The Underlying Reactive Transport of Branching
Morphogenesis
… and what it means for chalk dissolution
Abstract
The increase in the atmospheric level of CO2 is believed to be a major contributor to global climate change. Geologic carbon storage (GCS) is a promising means to reduce the emission of anthropogenic CO2.
In Denmark, the feasibility of sequestering CO2 in chalk formations remains a topic of heated debate. To advance the discussion, a key issue is to address the fate and transport of injected CO2 in natural porous structures.
CO2 is acidic and significantly changes the interactions between formation fluid and minerals. Rock dissolution and reprecipitation modify the macroscopic flow field over time.
The instability of the migrating reactive plume routinely amplifies microscopic physical and chemical heterogeneities, making the prediction of flow structure evolution very challenging.
We are urged to re-examine the homogeneity assumption stemming from the use of representative elementary volumes (REV), and to better understand how small differences in petrophysical properties are enhanced by flow-reaction coupling and ultimately define a macroscopic flow field.
I address two scientific issues in this talk. First, is the microscopic dissolution of chalk homogeneous? Second, how do we account for the evolution of geometric boundary in flow simulations?
I show that, using in situ X-ray computerised tomography (CT), one can record microstructure dissolution with sub-m resolution in 4D. I also present a reactor
network model that fully utilizes the geometric information in greyscale CT data and quantifies the flow structure evolution stemming from fluid-mineral interactions.
Combining both techniques, we found that: i) the presence of CO2 does affect the homogeneity of chalk microstructure dissolution. However, this effect is attributable to the relative size of reactive subvolume vs. the domain of interest, and is therefore not fundamental. ii) Cumulative surface (CS), a convolution between geometric surface and fluid residence time, defines the decay of fluid reactivity along streamlines and thus the size of reactive subvolume. iii)
The positive feedback between flow and dissolution indicates that pore develops along streamlines with the smallest CS. I also show an example in which the dissolution front migrates against flow direction. This counter-intuitive phenomenon was first predicted by the model and later verified using synchrotron radiation based X-ray CT.
The results have two important implications. First, the development of pore structures in GCS is inherently a dynamic process. Therefore, using static parameters – such as the Damkhöler and Péclet numbers, or the petrophysical properties based on REV – to predict system evolution might not be the optimal strategy. Second, CS is a good indicator for path election during pore structure development.
A better understanding of CS distribution in natural porous media will greatly benefit the prediction of flow structure evolution. We expect the synergy between in situ X-ray imaging and reactor network model to play a pivotal role in understanding nonlinear underground processes in the near future.