Segmented 3D Images of Supercritical CO2-Brine Draiange into Bentheimer Sandstones

Full description

In this experiment, we investigated the coupling effect of sub-core-scale heterogeneity and miscibility on the development of flow pattern under drainage in Bentheimer sandstone cores. The multiphase fluid-flow experiments were conducted using the high-temperature, high-pressure (HTHP) tri-axial core flooding system built in-house at the National Laboratory for X-ray Micro Computed Tomography (CTLab), Australian National University. Two sister cores drilled from the same Bentheimer sandstone block were used in this experiment and the porosity profile for the two cores are significantly different with two lower porosity bands being transverse to the direction of flow injection presented in both cores. For the drainage experiments, two series of sub-experiments were performed for each core. In series S ("S" for CO2-saturated), supercritical CO2 (scCO2) was injected into core saturated with CO2-saturated brine to investigate brine displacement by scCO2. After sub-experiment S, unsaturated brine without any dissolved CO2 was flushed through the core to dissolve any trapped scCO2. X-ray fast scans were taken to monitor if all trapped scCO2 was removed, and to ensure the core was reset to fully saturated by unaturated brine. To investigate the coupling of brine displacement by scCO2 and the dissolution of the invading scCO2 into brine, sub-experiment series US (“US” for unsaturated) was performed by injecting scCO2 into the core filled with unsaturated brine. In both the sub-experiments S and US, scCO2 was injected at flow rate equaling 0.01 ml/min, and the corresponding capillary number (Ca) was at the order of 10^-9. Before any fluid flow experiment was conducted, a high-resolution base scan was acquired for core 01 and 02 at voxel size equaling 3.85 and 3.78 micron, respectively. For each of the four sub-experiments (S01, US01, S02, and US02), fast MCT scans were taken simultaneously to track the invasion pattern of scCO2. The resultant voxel size and acquisition time for core 01 (i.e. S01 and US01) equals 15.98 micron and approximately 30 mins, respectively. For core 02, the voxel size was reduced to 13.72 micron by moving the HTHP cell slightly closer to the X-ray source and increasing the voltage for the source. The acquisition time for each fast scans of core 02 (i.e. S02 and US02) were thereby reduced to approximately 16.5 mins. Subsequent image segmentation of the high-resolution dry scan was accomplished via a “Converging Active Contours” (CAC) routine, where the two physical phases, i.e. the air and the grain phase was identified by CAC based on chosen intensity thresholds. For the segmentation of the fast scans acquired during the drainage experiments, each scan was firstly partially segmented as the scCO2 phase and the combined brine and grain phase. The partially segmented fast scans were then overlaid with the segmented high-resolution dry scan for the identification of the brine and the solid phase. Subsequent noise removal of the segmented fast scans including removing scCO2 clusters that were smaller than sphere with radius euqaling 15 micron. In this dataset, data are in .nc format and can be read in ImageJ software via NetCDF plugins (http://www.unidata.ucar.edu/software/netcdf/). Datasets labelled as S0X and US0X were drainage experiments conducted using the same core under immiscible and partially miscible conditions, respectively, with X=1 or 2 labelling experiments conducted with core 01 or 02. Here, the numerals label datasets for each sub-experiments until the breakthrough of scCO2, with datasets in which scCO2 was firstly oberved labelled as 1, e.g. S01_1; "endpoint" labels datasets acquired at the end of each sub-experiment. This dataset contains data acquired until the breakthrough of scCO2 and at the end of each sub-experiment. The remaining datasets are available on request.

Notes

40.5 GB.