- Email: [email protected]
Lifus (lithium for fusion) 6 loop design and construction
Fusion Engineering and Design 88 (2013) 769–773
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Lifus (lithium for fusion) 6 loop design and construction A. Aiello ∗ , A. Tincani, P. Favuzza, F.S. Nitti 1 , L. Sansone, G. Miccichè, M. Muzzarelli, G. Fasano, P. Agostini ENEA C.R. Brasimone, 40032 Camugnano (BO), Italy
h i g h l i g h t s The design of a lithium loop to perform corrosion tests is presented. General design parameters are 350 ◦ C, 16 m/s in the test section, isothermal configuration. The chemistry of lithium, which affects the ferrous materials compatibility with lithium, is controlled with a double trapping system: cold trap continuously operated and hot trap in the storage tank. An online monitoring of impurities is realized by mean of a resistivity meter. An offline monitoring of nitrogen content is made using sampling of lithium and successive ammonia titration technique.
a r t i c l e
i n f o
Article history: Available online 12 April 2013 Keywords: IFMIF Lithium corrosion Lithium chemistry
a b s t r a c t The Lifus 3 loop was designed and constructed in 2001/2002 in ENEA, to evaluate the corrosion behaviour of Fusion reference structural materials, and in particular AISI 316 and Eurofer, in flowing lithium environment. To fully accomplish the experimental program originally foreseen for Lifus 3 and not respected it was decided in 2010 to build a new experimental facility, simplified from the hydraulic point of view, capable of performing the set of tests required by the IFMIF community. The operating parameters of this loop, named Lifus 6 [1], are: isothermal loop with operating temperature of 350 ◦ C, maximum velocity in the test section of about 16 m/s, cold trap continuously operated to remove carbon and oxygen impurities, hot trap operated in batch to remove nitrogen ones. An integral measurement of impurities is made on line by using a resistivity meter. The Lifus 6 loop design is presented in the paper. © 2013 Elsevier B.V. All rights reserved.
1. Introduction The Lifus 3 loop was designed and constructed in 2001/2002 in ENEA, to perform compatibility tests of IFMIF reference materials in IFMIF relevant conditions. The original mission was to examine how the corrosion behaviour of the selected materials is affected by the impurities. The operation of the loop was affected by several problems, coming in particular from the purification system and the hydraulic control, and the original commitments were not fully accomplished. For this reason it was decided in 2010 to build a new experimental facility, simplified from the hydraulic point of view, capable to perform the set of tests required by the IFMIF community. A series of erosion corrosion tests, having different duration, will be performed in the LIFUS 6 test loop (see Fig. 1) fixing the basic parameters:
∗ Corresponding author. Tel.: +39 0534801380. E-mail address: [email protected] (A. Aiello). 1 Temporarily Seconded to F4E Garching, Germany. 0920-3796/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fusengdes.2013.02.129
• • • •
temperature of 350 ◦ C; Li flow velocity up to 16 m/s; impurity concentration lower than 10 ppm; F82H and Eurofer as specimen materials;
with the goal to validate the choice of material made in the IFMIF reference design. The materials to be tested are mounted in a test section specially designed to achieve the specified Li flow velocities. The Lifus 6 loop, in its operational configuration appears as a ring installed on a steel frame. For safety reasons the loop is not accessible during the operational phases and it is remotely controlled. 1. Drain/storage tank mounted on mobile supports to compensate the thermal expansion of the pipes. It is complete of electric tracing, thermocouples for regulation, safety and measure of the wall temperature, nozzles for connection to the piping. The tank also acts as hot trap at a temperature of 600 ◦ C during the off line purification of lithium from nitrogen impurities. In normal operation the storage lithium is maintained melted at about 300 ◦ C. 2. Electromagnetic pump type GAAA IP 121 provided of air cooling system.
770
A. Aiello et al. / Fusion Engineering and Design 88 (2013) 769–773
According to Fig. 2, the lithium flow at the pump outlet is divided in two parts. The main flow is driven to the test section and come back to the pump. A limited flow, about 1% of the total flow, enters in the dedicated purification loop, where it passes through the cold trap (operated at 200 ◦ C) to be continuously purified, essentially from carbon and oxygen. In the purification branch it is also installed the resistivity meter to perform an integral measurement of impurities. 2. Lifus 6 loop design
Fig. 1. Lifus 6 loop.
3. Two flow meters: a. The first one is a Vortex volumetric type, calibrated for a minimum flow velocity of 0.21 m/s and maximum 7.6 m/s corresponding to 0.37–13.4 m3 /h. b. The second one is a Coriolis mass flow meter, with a range of 0–4000 kg/h of liquid metal. 4. Test section, in which the liquid metal reaches the foreseen testing conditions of velocity.
As mentioned, the plant has a ring structure whose Piping and Instrumentation Diagram (P&ID) is shown in Fig. 2. Purple lines define the heated part of the loop. Two separate circuits can be easily identified: the liquid metal circuit, piping 1 in. and the gas lines made entirely with Swagelok pipes ½ in. The liquid metal circuit is characterized by the following components. Electromagnetic pump model IP 121 manufactured by the GAAA Company. The best operational point is reached at 12 m3 /h of volumetric flow rate with a head of 2.5 bar; Pneumatic Control Valve 1 ; Vortex Flowmeter, Rosemount model 8800; Test section, in which are reached the test conditions of fluid velocity; Coriolis mass flowmeter, Emerson Micro Motion DT65; Pressure gauges Gefran FO53732 in direct contact with liquid metal; Piping in AISI 316 L 1 Schedule 40; Storage tank. The main purpose of the gas circuit is to maintain safety conditions during operation and plant stops, and therefore the strong
Fig. 2. P&ID of Lifus 6 loop.
A. Aiello et al. / Fusion Engineering and Design 88 (2013) 769–773
771
Fig. 4. Pressure drops versus volumetric flow rate. Fig. 3. Stress configuration at design temperature.
In which: control of oxygen and water in all possible loop conditions and in particular in loading and drainage phases. The liquid metal circuit is electrically heated and thermal insulated to guarantee a constant and uniform temperature of 350 ◦ C during operation. The system is entirely electrically traced to ensure the operating temperature and prevent freezing of the liquid metal in the circuit in any possible accidental and ordinary condition. To reduce as much as possible thermal losses the loop is thermal insulated. The thermal insulation is designed to maintain a maximum surface temperature lower than 50 ◦ C. The reference design parameters of the loop are: T = 400 ◦ C (650 ◦ C for the storage tank)
K = fM
L + Kr + Ka + Kc + KT D
1
fM
= −2 log10
ε 2.51 + 3.7 Re fM
(2)
(3) Colebrook
Pressure drops have been evaluated for different components and the circuit globally. In Fig. 4 the pressure drops versus volumetric flow rate are reported. This theoretical evaluation was necessary to verify the compatibility of the available pump with the circuit. From this analysis the pump is capable to guarantee at least the velocity of 15 m/s in the test section.
P = 0.4 MPa
4. Test section design
The dimensioning of pipes has been made according to ASME B31.3 while the storage tank has been designed applying 97/23/CE – Pressure Equipment Design rules. According to 1 in. pipe diameter, the reference schedule value given by ASME B31.3 is 10, corresponding to a thickness of 2.769 mm. We decided to use only sched 40 pipes, with a thickness 20% larger and adequate safety margin. The same evaluation can be made for the storage tank. Also in this case the thickness is largely higher than the minimum allowed value. The only fixed points of the loop are represented by the pump and the storage tank. Therefore, the pipes are free to expand, and mechanical stresses are negligible. To check in any case the real stress/strain situation an analysis of the longest pipe using Ansys was performed and the stress configuration at design temperature is in Fig. 3. The total elongation is about 23 mm with a stress of 7.21 MPa. This value is very low, one order of magnitude lower than the basic allowable stress given by ASME B31.3, and in any case pipes are free to move thanks to flexible suspensions.
For what concerns the test section, a detail of this is shown in Fig. 5. The specimens have a cylindrical shape and are mounted on a rod fixed in the upper part to a removable centering plate, and supported on top and bottom of specimens’ assembly with centering systems. The centering plate is completely uncoupled with respect to the upper flange, to maintain the possibility to easily remove the upper flange and to use an extraction system to remove the specimen assembly that will be surely stuck due to the presence of lithium. With this solution it is realized a meatus in which the lithium flows at the design speed of at least 15 m/s. The triple anchoring of the support rod prevents the occurrence of vibrations in turbulent
3. Fluid dynamic analysis of the loop The general equation to determine pressure drops in a hydraulic system is: p = K
v2 2
(1) Fig. 5. Test section design with a detail of specimens housing.
772
A. Aiello et al. / Fusion Engineering and Design 88 (2013) 769–773 Table 1 Fitting parameters from the versus x experimental curves. These values refer to the calibration performed by Nottingham University. T (◦ C)
Fig. 6. Sketch of the cold trap (drawing not to scale).
flow conditions. The outer diameter of the specimens is 20 mm, with a central hole of 8 mm, for the rod, and a height of 8 mm. The channel that hosts them has a diameter of 21 mm, with an external diameter corresponding to the one of a 1 in. standard pipe. The upper part of the test section, for a height of about 300 mm, is filled with Ar, in order to prevent any contact between the lithium and the sealing area of the flange. The pressure during operation will be about 2 bar, to compensate the pump head at test section inlet maintaining a constant level of lithium also during the start up of the pump. A double-level sensor is installed in the test section to control any accidental level climbs. The samples can be extracted by draining the system, opening the flange while maintaining the circuit in an inert atmosphere and removing the rod sample holder. This procedure ensures a limited contamination of the circuit and the maximum safety for the operators, reducing downtime for the replacement operations. 5. Purification system The purification of lithium from solved non metals is a key element in the evaluation of the corrosion resistance of iron alloys, since it has been shown how their presence could dramatically increase the corrosion rate exerted by the flowing metal; particularly, N is considered to be the most affecting one, mostly because of the formation of Li ternary nitrides (Li3 FeN2 , . . .) [2,3]. The main components employed in Lifu6 plant for the purification are the cold trap (CT), the resistivity meter (RM) and the hot trap (HT): the first two elements are part of a dedicated purification loop, working in parallel with the main loop, while the third one is linked directly to the main Lifus 6 loop and is coincident with the lithium storage tank, from which the loop is charged and discharged. Only about 1% (∼0.3 L/min) of lithium enters the purification loop, being split by a stacking valve in two branches; after passing through the RM and the CT, this lithium rejoins the larger metal mass into the primary loop. The purpose of the cold trap is to reduce C, O, H and N impurities content inside the flowing metal by precipitating their corresponding binary Li compounds. It consists of a cylindrical AISI container (inner volume ∼ 3 L) filled with an AISI mesh, which acts as an active surface for the nucleation of the solid particles and for their subsequent growth. The working temperature of the trap is 200 ◦ C, while the residence time of the metal results about 10 min, which, borrowing sodium experience, is believed to be a suitable time for an high purification efficiency (≥70% for each passage) [4–10]. The gettering volume is realized by disposing a high number (∼50) of AISI mesh sheets (MESH:60; wire diameter: 0.18 mm; hole length: 0.25 mm), parallel packed inside the body of the trap, according to Fig. 6. The distance between the sheets is ∼3 mm: in this way the AISI packing density results about 300 kg m−3 . Employing all these conditions, residual C, O, H and N is reduced respectively to 2, 7, 63 and 1461 wppm [11]. In order to further reduce nitrogen content beyond the CT limit, the hot trap is employed. This trap is actually filled with a titanium sponge getter (solid grains; d = 2–12 mm; apparent density ∼ 0.55 kg m−3 ; total Ti mass ∼ 12 kg), which is able to adsorb nitrogen from lithium, due to the formation of the quite stable titanium nitride. Clearly, the higher the temperature of the HT, the less
8
10 A/ m 108 B/ m (mol% N)
200
250
300
350
400
450
24.1 5.7
25.7 6.2
27.1 6.5
28.5 6.8
29.6 7.0
30.4 7.2
effective the getter from a thermodynamic point of view. However, the rate at which gettering takes place, which involves the diffusion of a non-metal into the getter, is higher at higher temperatures and surface films have a less of an inhibiting effect. The chosen temperature compromise is 600 ◦ C, at which it is possible to reduce nitrogen level to 2 wppm [12] in a time of some days [13]. The monitoring of the nitrogen content is charged to the resistivity meter (RM) and to the off-line analysis procedure. The RM, schematically depicted in Fig. 2, is a device able to online monitor the electric resistivity () of the flowing lithium. It was developed at the University of Nottingham in collaboration with ENEA Brasimone and relies on the variation of the metal resistivity produced by the dissolved anions [14,15]: the higher the impurities concentration in lithium, the higher the resistivity. A detailed description of the device is in [14]. It was possible to calibrate the RM, finding the linear relation which, at each different temperature, links measured to N concentration (x) (see Table 1). Even if the RM has the advantage of an immediate answer, it anyway suffers from a limited sensitivity, which follows from the minimum detectable significant variation of resistivity, which, with Lifus 6 apparatus, is ∼0.18 n m: the minimum detectable variation of nitrogen concentrations in lithium results 53 wppm at 350 ◦ C. The device can anyway be useful to indicate the reaching of an high impurities threshold, as an alarm indicator. To overcome the limit of the RM and to compare the results, an offline lithium analysis is also employed. It entails to take a lithium sample from the plant using a small metallic container (V = 10 mL) linked to the purification loop by a trivial volume duct. Operating under an Ar atmosphere in a glove box, this Li is transferred into a three necks glass vessel and let to react with water, added slowly and paying attention to the violence of the transformation. According to the reaction: Li3 N + 3H2 O 3LiOH + NH3 all the nitride inside lithium is converted into ammonia. Under an Ar flow, the ammonia is distilled away and then quantitatively collected into a water solution of boric acid in excess (Fig. 7).
Fig. 7. Scheme of the apparatus for the production and distillation of NH3 (MFC: mass flow controller).
A. Aiello et al. / Fusion Engineering and Design 88 (2013) 769–773
The final solution so obtained is analyzed for its NH4 + content by HPLC (ionic chromatography) able to measure down to few wppm of starting nitrogen in lithium and with a relative precision of some %. 6. Conclusion The construction of the loop was completed, coherently with the presented design except for the purification branch. The charging of lithium is foreseen in March 2013. Experimental activities will start at the end of the start up phase, approximately in April 2013. References [1] Lifus 6 Design and Erosion/Corrosion Experimental Program, Borader Approach Report BA D 22RUY8, 2012. [2] K. Natesan, Influence of nonmetallic elements on the compatibility of structural materials with liquid alkali metals, Journal of Nuclear Materials 115 (1983) 251–262. [3] O.K. Chopra, D.L. Smith, Influence of temperature and lithium purity on corrosion of ferrous alloys in a flowing lithium environment, Journal of Nuclear Materials 141–143 (1986) 584–591.
773
[4] B.R. Grundy, Experimental characterization of sodium cold traps and modelling of their behaviour, in: Proc. Int. Conf. on Liq. Met. Tech. in En. Prod, Champion, 1976, p. 650. [5] R.B. Hinze, Chemical Engineering Progress Symposium Series: Nuclear Engineering Part XXI 66 (104) 94–106. [6] J.W. Mausteller, F. Tepper, S.J. Rodgers, Alkali metal handling and systems operating techniques, in: ANS, Gordon & Breach, New York, 1967, pp. 39–76. [7] R.W. Dickinson, in: O.J. Foust (Ed.), Liquid Metals Handbook – Sodium and NaK Supplement. Chapter 6, Purification, 1967, June. [8] K.R. Kim, J.Y. Jeong, K.C. Jeong, S.W. Kwon, S.T. Hwang, Journal of Industrial and Engineering Chemistry (Seoul) 4 (1998) 113–121. [9] M.M. Osterhout, Control of Oxygen, Hydrogen and Tritium in Sodium Systems at Experimental Breeder Reactor II – UAC – 41069. [10] W.H. Bruggeman, Purity control in sodium-cooled reactor systems, AIChE Journal 2 (2) (1978) 153. [11] M.G. Down, Proc. 2nd Int. Conf. Liquid Metals Technology and Energy Production, vol. 2, Richland, WA, 1980, pp. 14–16. [12] P. Hubberstey, Proc. Int. Conf. on Liquid Metal Eng. and Tech., vol. 2, BNES, London, 1984, p. 85. [13] A. Anttila, J. Raisanen, J. Keinonen, Diffusion of nitrogen in ␣Ti, Applied Physics Letters 42 (1983), http://dx.doi.org/10.1063/1.93981. [14] A.S. Baley, D.H. Gregory, P. Hubberstey, Development of a Monitoring System, Technical Note No. 4, School of Chemistry, University of Nottingham, 2004. [15] G.K. Creffrey, M.G. Down, R.J. Pulham, Electrical resistivity of liquid and solid lithium, Journal of the Chemical Society, Dalton Transactions (21) (1974) 2325–2329.
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Lifus (lithium for fusion) 6 loop design and construction A. Aiello ∗ , A. Tincani, P. Favuzza, F.S. Nitti 1 , L. Sansone, G. Miccichè, M. Muzzarelli, G. Fasano, P. Agostini ENEA C.R. Brasimone, 40032 Camugnano (BO), Italy
h i g h l i g h t s The design of a lithium loop to perform corrosion tests is presented. General design parameters are 350 ◦ C, 16 m/s in the test section, isothermal configuration. The chemistry of lithium, which affects the ferrous materials compatibility with lithium, is controlled with a double trapping system: cold trap continuously operated and hot trap in the storage tank. An online monitoring of impurities is realized by mean of a resistivity meter. An offline monitoring of nitrogen content is made using sampling of lithium and successive ammonia titration technique.
a r t i c l e
i n f o
Article history: Available online 12 April 2013 Keywords: IFMIF Lithium corrosion Lithium chemistry
a b s t r a c t The Lifus 3 loop was designed and constructed in 2001/2002 in ENEA, to evaluate the corrosion behaviour of Fusion reference structural materials, and in particular AISI 316 and Eurofer, in flowing lithium environment. To fully accomplish the experimental program originally foreseen for Lifus 3 and not respected it was decided in 2010 to build a new experimental facility, simplified from the hydraulic point of view, capable of performing the set of tests required by the IFMIF community. The operating parameters of this loop, named Lifus 6 [1], are: isothermal loop with operating temperature of 350 ◦ C, maximum velocity in the test section of about 16 m/s, cold trap continuously operated to remove carbon and oxygen impurities, hot trap operated in batch to remove nitrogen ones. An integral measurement of impurities is made on line by using a resistivity meter. The Lifus 6 loop design is presented in the paper. © 2013 Elsevier B.V. All rights reserved.
1. Introduction The Lifus 3 loop was designed and constructed in 2001/2002 in ENEA, to perform compatibility tests of IFMIF reference materials in IFMIF relevant conditions. The original mission was to examine how the corrosion behaviour of the selected materials is affected by the impurities. The operation of the loop was affected by several problems, coming in particular from the purification system and the hydraulic control, and the original commitments were not fully accomplished. For this reason it was decided in 2010 to build a new experimental facility, simplified from the hydraulic point of view, capable to perform the set of tests required by the IFMIF community. A series of erosion corrosion tests, having different duration, will be performed in the LIFUS 6 test loop (see Fig. 1) fixing the basic parameters:
∗ Corresponding author. Tel.: +39 0534801380. E-mail address: [email protected] (A. Aiello). 1 Temporarily Seconded to F4E Garching, Germany. 0920-3796/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fusengdes.2013.02.129
• • • •
temperature of 350 ◦ C; Li flow velocity up to 16 m/s; impurity concentration lower than 10 ppm; F82H and Eurofer as specimen materials;
with the goal to validate the choice of material made in the IFMIF reference design. The materials to be tested are mounted in a test section specially designed to achieve the specified Li flow velocities. The Lifus 6 loop, in its operational configuration appears as a ring installed on a steel frame. For safety reasons the loop is not accessible during the operational phases and it is remotely controlled. 1. Drain/storage tank mounted on mobile supports to compensate the thermal expansion of the pipes. It is complete of electric tracing, thermocouples for regulation, safety and measure of the wall temperature, nozzles for connection to the piping. The tank also acts as hot trap at a temperature of 600 ◦ C during the off line purification of lithium from nitrogen impurities. In normal operation the storage lithium is maintained melted at about 300 ◦ C. 2. Electromagnetic pump type GAAA IP 121 provided of air cooling system.
770
A. Aiello et al. / Fusion Engineering and Design 88 (2013) 769–773
According to Fig. 2, the lithium flow at the pump outlet is divided in two parts. The main flow is driven to the test section and come back to the pump. A limited flow, about 1% of the total flow, enters in the dedicated purification loop, where it passes through the cold trap (operated at 200 ◦ C) to be continuously purified, essentially from carbon and oxygen. In the purification branch it is also installed the resistivity meter to perform an integral measurement of impurities. 2. Lifus 6 loop design
Fig. 1. Lifus 6 loop.
3. Two flow meters: a. The first one is a Vortex volumetric type, calibrated for a minimum flow velocity of 0.21 m/s and maximum 7.6 m/s corresponding to 0.37–13.4 m3 /h. b. The second one is a Coriolis mass flow meter, with a range of 0–4000 kg/h of liquid metal. 4. Test section, in which the liquid metal reaches the foreseen testing conditions of velocity.
As mentioned, the plant has a ring structure whose Piping and Instrumentation Diagram (P&ID) is shown in Fig. 2. Purple lines define the heated part of the loop. Two separate circuits can be easily identified: the liquid metal circuit, piping 1 in. and the gas lines made entirely with Swagelok pipes ½ in. The liquid metal circuit is characterized by the following components. Electromagnetic pump model IP 121 manufactured by the GAAA Company. The best operational point is reached at 12 m3 /h of volumetric flow rate with a head of 2.5 bar; Pneumatic Control Valve 1 ; Vortex Flowmeter, Rosemount model 8800; Test section, in which are reached the test conditions of fluid velocity; Coriolis mass flowmeter, Emerson Micro Motion DT65; Pressure gauges Gefran FO53732 in direct contact with liquid metal; Piping in AISI 316 L 1 Schedule 40; Storage tank. The main purpose of the gas circuit is to maintain safety conditions during operation and plant stops, and therefore the strong
Fig. 2. P&ID of Lifus 6 loop.
A. Aiello et al. / Fusion Engineering and Design 88 (2013) 769–773
771
Fig. 4. Pressure drops versus volumetric flow rate. Fig. 3. Stress configuration at design temperature.
In which: control of oxygen and water in all possible loop conditions and in particular in loading and drainage phases. The liquid metal circuit is electrically heated and thermal insulated to guarantee a constant and uniform temperature of 350 ◦ C during operation. The system is entirely electrically traced to ensure the operating temperature and prevent freezing of the liquid metal in the circuit in any possible accidental and ordinary condition. To reduce as much as possible thermal losses the loop is thermal insulated. The thermal insulation is designed to maintain a maximum surface temperature lower than 50 ◦ C. The reference design parameters of the loop are: T = 400 ◦ C (650 ◦ C for the storage tank)
K = fM
L + Kr + Ka + Kc + KT D
1
fM
= −2 log10
ε 2.51 + 3.7 Re fM
(2)
(3) Colebrook
Pressure drops have been evaluated for different components and the circuit globally. In Fig. 4 the pressure drops versus volumetric flow rate are reported. This theoretical evaluation was necessary to verify the compatibility of the available pump with the circuit. From this analysis the pump is capable to guarantee at least the velocity of 15 m/s in the test section.
P = 0.4 MPa
4. Test section design
The dimensioning of pipes has been made according to ASME B31.3 while the storage tank has been designed applying 97/23/CE – Pressure Equipment Design rules. According to 1 in. pipe diameter, the reference schedule value given by ASME B31.3 is 10, corresponding to a thickness of 2.769 mm. We decided to use only sched 40 pipes, with a thickness 20% larger and adequate safety margin. The same evaluation can be made for the storage tank. Also in this case the thickness is largely higher than the minimum allowed value. The only fixed points of the loop are represented by the pump and the storage tank. Therefore, the pipes are free to expand, and mechanical stresses are negligible. To check in any case the real stress/strain situation an analysis of the longest pipe using Ansys was performed and the stress configuration at design temperature is in Fig. 3. The total elongation is about 23 mm with a stress of 7.21 MPa. This value is very low, one order of magnitude lower than the basic allowable stress given by ASME B31.3, and in any case pipes are free to move thanks to flexible suspensions.
For what concerns the test section, a detail of this is shown in Fig. 5. The specimens have a cylindrical shape and are mounted on a rod fixed in the upper part to a removable centering plate, and supported on top and bottom of specimens’ assembly with centering systems. The centering plate is completely uncoupled with respect to the upper flange, to maintain the possibility to easily remove the upper flange and to use an extraction system to remove the specimen assembly that will be surely stuck due to the presence of lithium. With this solution it is realized a meatus in which the lithium flows at the design speed of at least 15 m/s. The triple anchoring of the support rod prevents the occurrence of vibrations in turbulent
3. Fluid dynamic analysis of the loop The general equation to determine pressure drops in a hydraulic system is: p = K
v2 2
(1) Fig. 5. Test section design with a detail of specimens housing.
772
A. Aiello et al. / Fusion Engineering and Design 88 (2013) 769–773 Table 1 Fitting parameters from the versus x experimental curves. These values refer to the calibration performed by Nottingham University. T (◦ C)
Fig. 6. Sketch of the cold trap (drawing not to scale).
flow conditions. The outer diameter of the specimens is 20 mm, with a central hole of 8 mm, for the rod, and a height of 8 mm. The channel that hosts them has a diameter of 21 mm, with an external diameter corresponding to the one of a 1 in. standard pipe. The upper part of the test section, for a height of about 300 mm, is filled with Ar, in order to prevent any contact between the lithium and the sealing area of the flange. The pressure during operation will be about 2 bar, to compensate the pump head at test section inlet maintaining a constant level of lithium also during the start up of the pump. A double-level sensor is installed in the test section to control any accidental level climbs. The samples can be extracted by draining the system, opening the flange while maintaining the circuit in an inert atmosphere and removing the rod sample holder. This procedure ensures a limited contamination of the circuit and the maximum safety for the operators, reducing downtime for the replacement operations. 5. Purification system The purification of lithium from solved non metals is a key element in the evaluation of the corrosion resistance of iron alloys, since it has been shown how their presence could dramatically increase the corrosion rate exerted by the flowing metal; particularly, N is considered to be the most affecting one, mostly because of the formation of Li ternary nitrides (Li3 FeN2 , . . .) [2,3]. The main components employed in Lifu6 plant for the purification are the cold trap (CT), the resistivity meter (RM) and the hot trap (HT): the first two elements are part of a dedicated purification loop, working in parallel with the main loop, while the third one is linked directly to the main Lifus 6 loop and is coincident with the lithium storage tank, from which the loop is charged and discharged. Only about 1% (∼0.3 L/min) of lithium enters the purification loop, being split by a stacking valve in two branches; after passing through the RM and the CT, this lithium rejoins the larger metal mass into the primary loop. The purpose of the cold trap is to reduce C, O, H and N impurities content inside the flowing metal by precipitating their corresponding binary Li compounds. It consists of a cylindrical AISI container (inner volume ∼ 3 L) filled with an AISI mesh, which acts as an active surface for the nucleation of the solid particles and for their subsequent growth. The working temperature of the trap is 200 ◦ C, while the residence time of the metal results about 10 min, which, borrowing sodium experience, is believed to be a suitable time for an high purification efficiency (≥70% for each passage) [4–10]. The gettering volume is realized by disposing a high number (∼50) of AISI mesh sheets (MESH:60; wire diameter: 0.18 mm; hole length: 0.25 mm), parallel packed inside the body of the trap, according to Fig. 6. The distance between the sheets is ∼3 mm: in this way the AISI packing density results about 300 kg m−3 . Employing all these conditions, residual C, O, H and N is reduced respectively to 2, 7, 63 and 1461 wppm [11]. In order to further reduce nitrogen content beyond the CT limit, the hot trap is employed. This trap is actually filled with a titanium sponge getter (solid grains; d = 2–12 mm; apparent density ∼ 0.55 kg m−3 ; total Ti mass ∼ 12 kg), which is able to adsorb nitrogen from lithium, due to the formation of the quite stable titanium nitride. Clearly, the higher the temperature of the HT, the less
8
10 A/ m 108 B/ m (mol% N)
200
250
300
350
400
450
24.1 5.7
25.7 6.2
27.1 6.5
28.5 6.8
29.6 7.0
30.4 7.2
effective the getter from a thermodynamic point of view. However, the rate at which gettering takes place, which involves the diffusion of a non-metal into the getter, is higher at higher temperatures and surface films have a less of an inhibiting effect. The chosen temperature compromise is 600 ◦ C, at which it is possible to reduce nitrogen level to 2 wppm [12] in a time of some days [13]. The monitoring of the nitrogen content is charged to the resistivity meter (RM) and to the off-line analysis procedure. The RM, schematically depicted in Fig. 2, is a device able to online monitor the electric resistivity () of the flowing lithium. It was developed at the University of Nottingham in collaboration with ENEA Brasimone and relies on the variation of the metal resistivity produced by the dissolved anions [14,15]: the higher the impurities concentration in lithium, the higher the resistivity. A detailed description of the device is in [14]. It was possible to calibrate the RM, finding the linear relation which, at each different temperature, links measured to N concentration (x) (see Table 1). Even if the RM has the advantage of an immediate answer, it anyway suffers from a limited sensitivity, which follows from the minimum detectable significant variation of resistivity, which, with Lifus 6 apparatus, is ∼0.18 n m: the minimum detectable variation of nitrogen concentrations in lithium results 53 wppm at 350 ◦ C. The device can anyway be useful to indicate the reaching of an high impurities threshold, as an alarm indicator. To overcome the limit of the RM and to compare the results, an offline lithium analysis is also employed. It entails to take a lithium sample from the plant using a small metallic container (V = 10 mL) linked to the purification loop by a trivial volume duct. Operating under an Ar atmosphere in a glove box, this Li is transferred into a three necks glass vessel and let to react with water, added slowly and paying attention to the violence of the transformation. According to the reaction: Li3 N + 3H2 O 3LiOH + NH3 all the nitride inside lithium is converted into ammonia. Under an Ar flow, the ammonia is distilled away and then quantitatively collected into a water solution of boric acid in excess (Fig. 7).
Fig. 7. Scheme of the apparatus for the production and distillation of NH3 (MFC: mass flow controller).
A. Aiello et al. / Fusion Engineering and Design 88 (2013) 769–773
The final solution so obtained is analyzed for its NH4 + content by HPLC (ionic chromatography) able to measure down to few wppm of starting nitrogen in lithium and with a relative precision of some %. 6. Conclusion The construction of the loop was completed, coherently with the presented design except for the purification branch. The charging of lithium is foreseen in March 2013. Experimental activities will start at the end of the start up phase, approximately in April 2013. References [1] Lifus 6 Design and Erosion/Corrosion Experimental Program, Borader Approach Report BA D 22RUY8, 2012. [2] K. Natesan, Influence of nonmetallic elements on the compatibility of structural materials with liquid alkali metals, Journal of Nuclear Materials 115 (1983) 251–262. [3] O.K. Chopra, D.L. Smith, Influence of temperature and lithium purity on corrosion of ferrous alloys in a flowing lithium environment, Journal of Nuclear Materials 141–143 (1986) 584–591.
773
[4] B.R. Grundy, Experimental characterization of sodium cold traps and modelling of their behaviour, in: Proc. Int. Conf. on Liq. Met. Tech. in En. Prod, Champion, 1976, p. 650. [5] R.B. Hinze, Chemical Engineering Progress Symposium Series: Nuclear Engineering Part XXI 66 (104) 94–106. [6] J.W. Mausteller, F. Tepper, S.J. Rodgers, Alkali metal handling and systems operating techniques, in: ANS, Gordon & Breach, New York, 1967, pp. 39–76. [7] R.W. Dickinson, in: O.J. Foust (Ed.), Liquid Metals Handbook – Sodium and NaK Supplement. Chapter 6, Purification, 1967, June. [8] K.R. Kim, J.Y. Jeong, K.C. Jeong, S.W. Kwon, S.T. Hwang, Journal of Industrial and Engineering Chemistry (Seoul) 4 (1998) 113–121. [9] M.M. Osterhout, Control of Oxygen, Hydrogen and Tritium in Sodium Systems at Experimental Breeder Reactor II – UAC – 41069. [10] W.H. Bruggeman, Purity control in sodium-cooled reactor systems, AIChE Journal 2 (2) (1978) 153. [11] M.G. Down, Proc. 2nd Int. Conf. Liquid Metals Technology and Energy Production, vol. 2, Richland, WA, 1980, pp. 14–16. [12] P. Hubberstey, Proc. Int. Conf. on Liquid Metal Eng. and Tech., vol. 2, BNES, London, 1984, p. 85. [13] A. Anttila, J. Raisanen, J. Keinonen, Diffusion of nitrogen in ␣Ti, Applied Physics Letters 42 (1983), http://dx.doi.org/10.1063/1.93981. [14] A.S. Baley, D.H. Gregory, P. Hubberstey, Development of a Monitoring System, Technical Note No. 4, School of Chemistry, University of Nottingham, 2004. [15] G.K. Creffrey, M.G. Down, R.J. Pulham, Electrical resistivity of liquid and solid lithium, Journal of the Chemical Society, Dalton Transactions (21) (1974) 2325–2329.