CO2 exsolution -- challenges and opportunities in subsurface flow management

Advisors

Sally M. Benson, primary advisor
Roland N. Horne, advisor
Anthony Kovscek, advisor

Abstract

Carbon dioxide is known to be highly soluble in water/brine, up to 5% mass fraction under reservoir conditions. In geological carbon sequestration, a large amount of injected CO₂ will dissolve in brine over time. Exsolution occurs when pore pressures decline and CO₂ solubility in brine decreases, resulting in the formation of a separate CO₂ phase. This scenario occurs in carbon sequestration reservoirs by upward migration of CO₂ saturated brine, through faults, leaking boreholes or even seals. In this way, dissolved CO₂ could migrate out of storage reservoir and form a gas phase at shallower depths. Questions such as how exsolved CO₂ distributes and transports, and how multiphase flows and trapping are altered in a reservoir undergoing exsolution need to be answered to achieve better subsurface flow management and risk evaluation. This study summarizes the results regarding the implications of exsolution on storage security, including pore-scale and core-scale experiments, pore-scale modeling, and numerical simulations. Applications of CO₂ exsolution in Enhanced Oil Recovery are also explored. Microscopic observation of CO₂ exsolution in porous media under reservoir conditions have shown that, different from an injected CO₂ phase, where the gas remains interconnected and distributes at capillary equilibrium, exsolved CO₂ nucleates in various locations of a porous medium, forms disconnected bubbles and propagates by repeated expansion-snap off process under capillary instability. A good correlation between bubble size distribution and pore size distribution is observed, indicating that geometry of the pore space plays an important role in controlling the mobility of brine and exsolved CO₂. Core-scale multiphase flow experiments demonstrate that in the process where growing gas bubbles displace water (drainage), the water relative permeability drops significantly and is disproportionately reduced compared to gas injection, and the CO₂ relative permeability remains very low, 10^-5 to 10^-3, even when the exsolved CO₂ saturation increases to over 40%. Furthermore, during imbibition, exsolved CO₂ remains trapped even under relatively high capillary numbers (~ 10^-6), and the water relative permeability at the imbibition endpoint is one third to one half of that with water displacing injected CO₂. A model is developed to simulate the growth of exsolved gas phase in porous media under capillarity. Results are compared with experimental observations using three dimensional micro X-ray tomography. Convective transfer in the aqueous phase has been demonstrated to play an important role in controlling bubble growth and accumulation. With a Stokes flow simulator, water relative permeability curves are estimated for various sedimentary rocks under different conditions. We are capable of matching modeled gas distribution and relative permeabilities with experimental data, and extrapolating expected phase mobility reductions under reservoir conditions. CO₂ exsolution does not appear to create significant risks for storage security. Due to the low mobility of exsolved CO₂ and its large impact on reducing water flow, if carbonated brine migrates upwards and exsolution occurs, brine migration would be greatly reduced and limited by the presence of exsolved CO₂ and the consequent low relatively permeability to brine. Similarly, if an exsolved CO₂ phase were to evolve in the seal, for example, after CO₂ injection stops, the effect would be to reduce the permeability to brine and the CO₂ would have very low mobility. It is also possible that CO₂ exsolution could have an effect on CO₂-EOR recovery. This flow blocking effect is studied in experiments with water/oil/CO₂ for the purpose of water conformance and oil recovery enhancement. Experiments show that exsolved CO₂ performs as a secondary residual phase in porous media that effectively blocks established water flow paths and deviates water to residual oil zones, thereby increasing recovery. Overall, our studies suggest that CO₂ exsolution provides an opportunity for mobility control in subsurface processes. For example, CO₂ exsolution generated intentionally increases water sweep efficiency in oil reservoirs and forms gas barriers to seal high permeability zones. However, while the experimental evidence for dramatic mobility reduction is clear, the lack of simulation capability that accounts for differences between the CO₂ phase morphology resulting from gas injection and gas exsolution creates challenges for modeling and hence, designing studies to exploit the mobility reduction capabilities of CO₂ exsolution. Not only is history dependent behavior (hysteresis) important, but also process dependent behavior is needed. Using traditional drainage multiphase flow parameterization in simulations involving exsolution will lead to overestimates of flows and large errors in transport rates. Development of process dependent parameterizations of multiphase flow properties will be a key next step and will help to unlock the benefits from gas exsolution.