Prediction of phase behaviors of acetic acid containing fluids

• Prediction of VLE and LLE of acetic acid containing mixtures from PR + COSMOSAC EOS.
• The association of acetic acid to different forms of dimers is explicitly considered.
• Transition between monomer and dimers is taken into account via chemical reactions.
• Equilibrium constants obtained from the vapor pressure of acetic acid.
• Accurate phase behavior of mixtures can be made without further parameterization.

The modeling and prediction of fluid phase equilibria of acetic acid and its mixtures with other chemicals have long been a focus of research. The formation of specific local fluid structures, such as the dimers and/or hydrogen bonding network, makes it a challenging issue. The successful modeling of such systems requires a proper description of the distribution and transition among different fluid structures as the temperature, pressure and/or composition of the system changes. In this work, we identify three important fluid structures in mixture containing acetic acid: the monomer of acetic acid, the cyclic dimer, and the chain fragment. The distribution and interchange of acetic acid among the three molecular arrangements is considered by chemical reactions. With the explicit inclusion of these three important local fluid structures, the PR + COMOSAC equation of state (EOS) is capable of predicting the phase behaviors of pure acetic acid, and its mixtures with other organic compounds. The proposed method is as accurate as the Hayden-O’Connell EOS in vapor phase and is readily applicable to the liquid phase and mixtures. The prediction accuracy in describing the vapor–liquid equilibrium of 15 binary mixtures (388 data points, temperature range from 293.15 to 373.2) is 7.10% (AARD-P%) in pressure and 2.38% (AAD-y%) in vapor composition, which is similar to those from the most accurate group contribution method (e.g. the mod-UNIFAC + HOC model, 5.14% and 1.85%). The root-mean square error in predicting liquid composition in liquid–liquid equilibrium of 5 binary systems (51 data points, temperature range from 289.15 to 337 K) is found to be 0.091, which is more accurate than that from the group contribution method (e.g. the mod-UNIFAC model, 0.173).