| dc.description.abstract | Characterization of carbonate reservoirs is challenging as the result of heterogeneous distribution of petrophysical properties and mineralogy. Common rock classification techniques are strongly dependent on core measurements, whereas core data are usually sparse and inadequate for reliable heterogeneity analysis. I introduce an integrated rock classification workflow, based on conventional well logs and core data, which incorporates geological attributes and petrophysical, compositional, and elastic properties estimated from conventional well logs. The proposed rock classification method may enhance (a) assessment of petrophysical and compositional properties, (b) prediction of acid stimulation performance, (c) selection of completion depth intervals, and (d) production enhancement strategies in carbonate formations.
In the proposed workflow, I incorporate diagenetic and depositional attributes by taking into account the impact of shapes of different pore types and minerals in each geological facies on resistivity logs and elasticity. Rock quality and rock-fluid quality indices are introduced to take into account the impact of dynamic petrophysical properties, rooted in diagenesis, for real-time rock classification. Furthermore, I apply an analytical technique for the depth-by-depth assessment of elastic rock properties, as well as interparticle and intraparticle porosity, in a limiting case where shear-wave slowness measurements are not available. In addition, I take advantage of Mercury Injection Capillary Pressure (MICP) measurements, where available, to characterize pore-throat radius distribution and modality using a multi-modal Gaussian function. I finally use supervised and unsupervised learning techniques to classify rock types based on static and dynamic petrophysical, compositional, and elastic properties.
I successfully applied the proposed workflow in four carbonate formations, Hugoton, Happy Spraberry, Veterans, and SACROC fields. Although the main focus of this dissertation is carbonate rock classification, I also introduced rock classification techniques based on conventional well logs in organic-rich shale formations. The reliability of these techniques was investigated in the Haynesville Shale. The identified petrophysical rock classes in all field cases were validated using core-derived rock classes, lithofacies descriptions, and thin-section images, where available. The results showed improvement in the assessment of petrophysical properties, compared to the conventional assessment techniques.
The contributions of the proposed techniques include (a) incorporation of geological and petrophysical attributes for an integrated rock classification, (b) application of conventional well logs for the depth-by-depth assessment of elastic moduli and interparticle and intraparticle porosity, applicable where acoustic well logs are not available, (c) simultaneous characterization of pore modality and pore-throat radius distribution for rock classification in carbonate formations. | en |
