Frequency-Dependent Effective Electrical Properties of Suboptimal Rock Samples through Digital Rock Physics

Geothermal exploration drilling plays a crucial role in advancing the energy transition. To make the prospecting process more economical and time-efficient, the EU-funded GeoHEAT project has set the goal of developing methods that allow for obtaining the maximum amount of information about the subsurface with as little expense as possible. The project includes the development of a ground-penetrating radar (GPR) that can be used during borehole logging at elevated ambient temperatures. For the accurate interpretation of the GPR data, as well as for estimating the porosity and water content of the rocks surrounding the borehole, properties such as permittivity and electrical conductivity of these rocks are required. Here, we want to present the current progress of our research, which aims to determine the aforementioned characteristic properties using digital rock physics (DRP). Moving away from standardized cylindrical samples and using irregularly shaped by-products of the drilling process, known as drill cuttings, we can provide a more comprehensive understanding of the subsurface, thereby improving the characterization of potential geothermal reservoirs.

Core sampling during exploration drilling is costly and time-consuming, as it interrupts the operation. In addition, cores are often taken from only a few sections in order to keep the added costs low. However, these samples are necessary for laboratory testing, as sufficiently large and smooth contact surfaces must be available to ensure that the respective measurement devices deliver accurate results. No such requirements exist in DRP, as simulations are performed at the pore scale and therefore very small samples without flat surfaces, such as irregular drill cuttings, can be used.

The DRP workflow consists of three main steps. First, high-resolution computed tomography scans are taken of a small sample. These are then processed into a digital twin using segmentation, where the individual phases, such as minerals or pores, are distinguished from one another so that specific properties can later be assigned to them in this location-dependent volume. In combination with our finite volume method code, which solves a stationary potential equation, this digital model can be used to simulate the desired effective properties.

In previous studies, an early implementation of our code demonstrated reliable results for frequencies greater than 1 MHz. By implementing preconditioners, we now simulate lower frequencies with highly accurate results where before the increase in polarization led to code instabilities. Additionally, we fully validated the code on comparative data, such as analytical solutions and laboratory measurements of a high-porosity sandstone and a low-porosity granite. In the future, we will investigate how changes during the creation and transport of drill cuttings influence the accuracy of the results, thereby contributing further to a more efficient approach to geothermal exploration.