Using Neural Networks for Modeling Dehydration of Organic Solutions Using Pressure Swing Adsorption Process

Abstract

A major challenge in the production of amines is the high energy cost associated with the separation of amines from the large excess of water. The Pressure Swing Adsorption (PSA) process is attractive for the final separation since it requires little energy input and is capable of producing a very pure product. At this research, a pilot-scale PSA process was designed and manufactured. The possibility separation of amines -water mixtures by adsorption on 3A Nano pore zeolites has been investigated. The column was operated under the following operating variables: feed concentration (85-90-95%), adsorption temperature (85-125◦C), and adsorption pressures (0, 0.5, and 1.5 bar). The water concentration, temperature, and pressure of process parameters have been varied to establish their influence on the PSA performance. The effect of these variables on the enrichment percentage of the product was studied. In this study, neural network modeling was used to predict the experimental results and showed that the genetic algorithm can predict the experimental results with the least error.