|
Abstract
|
The aim of this study is to investigate the effects of key operating parameters of packed absorption columns on
the performance of mass transfer considering the volumetric overall mass transfer coefficient (K aG V). The effects
were studied for CO2 absorption by using 4-diethylamino-2-butanol (DEAB) and N,N-Diethylethanolamine
(DEEA) mixed with monoethanolamine (MEA) as novel amine solutions. In doing so, an optimization–simulation
framework was developed based on the two-film theory model, thermodynamic model, multi-layer perceptron
neural network (MLPNN), and statistical technique. To predict the CO2 loading as one of the parameters in input
of MLPNN model, the Deshmukh–Mather model, as an electrolyte thermodynamic model, was developed for
CO2 DEAB H2O and CO2 DEEA MEA H2O systems. The effect of enhancement factor on the K aG V
was considered based on the series resistances model including pseudo-first order enhancemnt factor and instantaneous
enhancemnt factor. To optimize and rank the key operating factors that simultaneously affect the
K aG V, the Taguchi method was used. Statistical indices showed that our model could efficiently predict the
experimental data with AARDs of 5.6%, 0.43% and 4.96%, respectively, for K aG V data, CO2 loading data of
DEAB and DEEA MEA. A significant order of process variables affecting the K aG V values was as CO2 mole
fraction > amine temperature > amine flow rate > gas temperature > packing type > CO2 loading >
amine concentration. Moreover, the sensitivity analysis results showed that by increasing the CO2 mole fraction
in gas feed, gas temperature, and CO2 loading, the K aG V values decreased, and by increasing the amine concentration,
amine flow rate, and amine temperature, the K aG V values increased.
|