Intelligent model for prediction of CO2 – Reservoir oil minimum miscibility pressure

Abstract

Multiple contact miscible floods such as injection of relatively inexpensive gases into oil reservoirs are considered as well-established enhanced oil recovery (EOR) techniques for conventional reservoirs. A fundamental factor in the design of gas injection project is the minimum miscibility pressure (MMP), whereas local sweep efficiency from gas injection is very much dependent on the MMP. Slim tube displacements, and rising bubble apparatus (RBA) are two main tests that are used for experimentally determination of MMP but these tests are both costly and time consuming. Hence, searching for quick and accurate mathematical determination of gas–oil MMP is inevitable. The objective of this study is to present a reliable, and predictive model namely, Least-Squares Support Vector Machine (LSSVM) to predict pure and impure CO2 MMP. To this end, about 147 data sets belonging to experimental CO2 MMP values from the literature and corresponding gas/oil compositional information was used to construct and evaluate the reliability of the model. The results show that the proposed model significantly outperforms all the existing methods and provide predictions in acceptable agreement with experimental data. Moreover, it is shown that the proposed model is capable of simulating the actual physical trend of CO2 MMP versus five most important input parameters: reservoir temperature, molecular weight of pentane plus, hydrogen sulfide and nitrogen concentration. Finally, for detection of the probable doubtful CO2 MMP data, outlier diagnosis was performed on the data sets.