This study is concerned with the water coning phenomenon that takes place around production wells of hydrocarbon reservoirs. In this paper, the development of artificial neural networks to predict the water saturation buildup around vertical and horizontal wells with a good level of accuracy is described. In the development of expert systems, it is assumed that water encroachment originates from an active aquifer which is located under the hydrocarbon reservoir (reservoir with bottom water drive). A high-fidelity numerical model is utilized in generating training data sets that are used in structuring and training the artificial neural networks. The artificial expert systems that are developed in this paper are universal and are capable of predicting the change of water saturation around the wellbore as a function of time and the prediction process is faster than a reservoir simulator and requires less data, which saves time and effort. With the help of these models, it will be possible to predict the position of high water saturation zones around the wellbore ahead of time so that remedial actions such as closing the perforations that produce the water can be implemented on a timely basis.
Key words: Bottom water drive, water coning, neural network, water saturation, vertical well, horizontal well.
Copyright © 2019 Author(s) retain the copyright of this article.
This article is published under the terms of the Creative Commons Attribution License 4.0