Alkaline flooding is a low-cost, promising chemical enhanced oil recovery (EOR) method when crude oil contains sufficient saponifiable (acidic) components. In this technique, in-situ surfactants are generated as the high-pH alkaline solution reacts with the acidic components of the oil at the oil–water interfaces. As a consequence, interfacial tension (IFT) reduces, and both phases emulsify, possibly leading to additional recovery. Since emulsification and initial oil mobilization takes place on the scale of individual pores, observations on the pore scale will provide insights into the physics and displacement mechanisms of alkali-based methods that can be used for the chemical optimization of injection fluids. Typically, the chemical composition of the injection solution is examined and designed by conducting classical phase behavior experiments through optical inspection of the microemulsion formation in test tubes. However, the dark color of heavy crudes makes the fluid and microemulsion phase identification and quantification challenging, if not impossible. This thesis investigates the effect of alkaline solutions on the displacement of and the emulsification with high total acid number (TAN) crude oil from the Vienna basin. Microemulsion formation is evaluated in conventional phase behavior experiments using the linear mass absorption coefficient of micro-x-ray tomography for quantifying the phase compositions. The predictive strength and the representativeness of these ex-situ experiments are investigated by flooding experiments and imaging emulsification under flow conditions. Flooding experiments are executed in microfluidics and micro-computed tomography (CT)-based core flood experiments to evaluate the alkaline flooding efficiency and microemulsion formation under flow conditions. Using x-ray methods makes it necessary to study the possible effects of x-ray contrast-enhancing additives on the complex fluid-phase behavior. This work shows that the typical visual assessment of classical phase behavior experiments disregards that a little oil contamination in water may lead to a substantial coloring of the aqueous phase and can be misleading. The implementation of x-ray attenuation leads to conclusive evaluations and the calculation of the exact phase composition and material balance. Using an x-ray technique, the same signature in phase behavior is found in micro-CT-based core flood experiments under flow conditions (as in the test tubes). The results indicate that minimal mutual emulsification can be identified as optimum, leading to the best oil recovery. Pore-scale EOR potential is evaluated by statistical and topological means in two-dimensional (2D) and three-dimensional (3D) porous domains. The size and fluid content of pores are monitored individually after each flooding experiment and the pore-filling trends are presented. The earlier notified discrepancy of the pore-scale fluid distribution in 2D and 3D systems is circumvented by the identification of proper and comparable quantities, demonstrating that 2D microfluidics can be used as a screening method for injection water optimization.