Análise de unidades de pré-purificação de ar por adsorção

Abstract

The main constituents of air, nitrogen, oxygen and argon, are increasingly present in industries where they are employed in chemical processes, to transport food and waste processing. The two main technologies for the separation of air components are the adsorption and cryogenic distillation. However, for both processes is necessary to remove air contaminants, such as carbon dioxide, water vapor and hydrocarbons, to avoid operational problems and safety concerns. Thus, this work deals with the study of air pre-purification using adsorption. In this system the air current flows alternately between two adsorbing beds to produce clean air continuously. More specifically, the focus of the dissertation is to investigate the behavior of pre-purification units, PSA (pressure swing adsorption), where the desorption step is accomplished by reducing the pressure. The analysis of pre-purification unit begins with adsorption beds modeling through a partial differential equations system of mass balance in the gas stream and in the bed. In this model, the adsorption equilibrium is described by the Dubinin-Astakhov isotherm extended to multicomponent mixtures. For the simulation model, the spatial derivatives are discretized via finite differences and the ordinary differential equations system resultant is solved by an appropriate solver (method of lines). To the operating unit simulation, this model is coupled to a convergence algorithm on the four phases of the operation: adsorption, depressurization, purge and desorption. The algorithm in question must ensure that the final terms of the final stage are equivalent to the initial conditions of the first stage (cyclic steady state). Thus, the simulation was implemented in the form of a computational code based programming environment Scilab (Scilab 5.3.0, 2010), which is a program of free distribution. The simulation algorithms of each individual step and the complete cycle are finally used to analyze the behavior of pre-purification unit, checking how their performance is affected by changes in design variables or operational. For example, we have investigated the charging system of the bed and showed that the optimal configuration of the bed is 50% alumina followed by 50% of zeolite. Process variables were also examined, the adsorption pressure, the flow rate and cycle time of adsorption, showing that increasing the feed flow rate leads to a loss of specification that can be taken up by reducing the cycle time of adsorption. It was also shown that a higher pressure adsorption leads to a greater removal of contaminants.