Comparing classical and neural regression techniques in modeling crude oilviscosity

The importance of crude oil viscosity makes its accurate determination nece ssary for reservoir performance calculations, evaluation of hydrocarbon res erves, planning thermal methods of enhanced oil recovery, and designing pro duction equipment and pipelines. Viscosity data are also involved in severa l dimensionless parameters to calculate dow regimes, friction factors and p ressure gradients in multiphase flow problems. Numerous research efforts ha ve been directed towards the development of viscosity models that are capab le of accurately predicting crude oil viscosity as a function of production data, and/or composition of well stream fluids, if available, using equati on of State. Since fluid compositions are not always available, most of the efforts were focused on developing viscosity correlations using classical regression techniques. The study presents, for the first time, a comparison among several models d eveloped using both classical regression techniques (CRT) and neural regres sion techniques (NRT). These models are developed in this study from viscos ity data collected from different oil fields. The models have also been tes ted using another collection of viscosity data that was not used before in the development phase. Results show that viscosity models developed using N RT were more accurate than viscosity models developed using CRT. Based on t his comparison, a viscosity model is therefore presented, which uses stock- tank oil API gravity, gas gravity, pressure(s), and temperature(s) to predi ct crude oil viscosity. The model was developed using General Regression Ne ural Network algorithm. (C) 2001 Elsevier Science Ltd. All rights reserved.