In this work, GERG2008 EoS embedded in volumetric-gravimetric system was developed to allocate multicomponent partial/individual components uptakes in the mixture. The refined arrangement may overlay the current theoretical anticipated outcomes and interchange it with experimental and more trustworthy selective adsorption outcomes. 13X zeolite was utilized as a solid adsorbent for binary and ternary CO2:CH4:H2O mixtures adsorption. Premixed and preloaded water vapor was studied at 343 K and up to 10 bar. Artificial neural network (ANN) modeling was engaged to predict binary and ternary mixtures. ANN results disclosed decent promise with experimental data. Besides, simulated formations utilizing ANN model replicated high consistency. The testified outcomes magnificently identified particular components behavior in ternary and higher multicomponent mixtures. © 2017 Elsevier Inc.