Integrated Rock Classification in the Eagle Ford Shale Formation Using Well Logs and Geological Evaluation

Abstract

Formation evaluation and rock classification in organic-rich mudrocks are challenging due to heterogeneity and complex rock physics of these reservoirs. Previous rock classification methods for unconventional formations required extensive core data, and often neglected the importance of mechanical rock properties, which contribute to the ability to initiate and maintain hydraulic fractures. Furthermore, the rock classes obtained from well logs are not always consistent with geological facies. We introduce an integrated rock classification method which takes into account petrophysical, compositional, and mechanical properties of the fluid-bearing rocks as well as geological attributes. Objectives of this paper include (a) to estimate depth-by-depth petrophysical, compositional, and elastic properties using well logs, geological information, and X-ray mineral maps and (b) to provide an optimum number of rock classes based on geological facies as well as petrophysical, compositional, and elastic properties which correspond to the productivity and fracability of the formation. We jointly interpret well logs (i.e., triple-combo and elemental capture spectroscopy logs) to assess petrophysical and compositional properties of the formation. We then apply the Self Consistent Approximation method to calculate depth-by-depth effective elastic properties. Next, we classify rocks based on well-log-based estimates of porosity, TOC, volumetric concentrations of minerals, fluid saturation, and elastic properties using unsupervised neural network. We independently classify rocks from the same well into lithofacies and chemofacies by visual inspection, petrographic analysis, and high-resolution x-ray fluorescence spectroscopy (measurements collected every 5 mm at 100 um resolution). We finally integrate the two petrophysical and geological rock classification schemes and provide an optimum number of rock classes. We successfully applied the method on a well in the oil window of Eagle Ford shale located in McMullen County in south Texas. Results show a good agreement between the rock classes identified using well-log-based and core-based rock classification, as well as geological lithofacies and chemofacies. The integrated well-log-based rock classification revealed finer rock classes compared to other techniques. The finer rock classes can help in minimizing the number of required fracture stages and selecting the best candidate zones for horizontal well placement.

AAPG Datapages/Search and Discovery Article #90217 © 2015 International Conference & Exhibition, Melbourne, Australia, September 13-16, 2015