Scientists from Kuwait University and Kuwait Scientific Research Institute developed a new model allowing for much more precise forecasting the heavy crude viscosity. Different to the existing methods, it takes into account not only the crude temperature and density, but also the contents of resins and asphaltenes – complex compounds determining its fluidness. This study provides the engineers a real tool for planning oil production, transportation and refining.
The viscosity problem is especially relevant for heavy crude grades containing big amounts of high-molecular compounds, due to which these grades are thick and flow poorly. Hence, special technologies are required for their production and transportation. Until now, simplified formulas were used for calculations linking the viscosity only with temperature and API-gravitation – the oil density indicator. However, these models often had very low accuracy, because they did not take into account the chemical formulations and real mechanisms impacting the oil fluidity.
When developing the new model, the scientists from Kuwait took three heavy crude samples from the oilfields in Northern Kuwait, separated clean resins and asphaltenes from them, and then added them back in different proportions forming 357 artificially created crude samples with different formulations. For each of them, viscosity, density and other physical and chemical parameters were precisely measured.
Then they applied the nonlinear regression analysis to the obtained date and derived a new formula linking the crude viscosity with temperature, density (expressed in terms of API), and the contents of asphaltenes and resins. Introducing a new parameter accounting for the ration between asphaltenes and resins (As/Rs) was a new feature of the model allowing for the mathematical description of their synergy impact.
The new model demonstrated very high accuracy. When testing it on the independent data, its forecasts turned out to be much closer to real measurements vs the results of other 19 known correlations. The average error mean was reduced down to 18–20%, and the determination coefficient reached 0.97 indicating the highest accuracy and reliability.
Now the researchers propose to use the developed model in hydrodynamic simulation of oilfields, for designing pipelines and pumping equipment, as well as for planning the enhanced oil recovery methods. This will allow to reduce the dependency on expensive laboratory tests.
