An evaluation of the performance of derived empirical correlations for predicting the oil viscosity of light Omani crude

In the past decades, many correlations for predicting the viscosity of crude oil were proposed. A review of the literature has showed that there is no study aims to investigate the performance of these correlations for predicting the oil viscosity of light Omani crude. Thirty eight oil viscosity correlations describing various conditions are evaluated with Omani crude oil viscosity database using statistical analysis. Twelve of these correlations are for dead oil viscosity, 14 are for saturated oil viscosity and 12 for under saturated oil viscosity. Error estimates for the predicted and measured oil viscosity together with standard deviation for each correlation are presented. Statistical analysis showed that none of the dead oil viscosity correlations was able to satisfactorily predict the Omani dead oil viscosity. The dead oil viscosity correlations were modified by exponentially fitting the viscosity data. For the saturated oil viscosity, the best agreement is obtained with the correlations of Whitson and Brule and Dindoruk and Christman while the worst agreement is obtained with the correlations of Labedi) and Almehaideb. For the under-saturated oil viscosity, Standing, Khan et al. and Kartoatmodjo and Schmidt correlations gave the best agreement with Omani crude oil while Vazquez and Beggs correlation attained the worst agreement.