SARA analysis and Conradson carbon residue prediction of Colombian crude oils using PLSR and Raman spectroscopy

Raman spectroscopy generates a large volume of information at the molecular level of crude oils, which processed using chemometric methods allows building models to predict their physicochemical properties, facilitating to make decisions in refining processes for obtain high-value products efficiently and minimal environmental impact. In this work, the Raman spectra of crude oils from different regions of Colombia and Partial Least Square Regression (PLSR) were used to determine the fractions of saturates, aromatics, resins, asphaltenes (SARA analysis) and Conradson carbon residue (CCR), which were determined complying with the standards established by the American Society for Testing and Materials (ASTM). The dimensionality of the model was determined according to the root mean square error of cross validation (RMSECV), the coefficient of determination (R2) and the correlation between consecutive pairs of regression vectors.