Digital SCAL in Complex Carbonates by the Density Functional Hydrodynamics Method

Abstract

Pore-scale modeling has seen tremendous development in understanding petrophysical rock properties and reducing uncertainties in reservoir characterization where experimental measurements may be lacking. The aim of this work was to validate and verify the accuracy of the density functional hydrodynamics (DFH) method in predicting reliable relative permeability data in mixed-wet carbonate reservoir rock types. Four plugs were selected to represent dolomite and limestone rock types from an offshore field in Abu Dhabi. The samples’ porosity and permeability values ranged from 0.14 to 0.30 (fractional) and from 40 to 60 md, respectively. The following workflow was implemented to obtain representative three-dimensional (3D) digital rock models (DRMs): • Plug-scale X-ray computed tomography (CT) at a resolution of 50 × 50 × 300 µm/voxel (X, Y, Z) was used to assess plug heterogeneity and select representative subsamples for higher-resolution scanning and mercury injection capillary pressure (MICP) analysis. • Selection and cutting of the subsample as a (4 mm × 15 mm) cylindrical mini-plug. • Micro-CT of the mini-plug at around 1 µm/voxel resolution. • High-resolution scanning electron microscope (SEM) mosaic imaging at 100 nm/pixel on ion-milled surfaces of the mini-plug and sample mineralogy classification with energy-dispersive X-ray spectroscopy (EDS) analysis using the Advanced Mineral Identification and Characterization System (AMICS) analysis. Construction of a multiscale 3D digital rock model (DRM) from the micro-CT and the SEM data. Capillary pressure and relative permeability curves were simulated in the DRM using the direct modeling method implemented in the Direct Hydrodynamic (DHD) simulator. Porosity, permeability, and MICP data were obtained from simulations of the multiscale-constructed DRMs, which compared reasonably well with the unimodal and bimodal pore-throat-size distributions of the studied plug samples. Wettability input to the rock model was distributed as contact angles between the fluid pair and the rock surface. The wettability distribution was constrained by the experimental drainage and imbibition capillary pressure properties of the samples. Waterflood relative permeability curves were predicted in the validated DRMs with assigned saturation-driven mixed wettability using the DHD simulator. The digital curves compared very well with the experimental measurements performed on the corresponding special core analysis (SCAL) plugs. This work was part of a validation and verification project to better understand digital SCAL using the DFH method. The study was conducted as a blind test, in which the experimental relative permeability curves were not disclosed until after the pore-scale simulations were reported. The DFH modeling technology offers a great opportunity to provide digital SCAL as a complementary approach to conventional SCAL, helping reduce costs and fill gaps in complex flow scenarios such as enhanced oil recovery (EOR).