Cubic equations-of-state (EOS) mathematically relate pressure, volume, and temperature. These equations describe the volumetric and phase behavior of single components and mixtures requiring only the critical properties and acentric factor of each component including the plus-fraction. For pure compounds, the required properties are well defined; however, nearly all naturally occurring gas and crude oil fluids contain heavy fractions that are not well defined. These heavy plus-fractions are mixtures of hundreds of paraffinic, naphthenic, aromatic, and other organic compounds that cannot be discretely identified and described. These heavy fractions are often lumped together and called the "plus-fraction" (e.g. C7+ fractions). Adequately characterizing these undefined plus fractions in terms of their critical properties and acentric factors has been long recognized as a challenge in compositional analysis using EOS models. Changing the characterization of the plus fraction can have a significant impact on the volumetric and phase behavior of a hydrocarbon mixture predicted by EOS models. In applying EOS in compositional modeling and PVT analysis, empirical correlations are used to estimate critical properties and acentric factors of the plus-fractions. Coefficients for these empirical expressions were generated by matching critical properties of pure components; however, the applications of these empirical relationships are extended to estimate the critical properties the plus-fractions.
This paper presents a practical approach for calculating EOS parameters "a, b, and α" of the plus-fraction from its readily available measured physical properties, i.e., molecular weight "M" and specific gravity "γ". Our objective is to improve the predictive capability of EOS models. A new model was developed based on the Peng-Robinson Equation of State "PR EOS" and with the modification as outlined in this paper. The predictive capability of the modified PR EOS is displayed by matching a set of laboratory data on several crude oil and gas-condensate systems. Performance of the proposed method was also compared with predictive PVT results generated by two commercial compositional simulation models. Additional validations were made by comparing results from applying the proposed approach with those of Coats and Smart4 regression methodology with PR EOS.