Calculating structural and geometrical parameters by laboratory measurements and X-ray microtomography: a comparative study applied to a limestone sample before and after a dissolution experiment

Abstract

Abstract. The aim of this study is to compare the structural, geometrical and transport parameters of a limestone rock sample determined by X-ray microtomography (XMT) images and laboratory experiments. Total and effective porosity, pore-size distribution, tortuosity, and effective diffusion coefficient have been estimated. Sensitivity analyses of the segmentation parameters have been performed. The limestone rock sample studied here has been characterized using both approaches before and after a reactive percolation experiment. Strong dissolution process occurred during the percolation, promoting a wormhole formation. This strong heterogeneity formed after the percolation step allows us to apply our methodology to two different samples and enhance the use of experimental techniques or XMT images depending on the rock heterogeneity. We established that for most of the parameters calculated here, the values obtained by computing XMT images are in agreement with the classical laboratory measurements. We demonstrated that the computational porosity is more informative than the laboratory measurement. We observed that pore-size distributions obtained by XMT images and laboratory experiments are slightly different but complementary. Regarding the effective diffusion coefficient, we concluded that both approaches are valuable and give similar results. Nevertheless, we concluded that computing XMT images to determine transport, geometrical, and petrophysical parameters provide similar results to those measured at the laboratory but with much shorter durations.

Digital core repository coupled with machine learning as a tool to classify and assess petrophysical rock properties

“…that ensures that the porosity computed from the segmentation mask equals the predicted porosity obtained from the bottleneck features. Final loss function is the sum of the previously defined loss functions (1), (2) and (3):…”

“…They are computed from segmented images obtained by binary thresholding. Porosity estimation obtained by Hysteresis segmentation is benchmarked and validated in several studies with diverse partners (IFPEN, Geosciences Montpellier, [2]). Figure 8 illustrates the porosity distribution obtained by these 3 methods on 1,000 juxtaposed blocks of size 100 x 100 x 100 voxels belonging to the same 1000 x 1000 x 1000 voxels image.…”

“…Rock images with voxel size down to the micrometer resolution allow to extract the topology of the pore space. Once segmented in two (usually pore and solid) or more phases of interest, the MCT images can be used to simulate porous media physical properties such as fluids transport, electrical and geo-mechanical properties as well as processes like enhanced oil recovery simulation [1,2]. Ongoing improvements in MCT systems and image analysis software quickly provide increasing amounts of data.…”

Pore-scale tomography and imaging: applications, techniques and recommended practice

“…Followed by Luquot et al (2016), who calculated structural and geometrical features from a limestone sample before and after a dissolution experiment, Sell et al (2016) present a case study of processing Sr-CT data sets of gashydrate-bearing sediments. Khan et al (2016) and Chauhan et al (2016) support the field of imaging data processing by presenting new research on single-and multi-phase segmentation and classification.…”

Elastic Anisotropy and Internal Structure of Rocks from the Uranium Ore Occurrences of the Litsa Ore Area (Kola Region, Russia)

“…Recently, a number of papers, discussing application of the X-ray computer tomography (X-ray CT) method to solving different geological problems has been published (see e.g., reviews [3,4]). This non-destructive method was so far used, in particular, in petroleum geology-when studying reservoir properties [5,6], in geomechanics-to study rock disintegration [7] and in microstructural analysis [8,9]. Apparently, this method can also be efficiently used in the study of ore deposits, since it allows the transition from a two-dimensional image of textural and structural features of rocks and ores (and distribution of the components of interest) to three-dimensional one [10], thus leading to gaining important information about the mechanism of ore formation [11].…”