Prediction of hydrogen solubility in PYGAS with equations of state

Abstract

The objective of this contribution was to explore the effect of thermodynamic model selection on predicting the solubility of hydrogen in Pyrolysis Gasoline (PYGAS). In order to do this, different combinations of cubic equations of state (EOS), cohesion functions and mixing rules were compared, using both own algorithms and those implemented in the PRO/II 8.0 process simulation software. The models were selected following the recommendations made by some authors for H₂ + hydrocarbons systems similar to the studied here. The cubic EOS of Redlich-Kwong (RK) and Peng-Robinson (PR) were considered using: five different cohesion functions, and the classical quadratic mixing rule (CQMR) with five single correlations to estimate the binary interaction parameters (BIP). Additionally, the system was simulated with PRO/II employing the Panagiotopoulos-Reid Modified (PA-RE-Mod) and the modified PA-RE mixing rule implemented by SimSci. The results were validated with experimental data reported in the literature for four temperatures 298 K, 313 K, 323 K, 343 K and pressures up to 6 MPa. The model composed by the RK EOS, Soave (1993) cohesion function, the CQMR with the BIP calculated with the Gray et al. correlation is the recommended selection to predict the solubility of H₂ in PYGAS, under the considered conditions, with a 3.1% AARD (average absolute relative deviation).