H54C-07
3D Observations of Dispersion, Mixing and Reaction in Heterogeneous Rocks

Friday, 18 December 2015: 17:30
3016 (Moscone West)
Maartje Boon1, Branko Bijeljic1 and Samuel C Krevor2, (1)Imperial College London, London, SW7, United Kingdom, (2)Imperial College London, Department of Earth Science & Engineering, London, SW7, United Kingdom
Abstract:
Rock structure heterogeneity can have a significant effect on dispersion, mixing and reaction of aqueous components in porous media. To observe the effect of pore structure heterogeneity on reactive transport, core flooding experiments were carried out for a sandstone and two carbonate rocks of different heterogeneity for eight different Peclet numbers ranging from 0.5 to 100. The rock cores were 20cm long and had a diameter of 7.62cm. A device consisting out of three annular regions was used for injection (fig. 1A). An analytical solution to the flow and transport equations for this new inlet configuration was derived to design the experiments (fig. 1B).The dispersion, mixing and reaction of chemical components were visualised in 3D with the use of chemical dopants and a medical CT scanner (fig. 1C). Local transverse dispersion coefficients (Dt) were calculated from the change in variance of the transverse distance travelled by the chemical dopant along the core. The change in variance along the core showed a characteristic pattern for each rock that was independent of spatial location. Heterogeneity was characterized by the spread in local transverse dispersion coefficients. For Peclet number 2, for the homogenous rock the local transverse dispersion coefficients ranged from 4.1x10-4 cm2 min-1 to 5.9x10-4 cm2 min-1 and for the most heterogeneous rock from 2.5x10-3 cm2 min-1 to 7.2x10-3 cm2 min-1. For the reactive transport experiments, an ICP-MS was used to measure the effluent. The core flooding experiments were modelled using both, the CrunchFlow and ToughReact reactive transport codes. High quality data sets of the space and time evolution of the concentration in non-reactive and reactive core-flooding experiments like these can be used as future benchmark test for numerical models. Furthermore, these observations can be used in the development of upscaling techniques for accurate and efficient modelling of chemical processes during flow in porous media.