See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/334454536 Modeling and Simulation of a Water Gas Shift Reactor operating at a low pressure Article - July 2019 CITATIONS 3 3 authors: Wail El Bazi ENSA khouribga Maroc 14 PUBLICATIONS 33 CITATIONS SEE PROFILE M.S. KADIRI Ecole Nationale des Sciences Appliquées de Khouribga, Morocco 21 PUBLICATIONS 318 CITATIONS SEE PROFILE All content following this page was uploaded by Wail El Bazi on 12 January 2021. The user has requested enhancement of the downloaded file. READS 4,020 ResearchGate A. El Abidi Cadi Ayyad University - Ecole Nationale des Sciences Appliquées - Safi - Morocco 19 PUBLICATIONS 77 CITATIONS SEE PROFILE JOURNAL INTERNATIONAL LOF ENGINERING BESR International Journal of Innovation Engineering and Science Research AND SO INOVATION SCIENCE R SEARCH Modeling and Simulation of a Water Gas Shift Reactor operating at a low pressure Wail El Bazi¹*, Abderrahim El-Abidi²,³, Moulay Saddik Kadiri¹, Said Yadir²,3 ¹ Laboratory of Process Engineering and Industrial Systems Optimization (LIPOSI) ENSA Khouribga, Sultan Moulay Slimane University, Béni Mellal, Morocco *corresponding author: w.elbazi@usms.ma ² Laboratory of Materials, Processes, Environment and Quality (LMPEQ), ENSA Safi, Cadi Ayyad University, Marrakech, Morocco 3 Laboratory of Electronics, Instrumentation and Energy (LEIE), Faculty of Science, Chouaib Doukkali University, El Jadida, Morocco Open Access ABSTRACT In order to study the WGS on an industrial scale at a low pressure, the modeling and simulation of a WGS reactor operating at a pressure close to Patm and processing an industrial charge in the presence of a high temperature shift catalyst (Fe2O3/Cr2O3) were performed. The Profiles of the carbon monoxide conversion, temperature and pressure along the reactor were obtained. The effect of several operating parameters (inlet temperature, H₂O/CO ratio) on the conversion of carbon monoxide along the reactor has been determined. The estimated catalytic mass to convert 60.5% of the carbon monoxide contained in the inlet is 170.76 t. The pressure drops in the reactor are not negligible and the maximum temperature reached is without any harmful effect on the catalyst. The choice of an optimal inlet temperature and a high H₂O/CO ratio improves the conversion of carbon monoxide. Keywords- Packed bed reactor, catalyst, water gas shift reaction I. INTRODUCTION Hydrogen is an important source of energy and is involved in various industrial processes such as: ammonia synthesis, methanol synthesis, etc. [1]. The production of this molecule can be carried out through several ways: Production from methane [2], biological production [3], water electrolysis [4], chemical production of aluminum and sodium hydroxide [5], or WGS (water gas shift reaction) which is a chemical reaction converting a mixture of carbon monoxide (CO) and water vapor (H₂O) into a mixture of carbon dioxide (CO₂) and hydrogen (H₂): CO+H2O = CO2 + H2 AHR (298K) This slightly exothermic reaction discovered by the Italian physicist Felice Fontana in 1780 [1] can occur in the presence of catalysts based on several metals such as Cu, Fe, Ni, Pd, PT, Rh, Ru [6] or even metal oxide [7]. At high temperatures (350-450°C) catalysts based on iron oxides and chromium oxide III (Cr2O3) can be used [7]. At low temperatures (120-240°C) copper or copper oxide catalysts can be used with promoters of alumina oxide (Al2O3) [7] and zinc oxide (ZnO) [8]. WGS can occur at pressures ranging from 1 to 83.75 bars [ 7-10]. but often industrial reactors operate at high pressures [ 8,11-12]. Numerous models of the water gas shift reactors have been published to date. Elnashaie et al developed a steady-state one-dimensional heterogeneous model to study the behavior of industrial reactors operating at high temperatures [11]. Their work also focused on the effect of temperature on the conversion of carbon monoxide. Adams et al used a dynamic two-dimensional heterogeneous model to study the behavior of reactors operating at both high and low temperatures [13]. Adams was also Volume 2 Issue 6 November - December 2018 == 41.2 KJ/mol 47 | Wail El Bazi et al. "International Journal of Innovation Engineering and Science Research" interested in the effect of important parameters on Xco (H₂O/CO - temperature ratio). Falleiros Barbosa Lima et al investigated an industrial reactor operating at high temperatures, using different one- dimensional pseudo homogeneous models [14]. The effect of catalyst deactivation on reactor performance was also investigated. A steady-state one-dimensional pseudo-homogeneous model was also used by Shokry et al to predict the evolution of molar flow rate along and at the outlet of an industrial WGS reactor operating at low temperatures [12]. These authors also studied the effect of pressure, inlet temperature, H₂O/CO ratio on Xco. The major disadvantage of operating at high pressures is the enormous energy costs that it would be interesting to reduce them by carrying out the WGS at low pressures. In addition, the operating life of the catalyst can reach 15 years if the reactor operates at a low pressure [15]. Indeed, several papers have been carried out to study this reaction and to determine its kinetic expressions in the presence of catalysts allowing its activation at pressures close to the atmospheric pressure [8,16-23]. It would therefore be interesting to use these expressions in the simulation and study of the behaviour of reactors operating at pressures on the order of 1 atm. In fact, the work of Maklavany et al has been interested in this topic. These authors used the kinetic expression of WGS in the presence of a low temperature shift catalyst (CuO/ZnO/Al₂O3) to simulate a laboratory reactor operating at 1.2 bar. Two models were used: a steady state 2D homogeneous isothermal model [24] and a 1D homogeneous isothermal model with axial dispersion [25]. The effect of temperature on the variation of several parameters along the reactor (CO concentration, pressure, reaction rate, superficial velocity) was also studied. In our study, we also carried out the simulation of a reactor operating at a low pressure. But in our case, it is in the presence of a high temperature shift catalyst (Fe2O3 /Cr₂O3) and for a large reactor. This will allow the industrial-scale study of the reaction at a low pressure and clarified the strengths and weaknesses of the realization of WGS under this condition. The model we used to simulate the fixed catalytic bed is the steady state one-dimensional pseudo homogeneous model that is widely used in the study of the behaviour of industrial catalytic convertors, sites of the WGS [12,14,26] or other gas phase reactions [27,28,29]. In the first part of this study, we used the kinetic rate expression corresponding to this catalyst in the modeling and simulation of an industrial WGS reactor. This allowed the prediction of profiles of the carbon monoxide conversion, temperature and pressure along the reactor. Then, the work was continued by studying the effects of the operating parameters (inlet temperatures, H₂O/CO ratios) on the conversion of carbon monoxide in order to define the optimal conditions of the reaction realization. II. DESCRIPTION A. Description of the studied catalyst and the operating conditions for establishing the kinetic rate expression The kinetic rate expression corresponding to the chosen catalyst was carried out by Keiski et al [18]. The characteristics of the catalyst, as well as the operating conditions under which the kinetic rate expression was established, are presented in Table I [7,13,18]. TABLE I. Catalyst CCE C12 Ref: [7,13,18] CATALYST CHARACTERISTICS AND OPERATING CONDITIONS FOR THE ESTABLISHMENT OF THE WGS KINETICS [7,13,18] Equivalent spherical diameter, dp (μm) Composition Fe₂O3/Cr₂O3 (89/9%) Operating conditions for the establishment of WGS kinetics T: [575-675 K], P : 1 atm H₂O/CO (molar) : [2.4-12.1] Gas mixture of different fractions of: CO, CO2, H2O, Hz, Nz 2800 Volume 2 Issue 6 November December 2018 Shape Cylinder Catalyst density, Pc (kg/m³) B. Description of the reactor and the operating conditions considered in the simulation The characteristics of the simulated reactor are presented in Table II. 3730 48 | Wail El Bazi et al. "International Journal of Innovation Engineering and Science Research" Packed bed diameter D (m) 0.09 Packed bed length, L (m) 2.2 TABLE II. REACTOR CHARACTERISTICS Number of packed beds (tubes) 6000 The first simulations were run to predict the profiles of carbon monoxide conversion, temperature and pressure along the reactor operating at a low pressure and for a feed flow of an industrial nature. We ran these Simulations under the operating conditions presented in Table III. TABLE III. OPERATING CONDITIONS USED FOR THE PREDICTION OF X(Z), T(Z) AND P(Z) Molar flow rate of the feed gas mixture, F₁,0 Inlet pressure, Po Inlet temperature, To (mol. s-¹) (atm) (K) CO: 23.28 CO₂: 94.19 H₂ : 364.149 H₂O: 228.93 N₂: 134.354 Keq III. EQUATIONS Thermal property of the packed bed Adiabatic The compositions and the inlet pressure of the gas mixture of table 3 are close to those studied when establishing the kinetic rate expression of the WGS [18]. The temperatures remain within the temperature range of the kinetic study [18]. At the same time, in order to evaluate the WGS at low pressures in a situation close to an industrial case, each partial molar flow rate of WGS reagent (CO, CO2, H₂O, H₂) presented in Table 3 is equal to 30% of the actual partial molar flow rate of the same reagent feeding the high pressure WGS reactor of the Alexandria Fertilizers Company (AlexFert) [12]. The molar flow rate of nitrogen shown in this table is equal to 30% of the real molar flow rate of the inerts feeding the company's reactor. A. kinetic expression The kinetic expression used is a simple power-law model [7,18]: -rco (mol.kg cata¯¹. h¯¹) = 2623447 exp -79759 RT В = ex p To study the effect of temperature on the carbon monoxide conversion along the reactor, other simulations were performed for other inlet temperatures (575 K, 605 K, 620 K) while keeping the other operating conditions (Table 3) and the same reactor characteristics (Table 2). Finally, the study of the effect of the H₂O/CO ratio on the Xco profile along the reactor required further simulations for [H₂O]/[CO] = 3 which corresponds to Fo(H₂O) = 69.84 mol.s-¹ and Fo(CO) = 23.28 mol.s 1 and for [H₂O]/[CO] = 5 which corresponds to Fo(H₂O) = 116.4 mol.s-¹ and Fo(CO)= 23.28 mol.s-1. These last simulations were established by varying only the molar flow rate of the water vapor. While the other operating conditions (Table III) and reactor characteristics (Table II) have not been changed. 1.12 Volume 2 Issue 6 November December 2018 Сco2 CH2 1 Ссо Снчо Ке void fraction in the packed bed The equilibrium constant, Keq, is given by the equation 2 [7,18]: 4577.8 4.33) T Where -rco is the CO conversion rate, Ci is the molar concentration of species i (mol.dm-³) and ß is the reversibility factor: દ 0.4 590 0.74 C047C0218[1 − B] (2) (1) 49 | Wail El Bazi et al. "International Journal of Innovation Engineering and Science Research" B. Process modeling and numerical solution The following assumptions were made to develop the used mathematical model for the packed bed reactor simulation: Adiabatic reactor. ● ● Steady state condition. Axial dispersion is neglected, because the flow rate is sufficiently high to create a turbulent flow G.dp (Re: = > 40) and (L/dp>150) [30]. Where Re is the Reynolds number, G is the superficial mass velocity (kg.m-². s-¹) and µ is the dynamic viscosity of the mixture (Pa.s). μl Radial dispersion is neglected, because the tube diameter is narrow, the reactor is adiabatic and the WGS is a moderately exothermic. Under these conditions, radial gradients of concentrations and temperatures are not important [30,31,32]. The system of differential equations used is as follows [14]: dXco -coFPresPBS Fco,0 dz The heat and mass transfer as well as the diffusion in the catalyst were lumped in the rate constant. ● ● dT ΣFiCpi dz For the reagent: dp dz C₁ is expressed using the perfect gas equation: Pi RT For the product: For the inert: C₁ = -f² = Where Pi is the partial pressure (Pa), expressed as follows: P₁ = -fPgu ²³ dp - AHRTCO FPresPBS With yi is the molar fraction of the species i The expression of the molar flow rate, Fi (mol. s-1), of each component depends on its nature, be it a reagent, a product or an inert: F₁ = F₁,0 — Fco,o * Xco F₁ = F₁,0 + Fco,o * Xco F₁ = F₁,0 (3) Volume 2 Issue 6 November December 2018 (4) (5) (6) F₁. P = y₁.P i=n Σ=0 Fi (7) (8) (9) 50 |