Comparing fluid flow mechanisms in reservoir rocks to improve subsurface pollutant remediation

In order to reach carbon neutrality, a drastic increase in subsurface carbon storage is essential. During the storge site selection process, sites are assessed based on overall security for long-term storage, and any remaining risks require a risk assessment with leakage mitigation plan. With an increase in large scale carbon storage, there is an increased chance of leakage and the potential for contamination of drinking water aquifers above the storage reservoir. In such a scenario, subsurface sweeping via the pumped injection of large volumes of water would be the most suitable method for remediation, yet this comes with issues of tailing and rebound occurring after the treatment has stopped, necessitating the injection of large volumes of water, over long time-scales, which significantly increases both the cost and carbon footprint of the remediation operation. With this study we look at how modified pumping techniques may improve sweeping to minimise rebound, but also how to minimise use of water.

The pumping techniques used in this study were continuous flow and pulsed flow. The pulsed flow was divided into two types, cyclic flow and rapid pulses. Cyclic flow has long breaks and relies on diffusion, whereas rapid pulses rely on vortices created in the pore spaces to increase mixing. The rock types used for the experiments were Clashach sandstone which represented a homogeneous rock, and Wattscliffe Lilac sandstone, representing a heterogeneous rock. The Clashach sandstone is quartz dominated and has a porosity of ca. 24 % with an overall uniform grain size and pore size, as well as well-connected pores. The Wattscliffe Lilac sandstone has a more heterogeneous mineralogy with a porosity of ca. 21 % and varied grain size, pore sizes, and pore connectivity. The different rock types were chosen based on their distinctly different pore structure to aid in understanding how heterogeneity impacts the different flow mechanisms. The cores were prepared at 4 cm long and just under 5 mm wide, covered in heat shrink tubing then cast in resin in order to prevent bypass flow, and optimised for pore-scale XCT imaging at injection pressures of up to 2 MPa. To study the recovery behaviour of the different flow types through the rock cores, fluorescein was used as a tracer and measured in a flow-through fluorometer. The resulting breakthrough curve, tailing, and total recovery was then used to determine which pumped flow method was the most efficient in terms of: 1) remediation of fluorescein, 2) water usage during remediation, and 3) speed of remediation. These different factors of efficiency can be crucial in determining which pumping method is used to remediate a contamination site.