Reservoir Characterization Using Neural and Fractal Analysis: Lithofacies Classification Using Artificial Neural Networks Combined with Fractal Analysis. Applications

Most of the world’s oil reserves come from improved exploitation of already known fields using increasingly sophisticated methods. Recoverable reserves are therefore estimated based on the various methods of field exploitation. Given this trend, every oil company must have a detailed understanding of the internal architecture of its underground reservoirs. To this end, identifying a reservoir’s lithofacies allows us to drill in porous areas with lower clay content. Identifying the lithofacies is the first step in reservoir characterization. The lithofacies consists of a description of the various geological formations comprising a well. To solve our most complex problems, we need to go beyond standard mathematical techniques. As an alternative, we need to supplement conventional analysis methods with several emerging methodologies, such as artificial intelligence. This book reports on and contributes to investigating the potential of new connectionist techniques in reservoir characterization.