Application of Machine Learning Techniques for EOR Screening in Heavy Oilfields in Peru

Huerta V.; Meza K. · 4th HGS/EAGE Conference on Latin America (2022) · DOI: 10.3997/2214-4609.202282027

Resumen

In general terms, the selection of EOR methods has been carried out through the use of comparative tables, where the application of each of these methods is determined by specific ranges of reservoir parameters. However, this selection technique can generate biased results due to its qualitative nature and interpretation depending on the criteria of the professional in charge of the process.. In this study, we propose to solve the EOR detection problem using machine learning or pattern recognition methods, which are well established in the literature. An exhaustive study was carried out on the application of various machine learning methods (classifiers) and artificial neural networks. The proposed data-based detection algorithm is a high-throughput tool to select the most appropriate EOR method. This innovative approach integrates theoretical screening principles, such as the criteria developed by Taber, and successful EOR project practices in fields around the world. The proposed algorithm allows the integration of different types of data, eliminates the arbitrary approach in decision making and provides accuracy and speed of calculation. We show that the proposed EOR detection algorithm can correctly predict the appropriate EOR method in more than 90% of the test cases.