University of Isfahan Journal of Stratigraphy and Sedimentology Researches 2008-7888 42 1 2026 03 21 Petrophysical Zonation through Integration of Well-Logging and Core Data in Carbonate Reservoirs: A Case Study from the Dalan and Kangan Formations, Central Persian Gulf Petrophysical Zonation through Integration of Well-Logging and Core Data in Carbonate Reservoirs: A Case Study from the Dalan and Kangan Formations, Central Persian Gulf 29 54 29805 10.22108/jssr.2025.145401.1314 FA Vahid Tavakoli Professor, School of Geology, College of Science, University of Tehran, Tehran, Iran Saman Darvish-Bastami MSc Student, School of Geology, College of Science, University of Tehran, Tehran, Iran Journal Article 2025 05 27 . Abstract The Permian–Triassic carbonates of the Kangan–Dalan formations in the central Persian Gulf represent one of the largest gas reservoirs in Iran and worldwide. Integrated petrophysical–core zonation, supported by multivariate cluster analysis, was applied to reduce reservoir heterogeneity and improve the understanding of reservoir properties. The results were compared with petrophysical logs, scanning electron microscopy, and pore-throat size distribution. Diagenetic processes were found to play a key role in reservoir quality. Fabric-destructive dolomitization in the upper K2 and lower K4 units generated micro-scale pathways connecting primary pores. Dissolution in the lower K2 and upper K4 units further enhanced pore connectivity, resulting in high permeability within these zones. In contrast, selective dolomitization and dissolution in K1 and K3 did not produce effective flow pathways. These processes led to the development of distinct reservoir zones, including high-porosity/low-permeability (Zone 2), high-permeability/low-porosity (Zone 3), and non-reservoir zones (Zones 4 and 5). Additionally, a thin interval with both high porosity and permeability (Zone 1) occurs as a transitional layer within the lower K4. Although diagenesis is the dominant control, primary depositional facies also influenced reservoir characteristics. All zones were successfully identified by the proposed algorithm. Keywords: Diagenesis, Petrophysical Zonation, Clustering, Permian–Triassic, Dolomitization Introduction Reservoir quality is one of the most critical parameters influencing the performance of hydrocarbon reservoirs, with pore-throat size distribution acting as a key controlling factor. This distribution is governed by both primary depositional attributes and secondary diagenetic processes (Tucker and Bathurst 1990; Cerepi et al. 2003; Baron et al. 2008). Diagenesis plays an especially significant role in carbonate reservoirs, where processes such as dolomitization, cementation, and dissolution can substantially alter porosity and permeability (Anselmetti and Eberli 1999). Studies have shown that reservoir zones with similar petrophysical behavior often reflect comparable diagenetic histories (Ehrenberg 2006). In many cases, core data are limited, highlighting the need for well-log analysis as an alternative tool. Several log responses are sensitive to diagenetic alterations, making them useful for reservoir characterization when integrated with petrographic observations. Among these, sonic logs and derived velocity-deviation curves have proven effective in distinguishing pore types and depositional–diagenetic trends, thereby enhancing geological and petrophysical interpretations. Multivariate cluster analysis is among the most powerful approaches for reservoir zonation. While it has been widely applied for electrofacies classification in both clastic and carbonate settings (Gill et al. 1993; Ye and Rabiller 2000), its large-scale application for defining reservoir zones remains limited. Combining well-log responses with petrographic and petrophysical parameters provides a more robust basis for identifying reservoir units. This study evaluates the application of multivariate cluster analysis for integrated petrophysical zonation in the Kangan and Dalan formations of the central Persian Gulf. The proposed workflow aims to establish a practical framework for accurate reservoir characterization by linking log responses, diagenetic features, and pore system evolution. Materials and Methods A key well with 420 m of continuous core from the Kangan and Dalan formations in the central Persian Gulf was selected. Core plugs were taken every 30 cm, cut at both ends, and had thin sections prepared. One-third of each section was stained with Alizarin Red-S (Dickson 1966) to distinguish calcite from dolomite. Thin sections were studied under a polarizing microscope to record depositional facies and diagenetic features. Twenty representative samples were selected for scanning electron microscopy (SEM) and pore-throat size distribution was determined by mercury injection up to 60,000 psi. Core plugs were cleaned, dried, and analyzed for porosity (Boyle’s law) and permeability (Darcy’s law). Petrophysical data from neutron porosity, bulk density, photoelectric factor, and sonic logs were integrated with petrographic porosity estimates to assess the effects of dolomitization and dissolution. For data analysis, customized programming in MATLAB was applied instead of conventional software, providing flexibility in clustering, optimization, and algorithm testing. Hierarchical clustering was employed due to its ability to handle diverse datasets and dynamically determine the number of clusters (Xu and Tian 2015). To further constrain diagenetic effects, velocity-deviation logs were calculated by comparing measured and predicted compressional velocities. Positive deviations indicate cementation or compaction, while negative values reflect enhanced porosity from dissolution or fracturing. This integrated workflow allowed for precise petrophysical zonation by linking log responses, petrographic features, and diagenetic alterations. Discussion of Results & Conclusions The integrated zonation results reveal that intervals with similar porosity and permeability values can be effectively distinguished, reflecting the diverse impact of pore types on flow properties. Each identified zone shows a characteristic porosity–permeability distribution controlled by diagenetic processes. Zone 1 exhibits both high porosity and permeability, largely associated with touching-vug porosity (Lucia 1995) where dissolution, fracturing, and fabric-destructive dolomitization enhanced pore connectivity. This thin interval acts as a transitional layer between the upper dissolution-dominated K4 and the lower dolomitized K4. Zone 2 displays slightly lower porosity but still retains effective reservoir quality, corresponding mainly to the lower K2 and upper K4 where dissolution was the dominant diagenetic process. Zone 3 includes the largest number of samples, with moderate porosity but significantly enhanced permeability due to well-connected pore systems created by fabric-destructive dolomitization. In contrast, Zone 4, representing parts of K1 and K3, shows reduced reservoir quality as porosity is partly occluded by anhydrite cement, leading to low permeability despite moderate porosity values. Zone 5 is non-reservoir, characterized by very low porosity and permeability due to pervasive anhydrite cementation and loss of primary pores. Velocity-deviation analysis confirms these patterns. Negative deviations in Zones 1 and 3 indicate dissolution-related microporosity and microfractures, producing higher-than-expected permeability. Zone 2 shows values close to zero with a slight negative trend, reflecting micropores with limited connectivity. In contrast, positive deviations in Zone 4 clearly point to cementation and compaction, while Zone 5 shows weakly negative trends but insufficient pore connectivity to sustain flow. Thus, negative deviations are reliable indicators of dissolution-enhanced reservoirs, whereas positive values reflect cementation-dominated intervals. Mercury injection capillary pressure (MICP) analysis further supports these findings. Dolomitized samples with preserved primary porosity show relatively uniform pore-throat distributions and moderate permeability, while fabric-destructive dolomitization and dissolution create wider throat-size spectra and higher permeability (Zones 1–3). In contrast, limestone samples with moldic pores sealed by anhydrite exhibit poor pore connectivity and reduced flow potential (Zones 4–5). SEM observations confirm these relationships, showing dissolution-enhanced pore networks in productive zones versus anhydrite-filled throats in non-reservoir intervals. Overall, the Permian–Triassic carbonates of the Kangan and Dalan formations exhibit a complex diagenetic history that strongly influences reservoir quality. Five reservoir zones were defined by hierarchical clustering of well-log data (NPHI, RHOB, Sonic, PEF) calibrated against petrographic and petrophysical observations. Among them, Zones 1–3 represent the productive units, with Zone 3 being the most significant due to widespread fabric-destructive dolomitization and pore connectivity. Zones 4 and 5, affected by anhydrite cementation, represent poor-quality or non-reservoir intervals. This integrated approach demonstrates that multivariate clustering, when combined with petrography, SEM, and MICP data, provides a robust framework for characterizing carbonate heterogeneity. The methodology is flexible, does not require prior training datasets, and can be applied to other fields. Importantly, zones with similar diagenetic histories and pore characteristics are shown to share comparable flow properties, offering a reliable basis for linking porosity, permeability, and reservoir performance. . Abstract The Permian–Triassic carbonates of the Kangan–Dalan formations in the central Persian Gulf represent one of the largest gas reservoirs in Iran and worldwide. Integrated petrophysical–core zonation, supported by multivariate cluster analysis, was applied to reduce reservoir heterogeneity and improve the understanding of reservoir properties. The results were compared with petrophysical logs, scanning electron microscopy, and pore-throat size distribution. Diagenetic processes were found to play a key role in reservoir quality. Fabric-destructive dolomitization in the upper K2 and lower K4 units generated micro-scale pathways connecting primary pores. Dissolution in the lower K2 and upper K4 units further enhanced pore connectivity, resulting in high permeability within these zones. In contrast, selective dolomitization and dissolution in K1 and K3 did not produce effective flow pathways. These processes led to the development of distinct reservoir zones, including high-porosity/low-permeability (Zone 2), high-permeability/low-porosity (Zone 3), and non-reservoir zones (Zones 4 and 5). Additionally, a thin interval with both high porosity and permeability (Zone 1) occurs as a transitional layer within the lower K4. Although diagenesis is the dominant control, primary depositional facies also influenced reservoir characteristics. All zones were successfully identified by the proposed algorithm. Keywords: Diagenesis, Petrophysical Zonation, Clustering, Permian–Triassic, Dolomitization Introduction Reservoir quality is one of the most critical parameters influencing the performance of hydrocarbon reservoirs, with pore-throat size distribution acting as a key controlling factor. This distribution is governed by both primary depositional attributes and secondary diagenetic processes (Tucker and Bathurst 1990; Cerepi et al. 2003; Baron et al. 2008). Diagenesis plays an especially significant role in carbonate reservoirs, where processes such as dolomitization, cementation, and dissolution can substantially alter porosity and permeability (Anselmetti and Eberli 1999). Studies have shown that reservoir zones with similar petrophysical behavior often reflect comparable diagenetic histories (Ehrenberg 2006). In many cases, core data are limited, highlighting the need for well-log analysis as an alternative tool. Several log responses are sensitive to diagenetic alterations, making them useful for reservoir characterization when integrated with petrographic observations. Among these, sonic logs and derived velocity-deviation curves have proven effective in distinguishing pore types and depositional–diagenetic trends, thereby enhancing geological and petrophysical interpretations. Multivariate cluster analysis is among the most powerful approaches for reservoir zonation. While it has been widely applied for electrofacies classification in both clastic and carbonate settings (Gill et al. 1993; Ye and Rabiller 2000), its large-scale application for defining reservoir zones remains limited. Combining well-log responses with petrographic and petrophysical parameters provides a more robust basis for identifying reservoir units. This study evaluates the application of multivariate cluster analysis for integrated petrophysical zonation in the Kangan and Dalan formations of the central Persian Gulf. The proposed workflow aims to establish a practical framework for accurate reservoir characterization by linking log responses, diagenetic features, and pore system evolution. Materials and Methods A key well with 420 m of continuous core from the Kangan and Dalan formations in the central Persian Gulf was selected. Core plugs were taken every 30 cm, cut at both ends, and had thin sections prepared. One-third of each section was stained with Alizarin Red-S (Dickson 1966) to distinguish calcite from dolomite. Thin sections were studied under a polarizing microscope to record depositional facies and diagenetic features. Twenty representative samples were selected for scanning electron microscopy (SEM) and pore-throat size distribution was determined by mercury injection up to 60,000 psi. Core plugs were cleaned, dried, and analyzed for porosity (Boyle’s law) and permeability (Darcy’s law). Petrophysical data from neutron porosity, bulk density, photoelectric factor, and sonic logs were integrated with petrographic porosity estimates to assess the effects of dolomitization and dissolution. For data analysis, customized programming in MATLAB was applied instead of conventional software, providing flexibility in clustering, optimization, and algorithm testing. Hierarchical clustering was employed due to its ability to handle diverse datasets and dynamically determine the number of clusters (Xu and Tian 2015). To further constrain diagenetic effects, velocity-deviation logs were calculated by comparing measured and predicted compressional velocities. Positive deviations indicate cementation or compaction, while negative values reflect enhanced porosity from dissolution or fracturing. This integrated workflow allowed for precise petrophysical zonation by linking log responses, petrographic features, and diagenetic alterations. Discussion of Results & Conclusions The integrated zonation results reveal that intervals with similar porosity and permeability values can be effectively distinguished, reflecting the diverse impact of pore types on flow properties. Each identified zone shows a characteristic porosity–permeability distribution controlled by diagenetic processes. Zone 1 exhibits both high porosity and permeability, largely associated with touching-vug porosity (Lucia 1995) where dissolution, fracturing, and fabric-destructive dolomitization enhanced pore connectivity. This thin interval acts as a transitional layer between the upper dissolution-dominated K4 and the lower dolomitized K4. Zone 2 displays slightly lower porosity but still retains effective reservoir quality, corresponding mainly to the lower K2 and upper K4 where dissolution was the dominant diagenetic process. Zone 3 includes the largest number of samples, with moderate porosity but significantly enhanced permeability due to well-connected pore systems created by fabric-destructive dolomitization. In contrast, Zone 4, representing parts of K1 and K3, shows reduced reservoir quality as porosity is partly occluded by anhydrite cement, leading to low permeability despite moderate porosity values. Zone 5 is non-reservoir, characterized by very low porosity and permeability due to pervasive anhydrite cementation and loss of primary pores. Velocity-deviation analysis confirms these patterns. Negative deviations in Zones 1 and 3 indicate dissolution-related microporosity and microfractures, producing higher-than-expected permeability. Zone 2 shows values close to zero with a slight negative trend, reflecting micropores with limited connectivity. In contrast, positive deviations in Zone 4 clearly point to cementation and compaction, while Zone 5 shows weakly negative trends but insufficient pore connectivity to sustain flow. Thus, negative deviations are reliable indicators of dissolution-enhanced reservoirs, whereas positive values reflect cementation-dominated intervals. Mercury injection capillary pressure (MICP) analysis further supports these findings. Dolomitized samples with preserved primary porosity show relatively uniform pore-throat distributions and moderate permeability, while fabric-destructive dolomitization and dissolution create wider throat-size spectra and higher permeability (Zones 1–3). In contrast, limestone samples with moldic pores sealed by anhydrite exhibit poor pore connectivity and reduced flow potential (Zones 4–5). SEM observations confirm these relationships, showing dissolution-enhanced pore networks in productive zones versus anhydrite-filled throats in non-reservoir intervals. Overall, the Permian–Triassic carbonates of the Kangan and Dalan formations exhibit a complex diagenetic history that strongly influences reservoir quality. Five reservoir zones were defined by hierarchical clustering of well-log data (NPHI, RHOB, Sonic, PEF) calibrated against petrographic and petrophysical observations. Among them, Zones 1–3 represent the productive units, with Zone 3 being the most significant due to widespread fabric-destructive dolomitization and pore connectivity. Zones 4 and 5, affected by anhydrite cementation, represent poor-quality or non-reservoir intervals. This integrated approach demonstrates that multivariate clustering, when combined with petrography, SEM, and MICP data, provides a robust framework for characterizing carbonate heterogeneity. The methodology is flexible, does not require prior training datasets, and can be applied to other fields. Importantly, zones with similar diagenetic histories and pore characteristics are shown to share comparable flow properties, offering a reliable basis for linking porosity, permeability, and reservoir performance.

https://jssr.ui.ac.ir/article_29805_60faaf8d4fed74cf2594c0f1d85765ac.pdf