The burning of fossil fuels leads to an increasing concentration of carbon dioxide (CO2) in the atmosphere and consequently to global climate change. Despite the shift towards renewable energy, the dominant role of fossil fuels in global energy consumption necessitates solutions like Carbon Capture and Storage (CCS). CCS involves capturing CO2 from large emission sources and storing it deep underground, where CO2 displaces native fluids, such as brine. The efficiency of this displacement process is influenced by various geological and physical factors. Understanding and optimizing these factors is crucial. This work provides a comprehensive investigation of CO2 brine displacement in porous rock, using both experimental and numerical methods. The experimental data are analyzed more rigorously than in previous studies, leading to a robust stochastic description of the two-phase flow in heterogeneous porous media. Additionally, numerical experiments were conducted to investigate the displacement stability, providing a new and unexpected scaling for viscous instabilities. Thus, this work provides a comprehensive and solid basis for risk analysis of CO2 plume migration in CCS processes. The migration of the CO2 plume and the efficiency of CO2 displacement are primarily determined by multiphase flow parameters, namely relative permeability, and capillary pressure saturation functions, which are usually derived experimentally. In the frame of this work, the underlying numerical data analysis was developed based on the solid foundation of a combined stochastic interpretation of complementary experimental data sets. In the developed approach, data from different experimental methods (Special Core Analysis - SCAL) are analyzed simultaneously, and their uncertainty is rigorously determined by state-of-the-art stochastic methods. The resulting uncertainty intervals of the saturation functions refer to the intrinsic uncertainty of SCAL experiments, however, not to variations in rock properties. By interpreting and combining experiments derived from various methods and on various samples, the analyses provide a certain access to the heterogeneity of the rock formation.
To underscore the impact of rock heterogeneity on CO2 migration, the approach was lifted to a larger length scale at which rock heterogeneity cannot be ignored anymore. Traditional SCAL methodologies typically do not account for heterogeneity, leading to discrepancies in measurements and field observations in terms of multiphase-flow saturation functions. Heterogeneity is vital for understanding the dynamics of plume migration and is explored in depth in this thesis. The thesis introduces an upscaling workflow that combines SCAL interpretations with continuum-scale experiments, emphasizing the need for rigorous upscaling procedures for CO2 storage in heterogeneous formations, such as carbonates.
Plume migration in heterogeneous formations is particularly affected when the mobility of the displacing fluid is higher than that of the displaced fluid. In this situation, viscous instabilities are to be expected, which can enhance fluid bypassing (sweep efficiency) in heterogeneous rock, depending on the characteristic length scale of the perturbation versus the finger width of the unstable front. This research challenges and extends existing theories on viscous fingering and its relation to interfacial tension and formation permeability. It further elucidates the findings from Darcy-scale numerical simulations that reveal finger wavelengths ranging from tens to a hundred meters under. Such a scale contrasts sharply with traditional predictions based on the Saffman & Taylor model, which significantly underpredicts the wavelengths. This insight is crucial for accurately predicting plume migration in CCS projects, as it accounts for the substantial deviation from expected behavior based on conventional models. The findings offer a novel perspective on the complexities of viscous-unstable displacement, challenging existing theories and providing a more accurate framework for understanding and predicting CO2 plume migration in CCS scenarios.
This thesis substantially advances our understanding of CO2 plume migration, addressing critical aspects of CO2-brine displacement, uncertainties in ideal homogeneous and stabilized systems, and the effects of laboratory-scale heterogeneity and viscous instability. By rigorously investigating the scaling of finger wavelengths and their implications, this work reveals the significant impact of viscous-unstable displacement on plume migration, reshaping our approach to CCS modeling and implementation. The research's holistic examination, spanning from traditional measurements to advanced numerical methodologies, elevates the field's understanding of geological CO2 storage, paving the way for more informed and effective carbon sequestration strategies and the associated risk assessment.