LAUSR

Although in Part I of this study (Chen and Yang 2020) we developed a tangent-line approach for effective density that is more general, robust, and flexible than the methods proposed by Saryazdi (2012) and Saryazdi et al. (2013), its application is only limited to heavy-oil/bitumen-associated mixtures [i.e., specifically, it has only been applied to bitumen-rich liquid phase (denoted as L2)]. As indicated in Part I, the density of nitrogen (N2)/hydrocarbon mixtures cannot be accurately predicted by using the ideal mixing rule (IM) with either real density or effective density. Not only do we need to explain and evaluate the observed deviations and patterns, but also the density prediction of solvent/Fraction 1 systems [i.e., Fraction 1 of the Athabasca bitumen, which has a molecular weight (MW) of 268.8 g/mol, as reported in Azinfar et al. (2018a, 2018b, 2018c)] needs to be improved for practical use. In this study, we evaluate the mixing characteristics of different molecules in a mixture using the tangent-line approach. By evaluating and comparing performances of the IM with effective density (IM-E) and the IM with real density (IM-R), the observed patterns and deviations together with those calculated from the Westman equation indicate that the oil/gas molecules somewhat behave like solid particles in mixing. Accordingly, we further modify the effective density used in the IM to bridge the gap between the IM-E and the IM-R. The database has been extended to light-oil/gas systems such as black oils, volatile oils, gas condensates, carbon dioxide (CO2) miscible fluids, sour gases, and wet/dry gases. The IM with modified effective density (IM-ME) has also been applied to solvent/Fraction 1 systems and the C2 or C3 or n-C4-extraction L1 phase (bitumen-related mixtures) with better accuracy. Also, we develop new criteria for the uses of the IM-E, IM-ME, and IM-R that can cover the density predictions for almost all types of oil/gas systems in the petroleum industry with high accuracy. The performances of the IM are thoroughly evaluated and compared with the volume-translated (VT) Peng-Robinson equation of state (EOS) (VT PR EOS), from which the deviations provide new insights for accurately quantifying the mixture density in a more robust and reliable manner.