Modeling an industrial dea unit with the CPA EoS: a parameter estimation study and data analytics

Abstract

In this work, two alkanolamines, DEA and MEA, were studied, respectively, to model an acid gas treatment system to reduce H2S content and to select the most suitable Equation of State (EoS) to be used in this system modeling. The EoS SRK, PR and CPA were applied for pure MEA parameter estimation using pure substance data. The EoS CPA presented better results than the other two, with a mean absolute deviation from the vapor pressure of 1.694% and, therefore, was chosen as the main thermodynamic model for the process modeling. The CPA was then applied for pure DEA parameter estimation, generating mean absolute deviation values in relation to vapor pressure of 0.3%, which indicates proximity of the values calculated by the model with the experimental data. It was proceeded to estimate the parameters of the main binary systems present in an acid gas treatment unit and these were then inserted in the Petrox simulator. For the parameters estimation the ThermOptimizer software was used, applying the optimization methods PSO and Simplex. The results of the simulations were evaluated via comparison with operational data and results of available laboratory analyzes.