(267f) Prediction of the Density and Viscosity of Visbroken Heavy Oils Using Regular Solutions and Expanded Fluid Models. | AIChE

(267f) Prediction of the Density and Viscosity of Visbroken Heavy Oils Using Regular Solutions and Expanded Fluid Models.

Authors 

Lopez Lago, C. A. - Presenter, University of Calgary
Yarranton, H. W., University of Calgary
Schoeggl, F. F., University of Calgary
Visbreaking is a mild thermal cracking process that significantly reduces oil viscosity and moderately reduces oil density. The higher the severity (temperature and residence time), the greater the viscosity reduction. However, the practical extent of viscosity reduction is limited by the final product stability; that is, its tendency to precipitate asphaltenes or form coke. Hence, predictions of product properties and stability are required to design and optimize visbreaking processes. The oil can be characterized in different pseudo-components. The most common strategies include the use of boiling points or solubility classes. The composition of the oil can later be used in different models to predict their density and viscosity.

Regular solution and Expanded Fluids are an example of models that can be used to predict the density and viscosity of heavy oils. The regular solution mixing (or volumetric) rule assumes that the intermolecular forces between like and unlike molecules are the same; hence, there is no volume change and or enthalpy change upon mixing. The mixture density is then the volumetric average of the component densities. Expanded Fluid (EF) model correlates viscosity to density. The fluid density is determined from the dilute gas viscosity and a departure function for the liquid viscosity. The departure function is similar to free volume theory except that relationship of liquid viscosity to inverse density is similar to a double exponential function instead of an exponential function. This correlation traces continuously the full range of single-phase viscosity in the gas, liquid, and critical or supercritical regions.

Recently, the density and viscosity of visbroken products from a Western Canadian bitumen were matched with correlations based on a DSARA characterization using Regular Solutions and EF. DSARA (distillates, saturate, aromatic, resins and asphaltenes) separates the oil into the distillable fraction (boiling point < 367 °C AET) and the SARA components of the residue. An advantage of this approach is that the volatile components are captured with the distillate fraction and the residue has a low volatility, providing a more consistent SARA assay. In addition, the two fractions that exhibit the largest changes with visbreaking (distillates and asphaltenes) are represented directly in the characterization. A significant limitation of this approach was that the DSARA composition of the visbroken products were necessary to be measured. Lately, a model to correlate DSARA composition to visbreaking conversion was developed.

The purpose of this study was to evaluate the Regular Solutions and EF models to predict the density and viscosity of visbroken oils using only the conversion and initial feed composition as input. Nine different feeds including a Colombian heavy crude, three Western Canadian bitumens, two partially deasphalted bitumens, a fully deasphalted bitumen, and a vacuum residue fraction were thermally cracked in a continuous visbreaker at constant pressure over a range of conversions. The Simulated Distillation assay, gas yield, pentane-insoluble asphaltene content, toluene insoluble (TI) content, density, viscosity, distillate content, and residue SARA composition of each feed and product were measured. Conversions from 5 to 46% were explored. The gas yield was below 1.5 wt% and the TI content was negligible in all cases. The density and viscosity of the visbroken products were predicted with Regular Solutions and EF fluids model and compared to the experimental values. The average absolute deviation for the density was below to 2.5 kg/m³ and the viscosity average relative deviation was below to 30%.