Predicting reservoir quality in sandstones through neural modeling

Abstract

Due to limited understanding of the details of many diagenetic processes, mathematical models become a very useful tool to predict reservoir quality prior to drilling. Porosity prediction is an important component in pre-drill and post-drill evaluation of reservoir quality. In this context, we have developed a mathematical model to predict porosity of sandstones reservoir systems. This model is based on artificial neural networks techniques. We propose a score to quantify their importance of each feature in prediction process. This score allows creating progressive enhancement neural models, which are simpler and more accurate than conventional neural network models and multiple regression. The main contribution of this paper is the building of a reduced model just with the most relevant features to porosity prediction. A dataset about Uerê formation sandstone reservoir was investigated. This formation is an important oil exploration target in Solimões Basin, western Brazilian Amazonia. Study results show that progressive enhancement neural network is able to predict porosity with accuracy near 90%, suggesting that this technique is a valuable tool for reservoir quality prediction.