DOI: 10.3303/CET25117146 Paper Received: 27 December 2024; Revised: 4 February 2025; Accepted: 24 May 2025 Please cite this article as: Mancera-Apolinar J.A., Mendoza D.F., Andersson R., Riascos C.A.M., 2025, Heat and Mass Transfer Study of Top- Heat-Integrated Distillation Column (T-HIDiC) using CFD , Chemical Engineering Transactions, 117, 871-876 DOI:10.3303/CET25117146 CHEMICAL ENGINEERING TRANSACTIONS VOL. 117, 2025 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it Guest Editors: Fabrizio Bezzo, Flavio Manenti, Gabriele Pannocchia, Almerinda di Benedetto Copyright © 2025, AIDIC Servizi S.r.l. ISBN 979-12-81206-17-5; ISSN 2283-9216 Heat and Mass Transfer Study of Top-Heat-Integrated Distillation Column (T-HIDiC) using CFD Javier A. Mancera-Apolinara,*, Diego F. Mendozab, Ronnie Anderssonc, Carlos A. M. Riascosd aUniversidad de América. Departamento de Ingeniería Química y Ambiental; tel. +57 3003865231, Avda Circunvalar No. 20- 53. Bogotá-Colombia. bUniversidad de Antioquia. Departamento de Ingeniería Química. tel. +57 3216451100. Calle 70 # 52-21, Medellín, Antioquia. cChalmers Technology University. Department of Chemical Engineering. tel. +46 317722941. Gothenburg-Sweden. dUniversidad Nacional de Colombia; sede Bogotá. Departamento de Ingeniería Química y Ambiental-Grupo de Investigación de Bioprocesos; tel. +57 3125830176. Carrera 30 # 45-03; Edificio 453. Bogotá-Colombia. javier.mancera@profesores.uamerica.edu.co Distillation is one of the most widely used separation techniques in chemical processes, this separation method accounts for a significant portion of global energy consumption. Process intensification strategies have led to alternative technologies such as Heat Integrated Distillation Columns (HIDiC), these units can significantly reduce energy consumption. The goal of this study is to develop a CFD model to analyze momentum, heat and mass transfer in the separation of propylene/propane system in a Top HIDiC. The initial and boundary conditions are based on data from Aspen plus process simulations. Ranz-Marshall approaches are used for the representation of the heat transfer between phases and resistance in the gas phase. The mass transfer coefficient was calculated using Higbie’s model implemented via UDF. The Peng-Robinson thermodynamic equilibrium model was implemented to estimate equilibrium ratio at the interface by UDF. The developed CFD model was used to evaluate three pairs of internal rectification stage plus external stripping stage, that are thermally integrated. These stages were selected as representative of the column. Phases distribution, temperature and vapor phase Murphree efficiency were obtained as results, showing the potential of CFD models to analyze new and structurally complex units. 1. Introduction Distillation is a widely used separation technique in the chemical industry, but it is also one of the most energy- intensive processes, significantly impacting operating costs and carbon footprint. To address this issue, energy integration has been proposed as a process intensification strategy, reducing equipment and utility requirements through heat and mass integration. In this context, the Heat-Integrated Distillation Column (HIDiC) concept has been developed, achieving diabatic operation by vapor recompression at a low compression ratio, enabling substantial energy savings. HIDiC is particularly effective for separating close-boiling mixtures, such as the propylene-propane system (Olujic et al., 2003). Previous studies demonstrated that the Top HIDiC (T-HIDiC) configuration, which thermally integrates the upper rectification stages with the stripping section, is more energy- efficient (Mancera et al., 2018). Research at institutions like the University of Delft and simulations using tools such as Aspen Plus® and Aspen Hysys® have evaluated these columns, but they often overlook critical phenomena such as fluid dynamics and heat and mass transfer effects. Computational Fluid Dynamics (CFD) approaches have been used to analyze these aspects, although prior models focused primarily on heat transfer within solid structure. This study aims to develop an advanced CFD model to analyze the operation of a T-HIDiC column for propylene/propane separation, as suggested in previous work (Mancera et al., 2018). Conventional process simulations provided temperature and concentration profiles, as well as vapor and liquid flow data, to define boundary and initial conditions for CFD analysis. 871 2. CFD and mass transfer model The approach of continuous liquid phase based on the small density and small spatial scales of the dispersed gas was used for the multiphase behavior. The two-equation turbulence RANS (Reynolds Averaged Navier Stokes) model for the liquid and vapor phase was used (Colombo & Fairweather, 2015). The momentum exchange is modeled to consider drag, lift, and virtual mass. 2.1 Flow equations The implemented two-fluid approach considers both, vapor and liquid phases as continuous and their models are based on two sets of conservation equations which estimate mass, momentum, and energy. These equations are written for each phase in different works (Versteeg & Malalasekera, 1995) 2.2 Modified drag model In this research, we use a modified drag coefficient (𝐶𝐶𝐷𝐷,𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐), based on the Grace model and developed by Mancera et al. (2023). This correction on drag coefficient is supported on the idea that Grace model considers the balance forces acting on a single bubble moving into the liquid, instead of a multi-bubble system where each bubble interacts with the liquid and with other bubbles. The equation used is 𝐶𝐶𝐷𝐷,𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 = 𝐹𝐹𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝐶𝐶𝐷𝐷, here, 𝐶𝐶𝐷𝐷 is the drag coefficient calculated with Grace’s model (Grace et al., 1976), and 𝐹𝐹𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 is the correction factor. 2.3 Heat exchange coefficient The rate of energy transfer between phases, 𝑄𝑄𝐿𝐿𝐿𝐿, is a function of the temperature difference and the interfacial area, 𝐴𝐴. The equation is 𝑄𝑄𝐿𝐿𝐿𝐿 = ℎ�𝐴𝐴(𝑇𝑇𝐿𝐿 − 𝑇𝑇𝐿𝐿), where ℎ�, represents the heat transfer coefficient between phases and is determined by Ranz-Marshall’s correlation (Ranz & Marshall, 1952), based on Nusselt number, 𝑁𝑁𝑁𝑁: 𝑁𝑁𝑁𝑁 = ℎ�𝐿𝐿𝐿𝐿 𝑘𝑘𝐿𝐿 = 2 + 0.6𝑅𝑅𝑅𝑅𝐿𝐿 1/2𝑃𝑃𝑃𝑃1/3 (1) here 𝑘𝑘𝐿𝐿 is the thermal conductivity of the gas phase, 𝐿𝐿𝐿𝐿 is a characteristic length (such as the bubble diameter), 𝑃𝑃𝑃𝑃 and 𝑅𝑅𝑅𝑅𝐿𝐿 are the Prandtl number and Reynolds number of the gas phase 2.4 Species Mass Transfer To model species mass transfer at the interphase, transport equations for each specie at each phase must be solved along with the mass, momentum, and energy equations. The transport equations for the mass fraction for light component (LK), are (similar equations are developed for the heavy component): Vapor Phase 𝜕𝜕(𝛼𝛼𝐿𝐿𝜌𝜌𝐿𝐿𝑌𝑌𝐿𝐿𝐿𝐿) 𝜕𝜕𝜕𝜕 + ∇ ∙ (𝛼𝛼𝐿𝐿𝜌𝜌𝐿𝐿𝑁𝑁𝐿𝐿𝑌𝑌𝐿𝐿𝐿𝐿) = ∇ ∙ �𝛼𝛼𝐿𝐿𝜌𝜌𝐿𝐿𝐷𝐷𝐿𝐿𝐿𝐿,𝐿𝐿∇𝑌𝑌𝐿𝐿𝐿𝐿� − 𝑆𝑆𝐿𝐿𝐿𝐿,𝐿𝐿𝐿𝐿 (2) Liquid Phase: 𝜕𝜕(𝛼𝛼𝐿𝐿𝜌𝜌𝐿𝐿𝑋𝑋𝐿𝐿𝐿𝐿) 𝜕𝜕𝜕𝜕 + ∇ ∙ (𝛼𝛼𝐿𝐿𝜌𝜌𝐿𝐿𝑁𝑁𝐿𝐿𝑋𝑋𝐿𝐿𝐿𝐿) = ∇ ∙ �𝛼𝛼𝐿𝐿𝜌𝜌𝐿𝐿𝐷𝐷𝐿𝐿𝐿𝐿,𝐿𝐿∇𝑋𝑋𝐿𝐿𝐿𝐿� + 𝑆𝑆𝐿𝐿𝐿𝐿,𝐿𝐿𝐿𝐿 (3) The interphase mass transfer term 𝑺𝑺𝑳𝑳𝑳𝑳,𝑳𝑳𝑳𝑳 is the interphase mass transfer rate of the light component per unit volume. It is given by 𝑆𝑆𝐿𝐿𝐿𝐿,𝐿𝐿𝐿𝐿 = 𝑘𝑘𝐿𝐿𝐿𝐿𝐴𝐴(𝐾𝐾𝜌𝜌𝐿𝐿𝐿𝐿𝐿𝐿 − 𝜌𝜌𝐿𝐿𝐿𝐿𝐿𝐿), where 𝐴𝐴 is the interfacial area per unit volume, 𝐾𝐾 is the mole fraction equilibrium ratio (𝑦𝑦𝑒𝑒 = 𝐾𝐾/𝑥𝑥), 𝜌𝜌𝑝𝑝𝐿𝐿𝐿𝐿 is the mass/volume concentration of the light component in the phase p (this is the nomenclature used in Fluent, and 𝜌𝜌𝑝𝑝𝐿𝐿𝐿𝐿 = 𝜌𝜌𝐿𝐿 ∗ 𝑥𝑥𝐿𝐿𝐿𝐿), and where: 1 𝑘𝑘𝐿𝐿𝐿𝐿 = 1 𝑘𝑘𝐿𝐿 + 𝐾𝐾 𝑘𝑘𝐿𝐿 (4) Here, 𝑘𝑘𝐿𝐿 and 𝑘𝑘𝐿𝐿 are the volumetric individual film mass transfer coefficients of gas and liquid, respectively 2.5 Mass transfer model Interphase species mass transfer is modeled using the two-resistance model from Ansys Fluent®. There are many theories available in the literature for the calculation of the mass transfer coefficient 𝑘𝑘𝐿𝐿 and 𝑘𝑘𝐿𝐿 (Lamont & Scott, 1970), however the most used are: • Penetration theory (Higbie, 1935): 𝑘𝑘𝐿𝐿 = 2 √𝜋𝜋 � 𝐷𝐷𝐿𝐿,𝐿𝐿𝐿𝐿 𝜕𝜕𝑒𝑒,𝐿𝐿 𝑘𝑘𝐿𝐿 = 2 √𝜋𝜋 � 𝐷𝐷𝐿𝐿,𝐿𝐿𝐿𝐿 𝜕𝜕𝑒𝑒,𝐿𝐿 (5) 872 where 𝜕𝜕𝑒𝑒,𝐿𝐿 and 𝜕𝜕𝑒𝑒,𝐿𝐿 are the exposure time in each phase, calculated as 𝜕𝜕𝑒𝑒,𝐿𝐿 = 𝑑𝑑𝑏𝑏 𝑢𝑢𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 and 𝜕𝜕𝑒𝑒,𝐿𝐿 = 𝑑𝑑𝑏𝑏 𝑢𝑢𝐻𝐻 , where 𝑑𝑑𝑏𝑏 is the bubble diameter, 𝑁𝑁𝑐𝑐𝑟𝑟𝑟𝑟𝑒𝑒 is the bubble slip velocity and 𝑁𝑁𝐻𝐻 is the vapor velocity through the tray holes. Therefore, the expression for the local mass transfer coefficient is: 𝑘𝑘𝐿𝐿 = 2Υ� 𝐷𝐷𝐿𝐿,𝐿𝐿𝐿𝐿𝑁𝑁𝑐𝑐𝑟𝑟𝑟𝑟𝑒𝑒 𝜋𝜋𝑑𝑑𝑏𝑏 𝑘𝑘𝐿𝐿 = 2Υ� 𝐷𝐷𝐿𝐿,𝐿𝐿𝐿𝐿𝑁𝑁𝐻𝐻 𝜋𝜋𝑑𝑑𝑏𝑏 (6) where Υ is a constant. • Ranz-Marshall Model (Ranz & Marshall, 1952): The Ranz-Marshall model uses an analogous approach to that for the Ranz-Marshall heat transfer coefficient model, and it is based on the Sherwood number: 𝑆𝑆ℎ = 𝑘𝑘𝑚𝑚𝐿𝐿𝐿𝐿 𝐷𝐷𝐿𝐿 = 2 + 0.6𝑅𝑅𝑅𝑅𝐿𝐿 1/2𝑆𝑆𝑆𝑆𝐿𝐿 1/3 (7) where 𝑘𝑘𝑚𝑚 is the mass transfer coefficient in the gas phase, 𝐷𝐷𝐿𝐿 is the diffusivity of the gas phase, 𝐿𝐿𝐿𝐿 is a characteristic length (such as a bubble diameter), 𝑆𝑆𝑆𝑆𝐿𝐿 and 𝑅𝑅𝑅𝑅𝐿𝐿 are the Schmidt and Reynolds number of the gas phase. 3. Simulation details Two types of simulation were performed in this study to quantitatively analyze the operation of the T-HIDiC unit: • Conventional or classic process simulation (Aspen Plus®) • CFD Simulations (Ansys Fluent®) 3.1 Simulation in Aspen Plus Conditions used for this setup of the process simulation are based on our previous work (Mancera et al., 2018b). The T-HIDiC unit was simulated using Aspen Plus 10. The Propylene-Propane system properties including vapor-liquid equilibrium data were estimated with Peng-Robinson’s model. The simulations were performed in steady state, using isentropic efficiency in the compressor, and avoiding heat flow between system and surroundings. 3.2 CFD Simulation For the geometrical configuration of the tray, concepts of distillation column design and data available in the literature for calculating the diameter of holes and landfills were used (Lockett, 1986). The mains details of the sieve tray used are reported in Table 1. Table1. Geometrical parameters of the T-HIDiC simulated trays Geometry Stripping section Rectification section Dimension Dimension Outside tray diameter 0.32 m 0.2 m Inside Diameter tray 0.22 m -- Tray spacing 0.2 m 0.2 m Hole diameter 0.005 m 0.005 m Weir height 10mm 10mm % Bubbling area 82.46 % 82.46 % % hole area 8.33 % 8.48 % Solid material Steel Steel The column section considered for the CFD simulation is shown in Figure 1a and 1b. For the calculation of the temperature profile of the liquid within the simulation domain, 7 equidistant points were established along a central line in each plate (annular and concentric) at a height of y = 4 mm (Figure 1c), where the flow direction is from 1 to 7. 4. Methodology In this research, for CFD simulations a two steps methodology was established: 1. Classical process simulation of the T-HIDiC unit under thermodynamic equilibrium conditions. The objective of this stage is to obtain the macroscopic conditions for each tray (flows, temperature, 873 concentrations, equilibrium ratio), these will be used to set up the boundary and initial conditions in CFD simulations. The T-HIDiC is initially simulated using the equilibrium conditions with the classical process simulation approach on Aspen Plus®. This process simulation was implemented aiming to get the input values for the CFD model. The results of this stage were also used for the estimation of the heat transfer rate from the rectification zone to the stripping zone. The results at stages 2, 30, and 59 in the stripping section and their respective rectifying stages (stages 3, 31, and 60) are used for the setting up of the boundary and initial conditions of the CFD simulations (Table 2), described in the following section. 2. CFD simulation of the T-HIDiC unit under non-equilibrium conditions. The objective of this second step is to obtain the conditions for three characteristic trays of the HIDiC column considering non-equilibrium conditions, these results are used to estimate phase volume fraction, temperature profile, and Murphree Efficiency. a) b) c) Figure 1: Geometry of the tray used for the CDF simulation of the T-HIDiC: a) Isometric view of the stage, b) Transversal section of the stage, c) Sampling points to determine temperature profile of liquid. Yellow points (rectification), red points (stripping). 5. Results 5.1 Classic process simulation Results obtained from Aspen plus to be used in the CFD simulation are shown in Table 2 for trays 3, 31, 60 of the rectification column and trays 2, 30, and 59 of the stripping column. Equilibrium ratios are close to unity, resulting in very low relative volatility (α= [1.1 – 1.2]), which is a sign of the difficulty in separating this system. Table 2. Values of boundary conditions for CFD simulation at selected HIDiC stages. Area Stag e Inlet liquid compositio n (C3= molar fraction) Inlet vapor compositio n (C3= molar fraction) Inlet velocit y liquid (m/s) Inlet velocit y vapor (m/s) Inlet liquid temperatur e (K) Inlet vapor temperatur e (K) Molar equilibrium ratio Vapor/Liquid Propylen e (C3=) Propan e (C3) upper 3 R 0,9956 0.9954 0.4781 1.3961 307.8913 307.8944 1.0003 0.9168 2 S 0,4406 0.4544 0.4602 3.6125 300.4481 300.6408 1.0887 0.9335 middle 31 R 0,9878 0.9881 1.2071 3.0453 307.9298 307.9338 1.0011 0.9166 30 S 0,1118 0.1156 0.2716 1.960 303.4091 303.5473 1.1734 0.9796 bottom 60 R 0,9666 0.9671 1.7241 4.1949 308.0373 308.0499 1.0030 0.9164 59 S 0,0161 0.0154 0.1403 0.8896 304.4227 304.4599 1.2022 0.9970 5.2 CFD Simulation The formulated CFD model for the simulations under non-equilibrium conditions of the HIDiC unit was used for a detailed analysis of the heat integration in near-real conditions of three representative trays of the column. Phase volume fraction Vapor phase volume fraction is shown in Figure 2, the larger volume inside the trays is take up by the vapor phase. 874 Phase volume fraction Temperature contours a) a) b) b) c) c) Figure 2. Phase volume fraction obtained from the CFD simulations at the three sections of HIDiC: a) upper section, b) middle section and c) bottom section. Figure 3. Temperature contours for the tray at: a) upper section, b) middle section and c) bottom section. Temperature Using the CFD results, temperature contours for solid structure and liquid phase in the three simulated trays under non-equilibrium conditions could be analysed. The temperature contours of the tray surface are shown in figure 3 and 4. Murphree tray efficiency To compare the efficiency results given by the CFD model, a simulation of the T-HIDiC in non-equilibrium was also carried out in Aspen Plus® using the Rate-Based module with the conditions defined in Table 2. Results of this simulation and its comparison with CFD results are shown in Table 3. 875 Figure 4. Temperature profiles of the liquid at 4mm on the trays (y=4mm). a) rectification trays, b) stripping trays Table 3. Murphree efficiency results, comparison between CFD results and Rate-based model with process simulation Tray Section Stage Murphree efficiency CFD model (%) Murphree efficiency Rate-Based model (%) upper Rectification 3 41.76 82.45 Stripping 2 42.66 72.80 middle Rectification 31 57.65 78.23 Stripping 30 55.27 72.23 bottom Rectification 60 37.48 76.51 Stripping 59 40.77 69.31 6. Conclusions The results indicated that CFD is a tool capable of predicting reliably effects of column internals geometry on the multiphase flow field and can be considered as a useful aid for the design and evaluation of the performance of the T-HIDiC column. Acknowledgments The authors acknowledge the financial support of Universidad Nacional de Colombia and Colciencias. References Colombo, M., & Fairweather, M. (2015). Multiphase turbulence in bubbly flows: RANS simulations. International Journal of Multiphase Flow, 77, 222–243. https://doi.org/10.1016/j.ijmultiphaseflow.2015.09.003 Higbie, R. (1935). The rate of absorption of a pure gas into a still liquid during short periods of exposure. Transactions of the American Institute of Chemical Engineers, 35, 365–389. Grace, J. R., Wairegi, T., & Nguyen, T. H. (1976). Shapes and velocities of single drops and bubbles moving freely through immiscible liquids. In Trans Inst Chem Eng (Vol. 54, Issue 3) Lockett, M. J. (1986). Distillation tray fundamentals. Cambridge University Press. Mancera, J. A., Mendoza, D. F., & Riascos, C. A. M. (2018). HIDiC configuration selection based on exergetic analysis. Chemical Engineering Transactions, 69, 901–906. https://doi.org/10.3303/CET1869151 Mancera-Apolinar, J. A., Mendoza, D. F., & Riascos, C. A. M. (2023). CFD Simulation and Verification of Distillation Sieve Tray Hydrodynamics by Clear Liquid Height. Chemical Engineering Transactions, 100, 379–384. https://doi.org/10.3303/CET23100064 Olujic, Z., Fakhri, F., De Rijke, A., De Graauw, J., & Jansens, P. J. (2003). Internal heat integration - The key to an energy-conserving distillation column. Journal of Chemical Technology and Biotechnology, 78(2–3), 241– 248. https://doi.org/10.1002/jctb.761 Ranz, W. E., & Marshall, W. R. (1952). Evaporation from drops. Parts I & II. Chem. Eng. Progr, 48(22). Versteeg, H. K., & Malalasekera, W. (1995). An Introduction to Computational Fluid Dynamics The Finite Volume Method. In Fluid flow handbook. McGraw-Hill. Longman Scitific & Technical. https://doi.org/10.2514/1.22547 307.85 307.90 307.95 308.00 308.05 308.10 0 1 2 3 4 5 6 7 Te m pe ra tu re (K ) Point on the tray Upper Rectification Middle Rectification 300.0 301.0 302.0 303.0 304.0 305.0 0 1 2 3 4 5 6 7 Te m pe ra tu re (K ) Point on the tray Upper Stripping Middle Stripping 876 CET-vol117-b.pdf 92manceraapolinar.pdf Heat and Mass Transfer Study of Top-Heat-Integrated Distillation Column (T-HIDiC) using CFD