455097 Predicting the Viscosity of Characterized Crude Oils
The Expanded Fluid (EF) viscosity model  is a semi-empirical approach based on the observation that viscosity decreases as a function of density as a fluid expands from an infinite viscosity “glassy state” to the dilute gas state. The inputs of the EF model are the density of the fluid, pressure, and three fluid-specific parameters: c2, ρso and c3. The EF model is continuous across the phase diagram and the critical point and has been successfully tested on conventional and heavy oils, distillation cuts and pure hydrocarbons. The model was extended to mixtures using mass-based mixing rules with non-zero viscosity binary interaction parameters  for the model fluid-specific parameters.
The Walther model  correlates the double log of viscosity to the log of temperature for liquids far from their critical point. While limited to the liquid phase, the Walther model does not require density (explicitly or implicitly) as an input. It has only two fluid-specific parameters (Walther parameters) calculated by fitting the correlation to experimental viscosity data, usually at atmospheric pressure. Yarranton et al.  developed a generalized version of the Walther model to predict the viscosity of liquid crude oils at any temperature and pressure as a departure from the viscosity calculated at atmospheric pressure. This generalized Walther model has two additional fluid-specific parameters and has been tested on conventional and heavy oils, pure hydrocarbon mixtures, diluted crude oils and distillation cuts.
To predict the viscosity of crude oils both models rely on the following characterization approach. The crude oil maltenes are divided into pseudo-components based on distillation assay data. The viscosity model parameters for the maltene pseudo-components are calculated from a set of correlations developed from experimental data. The pentane insoluble asphaltenes are treated as a single component with defined viscosity model parameters. These parameters were determined by fitting each model to liquid asphaltene viscosity data collected in the low shear Newtonian region. The viscosity parameters for the whole oil are determined from the model mixing rules and a correlation for the binary interaction parameters.
The performance of both predictive models was tested on pure hydrocarbons, conventional oils, heavy oils, bitumens, distillation cuts and mixtures of pure hydrocarbons and crude oil diluted with liquid and vapor solvents. Additionally, both approaches are used to model the viscosity of maltenes, a partially deasphalted bitumen, and an asphaltene/toluene mixture. This test dataset includes data collected at temperatures and pressures up to 175°C and 10 MPa, as well as, crude oil/solvent pseudo-pairs with solvent content up to 50 wt%. The EF and Walther models predicted the viscosity of the samples in the test dataset with and average absolute relative deviation (AARD) of 41 and 57%, respectively.
A tuning scheme is proposed for each model using a single viscosity data point at atmospheric pressure. Tuning with a single multiplier reduced the EF and Walther models AARDs to 8 and 7%, respectively. Both models provide similar accuracy after tuning. The EF model is applicable to a broader range of conditions and is suitable for process simulation but requires accurate input densities. The Walther model does not require a density input and is suited for reservoir simulation but is limited to reduced temperatures below 0.75.
 Yarranton, H. W. and Satyro, M. A. Expanded Fluid-Based Viscosity Correlation for Hydrocarbons. Ind. Eng. Chem. Res. 2009, 48, 3640-3648.
 Ramos-Pallares, F. et al. Predicting the Viscosity of Hydrocarbon Mixtures and Diluted Heavy Oils Using the Expanded Fluid Model. Energy and Fuels, 2016, DOI: 10.1021/acs.energyfuels.5b0195.
 Walther, C. The Evaluation of Viscosity Data. Erdol und Teer. 1931, 7, 382-384.
 Yarranton, H., van Dorp, J., Verlaan, M., Lastovka, V. Wanted Dead or Live: Crude Cocktail Viscosity-A Pseudo-Component Method to Predict the Viscosity of Dead Oils, Live Oils and Mixtures. J. Can. Pet. Techol. 2013, 52, 176-191.