Pore-scale imaging and analysis of low salinity waterflooding in a heterogeneous carbonate rock at reservoir conditions

X-ray micro-tomography combined with a high-pressure high-temperature flow apparatus and advanced image analysis techniques were used to image and study fluid distribution, wetting states and oil recovery during low salinity waterflooding (LSW) in a complex carbonate rock at subsurface conditions. The sample, aged with crude oil, was flooded with low salinity brine with a series of increasing flow rates, eventually recovering 85% of the oil initially in place in the resolved porosity. The pore and throat occupancy analysis revealed a change in fluid distribution in the pore space for different injection rates. Low salinity brine initially invaded large pores, consistent with displacement in an oil-wet rock. However, as more brine was injected, a redistribution of fluids was observed; smaller pores and throats were invaded by brine and the displaced oil moved into larger pore elements. Furthermore, in situ contact angles and curvatures of oil–brine interfaces were measured to characterize wettability changes within the pore space and calculate capillary pressure. Contact angles, mean curvatures and capillary pressures all showed a shift from weakly oil-wet towards a mixed-wet state as more pore volumes of low salinity brine were injected into the sample. Overall, this study establishes a methodology to characterize and quantify wettability changes at the pore scale which appears to be the dominant mechanism for oil recovery by LSW.


1.
A pressure of 2 MPa using deionized water was applied in the confining isolated space between the Viton sleeve and the carbon fibre sleeve and a dry (air) scan of the sample was taken. 2. The sample was initially flooded with brine solution made from deionized water with 20 weight% potassium iodide (KI). This brine can be distinguished from rock phases for effective image segmentation and characterization of the rock bimodal porosity 1 . 3. The sample was then cleaned with four-times diluted formation brine followed by isopropanol both injected for 10 pore volumes. It was dried with a gentle nitrogen flux for 24 hours then vacuumed for 3 hours. Formation brine was then injected to fully saturate the rock sample. 4. Drainage was performed with the injection of viscous synthetic oil to reach irreducible water saturation. This oil has a relative density and dynamic viscosity (measured at 20 °C) of 836 kg/m 3 and 12 mPa.s respectively. 5. Toluene was then injected for 10 PVs followed by the injection of crude oil at an increased temperature, 50 °C to avoid asphaltene precipitation. 6. The sample was then aged with a continuous injection of crude oil for three weeks at a low flow rate (2 µL/min) at high pressure (11 MPa) and temperature (80 °C) to change the rock wettability to a similar state found in oil reservoirs. The flow direction was reversed at mid-point during ageing. 7. After ageing, the crude oil was doped with 20 weight% 1-iododecane as a high contrast dopant to distinguish the oil phase from the brine and rock phases in the Xray images. 8. The temperature in the system was then changed to 70 °C using an Omega flexible heater and a PID controller as shown in Fig. S1. 9. Low salinity brine was then injected at a sequence of six waterfloods. At each step 10 pore volumes were injected at a different flow rate. The flow rates and corresponding capillary number are provided in Table S1. Figure S1. The experimental flow apparatus consisted of core holder assembly, syringe pumps to apply back pressure and flow rates and a PID controller to apply temperature on the rock and fluids inside the micro-CT. Table S1. Low salinity waterflooding injection steps with calculated capillary numbers. A total of 60 pore volumes were injected throughout waterflooding with 10 pore volumes at each rate.

Image acquisition, processing, and segmentation
Image acquisition was performed with 95 keV energy, exposure time of 0.8 s and 3600 projections with the continuous helical movement of the sample and 360° rotation. The resultant 3D-images (tomograms) were registered, filtered and segmented, then used for fluid saturation and pore occupancy analyses. A smaller sub-volume was extracted for the contact angle, curvature, and capillary pressure analyses, Fig. S2.
The images were filtered using non-local means edge-preserving filter to eliminate image noise 2 . A watershed algorithm was used for image segmentation. This algorithm is based on generating a seed using a two-dimensional histogram of both the greyscale and the greyscale gradient images 3,4 , as shown in Fig. S3. A mask of segmented pores from the dry scan was applied on all subsequent images to separate the macro-pores from the rest of the rock and to simplify the segmentation of the oil and brine within, Fig. S4. Figure S2. A tomogram of the sample used in this study with a total volume of 28.8 billion voxels. Contact angle and curvature measurements were performed on an extracted sub-volume (1.5 mm 3 )  Figure S3. An example showing the segmentation workflow using seeded watershed algorithm. A seed is generated using a two-dimensional histogram of both the greyscale (a) and the greyscale gradient (b) images to produce a labelled image (c) showing the different phases: brine, oil and rock. Figure S4. Image segmentation workflow. A filtered two-dimensional cross-section of the threedimensional micro-CT dry image (a), which is segmented into resolved pores (black) and grains (grey) in (b). An orthogonal slice from an image of the sample saturated with brine and doped oil at the end of low salinity waterflooding (c). The pores mask was applied to the saturated sample image to separate the pore space and segment it into oil and brine shown in red and blue respectively, and then the rock phase in grey was added (d).

Pore space characterization (differential imaging)
To accurately quantify sub-resolution porosity, a technique called differential imaging is used 1 . A brine solution made from de-ionized water doped with 20 weight% Potassium Iodide (KI), as a high contrast agent, was used in the characterization of sub-micron porosity. This dense brine can be distinguished from rock phases for effective image segmentation and characterization of the rock bimodal porosity. As mentioned previously in the experimental procedure steps, X-ray micro-CT was used to acquire a dry (air) image of the rock and an image after KI-brine injection. To maximize the phase contrast between the grain and pore phases, a differential image between the KI-saturated and dry images was obtained. This image was then segmented using the seeded watershed method into solid grains, subresolution pores and macro pores. This differential imaging workflow is shown in Fig. S5. The segmented label images combined with the KI-saturated image were used to obtain histograms of the grey-scale voxels for each phase (Fig. S6). The peak values (CT) of the histograms were used to calculate the micro-porosity fraction (ϕmicro), The total pore space comprises the macro-and macro-pores, where Vmicro and Vmacro are the total volume fractions for each phase obtained from the segmented image. The total porosity from differential imaging was 0.299 compared to a value of 0.293 measured by a Helium porosimeter (measurement conducted at Imperial College London). A detailed explanation of the differential imaging method can be found in ref. 1.

Pore occupancy analysis
The distribution of pore elements fluid occupancy, sizes and location were investigated using a generalized pore network extraction algorithm 5 . The algorithm divided the pore space into pores (wider regions) connected with throats (narrower regions) with voxels assigned to unique pores and throats (Fig. S7a). The volume-weighted fractions of pore elements whose centers were occupied mainly by either brine or oil were calculated to assess the fluid occupancy at different stages before and during low salinity waterflooding (Fig S7b-d).

Contact angle measurements
Contact angle measurements were conducted on the segmented images of a 1.5 mm × 1.5 mm × 1.5 mm sub-volume using an automated algorithm 6 . The number of contact points and values of contact angles measured before and during waterflooding is shown in Table. S2.

Curvature measurements
The curvatures were analyzed from the fluid interfaces extracted from the same sub-volume on which contact angles measurements were performed. The voxelized interfaces were smoothed using a volume preserving Gaussian smoothing 7 , Fig. S8. The mean of the two principal curvatures ( 1 and 2) was calculated as = ( 1 + 2)/2. Figure S8. Oil-brine interfaces were smoothed using volume preserving Gaussian smoothing (kernel size 5) to remove voxelization artefacts. Curvatures were measured on the smoothed interfaces. The interfaces were extracted using Avizo 9.5.