Simulation and Experimental Investigation of Different Phenomena in CO² Storage in the Saline Aquifers
MetadataVis full innførsel
Climate change is believed to be one the main challenges for human society. The emission of greenhouse gases is believed to increase global warming and CO2 is one of the most important greenhouse gases emitted by humans. A proposed option for the reduction of CO2 emissions to the atmosphere is CO2 capture and storage. CO2 storage in geological formations is considered to be an important solution for preventing CO2 emissions to the atmosphere; however, it is essential to perform it safely and securely. Success of the project is highly dependent on understanding the process at reservoir conditions. One of the suitable formations considered for CO2 storage is a saline aquifer. There are other potential geological storage possibilities including depleted oil and gas reservoirs and coal beds, but the main attraction of brine formations as suitable storage sites for CO2 is the availability and larger volumes. After injecting CO2 into the aquifer formation, it will be trapped in the reservoir by various mechanisms. Each trapping mechanism has generally a different timescale. At the beginning the structural and geological trapping mechanism dominates, which is trapping of CO2, below a sealing cap rock. During the movement of CO2 upward or along a dipping barrier, brine will be imbibed into the formation that already has CO2 inside the pores and some fraction of the CO2 will be trapped in the pore spaces. This is called residual trapping. In a long-term process, CO2 starts to be soluble in the brine and this makes a CO2-rich phase that is now dense and located on the top of the brine. As a result, the heavier saturated CO2 becomes unstable and will move into deeper parts of the reservoir and the convection mixing phenomena will be activated. This is called solubility trapping. In this research, geological, residual and solubility trapping mechanisms are studied but main focus is on the residual trapping mechanism through experimental and simulation activities. This thesis also presents numerical simulation investigations of the geological, residual and soluble trapping mechanisms involved in CO2 storage in the brine. As part of this research, 2D glass model experiments are carried out to demonstrate the effect of different parameters during CO2 injection into the reservoir and also during the imbibition process by the brine. This is done by use of proxy fluids that are chosen to represent the CO2-brine behavior at reservoir condition. Images from the 2D model setup are analyzed through an image analysis method based on red, green and blue (RGB) color concept. The results are compared with the fine scale numerical simulation results. The effect of the injection rate on both the drainage and imbibition processes is discussed. Using different fluid sets, the effect of wettability has also been studied. In connection with the 2D model experiment, relative permeability experiments with the same fluid property and porous material are carried out and the results are compared and used in the simulation study. Conceptual simulation models are constructed based on CO2 storage reservoir conditions and sensitivity analysis of the effective parameters is used to reveal the importance of them. In the study of residual trapping, the effect of wettability, dipping angle, imbibition modeling methods and grid refinement on the state of CO2 on the short- and long-terms are studied. The behavior of the CO2 after injection in different scenarios is compared and different reservoir conditions are examined. Finally, compositional simulations constructed for the purpose of studying solubility trapping mechanism and different reservoir parameters that affect it are investigated. The results of the relative permeability measurements and the 2D visual model verify that using two sets of fluids representing the CO2-brine systems in the reservoir can be a feasible method to measure the effect of different parameters in CO2 storage in the brine. Scaling the experiments by use of the dimensionless parameter shows that the designs are close to the real conditions. Relative permeability measurements show sensitivity of the residual CO2 proxy fluid volumes, saturation end points and corresponding relative permeability on both the rate and wettability. However, in the case of less water-wet systems, residual CO2 proxy trapping is more sensitive to the imbibition rate. Correlations are proposed to predict CO2 residuals based on the effect of the rate and wettability. The results of image analysis confirm that both imbibition and drainage rates affect the CO2 proxy fluid trapping saturations. It is observed that higher drainage rates created higher residual saturation, while higher imbibition rates caused lower residual saturation. Using a fluid with higher wettability, considerably lower CO2 proxy residuals are predicted. This conclusion is compared with the findings from balance readings and simulation studies and good agreement is found between them. Different simulation scenarios for conceptual models based on CO2 storage reservoir characteristics are constructed. Considering the base relative permeability model as representative of water-wet condition, the highest trapping volume is in a slightly waterwet condition and the lowest is the case of a strongly water-wet scenario. The results of the compositional modeling of the solubility trapping mechanism reveal that the CO2 state is sensitive to the reservoir permeability and to brine salinity. Higher permeability resulted in higher values of CO2 dissolved into the reservoir after the start of convection, while lower salinity values resulted in very sharp increases in the volume of CO2 dissolved in the brine. The onset time of convection for the base case simulation is higher than estimations by theoretical stability analysis, but by decreasing the grid sizes this difference is decreased. Considering capillary pressure in the modeling increased the predicted soluble CO2 in the brine considerably.