Resources / Publications
Kamila Scheffer (1), Yves Méheust (2), Marcio S. Carvalho (3), Marcos H.P. Mauricio (1), Sidnei
Journal of Petroleum Science and Engineering, 198, March 2021. DOI: 10.1016/j.petrol.2020.108134
EOR; Emulsion; Secondary recovery; Microfluidics; Micro-CT; 3D image analysis
The efficiency of emulsion injection as an enhanced oil recovery (EOR) method is investigated in a synthetic porous medium consisting of sintered bi-disperse glass beads, which emulates the porosity and permeability of real oil-bearing rocks. Synthetic seawater and an oil-in-water emulsion are successively used to displace a mineral oil which initially saturates the porous space. Micro-CT images with a 4 μm resolution are acquired at the end of the different stages of the process; the water phase is doped with KI to optimize the contrast between the liquid phases. Thus, three-dimensional (3D) images showing the beads, doped water and residual oil present a 3-modal histogram. After denoising with a non-local means filter, alignment and segmentation, the 3D images provide the spatial distributions of the water and oil phases, and thus allow comparing the populations of residual oil ganglia prior to and after the injection of the emulsion. Visual comparison of the spatial phase distributions show that the oil droplets of the oil-water emulsion divert the water path, mobilizing previously trapped oil ganglia. The probability density functions (PDFs) of different geometrical properties of the trapped oil ganglia (104 ganglia with volumes spanning 6 orders of magnitude) after water injection show well defined power law behaviors between a size corresponding to the typical pore throat and that typical of 10 typical pore volumes, and a few very large clusters of sizes between 10 and typical pore volumes. The largest of them alone accounts for 97% of the trapped oil. The same PDFs after emulsion injection demonstrate successful fragmentation of these few very large ganglia, which in this case is the key to efficient EOR procedure through diversion of the water flow by emulsion oil droplets to less permeable regions.
Dragonfly was used to process and analyze 3D images.
(1) Department of Chemical and Materials Engineering, Pontifícia Universidade Católica do Rio de Janeiro, Rio de Janeiro, Brazil.
(2) Univ. Rennes, CNRS, Géosciences Rennes, UMR 6118, 35000 Rennes, France.
(3) Department of Mechanical Engineering, Pontificia Universidade Católica do Rio de Janeiro, Rio de Janeiro, Brazil.
Return to Publications list.