A nonlinear approach for predicting pore pressure using genetic algorithm in one of the Iranian petroleum carbonate reservoirs

Abstract

The fluid pressure within a formation pores is called the pore pressure in petroleum engineering. The estimation of pore pressure is a challenging task during a reservoir’s life cycle. An impermeable rock, such as shale, confides the fluids which lead to anomalously high pressures. Besides, during the exploitation life of a reservoir, the pressure reduces in the reservoir. The estimation of these high-risk pore pressures is an essential task in planning for infill drilling and field development. Herein, we propose a nonlinear model for pore pressure estimation, using a genetic algorithm. We compare our method with two of the classical linear methods for pore pressure estimation, the modified Eaton method and the Bowers method, using the Modular Formation Dynamics Tester (MDT) and well logs data related to an Iranian oil-bearing carbonate reservoirs. The results of the nonlinear estimation models showed higher accuracy and less uncertainty than the other models. The studied oil field is in the development phase; therefore, a reliable estimation of pore pressure decreases the future drillings risks and can find application in hydraulic fracturing, completion, and cement works operations.