An understanding of multi-phase flow in porous media is of crucial
importance in many important industrial and environmental applications including the
recovery of oil and gas from sub-surface reservoirs, the uptake of water by plant roots and the transport of
organic pollutants in groundwater. This project uses a number of novel tools to exploit the convergence of imaging technology and
compute power to break new ground in the understanding of how fluids displace
one another inside the complex geometry of real world porous materials.
The continuously evolving Applied Mathematics X-ray micro-CT facility generates images that describe
the microstructure of materials down to the micron scale with a resolutions beyond
2000x2000x8000 voxels. We have developed a distributed-memory software toolkit
for the analysis and registration of these images, so that accurate mapping of fluid distributions in partially saturated media is now possible at the pore scale.
Continual advances in compute power now means that direct computation of flow properties, either by solution of Navier-Stokes equations, or by finding interfacial surfaces of constant hydraulic radius, can be performed on large, directly imaged volumes. These methods provide a perfect complement to the simplified geometry of "network" models in the study of multi-phase flow phenomena. Combining this modelling with 3D imaging opens up the possibility of developing models with true predictive power, something previously unattainable.