Machine Learning Assisted Fracture Network Characterization of A Naturally Fractured Reservoir Using Seismic Attributes

Abstract

Summary Characterizing the naturally fractured unconventional reservoirs is challenging because of the reservoir complexity and heterogeneity. Delineating Fault-fracture networks through seismically derived discontinuity attributes is an effective way to characterize these fractured reservoirs. As there are multiple discontinuity attributes containing meaningful information about the fault-fracture networks, it is important to integrate these attributes using a sophisticated machine learning technique for enhanced delineation of these fault-fracture networks. Self-Organizing Map (SOM), the latest and robust unsupervised classification technique, can extract the integrated information of these anomalous features from multiple seismic attributes. Automated clustering algorithms fall into two categories – supervised and unsupervised algorithms. Unsupervised machine learning algorithms are purely data-driven and help in recognizing and classifying the patterns from a dataset without any prior information. Posteriori information such as well data, is integrated into the results for recognizing the facies classification and calibrating the interpretation. Unsupervised learning methods also help to highlight subtle stratigraphic features that might otherwise be unnoticed using conventional analytical methods. In this study, we adopted a recent unsupervised classification technique called SOM. The technique has been applied successfully to a naturally fractured reservoir in an Onland field, in India. The objective of the study was to extract the subtle faults and fracture network information from the seismic-based attributes in order to characterize the naturally fractured reservoir. By use of the relevant seismic discontinuity attributes and application of the advanced SOM technique using optimum parameters set, subtle faults/fracture corridors could be mapped effectively using the SOM results. SOM Principal Axis derived from this technique, was a valuable attribute in characterizing the naturally fractured reservoir. This attribute was also a useful input during the Discrete Fracture Network (DFN) modeling stage, in guiding the natural fracture network propagation.