Skip to main content
The European High Performance Computing Joint Undertaking (EuroHPC JU)
Awarded Resources (in core hours)
System Partition
8 March 2023 - 7 March 2024
Allocation Period

Fossil fuels energy production is steeply increasing with correspondingly increasing atmospheric carbon dioxide (CO2) emissions. With this trend showing no sign of change, sequestration of liquid CO2 via injection into brine-filled large geological formations has been identified as one possible remedy.

Essential aspects of CO2 capture mechanisms are as follows: after injection, liquid CO2 is lighter than the resident fluid (brine) and migrates upwards, below the upper caprock of the reservoir. This situation is potentially hazardous, since the presence of fractures in the upper layer may let the CO2 escape and eventually return into the atmosphere. However, at the interface between the two fluid layers, supercritical CO2 dissolves in brine, forms a heavier mixture (CO2-brine), which flows downward making CO2 permanently trapped.

Therefore, it is crucial to evaluate adequately the vertical mass flux of CO2 because this macroscopic parameter determines the time required to achieve a complete dissolution of the volume of carbon dioxide injected. The dissolution time is crucial to determine the injection rate and to evaluate the suitability of potential injection sites.

The geological scale of the parameters involved makes precise flow predictions hard to obtain, and therefore an accurate modelling of the process is essential.

Predictions of this flow is made complex by a characteristic instability: the heavier layer of CO2-brine is unstable and gives rise to the formation of fingers, which tremendously increase the mixing rate of CO2 in brine.

As a result, these fingers make the sequestration process more efficient, but inevitably harder to predict.In this project we will perform high-resolution three-dimensional numerical simulations to analyze the dynamics of the CO2-brine mixture into deep saline aquifers. The parameter that characterizes the physical problem is the Rayleigh-Darcy number (Ra), which depends on the density difference between brine and CO2-brine mixture, on the domain size and on the porous medium properties.

To reproduce realistic situations, we will consider high Ra numbers flows, along with a realistic modeling of the underlying flow physics (including also a non-linear Equation of State for the density, which yields a deformable interface between the two fluids, i.e. brine and CO2-brine), and we will generate a reliable database upon which we will ground detailed analysis of transport phenomena in porous media.