Adaptation of Models for Estimating Viscosity in Multicomponent Mixtures with Colombian Heavy Crude

Abstract

Two new correlations for calculating the viscosity in mixtures of heavy Colombian crude with diluents are presented. The first one represents a modified Parkash model and predicts with errors lower than 0.5% the viscosity measured for mixtures including studies between 15 and 18 API and diluents (with a fraction in the mixture between 1.8 and 8.8 % w/w) of API gravity between 40 and 64. The second correlation is obtained from a linear regression of variables such as APL índex and volume of the mixture and represents the field data with a correlation coefficient (R²) of 0.95. Both correlations were validated with field data different to those used in the estimation of the models. In addition, the proposed models can be easily implemented for field calculations and estímate the viscosity of mixtures with equal or laugher accuracy than other existing correlations. The methodology used to establish the new models can be easily implemented to derive or adapt correlations to estímate different properües of interest m Colombian crude mixtures with diluents.