MICP-Based Petrophysical Classification of Complex Carbonate Reservoir Rocks

Carbonate reservoir rocks present highly heterogeneous pore structures due to their growth by mineral precipitations and diagenetic alterations. The resulting rock matrix properties, such as capillary pressure, porosity, and permeability, control the recoverable hydrocarbon storage, distribution, and flow mechanisms. In this work, we investigate the pore geometry in microscale to mesoscale and its control over the absolute hydraulic permeability. We examine different carbonate matrix facies in the Barra Velha and Itapema formations, both in the presalt Aptian section of the Mero Field, located at the ultradeep waters of Santos Basin, Brazil. The gathered data comprises 50 core rock samples with laboratory measurements of mercury injection capillary pressure (MICP), porosity, and permeability from routine core analysis and descriptions of petrographic thin sections. We also use 15 oil/water porous plate capillary pressure curves and associated electrical properties to assess irreducible water saturation, besides a complete suite of wireline well logs for petrophysical upscaling. We propose a nonlinear least-squares fit of a multimodal Weibull function to the MICP data by implementing the Levemberg-Marquardt algorithm. We, then, petrophysically interpret the multi-Weibull parameters and compute the absolute hydraulic permeability using these adjusted function parameters and Kozeny-Carman equations. By comparing the MICP-based absolute permeability to the corresponding lab-measured data and respective petrographic thin sections, we were able to perform quality control on the MICP results and detect the nonrepresentative measurements. We then use the MICP-based permeability to perform rock classification through Lucia’s rock fabric and Amaefule’s flow zone indicator methods, further improving these classification schemes with depositional and diagenetic properties from the thin sections, finally establishing a link between pore-scale properties and geological characteristics. This information can be used for later petrophysical upscaling and reservoir modeling. We developed a semi-automated analytical model well suited to characterize complex carbonate matrix reservoir rocks, delivering multiple cross-correlation charts and matrix-correlation tables and associating multiple geological and petrophysical properties. Our approach also allowed for detecting anomalous data, providing a way to quality control experimental results, therefore enhancing the petrophysical characterization from multiple data sources. The multi-Weibull inversion was shown to be highly versatile in modeling several non-symmetric pore-throat distributions, accurately probing porous volume partitioning, and properly detecting the length scales controlling permeability, mercury entry pressure, and fluid saturation. Our new method calculates MICP-based absolute hydraulic permeability with higher accuracy by using the multi-Weibiull model parameters and Kozeny-Carman equations. Correlation of petrophysical and geological properties from petrographical thin sections with our multi-Weibull models and classification through Lucia’s rock-fabric and Amaefule’s flow zone indicator methods led to a better comprehension of how depositional and diagenetic textures control the microscopic fluid flow inside the reservoir. The reservoir rock classification revealed that the rocks under analysis are well represented by general porosity-permeability trend lines of Lucias’ rock fabric numbers. Hydraulic flow units revealed strong correlations to characteristic pore-scale dimensions and to post-depositional diagenetic alterations, controlling the absolute hydraulic permeability behavior.