2013 International Nuclear Atlantic Conference - INAC 2013 Recife, PE, Brazil, November 24-29, 2013 ASSOCIAÇÃO BRASILEIRA DE ENERGIA NUCLEAR - ABEN ISBN: 978-85-99141-05-2 THERMAL HYDRAULIC MODELING OF THE LS-VHTR Maria E. Scari, Antonella L. Costa, Claubia Pereira, Maria A. F.Veloso, Clarysson A. M. Silva, Patrícia A. L. Reis Departamento de Engenharia Nuclear Escola de Engenharia Universidade Federal de Minas Gerais Av. Antônio Carlos, n 6627, Campus UFMG, PCA 1, CEP 31270-901, Belo Horizonte, MG, Brasil Instituto Nacional de Ciências e Tecnologia de Reatores Nucleares Inovadores/CNPq, Brasil Tel:: (+55) 31 3409-6688 melizabethscari@yahoo.com, antonella@nuclear.ufmg.br, claubia@nuclear.ufmg.br, dora@nuclear.ufmg.br, clarysson@nuclear.ufmg.br, patricialire@yahoo.com.br ABSTRACT An initial thermal hydraulic analysis of the liquid-salt-cooled very high-temperature reactor (LS-VHTR) has been performed in this study. Molten salts are being considered as coolants for the Next Generation Nuclear Plant (NGNP) because of their superior thermophysical properties compared to helium. Simulations using a RELAP5-3D model of the 2400 MW(t) LS-VHTR cooled by liquid Li2BeF4 (Flibe) salt are presented using a point kinetics model. The RELAP5-3D model was used to simulate the steady state behavior of a simple unit cell of the LS-VHTR. Each unit cell was considered as a composition of one simple coolant channel, filled with Flibe, and two fuel channels, with two gaps, immersed in a graphite moderator matrix. The main aim was to verify our model comparing the results with those obtained by (Davis and Hawkes, 2006) from the Idaho National Laboratory, also using a RELAP5-3D model. The results are satisfactorily approximated each other. However, more investigations are necessary to complete the model validation process. 1. INTRODUCTION The Liquid-Salt-Cooled Very-High-Temperature Reactor (LS-VHTR) is one of the IV generation reactors under development with the possible start running date of 2030. Its project is being conducted by the U. S. Department of Energy s (DOE) and it is an evolution of the Very-High-Temperature Reactor (VHTR) that is cooled by helium. This reactor uses coated-particle graphite-matrix fuels and a liquid-fluoride-salt coolant [3]. In the LS-VHTR there is a modification of the VHTR and a new way to combine four new technologies: (1) successful use of coated-particle graphite-matrix fuel in helium-cooled reactors; (2) reactor plant and safety systems similar to that developed to the liquid-metalcooled fast reactors; (3) low-pressure liquid-salt coolants studied and researched for liquidfuel reactors; and (4) Brayton power cycles at high-temperatures. The mix of these four technologies makes it possible to design a high-powered, high-temperature reactor that can economically produces hydrogen or electricity and has a fully passive safety mechanism [4]. The LS-VHTR has a cylindrical core and reflector. This cylindrical shape improved neutron economy, heat transfer, transport of liquid coolant and increase the total power output compared with the VHTR gas-cooled. LS-VHTR uses a closed primary cooled loop immersed in a tank containing a separate buffer salt. This design allows the use of a better salt in the primary coolant loop, that is, a salt that have better coolant properties but is not so good financially.
The use of liquid-salt as cooling gives a lot of vantages like: an efficient heat transfer; operation at low pressures, near the atmospheric, where its heat transfer properties are similar to the water heat transfer properties; very low vapor pressures; high volumetric heat capacity compared to gases and sodium, high Prandtl numbers that mitigate thermal shock phenomena, transparency similar to water and gases that enables optical inspection and simplify the refueling and the maintenance and very low corrosion rates [1]. There major disadvantages are: very high freezing temperatures (350 to 450 o C), that are resolved by the operation at high temperatures, and potential corrosiveness when used as solvents for molten salts fuels, which can be ignored because the LS-VHTR uses a solid fuel and clean salt coolant. In this study, the primary coolant was assumed to be Flibe, which consists of LiF- BeF 2, in a mixture of 66% LiF and 34% BeF 2. The Flibe has a small neutron cross section, which makes it relatively transparent to neutrons. [2]. In this study the baseline fuel block design considered is the same as that considered in the work of Davis and Hawkes, 2006 [2]. Each fuel block consists of a hexagonal element of 216 fuel channels, with diameter of 12.7 mm, 108 coolant channels with diameter 9.53 mm and a fuel handling hole. The flat-to-flat distance of the block is 360.0 mm and the channel pitch is 18.8 mm. The baseline block is shown in Fig. 1. The small hexagonal cells that can be seen inside the baseline block in the Fig. 1 represent the part that is being simulated in this work. Figure 1: Baseline LS-VHTR fuel block. The reactor core is formed of 265 fuel columns surrounded by reflector blocks. The geometrical parameters used in this study are given in the Table 1.
Table 1: Geometrical parameters of the LS-VHTR model [3] Parameter Value Coolant channel diameter, mm 9.53 Fuel compact diameter, mm 12.45 Fuel channel diameter, mm 12.7 Fuel channel pitch, mm 18.8 Number of coolant channels per block 108 Number of fuel channels per block 216 Number of fuel columns 265 Flat-to-flat distance of hexagonal blocks, mm 360.0 Gap between hexagonal blocks, mm 1.0 Heated length, m 7.93 In this work, RELAP5-3D, version 3.0, has been used to perform the simulations. The most prominent attribute that distinguishes the RELAP5-3D code from the previous versions is the fully integrated, multi-dimensional thermal hydraulic and neutron kinetic modeling capability [5]. There are two options for the computation of the reactor power in the RELAP5-3D code. The first option is the point reactor kinetics model that was implemented in previous versions of RELAP5. The second option is a multi-dimensional neutron kinetics model based on the NESTLE code developed at North Carolina State University. RELAP5-3D was modified to call the appropriate NESTLE subroutines depending upon the options chosen by the user. The neutron kinetics model in NESTLE and RELAP5-3D uses the few-group neutron diffusion equations. Two or four energy groups can be utilized, with all groups being thermal groups if desired. Core geometries modeling include Cartesian and hexagonal. Core symmetry options are available, including quarter, half and full core for Cartesian geometry and one-sixth, onethird and full core for hexagonal geometry. 2. RELAP5-3D MODELING Initially, a simple simulation of a unit cell was performed. Each unit cell is considered as a part of the hexagonal fuel block illustrated in the Fig. 1 being represented by one coolant channel, filled with Flibe, and two fuel channels, with two gaps, immersed in graphite moderator matrix (Fig. 2). The gap between the fuel channel and the graphite is assumed to be filled with helium gas. Figure 2 presents also the model for simulation in RELAP5-3D code using a one-dimensional heat structure with inner diameter as the diameter of the coolant channel. The cooling channel is surrounded by a graphite ring, a helium ring and a fuel ring. The volumes of the graphite, the helium and the fuel are been preserved and they corresponds to those from the original part simulated (left side in the Fig. 2). Point kinetics model was used in the simulations.
Figure 2: RELAP-3D model of an unit cell of the fuel block To model the unit cell, a single pipe was used to represent the coolant channel and a heat structure is used to represent the graphite, helium gap and the fuel rings. The pipe is divided in 24 axial volumes connected through single junctions to two time dependent volumes representing the inlet and outlet plena. The heat structure is used to simulate the power source and is divided in 24 axial nodes and 12 radial meshes. The radial meshes were divided being 4 intervals to the graphite region, 1 interval for the gap and 6 intervals representing the fuel region. The RELAP5-3D model is illustrated in Fig. 3, where the components TMDPVOL 500 and 600 are, respectively, the time dependent volume that represents the outlet and inlet plena, the SJ 400 and SJ 300 are single junctions, pipe 201 represents the coolant channel and the HS 201 is the heat structure. Figure 3: RELAP5-3D model of a unit cell The initial conditions used to simulate the unit cell channel are shown in Table 2. Obviously, as it was used only one coolant channel, the values has been divided to correctly represent the mass flow rate and the local power.
Table 2: Initial conditions for the LS-VHTR [2] Parameter Value Core total power, MW 2400 Core mass flow rate, kg/s 10,264 Core inlet temperature, o C 900 Core outlet temperature, o C 1000 Average fuel temperature, o C 1093 Maximum fuel temperature, o C 1329 Core pressure drop, MPa 0.211 Vessel pressure drop, MPa 0.276 3. RESULTS The average temperature along the heat structure is shown in the Fig. 4. Each point represents an axial node of the structure, from 1 up to 24. The temperature of each point is the average radial temperature of the corresponding axial node in the structure. The temperature increases along the channel and it reaches a maximum value in the axial node 21 and then decreases. Figure 4: Mean axial temperature along the heat structure The Fig. 5 shows the inlet and outlet pressure in the coolant channel. The pressure drop obtained by the calculation in this work is 0.19 MPa, which is very close to the pressure drop from the reference data (0.21 MPa as shown in the Tab. 2).
Figure 5: Pressure drop along the coolant channel The inlet and outlet coolant temperatures are shown in the Fig. 6. There is a temperature increase of 186 o C along the channel, and this value is much higher than the reference (100 o C, as it can be seen in Table 2). Investigations will be performed to verify the causes of such difference. Figure 6: Inlet and outlet temperatures along the coolant channel Comparison between the result of this study and those obtained by Davis and Hawkes [2] is shown in Fig. 7. The node 21 from the heat structure presented the highest average
temperature as it is shown in the Fig. 4. Considering the fuel region, this is the node where the temperature more approximated from the results obtained by Davis and Hawkes [2]. However, in the region of the gap and the graphite this node presents higher temperatures. In those regions, the results for the axial node 12 are better. The average temperature along all the heat structure is lower than that from the reference. In the reference work [2] it is not clear in which node exactly the temperatures were taken. Figure 7: Comparison between the results obtained in this work and those from [2] Fig. 8 shows some axial temperatures along the radial simulated unit cell beginning in the central parte corresponding to the coolant channel up to the center of the fuel element.
Figure 8: Radial temperature for 4 axial nodes In the Tab. 3, the results for the average temperature in each ring of the model are compared with those obtained by Davis and Hawkes [2]. In general, the average temperatures obtained in this study are underestimated in relation to those from [2]. For the graphite, the values are close each other. However, the value of the fuel temperature obtained is underestimated in comparison with the reference. Therefore, it is necessary more investigations to validate the developed model. Table 3: Mean temperatures in the materials Reference [2] ( o C) Simulation ( o C) Graphite 1048.0 1039.0 Helium gap 1083.0 1059.0 Fuel 1120.0 1082.0 4. CONCLUSIONS Thermal hydraulic analysis of a small part of the liquid-salt-cooled very high-temperature reactor (LS-VHTR) has been performed in this study. Simulations of pressures and temperatures using the RELAP5-3D model of the LS-VHTR cooled by liquid Li 2 BeF 4 (Flibe) salt were presented. In spite of the results obtained in this work presented similar behavior to those of the reference one, it is necessary to investigate in more details some results mainly those connected with the coolant temperature evolution that presented a very high value in the outlet channel. After the model validation, the idea is to simulate one of the 265 hexagonal fuel blocks and, finally, simulate the entire core with all reactor thermal-hydraulic parts.
ACKNOWLEDGMENTS The authors are grateful to CAPES, CDTN/CNEN, FAPEMIG and CNPq for the support. Thanks also to Idaho National Laboratory for the permission to use the RELAP5-3D computer software. REFERENCES 1. C. B. Davis, Implementation of Molten Salt Properties into RELAP-3D/ATHENA, INEEL/EXT-05-02658, Idaho Falls, USA, January (2005). 2. C. B. Davis and G. L. Hawkes, Thermal-Hydraulic Analyses of the LS-VHTR, Proceedings of ICAPP 2006, Reno, Nevada, USA, June 4-8 (2006). 3. D. T. Ingersoll, C. W. Forsberg, P. E. MacDonald, Trade Studies for the Liquid-SaltCooled Very High-Temperature Reactor: Fiscal Year 2006 Progress Report, Oak Ridge National Laboratory, ORNL/TM-2006/140, Tennessee, USA, February (2007). 4. D. T. Ingersoll, K. T. Clarno, C. W. Forsberg, J. C. Gehin, R. W. Christensen, C. B. Davis, G. L. Hawkes, J. W. Sterbenstz, T. K. Kim, T. A. Taiwo, and W. S. Yang. Status of Physics and Safety Analyses for the Liquid-Salt-Cooled Very High-Temperature Reactor (LSVHTR), ORNL/TM-2005/218, Tennessee, USA, December (2005). 5. INEEL, RELAP-3D Code Manual, Idaho National Laboratory, INEEL-EXT-98-00834, Idaho Falls, USA, September (2009).