Estimação de propriedades petrofísicas de rochas sedimentares a partir de imagens microtomográficas de raíos-x.

Abstract

An innovative technique currently under development is the construction of digital templates from a set of X-ray tomography, high-resolution images acquired in rock samples. This technique allows to define the 3D geometry of the pore space and mineral grains, as well as recognition of the minerals present in the rock. The objective of this research is to estimate the petrophysical properties (grain size distribution, mineral composition, porosity, permeability and P wave velocity) of sedimentary rocks from microtomography X-ray images. This work analyzed five samples of sandstone (A4, A7, A9, AM10H and AM14V), one of shale (F9), another of limestone (Carbonato Rosário) and two samples of carbonatic Tufa (Tufa CR and Tufa FG). Regarding the particle size distribution is possible to quantify the proportion of fines present in the samples, but such concentrations may be changed according to the resolution of the images analyzed. In this work we propose a new method for the analysis of mineral composition from microCT images. The results obtained for the simulations indicate that the mineral composition of the sandstone samples have approximately the same mineral composition, with different proportions of quartz and clay. Shale sample has an essentially clay composition and the carbonate samples are essentially formed by calcite and dolomite. DRX analyses have proved the mineral composition indicated by microCT analysis. In general, it is observed an approximately linear relationship between porosity values measured in the plugs and those determined by simulation of microCT images. The numerical simulation resulted in values of permeability compatible with permeabilities measured in the laboratory, except for the shale sample and for the Tufa CR sample. For the case of samples having an approximately monomineralic composition, acoustic velocity simulations showed excellent results compared with the physically measured velocities. But for the case of multimineralic rock samples the model generation not considered the different phases making up the mineral matrix, which resulted in simulated velocities far above the measured velocities.