| Abstract
|
:
|
Permeability and lithofacies are very important in the computations of reservoir engineering. Permeability, or flow capacity, is the ability of porous rock to transmit fluid while lithofacies represent the physical properties of rocks and are important components of hydrocarbon reservoir description. There are many empirical approaches for predicting permeability and lithofacies; such as linear/non-linear multiple regression or graphical techniques. Recently, researchers utilized artificial neural networks (ANNs) to achieve better predictions. These achievements of ANN open the door to both machine learning and soft computing techniques to play a major role in petroleum, oil, and gas industry. Unfortunately, the developed ANN model is bedeviled with several limitations especially in uncertain situations. Recently, Type-1 and Type-2 Fuzzy Logic Inference Systems have been introduced as novel computational intelligence approaches for both prediction and classification. They were successfully used in several areas of science and engineering, however, they have not been fully utilized in the oil and gas industry, particularly the type-2 FIS that was recently developed. To demonstrate the usefulness of the Type-1 and Type-2 Fuzzy Inference Systems (T1&T2 FIS) techniques in petroleum engineering area, we describe both their steps and their use for predicting carbonate lithofacies and permeability from well logs. This work explores the capability of both Type-1 and Type-2 Fuzzy Inference Systems as novel approaches for predicting carbonate lithofacies and permeability from Well Logs. As expected, these new computational intelligence approaches overcome the weaknesses of the standard neural networks limitations. In addition, we carried out a comparative study to compare their performances with those of the conventional artificial neural network models and other popular techniques. Empirical results show that the performance of T1&T2 FIS novel approaches outperform most of the common existing approaches, particularly in the area of ability to handle data in uncertain situations, which are the common characteristics of well logs data. Finally, for the first time in the history of permeability prediction, a Type-2 fuzzy logic based model has been developed that will generate, not only the target forecast, but also prediction intervals without additional effort. We have developed smart simulation structure to attack such problem with real-industry data to prove our contribution.
|