Optical Image Analysis Applied to Pore Network Quantification of Sandstones Under Experimental CO2 Injection

Abstract:

This research aims to propose a protocol for pore network quantification in sandstones applying the Optical Image Analysis (OIA) procedure, which guarantees the measurement reproducibility and its reliability. Two geological formations of sandstone, located in Spain and potentially suitable for CO₂ sequestration, were selected for this study: a) the Cretaceous Utrillas unit, at the base of the Cenozoic Duero Basin and b) a Triassic unit at the base of the Cenozoic Guadalquivir Basin. Sandstone samples were studied before and after the CO₂ experimental injection using Optical and scanning electronic microscopy (SEM), while the quantification of petrographic changes was done with OIA.

The first phase of the rersearch consisted on a detailed mineralogical and petrographic study of the sandstones (before and after CO₂-injection), for which we observed thin sections. Later, the methodological and experimental processes of the investigation were focused on i) adjustment and calibration of OIA tools; ii) data acquisition protocol based on image capture with different polarization conditions (synchronized movement of polarizers), using 7 images of the same mineral scene (6 in crossed polarizer and 1 in parallel polarizer); and iii) automated identification and segmentation of pore in 2D mineral images, generating applications by executable macros. Finally, once the procedure protocols had been, the compiled data was interpreted through an automated approach and the qualitative petrography was carried out.

The quantification of changes in the pore network through OIA (porosity increase ≈ 2.5%) has allowed corroborate the descriptions obtained by SEM and microscopic techniques, which consisted in an increase in the porosity when CO₂ treatment occurs. Automated-image identification and quantification of minerals, pores and textures together with petrographic analysis can be applied to improve pore system characterization in sedimentary rocks. This research offers numerical support to quantify geometric and densitometric mineralogical parameters that would complement the traditional petrographic procedures.