Use of Artificial Neural Network to Select the Appropriate EOR Method for a Reservoir Field

Abstract

Enhanced Oil Recovery (EOR) has gained great attention globally. Extensive researches have been conducted to develop various EOR methods, evaluate their applicability and optimize operation conditions. The aim of the study is to screen various EOR methods based on field characteristics and evaluate their technical/economic applicability efficiently instead of predicting field performances of all possible competing strategies and comparing their economics. The Artificial Neural Network (ANN) approach is presented to select an appropriate EOR method with given reservoir properties. The ANN developed is a four-layered feed-forward Back Propagation (BP) network consisting of one input and output layer with two hidden layers. The input layer is composed of key reservoir parameters (reservoir depth, temperature, porosity, permeability, initial oil saturation, oil gravity, and in-situ oil viscosity) while output layer is composed of methods to be evaluated (steam, CO2 miscible, hydrocarbon miscible, in-situ combustion, polymer flooding). The number of hidden layers and neurons are optimized during training by repeated trial and error. After trained successfully, the ANN is tested and applied to other fields. A series of results show that ANN developed can be used to select best EOR process according to reservoir rock and fluid characteristics in a time and cost-effective way.