Microfluidics Laboratory: Micromodel Studies of
CO2 Pore-Scale Trapping Processes
Key laboratory capabilities include:
To improve our understanding of geological carbon sequestration, pore-scale experimental and numerical studies of processes related to caprock-sealing efficiency and trapping are needed. Caprock-sealing efficiency is a measure of the capillary pressure at the caprock/reservoir interface that needs to be exceeded before supercritical carbon dioxide (scCO2) can move into the caprock. The major trapping mechanisms include storage of free-phase gas through hydrodynamic and capillary trapping, dissolution in formation brines, and mineral trapping through geochemical reactions. Many of these trapping processes are controlled by pore-scale mechanisms that can best be understand by pore-scale experiments and simulations.
PNNL has developed a high-pressure, pore-scale flow and transport capability for highly controlled displacement and dissolution experiments involving brine/scCO2 fluid pairs. The laboratory uses centimeter-scale microfluidics devices with porous structures etched into silicon wafers to study critical pore-scale multi-fluid flow processes.
Experiments conducted under a wide range of conditions have been used to quantify distinct displacement regimes (stable displacement, viscous fingering, capillary fingering) and their impacts on effective flow properties. Micromodel experiments and related analyses are being used to guide modeling at larger scales. For example, DeHoff et al. (2012) demonstrated that under unstable displacement conditions, the commonly used approach that couples constitutive relationships of relative permeability, saturation, and capillary pressure fails to provide good predictions. However, an approach that decouples relative permeability relationships can accurately predict average saturations. This result has concrete implications for how field-scale reservoir modeling should be performed under different injection conditions and will lead to improvements in field-scale STOMP modeling.
Variations in the pattern of liquid CO2 displacing water in a dual-permeability pore network under three different flow rates.