- Email: [email protected]
Thermal analysis of a cryogenic distillation column for hydrogen isotopes separation
Fusion Engineering and Design 146 (2019) 1868–1871
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Thermal analysis of a cryogenic distillation column for hydrogen isotopes separation
T
⁎
George Anaa, , Florian Vladulescub, Radu Anac, Gheorghe Pascad, Alina Niculescua National R&D Institute for Cryogenics and Isotopic Technologies – ICSI, Rm. Valcea, Romania INAS, Craiova, Romania c KIT – Tritium Laboratory, Karlsruhe, Germany d ISTECH, Timisoara, Romania a
b
A R T I C LE I N FO
A B S T R A C T
Keywords: Thermal Analysis Cryogenic
Cryogenic distillation (CD) of hydrogen in combination with Liquid Phase Catalytic Exchange (LPCE) or Combined Electrolytic Catalytic Exchange (CECE) process is used for tritium removal/recovery from tritiated water from fission rectors like CANDU and is the proposed technology for ITER/DEMO. Tritiated water is being obtained after long time operation of CANDU reactors, or in case of ITER mainly by the Detritiation System. The cryogenic system of the Experimental Pilot plant for tritium and deuterium separation from ICSI Rm. Valcea consists of four distillation columns and significant effort is required in various batch mode operations for achieving high tritium concentration. Some problems have been experienced with the fourth column of the cascade regarding the heat transient transfer during start-up. This paper intent is to present a CAE/FEA (Computer Aided Engineering/Finite Element Analysis) thermal simulation of the column during the transitory thermal regime in order to investigate the temperature distribution along the column and its connections to other equipment. The simulation is done on the as-built design of the column and takes in consideration the actual configuration. This work provides an important platform to understand the thermal phenomena during cool down of a cryogenic distillation column besides finding the failures in thermal insulation or design flaws. The results of the simulation can be used as lessons-learned for future design work of cryogenic systems like Tritium Removal Facility from Cernavoda Nuclear Power Plant or ITER/DEMO Isotope Separation System.
1. Introduction The most complex part of the ICSI detritiation plant is the cryogenic distillation system, which also has the largest tritium inventory and has undergone several stages of optimization during design, especially for hydrogen and tritium inventory minimization. The system consists of four distillation columns, heat-exchangers for heat recovery, pumps, instrumentation and discharge safety systems and a helium refrigeration unit to provide the cooling power for columns operation [1]. During start-up of the system, the first objective is cool down all columns in order to reach the operation parameters: temperature, pressure, gas and liquid inventories. As shown in ref. [1], all issues encountered with the cool down of the first three columns have been overcome and the required operating parameters have been reached. As for the fourth column, some problems have been experienced regarding start-up or cool down and it couldn’t be operated at nominal
⁎
parameters. The fourth column of the cascade which actually has the largest tritium inventory from the whole plant, has a inner diameter of 8 mm and a length of approximately 2.5 m (including condenser and reboiler). The condenser of the column provides a maximum calculated refrigeration power of 20 W needed to drive the column internal circulation of saturated vapor and liquid in order to facilitate isotope exchange between these two media around a temperature of 25 K. In order to thermally insulate the column, it is placed together with the other three columns inside a vacuumed cold box at 10−3 Pa and individually wrapped (over the whole length) in 30 layers of super insulation material (multilayer insulation – MLI). In order to cool-down the column and reach the nominal operating parameters in a reasonable period of time (36–48 h), both statically (gradually liquefying injected inventory) and dynamically procedures have been adopted, in which a continuous flow of gas has been fed to
Corresponding author. E-mail address: [email protected] (G. Ana).
https://doi.org/10.1016/j.fusengdes.2019.03.053 Received 5 October 2018; Received in revised form 8 March 2019; Accepted 8 March 2019 Available online 02 April 2019 0920-3796/ © 2019 Elsevier B.V. All rights reserved.
Fusion Engineering and Design 146 (2019) 1868–1871
G. Ana, et al.
Fig. 1. Discretized finite element model of the column. (Top left – helium domain; left – hydrogen domain).
Fig. 2. He circulation (velocity profile - left) and condenser temperature profile at beginning of cooling (right).
the cold condenser and passed through the column, down to the reboiler, keeping in the same time an overpressure of about 35 kPa. Potential causes for the failure of cool down of the column might be: poor insulation of the column and heat gain by radiation from all other system internals placed next to it; heat gain by conduction from the pipework connected to the column; large metal mass of the column which translates into a large quantity of heat to be removed to lower the temperature to 22–25 K; not enough condensation power of the condenser to overcome all heat fluxes, due to the configuration of the refrigerant feeding system [1]: the duty of the columns condenser is heavily dependent on the duty of the third column condenser. The main objective of this work is to undertake a fluid flow and thermal analysis of the as-built column using ANSYS Fluent to investigate the problems encountered with it and promote a potential solution to the issue in order to be operated at nominal parameters.
2. About ANSYS Fluent and FEA analysis Fig. 3. Hydrogen flow along the column (velocity profile – left) and temperature profile (right).
ANSYS Fluent is a computational fluid dynamics (CFD) tool which is the science of predicting fluid flow, heat transfer, mass transfer, chemical reactions, and related phenomena by solving the mathematical equations which govern these processes using computational methods. The technique used by ANSYS is based on Finite Element Analysis or 1869
Fusion Engineering and Design 146 (2019) 1868–1871
G. Ana, et al.
Fig. 4. Colum temperature profile (left); Temperature evolution of the lowest point on the column after aprox. 251,000 s (right).
Method (FEA/FEM). FEA is a mathematical representation of a physical system comprising a part/assembly (model), material properties, and applicable boundary conditions – collectively referred to as pre-processing, the solution of that mathematical representation – solving, and the study of results of that solution – post-processing. The premises of FEA analysis is to discretize a complex form into regular shapes (pyramids, cubes, tetrahedrons etc. – as shown in Fig. 1) for which mathematical models exist and can be applied for solving a given problem, analyze the solution and then concatenate the results to determine the overall solution for the initial form [2]. The applied model for the present analysis is the “k-epsilon” model, a two equation model of turbulence in which the solution of two separate transport equations allows the turbulent velocity and length scales to be independently determined. It is a semi-empirical model, and the derivation of the model equations relies on phenomenological considerations and empiricism and has a reasonable accuracy for a wide range of turbulent flows. The model has been chosen due to the turbulent flow on the helium side of the condenser, Re number of about 7.4E+5. 3. The geometric model and analysis The model of the column presents (Fig. 1) a condenser with the exterior coiled helium pipeline, the column itself and a reboiler area at the bottom. The equipment is foreseen with several nozzles for feeding and extracting process fluids. The main aspect during the design of the column was the minimization of the overall mass of the equipment, due to the small refrigeration power of the condenser. In order to reduce analysis time, the model of the column has been simplified and all components and design details with no or little influence on the anticipated analysis results have been removed and no internals (packing) have been considered. The analysis has been done using hydrogen instead of tritiated deuterium as for the real application. The desired cooldown temperature was 22 K for the entire column, which is considered conservative in view of the needed cool down temperature for tritiated deuterium between 24.5 and 26 K. Based on the design configuration of the column, two scenarios were considered when analyzing the thermal transients of the column during start-up/cooldown:
Fig. 5. Column with radiation shield (left); Cross section of the radiation shield (top-right); Column temperature profile when protected by a radiation shield (down-right).
column and adding a source of cold at the middle of the column, which will remove the heat gain by the circulated stream of gas from the upper part and further cooling the lower part of the column; Scenario 1 takes in consideration the circulation of cold helium (20 K) through the exterior mounted pipe (Fig. 2), which in turn cools the main body of the condenser, further cooling the stream of hydrogen fed to the top of the condenser (Fig. 3) from the third column of the cascade at a flow rate of 2.7e-5 kg/s. The hydrogen feed is assumed to be at a temperature of 22–23 K and it can be seen as a “carrier of cold” as it flows from the condenser down to the reboiler, gradually cooling the column. The hydrogen flow along the column simulates the downward flow of liquid hydrogen otherwise obtained during real operation of the column, when the condenser is cooled at 20 K. As the
• by cooling the condenser and circulating a stream of gas along the column from top to bottom; • by cooling the condenser, circulating a stream of gas along the 1870
Fusion Engineering and Design 146 (2019) 1868–1871
G. Ana, et al.
length of the column and the cooling capacity of the condenser does not allow the cooling of the entire column length to the necessary temperature (22 K). The heat gain by radiation can’t be overcome by the cold gas coming from the condenser and or from the middle feeding streams trough the columns nozzles. The proposed solution is to mount a thermal shield around the column in order to enhance insulation of the entire column from radiation heat emitted by the cold box wall and other equipment places next to the column. The shield, covering the whole column except the condenser and the reboiler, is to be cooled with 20–21 K helium used in the columns condenser and disposes of 10 W of cooling power. Fig. 5 top-left shows a cross section of the proposed radiation shield around the column with a small slit for piping connections to the other equipment within the system. The shield has been simulated by extracting 10 W of heat from the external surface of the column, by radiation. Fig. 5 (down-right) shows the temperature profile, actually the same temperature of 22 K for the entire column, while Fig. 6 shows the time required for the entire column to reach this temperature, 42 h. Fig. 5 (left) shows a 3D model of the column together with the radiation shield and the helium connection coming from the upper part of the column condenser.
Fig. 6. Temperature evolution of the lowest point on the column after 151,000 s, when protected by a radiation shield.
hydrogen flows along the column, it warms up (due to the initially warm column wall and heat gain by the column from the exterior through radiation) the stream velocity increases as seen in Fig. 3-left and is evacuated from the column through the lowest nozzle of the reboiler. The input data for the first scenario are as follows: helium flow rate of 0.023 kg/s at 20 K and 700 kPa, hydrogen feed rate of 2.7e-5 kg/ s at 23 K and 135 kPa, assumed heat flow rate to the column through radiation, 0.2 W/m2. The analysis shows that after aprox. 17 h the bottom of the column reaches only aprox. 254 K (Fig. 3-right) insufficient for nominal operation of the column at 25 K (or 22 K – the temperature. Scenario 2 takes in consideration the circulation of cold helium (20 K) through the exterior mounted pipe as for scenario 1, and supplementary cooling in the middle of the column arising from the feed of cold hydrogen at about 23 K. Adding a source of cold at the middle of the column brings a significant improvement to the column cooldown, but not good enough as shown in Fig. 4. The source of cold takes the form of a stream of hydrogen circulated in a closed loop and returned to the middle of the column after being cooled to aprox. 23. The temperature of the lowest point of the column reaches only 25 K after aprox. 70 h as shown in Fig. 5-right with a tendency of stabilization. Still, also in this way of operation the temperature of 22 K is not reached, implying additional measures to be implemented for the column cooldown.
5. Conclusion The analysis employed revealed that without a thermal shield, a column of these size and dimensions is hard or even impossible to cool down. In scenario 2, feeding a cold stream of H2 into the middle of the column was not sufficient to overcome the radiation heat flux. Hence, cooling the entire metal mass of the column down 22 K is not possible within reasonable time. Using a thermal shield permits cooling down of the entire column in a relatively short time by significantly reducing radiation heat (to some extent the shield acting as a pre-cooler). References [1] George Ana, et al., Construction and commissioning of a hydrogen cryogenic distillation system for tritium recovery at ICIT Rm. Valcea, Fusion Eng. Des. 106 (2016) 51–55. [2] S. Butnaru, Gh. Mogan, Analiza cu elemente finite in ingineria mecanica – Aplicatii practice in ANSYS, Ed. Universitatii Transilvania din Brasov, 2014 ISBN 978-606-190474-7.
4. Proposed solution The analysis of both scenarios shows that the ratio between the
1871
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Thermal analysis of a cryogenic distillation column for hydrogen isotopes separation
T
⁎
George Anaa, , Florian Vladulescub, Radu Anac, Gheorghe Pascad, Alina Niculescua National R&D Institute for Cryogenics and Isotopic Technologies – ICSI, Rm. Valcea, Romania INAS, Craiova, Romania c KIT – Tritium Laboratory, Karlsruhe, Germany d ISTECH, Timisoara, Romania a
b
A R T I C LE I N FO
A B S T R A C T
Keywords: Thermal Analysis Cryogenic
Cryogenic distillation (CD) of hydrogen in combination with Liquid Phase Catalytic Exchange (LPCE) or Combined Electrolytic Catalytic Exchange (CECE) process is used for tritium removal/recovery from tritiated water from fission rectors like CANDU and is the proposed technology for ITER/DEMO. Tritiated water is being obtained after long time operation of CANDU reactors, or in case of ITER mainly by the Detritiation System. The cryogenic system of the Experimental Pilot plant for tritium and deuterium separation from ICSI Rm. Valcea consists of four distillation columns and significant effort is required in various batch mode operations for achieving high tritium concentration. Some problems have been experienced with the fourth column of the cascade regarding the heat transient transfer during start-up. This paper intent is to present a CAE/FEA (Computer Aided Engineering/Finite Element Analysis) thermal simulation of the column during the transitory thermal regime in order to investigate the temperature distribution along the column and its connections to other equipment. The simulation is done on the as-built design of the column and takes in consideration the actual configuration. This work provides an important platform to understand the thermal phenomena during cool down of a cryogenic distillation column besides finding the failures in thermal insulation or design flaws. The results of the simulation can be used as lessons-learned for future design work of cryogenic systems like Tritium Removal Facility from Cernavoda Nuclear Power Plant or ITER/DEMO Isotope Separation System.
1. Introduction The most complex part of the ICSI detritiation plant is the cryogenic distillation system, which also has the largest tritium inventory and has undergone several stages of optimization during design, especially for hydrogen and tritium inventory minimization. The system consists of four distillation columns, heat-exchangers for heat recovery, pumps, instrumentation and discharge safety systems and a helium refrigeration unit to provide the cooling power for columns operation [1]. During start-up of the system, the first objective is cool down all columns in order to reach the operation parameters: temperature, pressure, gas and liquid inventories. As shown in ref. [1], all issues encountered with the cool down of the first three columns have been overcome and the required operating parameters have been reached. As for the fourth column, some problems have been experienced regarding start-up or cool down and it couldn’t be operated at nominal
⁎
parameters. The fourth column of the cascade which actually has the largest tritium inventory from the whole plant, has a inner diameter of 8 mm and a length of approximately 2.5 m (including condenser and reboiler). The condenser of the column provides a maximum calculated refrigeration power of 20 W needed to drive the column internal circulation of saturated vapor and liquid in order to facilitate isotope exchange between these two media around a temperature of 25 K. In order to thermally insulate the column, it is placed together with the other three columns inside a vacuumed cold box at 10−3 Pa and individually wrapped (over the whole length) in 30 layers of super insulation material (multilayer insulation – MLI). In order to cool-down the column and reach the nominal operating parameters in a reasonable period of time (36–48 h), both statically (gradually liquefying injected inventory) and dynamically procedures have been adopted, in which a continuous flow of gas has been fed to
Corresponding author. E-mail address: [email protected] (G. Ana).
https://doi.org/10.1016/j.fusengdes.2019.03.053 Received 5 October 2018; Received in revised form 8 March 2019; Accepted 8 March 2019 Available online 02 April 2019 0920-3796/ © 2019 Elsevier B.V. All rights reserved.
Fusion Engineering and Design 146 (2019) 1868–1871
G. Ana, et al.
Fig. 1. Discretized finite element model of the column. (Top left – helium domain; left – hydrogen domain).
Fig. 2. He circulation (velocity profile - left) and condenser temperature profile at beginning of cooling (right).
the cold condenser and passed through the column, down to the reboiler, keeping in the same time an overpressure of about 35 kPa. Potential causes for the failure of cool down of the column might be: poor insulation of the column and heat gain by radiation from all other system internals placed next to it; heat gain by conduction from the pipework connected to the column; large metal mass of the column which translates into a large quantity of heat to be removed to lower the temperature to 22–25 K; not enough condensation power of the condenser to overcome all heat fluxes, due to the configuration of the refrigerant feeding system [1]: the duty of the columns condenser is heavily dependent on the duty of the third column condenser. The main objective of this work is to undertake a fluid flow and thermal analysis of the as-built column using ANSYS Fluent to investigate the problems encountered with it and promote a potential solution to the issue in order to be operated at nominal parameters.
2. About ANSYS Fluent and FEA analysis Fig. 3. Hydrogen flow along the column (velocity profile – left) and temperature profile (right).
ANSYS Fluent is a computational fluid dynamics (CFD) tool which is the science of predicting fluid flow, heat transfer, mass transfer, chemical reactions, and related phenomena by solving the mathematical equations which govern these processes using computational methods. The technique used by ANSYS is based on Finite Element Analysis or 1869
Fusion Engineering and Design 146 (2019) 1868–1871
G. Ana, et al.
Fig. 4. Colum temperature profile (left); Temperature evolution of the lowest point on the column after aprox. 251,000 s (right).
Method (FEA/FEM). FEA is a mathematical representation of a physical system comprising a part/assembly (model), material properties, and applicable boundary conditions – collectively referred to as pre-processing, the solution of that mathematical representation – solving, and the study of results of that solution – post-processing. The premises of FEA analysis is to discretize a complex form into regular shapes (pyramids, cubes, tetrahedrons etc. – as shown in Fig. 1) for which mathematical models exist and can be applied for solving a given problem, analyze the solution and then concatenate the results to determine the overall solution for the initial form [2]. The applied model for the present analysis is the “k-epsilon” model, a two equation model of turbulence in which the solution of two separate transport equations allows the turbulent velocity and length scales to be independently determined. It is a semi-empirical model, and the derivation of the model equations relies on phenomenological considerations and empiricism and has a reasonable accuracy for a wide range of turbulent flows. The model has been chosen due to the turbulent flow on the helium side of the condenser, Re number of about 7.4E+5. 3. The geometric model and analysis The model of the column presents (Fig. 1) a condenser with the exterior coiled helium pipeline, the column itself and a reboiler area at the bottom. The equipment is foreseen with several nozzles for feeding and extracting process fluids. The main aspect during the design of the column was the minimization of the overall mass of the equipment, due to the small refrigeration power of the condenser. In order to reduce analysis time, the model of the column has been simplified and all components and design details with no or little influence on the anticipated analysis results have been removed and no internals (packing) have been considered. The analysis has been done using hydrogen instead of tritiated deuterium as for the real application. The desired cooldown temperature was 22 K for the entire column, which is considered conservative in view of the needed cool down temperature for tritiated deuterium between 24.5 and 26 K. Based on the design configuration of the column, two scenarios were considered when analyzing the thermal transients of the column during start-up/cooldown:
Fig. 5. Column with radiation shield (left); Cross section of the radiation shield (top-right); Column temperature profile when protected by a radiation shield (down-right).
column and adding a source of cold at the middle of the column, which will remove the heat gain by the circulated stream of gas from the upper part and further cooling the lower part of the column; Scenario 1 takes in consideration the circulation of cold helium (20 K) through the exterior mounted pipe (Fig. 2), which in turn cools the main body of the condenser, further cooling the stream of hydrogen fed to the top of the condenser (Fig. 3) from the third column of the cascade at a flow rate of 2.7e-5 kg/s. The hydrogen feed is assumed to be at a temperature of 22–23 K and it can be seen as a “carrier of cold” as it flows from the condenser down to the reboiler, gradually cooling the column. The hydrogen flow along the column simulates the downward flow of liquid hydrogen otherwise obtained during real operation of the column, when the condenser is cooled at 20 K. As the
• by cooling the condenser and circulating a stream of gas along the column from top to bottom; • by cooling the condenser, circulating a stream of gas along the 1870
Fusion Engineering and Design 146 (2019) 1868–1871
G. Ana, et al.
length of the column and the cooling capacity of the condenser does not allow the cooling of the entire column length to the necessary temperature (22 K). The heat gain by radiation can’t be overcome by the cold gas coming from the condenser and or from the middle feeding streams trough the columns nozzles. The proposed solution is to mount a thermal shield around the column in order to enhance insulation of the entire column from radiation heat emitted by the cold box wall and other equipment places next to the column. The shield, covering the whole column except the condenser and the reboiler, is to be cooled with 20–21 K helium used in the columns condenser and disposes of 10 W of cooling power. Fig. 5 top-left shows a cross section of the proposed radiation shield around the column with a small slit for piping connections to the other equipment within the system. The shield has been simulated by extracting 10 W of heat from the external surface of the column, by radiation. Fig. 5 (down-right) shows the temperature profile, actually the same temperature of 22 K for the entire column, while Fig. 6 shows the time required for the entire column to reach this temperature, 42 h. Fig. 5 (left) shows a 3D model of the column together with the radiation shield and the helium connection coming from the upper part of the column condenser.
Fig. 6. Temperature evolution of the lowest point on the column after 151,000 s, when protected by a radiation shield.
hydrogen flows along the column, it warms up (due to the initially warm column wall and heat gain by the column from the exterior through radiation) the stream velocity increases as seen in Fig. 3-left and is evacuated from the column through the lowest nozzle of the reboiler. The input data for the first scenario are as follows: helium flow rate of 0.023 kg/s at 20 K and 700 kPa, hydrogen feed rate of 2.7e-5 kg/ s at 23 K and 135 kPa, assumed heat flow rate to the column through radiation, 0.2 W/m2. The analysis shows that after aprox. 17 h the bottom of the column reaches only aprox. 254 K (Fig. 3-right) insufficient for nominal operation of the column at 25 K (or 22 K – the temperature. Scenario 2 takes in consideration the circulation of cold helium (20 K) through the exterior mounted pipe as for scenario 1, and supplementary cooling in the middle of the column arising from the feed of cold hydrogen at about 23 K. Adding a source of cold at the middle of the column brings a significant improvement to the column cooldown, but not good enough as shown in Fig. 4. The source of cold takes the form of a stream of hydrogen circulated in a closed loop and returned to the middle of the column after being cooled to aprox. 23. The temperature of the lowest point of the column reaches only 25 K after aprox. 70 h as shown in Fig. 5-right with a tendency of stabilization. Still, also in this way of operation the temperature of 22 K is not reached, implying additional measures to be implemented for the column cooldown.
5. Conclusion The analysis employed revealed that without a thermal shield, a column of these size and dimensions is hard or even impossible to cool down. In scenario 2, feeding a cold stream of H2 into the middle of the column was not sufficient to overcome the radiation heat flux. Hence, cooling the entire metal mass of the column down 22 K is not possible within reasonable time. Using a thermal shield permits cooling down of the entire column in a relatively short time by significantly reducing radiation heat (to some extent the shield acting as a pre-cooler). References [1] George Ana, et al., Construction and commissioning of a hydrogen cryogenic distillation system for tritium recovery at ICIT Rm. Valcea, Fusion Eng. Des. 106 (2016) 51–55. [2] S. Butnaru, Gh. Mogan, Analiza cu elemente finite in ingineria mecanica – Aplicatii practice in ANSYS, Ed. Universitatii Transilvania din Brasov, 2014 ISBN 978-606-190474-7.
4. Proposed solution The analysis of both scenarios shows that the ratio between the
1871