Experimental evaluation and neural networks modeling of removal efficiency and volumetric mass transfer coefficient for gas desulfurization in spray tower

Abstract

Spray towers to can be used to absorb pollutants gases. Due to the complexity of the process, to propose mathematical models for prediction of removal efficiency and volumetric mass transfer coefficient, is a difficult task. This work aimed to evaluate the influence of variables on the SO2 removal process in a spray tower and to obtain an artificial neural network (ANN) to predict the removal efficiency and the volumetric mass transfer coefficient for columns with different heights, gas and liquid flow rates. The tower built in this study reached a maximum efficiency of 99.47% and showed that it was possible to achieve high levels of removal without the need to build high towers. The best ANN obtained presented error of 2.76% for the removal efficiency of the SO2 and 6.79% for the volumetric mass transfer coefficient. The obtained results were very promising in the prediction of important parameters in the process of removing atmospheric pollutants via spray tower.