Solubility Parameter of Methane in Bitumen

Abstract

Methane solubility parameters are required when using a regular solution approach to predict asphaltene precipitation from crude oils that contain methane. To determine these parameters, asphaltene onsets and yields from mixtures of bitumen, methane, and n-pentane were measured at temperatures of 21 and 130 °C and pressures of 10 and 60 MPa. The onsets (solvent content at which asphaltene precipitation was first observed) were measured by titrating bitumen with a mixture of methane and n-pentane in a high-pressure microscope. Asphaltene yields (mass of asphaltenes divided by mass of bitumen in feed) were measured in a blind cell pressure-volume-temperature apparatus for mixtures with the same methane content as the onset measurements. Saturation pressures were also measured in the blind cell apparatus for the mixtures with methane and were used to ensure that the onset and yield measurements were performed in a vapor-free condition. The onset and yield data were fitted with a Modified Regular Solution (MRS) model. The inputs to the MRS model are temperature, pressure, and the mole fraction, molar volume, and solubility parameters of each component in the mixturein this case, methane, n-pentane, and the bitumen characterized into saturates, aromatics, resins, and asphaltenes. The properties of the nonasphaltene bitumen components and n-pentane were determined from previously established correlations, and the only unknowns were the asphaltene and methane solubility parameters. The respective unknown solubility parameters were determined by fitting the MRS model to asphaltene precipitation data from mixtures of bitumen and n-pentane and from mixtures of bitumen, n-pentane, and methane. The fitted methane solubility parameters ranged from 6.1 to 9.5 MPa0.5 depending on the temperature. A correlation for the parameter was proposed, and the MRS model with this correlation matched the measured onsets and yields with average deviations of 1 wt % solvent and 8 wt % compared with the experimental uncertainties of ±1.5 wt % solvent and ±6 to 11 wt %.