Comparative study of undissolved and karstified limestone based on microtomography

We develop methods for qualitative and quantitative assessment of the transformation of pore geometry of a rock as a result of karstification. We then apply these tools to characterize dissolution-induced changes in Miocene limestone samples collected from a quarry located near Smerdyna (Poland), where intense epikarst development is observed, with the formation of hundreds of solution pipes. Partially dissolved samples collected in the immediate vicinity of the pipes are compared with undissolved samples collected several meters away.

For both types of samples 26 micron resolution grayscale X-ray scans has been performed, and cubical regions of interest of size 506^3 voxels, which corresponds to (13,156 mm)^3, have been studied. Images have been segmented by tuning the grayscale threshold to match the experimentally measured porosity values of respective samples. Additionally, based on the segmented tomograph of undissolved sample another geometry was numerically created in order to mimic a uniform dissolution of the rock up to a porosity value equal to that of the dissolved sample.

The irregular geometry of the pore space, vast majority of which forms a single connected component, can be conveniently characterized by a local thickness function,  which corresponds to a diameter of the largest sphere that fits within the pore space and contains a given point. A similar measure can be introduced for the solid component (grains). We have compared thickness distributions  of undissolved and dissolved sample as well as numerically generated uniformly dissolved sample. Such a comparison allowed us to quantify the extent of homogeneity of the natural karstification process.

To further characterize pore geometry, we have calculated the ellipsoid factor, which – based on the axis lengths of the fitted ellipsoids – can be used to characterize how prolate or oblate the pore space locally is. Next, we have used (modified) Flinn diagram to quantify differences between undissolved, numerically eroded and naturally dissolved samples, especially those indicating pore merging and inhomogeneous dissolution.

The above analysis is complemented by calculation of connectivity density – a topological measure of the degree to which a structure is multiply connected. Values obtained for undissolved, numerically dissolved and naturally dissolved samples indicate on ‘excessive’ reduction of interconnections during natural dissolution, which may be understood on the basis of high degree of pore merging due to inhomogeneous dissolution.

Both methods: (generalized) thickness analysis and connectivity calculation emphasise the role of merging of pores and inhomogeneous dissolution in the process of natural dissolution for the analyzed  samples.