Core plug and 2D/3D-image integrated analysis for improving permeability estimation based on the differences between micro- and macroporosity in Middle East carbonate rocks

Abstract Carbonate rocks are porous systems, with pores and pore throats of varying morphologies that result from depositional and diagenetic processes. Such heterogeneity produces a complex arrangement between the grains and pores that affects the petrophysical properties while limiting the utility of measurement techniques. Petrophysical properties are generally acquired by conventional laboratory methods, although to provide accurate results, core plugs need to be recovered intact. Two-dimensional digital image analysis (2D DIA) enables the processing of any core cut and requires minimal data manipulation and computation when compared to three-dimensional approaches. In DIA, permeability is calculated using models described in the literature that often do not provide good predictions for carbonate rocks. Often, the permeability, which is controlled by the size and shape of the pores and pore throats, is related to porosity values; however, the porosity of a rock varies from the micro to macro level, resulting in enormous uncertainty in estimating permeability. In this article, we present a new strategy to improve the prediction of permeability by using pore-shape parameters from 2D DIA, which provides data related to the macropores resulting from the optical resolution. This gas technique measures the absolute permeability, which is used as a calibration parameter, and the total porosity, which is used to calculate the microporosity. The test samples used are from Oman outcrops of the Huqf Supergroup and Salalah Formation, which are analogous to the carbonates of the giant reservoirs in the Middle East. Microporosity was characteristic of all the samples due to the calcite mud matrix, recrystallized calcite cement, microcracks and the crystalline texture caused by dolomitization. The pore-shape parameters from the 2D DIA improved the permeability prediction and were found to relate to the pore types that make up the rock, whereas the 3D technique did not provide a good result. The R2 of the 2D data was 0.96, demonstrating the efficiency of the procedures applied in mitigating the uncertainties of the models for the set of samples studied.