High-pressure vapor-liquid equilibrium of hydrogen and diesel components: Maximum likelihood implicit estimation of equation-of-state binary parameters

Abstract

Due to increasing environmental requirements and efforts to improve fuel quality, hydro-processing has attained worldwide importance as a critical oil refining operation. Hydro-processing involve hydrogenation of oil fractions and subsequent separations at high-temperature, high-pressure and high hydrogen fugacity. Thus, effective thermodynamic modeling of hydro-processing systems allows accurate vapor and liquid property calculation to solve mass-energy hydro-processing balances. In this work, experimental data of high-pressure vapor-liquid equilibrium of binary systems consisting of hydrogen and Diesel hydrocarbons were compiled and used to statistically calibrate classical cubic equations-of-state and variants with different alpha functions. To accomplish this, the Maximum Likelihood Principle was used to conduct the implicit estimation of binary interaction parameters of hydrogen and each one of 22 hydrocarbons (paraffin, naphthenic and aromatic hydrocarbons) typical of Diesel fractions. Several statistical estimators were used to evaluate the estimated parameters as well as to test the insertion of a second binary interaction parameter or temperature dependence. In general, the FT Equation-of-State – i.e., the Peng-Robinson Equation-of-State using Fu & Tan alpha function for hydrogen – attained the best adherence to data and this adherence is improved by using two binary interaction parameters per pair or imposing temperature-dependence on the binary interaction parameter.