Undersaturated Oil Viscosity Based on Multi-Gene Genetic Programming

Abstract

Undersaturated oil viscosity represents an important physical property for reservoir simulation, enhanced oil recovery, and optimal production. It can be determined either by experimental measurements or by modeling through empirical correlations of appropriate accuracy. As a result of the high cost of its determination experimentally as well as its unavailable in most cases, looking for a high-reliability model is vital. Therefore, in this paper, a new undersaturated crude oil viscosity model using multi-gene genetic programming (MGGP) is presented. This model was built using several data points which are distributed as 528 experimental measurements for a broad range of reservoir pressure and oil properties and 276 points were used for validating and testing the new model. Furthermore, the new model was compared to 11 published correlations. The results indicated that the new MGGP-based model is the closest to the experimental measurements and yields a precise prediction of undersaturated oil viscosity.