Facies and Fracture Network Modeling by a Novel Image Processing Based Method

Abstract

A wide range of methods for geological reservoir modeling has been offered from which a few can reproduce complex geological settings, especially different facies and fracture networks. Multi Point Statistic (MPS) algorithms by applying image processing techniques and Artificial Intelligence (AI) concepts proved successful to model high-order relations from a visually and statistically explicit model, a training image. In this approach, the patterns of the final image (geological model) are obtained from a training image that defines a conceptual geological scenario for the reservoir by depicting relevant geological patterns expected to be found in the subsurface. The aim is then to reproduce these training patterns within the final image. This work presents a multiple grid filter based MPS algorithm to facies and fracture network images reconstruction. Processor is trained by training images (TIs) which are representative of a spatial phenomenon (fracture network, facies...). Results shown in this paper give visual appealing results for the reconstruction of complex structures. Computationally, it is fast and parsimonious in memory needs.