Assessment of Petrophysical and Textural Heterogeneity Using Core-Scale Images, Image Logs, and Nuclear Magnetic Resonance Measurements

Summary Heterogeneous carbonate reservoirs pose significant challenges for the reliable assessment of petrophysical properties, driven by complex pore systems and strong spatial variability. The limited core samples in such heterogeneous reservoirs undermine laboratory-calibrated rock physics models, underscoring the need for representative core sampling strategies. To optimize the selection of representative core sampling locations, we introduce a novel integrated workflow developed for near-real-time quantification of depth-by-depth petrophysical heterogeneity in geologically complex formations. The workflow includes (a) extracting nuclear magnetic resonance (NMR)-based pore-size distribution (PSD) parameters to quantify pore-structural heterogeneity (HTIpore), (b) extracting image-based rock textural features from acoustic image logs to assess rock textural heterogeneity (HTItextural), and (c) integrating pore- and well-log-scale information to compute depth-by-depth petrophysical heterogeneity index (HTI). The workflow begins by calibrating NMR transverse relaxation time (T2) distribution data quantified by fitting multimodal Gaussian functions to extract PSD parameters. These PSD parameters are subsequently used to compute pore-structural characteristic features [i.e., intrapeak pore-size variability (IPV) and weight factor (WF)], enabling quantification of within-peak variability through a variability index (VI). Then, the number of PSD peaks and VIs are used to compute HTIpore. In parallel, rock textural features are extracted from acoustic image logs using a grayscale co-occurrence matrix (GLCM) analysis applied across multiple sliding windows within the images. Subsequently, the extracted textural features are analyzed using principal component analysis (PCA) to obtain rock textural characteristic values (RTCVs) that capture maximum spatial variance within rock texture. These RTCVs are then used to estimate HTItextural. Finally, an analytical index, HTI, is introduced to integrate pore-structural and image-based textural heterogeneities for the depth-by-depth assessment of petrophysical heterogeneity. The proposed workflow is applied to a field data set from a well drilled in a Brazilian pre-salt carbonate sequence consisting of five major rock classes (shrubstone with calcite inclusion, shrubstone, spherulitic shrubstone, cemented packstone with shrub fragments, and cemented packstone). Among the evaluated rock classes, shrubstone with calcite inclusion exhibited the highest heterogeneity (median HTI value of 20.53), whereas cemented packstone showed the lowest heterogeneity (median HTI value of 8.40) within the evaluated depth interval, consistent with observations from high-resolution core images. The median HTI values were in agreement with the conventional heterogeneity approach, the coefficient of variation (Cv). The median difference between the proposed HTI and Cv ranges from 0.6% to 30% across the evaluated rock classes. While Cv quantifies overall within-rock-class variability, it does not preserve spatial continuity along the wellbore. In contrast, the computed HTI provides continuous depth-by-depth heterogeneity along the wellbore, enabling fine-scale heterogeneity in complex reservoirs. The proposed HTI-based approach integrates multiphysics data to identify geologically complex intervals and recommends narrower core sampling in spatially heterogeneous rocks relative to homogeneous rocks, enabling quantification of rapid vertical variations in petrophysical properties. This unique contribution can potentially enhance core sampling efforts, leading to reliable reservoir characterization at an optimal cost.