Effect of heavy-light oil viscosity ratio on prediction accuracy of mixed oil viscosity

Accurate calculation of mixed oil viscosity after heavy oil is mixed with light oil is essential for calculation of heavy oil surface facility and pipelining process. In order to analyze the effect of high viscosity ratio (> 10⁴) on prediction accuracy of mixed oil viscosity, 6 common calculation models of mixed oil viscosity are firstly selected. After Capella heavy oil is mixed with 4 types of light oil with different physical properties, 38 groups (including 114 mixtures) are tested and their measured viscosity and model prediction value are compared. Error analysis method is adopted to identify the effect of heavy-light oil viscosity ratio on prediction accuracy of different models. It is indicated that Cragoe correction model presents stable prediction error, averaging less than 10%, when viscosity ratio is less than 3 000. When viscosity ratio is more than 3 000, it is recommended to adopt Lederer improved model whose average prediction error is less than 15%. Attention shall be paid to the effect of mixed oil viscosity ratio on viscosity prediction errors of mixed oil viscosity models. It is strongly suggested to introduce viscosity ratio into correction formula when the existing models are to be corrected.