Abstract-- This work reports preliminary results on the fluid dynamic simulation using ANSYS CFD software of a pilot scale reactor experimentally tested for the removal of VOCs (toluene, ethylbenzene, methyl ethyl ketone and xylenes). The reactor is constructed from the modification of a commercial air stripper and has been successfully evaluated for the removal of VOCs using Mn, Ce and Pt based catalysts on ceramic monoliths (bentonite). By means of ANSYS CFD simulation it was possible to represent the flow in the reactor (validating with experimental data) and to simulate the flow in the monolith channels. This information will allow us to have a better understanding of the fluid dynamics in the catalysts and to simulate the reactions with different VOCs by combining ANSYS CFD with ChemKin-PRO.
Keywords-- ANSYS, Combustion, Monoliths
I. INTRODUCTION
Volatile organic compounds (VOCs) are important anthropogenic pollutants that tend to have high levels in urban and industrial areas. They are also important indoors because humans spend about 80% of our time indoors, and some authors (Guo, 2002; Liu et al., 2009; Nevers, 1997; Sun Kou et al., 2008) report that the concentration of VOCs can be several times higher than outdoor levels. Indoor sources of VOCs include glues, cleaning and pharmaceutical products, perfumes, cigarettes, food cooking, paints and solvents (Sun Kou et al., 2008).
At the industrial level different VOC removal technologies are used including thermal and catalytic incineration, biological, adsorption, condensation, etc., while to reduce the concentration of VOCs in indoor spaces there are three ways: (i) eliminate the source, (ii) increase the air flow, i.e. dilute the concentration of the pollutant, which in many situations is not possible and (iii) purify the air, usually by forcing it to pass through an adsorbent, which must be subsequently treated or recovered or eliminate the VOC by catalytic combustion (Nevers, 1997). Structured catalytic reactors are of increasing interest in VOC removal due to their high heat/mass transfer coupled with low pressure drops compared to packed bed reactors, coupled with a large geometric surface area and high catalyst efficiency factor (Liu et al., 2009; Mei et al., 2005; Miller and Han, 1971; Phan et al., 2011; Presti et al., 2002; Roy et al., 2004; Shah and London, 1978; Specchia et al., 2011).
Computational simulation of chemical processes in a catalytic reactor is a suitable tool and an easy way to study and predict the effect of the parameters that influence a process. Computational fluid dynamics (CFD) is a theoretical way to investigate and predict the performance of processes related to fluid, heat and mass transfer. Computational fluid dynamics (CFD) brings surprising flexibility, accuracy, and breadth of application. ANSYS CFD, although a commercial software, has significant flow analysis capabilities, with a high-level interface, linking all computational modules (geometry - meshing - solver - post-processing). ANSYS contains a powerful and reliable toolset that complements conventional reactor engineering practices. It has capabilities that span fluid mechanics, structural mechanics, impact and safety analysis, custom mixing tools, pressure vessel analysis, electromagnetic and control systems, and heat and mass transfer analysis. ANSYS can model a wide range of reactions, including gaseous and liquid, single and multiphase, homogeneous, and heterogeneous. The ANSYS Fluent module is the fluid simulation software used to predict fluid motion, heat and mass transfer, chemical reactions, and other related phenomena. Known for offering the industry's most accurate solutions without compromise, Fluent's advanced physical modeling capabilities include state-of-the-art turbulence models, multiphase flows, heat transfer, combustion, shape optimization, multiphysics and more. ANSYS CFD, can be used for a wide variety of single and multiphase problems, including flow through packed beds, filter papers, perforated plates, flow manifolds, and tube banks (Atmakidis and Kenig, 2009; Augier et al., 2010; Borbon et al., 2003; Pavlisic et al., 2018; Wehinger et al., 2017; US EPA 1996].
This paper presents preliminary results of the ANSYS CFD fluid dynamic simulation of a pilot scale reactor, experimentally tested for the removal of VOCs (toluene, ethylbenzene, methyl ethyl ketone and xylenes). The reactor is constructed from the modification of a commercial air stripper and has been successfully evaluated for the removal of VOCs using Mn, Ce and Pt based catalysts on ceramic monoliths (bentonite) (Colman Lerner, 2013).
Figure 1. Figure 1: (a) pilot scale reactor; (b) monolith and monolith support views.
Figure 2. Geometry of the computational fluid dynamic domain (a) full view, (b) zoom to monoliths.
II. METHODS
A. Pilot scale reactor
The simulated reactor is an air purification system (Spar brand), to which a chimney was adapted with a catalytic system heated by a system of electrical resistances in the exhaust duct (Fig. 1a). The construction of the system (iron duct with a surrounding insulating chamber to avoid heat losses) was carried out to achieve approximately 10% of the minimum flow rate in SMEs industries, which is estimated at 0.33 m3 s-1 (Colman Lerner, 2013), with a gas volumetric flow rate of 1.3 m3 min-1, calculated according to ASTM D3154 standards (Colman Lerner, 2013). At the outlet of the extractor and prior to the catalytic system, a honeycomb arrangement was incorporated to achieve a flow without preferential directions. The ceramic monoliths (Fig. 1b) used in this work were prepared from Al(OH)3, Alcan Gibbsite and a Patagonian bentonite in a 0.7:0.1:1 ratio. The monoliths were cylindrical (1.8 cm diameter and 2 cm height) with triangular channels (cell density of 3.1 cells cm2), impregnated with Pt, Mn, Ce, placed on aluminum support. When evaluated in the combustion of toluene, ethylbenzene, methyl ethyl ketone and xylenes, conversions (combustion to CO2) higher than 90% were achieved below 300°C (Colman Lerner, 2013).
Table 1. Physical properties of the materials that make up the reactor used to simulate.
B. Simulation with ANSYS CFD
The fluid dynamic simulation of the reactor (decoupled from the combustion reaction) under experimental operating conditions was carried out using ANSYS CFD 19.3 software (academic license) on a cluster of 1 server and 2 nodes (HP ProDesk 400 G3, Intel® CoreTM i5-6500, 8 GB SDRAM, 1Tb). The geometry was designed in SpaceClaim (Fig. 2). The mesh was constructed using ANSYS Meshing, applying a tetrahedral mesh of 79389 elements of good quality according to the Skewness and Orthogonal Quality statistics. The mesh was lifted with ANSYS Fluent, running a steady-state modeling with a RANS k-ɛ turbulence model realizable with the enhanced wall treatment option. A pressure-velocity coupling method with COUPLED algorithm, with second-order spatial discretization for all variables, was used to solve the equations. The duct inlet velocity profile (prior to the passage of the monoliths) was plotted parabolic according to the experimental data. The thermodynamic conditions were represented with convection models. The results obtained were validated with experimental measurements of velocity and temperature (before and after the monolith support) during the catalytic reaction tests. The physical properties of the different are shown in Table 1.
III. RESULTS AND DISCUSION
The simulation was carried out for two scenarios, (1) with reactor inlet temperature (Tin) of 100 °C and, (2) with Tin of 300 °C, both with experimentally measured parabolic inlet velocity profiles (max. 2.61 m/s) (Figure 3a). Figures 3b to 3d show the results obtained for the velocity distribution in both scenarios. Figures 4 and 5 show the results for the temperature distribution in scenarios 1 and 2, respectively. The results obtained with the simulation were compared with experimental values of velocity and temperature measured in the reactor, obtaining similar results that have been validated with the FB (Fractional Bias) = (Co-Ce)/ [0.5* (Co+Ce)], where Co represents the observed values and Ce the estimated values, values of FB=0 represent simulations in accordance with reality. They are presented in Table 2.
Vout represents the velocity observed in the central axis of the duct, 10 cm from the drum outlet. Tout was observed in the central axis of the duct, at the exit of the drum (monolith support). Vin and Tin were observed at the entrance of the duct, before crossing the monolith channels.
Inside the channels, the average velocity obtained in the simulation is 1.54 m/s, with a maximum value of 13.36 m/s, and temperature of 277 °C (Fig. 6).
Table 2. Experimental and simulation results: simulated (S), experimental (E)
|
|
E1 |
S1 |
E2 |
S2 |
|
Tin (°C) |
100 |
100 |
300 |
300 |
|
Tout (°C) |
75 |
77 |
230 |
242 |
|
FB |
-0.03 |
-0.05 |
||
|
Vin (m/s) |
2.70 |
2.61 |
2.70 |
2.61 |
|
Vout (m/s) |
1.20 |
1.24 |
1.20 |
1.24 |
|
FB |
-0.03 |
-0.03 |
||
Figure 3. (a) Input velocity profile; (b)-(d) velocity distribution results obtained by simulation with ANSYS CFD.
Figure 4. Temperature distribution obtained by simulation with ANSYS CFD, for scenario 1.
Figure 5. Temperature distribution obtained by simulation with ANSYS CFD, for scenario 2
Within the channels, velocity profiles are formed due to frictional effects with the walls, with an increase in velocity being observed in the central part of the channel section, minimized at the corners of the triangular channel. It is observed that passing the middle of the channel the maximum velocity decreases, obtaining a flat and uniform velocity profile towards the exit of the channel, agreeing with what were obtained by Iwaniszyn et al. (2017).
Figure 6. Velocity profile (m s-1) over different coss-seccctions along of monolith chanel
Figure 7. Temperature (k) profile over different coss-seccctions along of monolith chanel
According to Iwanisyn, the effects caused by the thickness of the channel walls at the inlet cause transverse components in the velocity vector contributing to the formation of stagnant zones and turbulence, which correspond to the velocity profiles obtained in this work and the experimental results of catalytic conversion.
The temperature profiles of the gas stream along different sections of the channel are shown in Figure 7, showing similar behavior to the velocity profile.
This preliminary study simulates and validates the fluid dynamics of the pilot scale reactor already successfully tested experimentally as a first step.
With these results (velocity and temperature profiles) we can deepen the simulations inside the channels and introduce the combustion reactions with their respective kinetic equations using the chem-kim PRO module of the ANSYS package.
V. CONCLUSIONS
Numerical simulation with ANSYS CFD was used to represent the thermal and fluid dynamics of a pilot scale reactor experimentally tested for the combustion of volatile organic compounds (VOCs) at two operating temperature conditions (100 and 300 °C).
The CFD simulation results showed excellent agreement with the experimental results for the fluid dynamic and thermal simulation of the reactor at operating conditions, with absolute values of FB between 0.03 and 0.05.
With the good agreement with the experimental data, it was possible to obtain velocity and temperature profiles inside the monolith channels that explain the good performance of the reactor as an intramural air purifier.
These advances will allow in the next steps to include kinetic parameters and to perform simulations of catalytic combustion reactions with the ChemKin PRO module and to study different support geometries to improve the reactor performance at pilot scale.
REFERENCES
Augier, F., Idoux, F. and Delenne, J.Y. (2010) Numerical simulations of transfer and transport properties inside packed beds of spherical particles. Chemical Engineering Science. 65, 1055-1064.
Atmakidis, T. and Kenig, E.Y. (2009) CFD-based analysis of the wall effect on the pressure drop in packed beds with moderate tube/particle diameter ratios in the laminar flow regime. Chemical Engineering Journal. 155, 404-410.
Borbon, A., Fontaine, H., Locoge, N., Veillort, M. and Galloo, J. (2003) Developing receptor-oriented methods for non-methane hydrocarbon characterisation in urban air—Part I: source identification. Atmosp. Environm. 37, 4051–4064.
Colman Lerner, J.E (2013) Contaminación Ambiental Análisis y Mitigación/Remoción de Material Particulado (Mp) y Compuestos Orgánicos Volátiles (Covs) Y Semivolátiles (COSVS). Tesis Doctoral, Facultad de Ciencias exactas, UNLP.
Guo, Z. (2002) Review of indoor emission source models. Part 1. Overview. Environmental Pollution. 120, 533-549.
Iwaniszyn, M., Piątek, M., Gancarczyk, A., Jodłowski, P. J., Łojewska, J. and Kołodziej, A. (2017) Flow resistance and heat transfer in short channels of metallic monoliths: Experiments versus CFD. International Journal of Heat and Mass Transfer. 109, 778-785.
Liu, W., Hu, J. and Wang, Y. (2009) Fischer-Tropsch synthesis on ceramic monolithstructured catalysts. Catal. Today. 140, 142–148.
Mei, H., Li, C. and Liu, H. (2005) Simulation of heat transfer and hydrodynamics for metal structured packed bed. Catal. Today. 105, 689–696.
Miller, R.W. and Han, L.S. (1971) Pressure losses for laminar flow in the entrance region of ducts of rectangular and equilateral triangular cross section. J. Appl. Mech. 38, 1083–1087.
Nevers, N. de (1997) Ingeniería de control de la contaminación del aire, Ed. Mc Graw Hill.
Pavlišič, A., Pohar, A. and Likozar, B. (2018) Comparison of computational fluid dynamics (CFD) and pressure drop correlations in laminar flow regime for packed bed reactors and columns. Powder Technology. 328, 130-139.
Phan, X.K., Bakhtiary-Davijany, H., Myrstad, R., Pfeifer, P., Venvik, H.J. and Holmen, A. (2011) Preparation and performance of Cu-based monoliths for methanol synthesis. Appl. Catal. A. 405, 1–7.
Presti, M., Pace, L., Hodgson, J., Bella, G. and De Maio, A. (2002) A computational and Experimental Analysis for Optimization of Cell Shape in High Performance Catalytic Converters. SAE Paper. 2002–01-0355.
Roy, S., Heibel, A.K., Liu, W. and Boger, T. (2004) Design of monolithic catalysts for multiphase reactions. Chem. Eng. Sci. 59, 957–966.
Shah, R.K. and London, A.L. (1978) Laminar Flow Forced Convection in Ducts. Academic Press, New York.
Specchia, S., Tacchino, S. and Specchia, V. (2011) Facing the catalytic combustion of CH4/H2 mixtures into monoliths. Chem. Eng. J. 167, 622–633.
Sun Kou, M., Montes, M., Picasso, G. and Sambeth, J. (2008) Eliminación de emisiones atmosféricas de COVs por catálisis y adsorción. Ed. CYTED.
US-EPA (1996) Consumer and Comercial Solvent Use. www.epa.gov/ttn/chief/techreport/volume3/ iii05.pdf
Wehinger, G.D., Klippel, F. and Kraume, M. (2017) Modeling pore processes for particle-resolved CFD simulations of catalytic fixed-bed reactors. Computers & Chemical Engineering. 101, 11-22.
Received: August 1, 2022
Sent to Subject Editor: August 1, 2022
Accepted: December 16, 2022
Recommended by Subject Editor Laura Briand