Application of the Modified Regular Solution Model to Crude Oils Characterized from a Distillation Assay

摘要

The modified regular solution model was developed to predict onset and amount (yield) of asphaltene precipitation from mixtures of crude oil and solvents. The most recent version of the model includes the partitioning of all components between a solvent-rich phase and an asphaltene-rich phase, where the solvents are pure components with known properties and the crude oil is represented as pseudo-components defined on the basis of a saturate, aromatic, resin, and asphaltene (SARA) assay. The molecular weight, density, and solubility parameter of each pseudo-component are determined from correlations. This model is sensitive to uncertainties in the SARA assay composition and is not compatible with the standard phase behavior modeling methodology based on pseudo-components defined from a distillation assay. In this contribution, the model was extended to crude oil pseudocomponents based on boiling cuts (TBP). The n-pentane-insoluble asphaltene fraction was characterized in the same way as the SARA-based model. The molecular weight and density of the TBP-based pseudo-components were predicted from well-established correlations. New correlations were proposed for the maltene pseudo-component solubility parameters as a function of the temperature and pressure. The TBP-based model was tested on two data sets: (1) onsets, yields, and phase compositions for a Western Canada bitumen mixed with n-alkanes from propane to n-heptane at temperatures and pressures up to 250 degrees C and 13.8 MPa, respectively, and (2) yields from eight oils from different geographical regions mixed with n-alkanes from n-pentane to noctane at temperatures and pressures up to 100 degrees C and 6.8 MPa, respectively. The overall absolute deviation was 2 wt % for both data sets, similar to the deviations found for the SARA-based model.