Pore-Scale Mechanism for CO₂ Storage in a Fracture-Matrix System, Physical and Geometrical

The aim of the project is to establish criteria for CO₂ transport in fractured and heterogeneous rock. Main activities will be development of a 3D pore-scale model and computation of capillary-controlled fluid displacements in porous rock.
A 2D semi-analytical model is developed for computing fluid configurations, capillary entry pressures, capillary pressure curves and saturation profiles evolution directly in 2D rock images at uniformly- and mixed-wet conditions.

The simulated results depend on initial saturation and wettability. Based on this model, a novel dimensionless capillary pressure function for mixed-wet conditions has been developed that could be implemented in reservoir simulation models and describe variations in formation wettability more accurately, and hence allow for improved evaluation of CO₂ storage processes.

Work is in progress to extend the semi-analytical model to three fluid phases, which will enable us to investigate how the presence of an oil phase affects the CO₂ capillary entry pressures.

A 3D level set based model has been developed for computing the evolution of fluid configurations and capillary pressure curves in 3D images of heterogeneous and fractured rock. A novel method to account for nonzero contact angles at the pore/solid boundaries has been developed and implemented in the model. Drainage and imbibition simulations in high-resolution 3D images of sandstone rock samples demonstrate that the model accounts for well-known pore-scale mechanisms, such as Haines jumps, interface splitting and coalescence, retraction and snap-off. The models are validated by comparing capillary entry pressure computations with results from the semi-analytical model.

Three-fluid displacements in 3D porous rocks will be simulated by combining a variational level set model with the nonzero contact angle model. Currently, formation and evolution of triple junctions in three-fluid systems are simulated successfully.

Micro-CT scanning of Bentheim sandstone has been performed at LBNL to generate data sets of realistic 3D rock samples, which will be taken as input to the computational models. 3D fractured rock samples have been generated by pulling grains apart along fracture planes using computer algorithms developed at LBNL. We are currently performing level set simulations on these samples to investigate effects of the presence of micro-fractures on the fluid configuration and capillary pressure properties.