Modeling and optimization of CO2 capture using 4-diethylamino-2-butanol (DEAB) solution

Abstract

A multi-layer perceptron neural network (MLPNN) model with Levenberg–Marquardt learning algorithm were applied to model CO2 capture by a novel amine solution called 4-diethylamino-2-butanol (DEAB). The MLPNN model predicted the CO2 concentration and temperature profiles along the height of the packed column as the model output. Inlet feed conditions of the absorber column (flue gas and amine) were selected as the inputs of the MLPNN model. Experimental data about random and structured packed columns were extracted from the literature and used to train the MLPNN model. In addition, a systematic procedure, i.e. Taguchi method, was applied to obtain the significant sequence of process parameters affecting CO2 removal efficiency and to optimize the variables in the absorber column. Five levels of five variables, including lean amine temperature, amine concentration, CO2 loading of amine, gas temperature, and amine flow rate, were used for the optimization of the absorber column. The average absolute relative deviations (AARD) between the predicted results and the experimental data suggested that our MLPNN model could predict CO2 concentration and temperature profiles along the packed column (AARD%=5.47 and 3.61, respectively). The signal to noise ratio analysis of the Taguchi method yielded a significant sequence of factors affecting CO2 removal efficiency in the packed column (CO2 loading>amine flow rate>amine concentration>gas temperature>amine temperature). This study demonstrated the acceptable accuracy of the MLPNN and Taguchi method in, respectively, the modeling and optimization of CO2 capture in amine solutions.