Otimização da produção de hidrogênio em um reator com membrana

• Ano de defesa: 2015
• Autor(a) principal: Souza Netto, Leonardo Dantas de
• Orientador(a): Pagano, Rogério Luz
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Programa de Pós-Graduação: Pós-Graduação em Engenharia Química
• Palavras-chave em Português: Reatores químicos; Hidrogênio; Ansys (Sistema de computador); Reator com membrana; Produção de hidrogênio; Reforma a vapor do metano; Convergência de malha; Computacional Fluid Dynamics (CFD); Particle Swarm Algorithms (PSO)
• Palavras-chave em Inglês: Reactor with membrane; Hydrogen production; Methane vapor reform; Mesh convergence
• Área do conhecimento CNPq: ENGENHARIAS::ENGENHARIA QUIMICA
• Link de acesso: http://ri.ufs.br/jspui/handle/riufs/17116

Hydrogen has its main production from steam reforming reaction. Due to the characteristics of this reaction, to achieve reasonable conversions are required high temperatures and pressures in conventional reactors. The literature presents proposals for using reactors with membrane in order to shift the thermodynamic equilibrium, favoring higher operating temperatures in smaller conversions. In this scenario, the present work proposes the simulation of a reactor with membrane for hydrogen production from steam reforming reaction through the software ANSYS/CFX® and the optimization performed by Particle Swarm method implemented via Fortran. Initially, the geometry at the CFX was elaborated, generated mesh that will present the reactor with membrane and then made the whole setup of the ballast conditions, dimensions, boundary conditions and expressions that describe the process. A study of convergence of mesh it was performed to determine the invariability of the results obtained, comparing meshes with 1944, 980, 480, 209 and 104 elements. The invariability of the meshes is quantified by the mesh convergence index, GCI, the lowest values of GCI obtained were between the meshes with 980 and 480 elements, with GCIs like 0.0128 and 0.0436 reaction zones and drag, respectively. The model was validated based on experimental data available in the literature, evaluating the variation of methane conversion as a function of pressure and operating temperature and molar ration of food between water vapor and methane (RVM). The results obtained corroborate with the experimental results presented in the literature. Then, the method of Particle Swarm Optimization (PSO) was implemented in the programming language Fortran and created an interactive method between Fortran and the CFX, where the great values are estimated by the algorithm and added to the CFX as process variables, determining that way, after the solution of the model, the objective function, which in this case was used to sum the conversion of methane to hydrogen recovery. Anyway, it was possible to determine the optimal operating conditions, P = 121.325 kPa, T = 826.32 K and RVM = 2.77, maximizing conversion of methane and hydrogen recovery reaching values of approximately of 100% for both.