During carbon dioxide (CO₂) sequestration, CO₂ injection suffers from viscous fingering and low sweep efficiency. In addition, the lower density of CO₂ compared to in-situ brine leads to the possibility of sequestered CO₂ rising up through the relatively permeable path in the cap rock and being emitted back out to the atmosphere. This research proposes a mechanism of CO₂-in-brine emulsion stabilization by surface-coated nanoparticles as a potential cure for these problems. This mechanism is studied in detail by conducting a series of core floods to investigate the interactions between nanoparticles and the surroundings such as fluids and rock surfaces during nanoparticle transport in sedimentary rocks. The experiments presented here use n-octane as a low-pressure analog fluid to supercritical CO₂ as they share several key characteristics. Comparisons of pressure drop and CT images from drainage displacement experiments with and without nanoparticles show that nanoparticle-stabilized emulsions were generated in-situ in highly permeable and homogeneous Boise sandstones tested in this study. Roof snap-off is proposed as the key mechanism for generating the emulsions. The imbibition experiment presents a case where Roof snap-off does not occur. The pressure drop for the control experiment and the nanoparticle experiments confirmed that without Roof snap-off nanoparticles do not affect the dynamics of the displacement except for the viscosity increase of the aqueous phase. However, it was inferred from the saturation profiles and effluent concentration history that nanoparticles were traveling faster than the aqueous phase in which they were dispersed and accumulating at the main displacement front. Inaccessible pore volume is proposed as a mechanism responsible for the accelerated transport of nanoparticles. The single-phase flow experiments demonstrate the accelerated transport of nanoparticles in porous media that was invoked to explain observations during imbibition displacement. During these experiments, tracer and nanoparticles were simultaneously injected into a porous medium and their effluent concentrations were monitored using a UV-Vis detector. The results show that nanoparticles traveled faster than the tracer in Boise and Berea sandstones studied in this research. Two-site model developed by Zhang (2012) was used to fit the data. Simulations suggested that the two-site model could replicate the overall shape of the experimental data when a slug of nanoparticle dispersion was injected, but it was not able to accurately predict the leading edge and the trailing edge of the effluent concentration history, where nanoparticles appeared before tracer due to accelerated transport. To account for the enhanced transport of nanoparticles, a modified two-site model with an acceleration factor, E, is proposed. The resulting fit matched the experimental data better than the original two-site model. / text
| Identifer | oai:union.ndltd.org:UTEXAS/oai:repositories.lib.utexas.edu:2152/22366 |
| Date | 21 November 2013 |
| Creators | Chung, Doo Hyun |
| Source Sets | University of Texas |
| Language | en_US |
| Detected Language | English |
| Format | application/pdf |
Page generated in 0.0017 seconds