62 (2002) 375–381. [15] A. Baba, M. Nishikawa, T. Eguchi, T. Kawagoe, Isotope exchange reaction on solid breeder materials, Fusion. Eng. Des. 49/50 (2000) 483–489. [16] T. Kawagoe, M. Nishikawa, A. Baba, S. Beloglazov, Surface inventory of tritium on Li2 TiO3 , J. Nucl. Mater. 297 (2001) 27–34. [17] M. Nishikawa, N. Nakashima, K. Hashimoto, S. Beloglazov, Isotope exchange capacity on Li4 SiO4 and comparison of tritium inventory in various solid breeder blankets, J. Nucl. Sci. Technol. 38 (2001) 944–951. [18] T. Kinjyo, M. Nishikawa, M. Enoeda, Estimation of tritium release behavior from solid breeder materials under the condition of ITER test blanket module, J. Nucl. Mater. 367-370 (2007) 1361–1365.