CT-Scan Image Dataset of Residual Fluid-Driven Fracture in a Molasse de Villarlod Sandstone Core - Post-Radial Hydraulic Fracture Experiment - M04 Sample

Dataset Description

This dataset contains high-resolution CT-scan images that capture the residual fracture surface within a core sample of Molasse de Villarlod Sandstone. The core sample was extracted after conducting a radial hydraulic fracture experiment on a 25 × 25 × 25 cm cubic block of sandstone (M04 Smaple). The experiment was designed to simulate fluid-driven fracture propagation and closure, and the resulting fracture path was preserved in the core sample.

Core Location in the M04 Cube Sample:

• Z: 12.5 cm
• South-North: 12.5 cm
• West-East: 13.5 cm to 23.3 cm (Coring direction)

This spatial information specifies the exact location and orientation of the core extraction within the M04 cube sample.

CT-scan instrument details:

The M04 sample was analyzed using an X-ray micro-CT scanner (RX-Solutions Ultratom) under consistent scanning protocols and parameters. A reflective 230 kV microfocus X-ray source (Hamamatsu L10801) equipped with a 0.2 mm thick copper filter, a tungsten cathode, and a tungsten target was employed for the imaging process. The scans were conducted with a voltage of 120 kV and a current intensity of 80 mA.

The volume data acquisition was performed in continuous helical mode, ensuring complete coverage of the sample’s height. For sample M04, 5 full rotations were executed, with 1312 projections captured for each 360° rotation, allowing for highly precise volume reconstruction. The X-ray beam attenuation was recorded by an XL Varex Paxscan 2530HE plane detector with a resolution of 2176 x 1792 pixels, and an exposure time of 0.50 seconds per projection.

The acquired projections were processed using RX-Solutions X-act software with Filtered Backprojection to reconstruct a corrected volume. This reconstruction yielded approximately 9000 slices in 16-bit TIFF format, with voxel dimensions of 10 x 10 x 10 microns, providing detailed insights into the internal structure of the sample.

Key Features:

• Fracture Characteristics: The fracture observed in the CT-scans represents a residual opening that remains post-fracturation. It is entirely contained within the core, showcasing the internal fracture geometry resulting from the hydraulic fracturing process.

• CT-Scan Details: The CT-scans were taken perpendicular to the fracture surface, offering a detailed cross-sectional view of the fracture at different depths. This orientation is critical for accurately capturing the fracture morphology and allows for the reconstruction of the fracture surface in 3D.

• Material Information: The core sample is composed of Molasse de Villarlod Sandstone, a sedimentary rock which is porous (18% porosity) and permeable. This material choice is relevant for studying fracture closure subjected to the leak-off of the fluid inside the porous medium.

• Experimental Context: The radial hydraulic fracture experiment aimed to simulate the propagation of hydraulic fracture and its closure due to the leakage of fluid inside fracture into the porous medium. The dataset provides valuable insights into fracture propagation patterns, surface roughness, and the effects of fluid-driven fractures in porous media.

Applications:

This dataset is particularly valuable for researchers and engineers involved in:

• Fracture mechanics and surface characterization
• 3D reconstruction and visualization of fracture surfaces
• Surface roughness analysis
• Hydraulic fracturing studies
• Geomechanical modeling

File Structure:

The dataset is organized into zip-folder contains .tif images corresponding to different depths within the core. Each tif-image is a CT-scan for that specific depth, labeled according to their position along the fracture path.

Processing code:

Follow the URL repository in the software section to access to the code for processing these images and reconstructing the fracture surfaces.

Acknowledgment:

We would like to extend our deepest thanks to Gary Perrenoud, Albert Taureg, and Lionel Pittet, the technical specialists of the PIXE platform at École Polytechnique Fédérale de Lausanne (EPFL). Their expertise and support in operating the CT-scan machine were important to the success of this research. We greatly appreciate their dedication and the high-quality work they provided.

Contact and Support:

Email:

Brice Lecampion: brice.lecampion@epfl.ch

Mohsen Talebkeikhah: m.talebkeikhah@gmail.com