The variety of available EOR techniques requires an in-depth screening to select a viable method that matches well to the reservoir rock and fluid parameters and still remains economically attractive. In the present paper a Comprehensive Integrated EOR Workflow is proposed that starts with an Advanced EOR Screening Method. This is comprised of both a Neural Network Part and an Operational Module. While the First part uses proven data mining techniques the Operational Module considers the specific features of the screened EOR Method influencing the field implementation.

The Neural Network Part is based on an exhaustive review and selection of successfully deployed literature case. It uses the rock, fluid and other reservoir parameters to screen various EOR methods considering their technical-economical applicability. This Artificial Intelligence approach utilizes data mining techniques in the form of a hybrid system that makes use of a neural network as a screening tool and the genetic algorithm as an optimization tool to land into the optimum recommendation. The Operational Part enables to evaluate the implementation capability on the given field based on the specific requirements of the preselected EOR Method.

The system works its way through the literature data of successful EOR projects trying to detect patterns and learning from the data the relationship between these characteristics and the feasibility of applying each EOR technique mimicking the ability of the human mind to learn from previous experience. The system is a multi-layers neural network whereby the input layer is composed of seven key reservoir parameters (depth, temperature, porosity, permeability, initial oil saturation, oil gravity and in-situ oil viscosity) while the output layer is composed of the probability of success of the evaluated EOR methods (steam, CO2 miscible, hydro-carbon miscible, in-situ combustion, polymer flooding). The number of hidden layers and neurons are optimized using genetic algorithm for best matching of the training data set and accurate prediction of the testing set.

Comparing the system output with the actual applied EOR techniques in the field shows a reliable result with only a 5% miss-prediction of the total test fields. The Operational Module determines the deployment capabilities in the given reservoir considering the specific parameters of the pre-selected EOR Method, production-pressure history, Formation fluid flow properties and the actual field and well set up, thus providing an advanced EOR Screening.

You can access this article if you purchase or spend a download.