One of the leading challenges in hydrocarbon recovery is predicting rock types/fluid content distribution throughout the reservoir away from the boreholes because rock property determination is a major source of uncertainty in reservoir modeling studies. Spatial determination of the lateral and vertical heterogeneities has a direct impact on a reservoir model because it will affect the property distributions. An inappropriate determination of the facies distribution will lead to unrealistic reservoir behavior. Because these data can take different forms (lithologs, cuttings, and for seismic, poststack, and prestack attributes) and have different resolutions, the manual integration of all the information can be tedious and is sometimes impractical. We developed a new neural network-based methodology called democratic neural network association (DNNA). The DNNA method was trained using lithology logs from wells simultaneously with prestack seismic data. This technique, using a probabilistic approach, aims to find patterns in seismic that will predict lithology distribution and uncertainty. IntroductionThe economic viability of a field is dependent on the quality and accuracy of lithology distribution prediction, as well as by the heterogeneity of a potential reservoir. These components are the keys to successful hydrocarbon exploration and production. The rise in unconventional resource prospecting and the increasing complexity of conventional plays have made accurate lithology prediction even more critical. All relevant data must be used optimally to determine lithology at the prospect scale with the highest degree of resolution, resulting in the most geologically meaningful lithology distribution. Risk increases with complexity, however, and the probability of success and the integration of uncertainty into the nature and distribution of lithology must be taken into account in any approach that tries to predict lithology.Conventional approaches are mainly based on 2D or 3D analyses of inverted data to describe the elastic properties of the reservoir. However, precise lithology description in such attribute spaces often overlaps. This makes it difficult to clearly differentiate, for example, intermediate-type facies such as thin interbedded layers. The result is nonunique and highly sensitive to facies interpretation. It, therefore, becomes critical to estimate reservoir connectivity, as some lower qual-

This article proposes a new approach to build a 3D geological model calibrated to well data using an adaptive neural network taking into account pre stack or post stack seismic behavior.

No abstract

No abstract

Automatic seismic facies classification is now common practice in the oil and gas industry. Unfortunately unsupervised seismic classification is often not optimal. The main criticism of unsupervised classification is the a priori nature of the seismic data set organization and the poor description of seismic due to data redundancy. Data reduction, such as Principal Component Analysis (PCA) is often used in association to reveal the principal characteristics of the geological system. The new clustering described here will with a dynamic process naturally search to fill the data space, and to describe the full variability of the seismic. The process can be imagined as a gas expanding in volume. Finally, the process details the anomalies which potentially correspond to hydrocarbon accumulations.