Reza Azin

As the ages of most oil fields fall in the second half of their lives, many attempts have been made to enhance oil recovery in an efficient way. Gas injection into oil reservoirs for enhanced oil recovery (EOR) purposes requires relative permeability as a crucial issue in reservoir engineering. In this study, a new method is applied to predict relative permeabilities of gas/oil system related to various rock and fluid types. For this reason, a soft computing technique -multi-gene genetic programming (MGGP) isemployed to develop tools for prediction of relative permeability. The newmethods are evaluated by experimental data extracted fromopen literature and are validated by extensive error analysis. The generated smart mathematical equations are able to predict relative permeabilities of gas/oil system with high accuracy and are applicable for various types of rock and fluid as well. In contrary to other reported correlations, the new novel equations require oil API and gas molecular weight as extra input variables to improve their estimating ability for every type of rock and fluid. The proposed technique is promising and encouraging for petroleum and reservoir engineers to be implemented for other gas/oil petro-physical properties