3D Electrical Resistivity Tomography of Karstified Formations Using Cross-Line Measurements

Limestone formations constitute a serious geological challenge for large building projects such as wind turbine farms. Indeed, limestone formations might be subject to karstic phenomena such as sinkholes or subsidence constituting a serious hazard for existing constructions and future civil engineering projects. In calcareous geological settings, a characterization of the subsurface to detect wealthy rocks is therefore mandatory. The classical approach consists in drilling boreholes and cone penetration tests to identify subsurface properties. However, they only provide punctual information whereas karstic environments show sharp variations with complex 3D geometry, making interpolation between boreholes relatively inefficient. In this context geophysical methods can provide spatially distributed information at a limited cost. In particular, surface electrical resistivity tomography (ERT) aims at mapping the distribution of electrical resistivity in the subsurface in a passive way using only surface sensors. The method is based on the measurement of electrical potential resulting from the injection of DC electrical current. In karstic contexts, weathered rocks generally show an increased porosity and water content compared to healthy limestones, leading to strong contrasts in their electrical resistivity. ERT is therefore particularly sound to investigate karstic phenomena. Most standard ERT applications use 2D profiles that are quick to acquire and interpret. However, in complex 3D geometries, the acquisition of 2D profiles is not sufficient to image correctly subsurface structures. 3D data sets require more efforts both for acquisition and interpretation, making their use more costly and therefore less common in practice. In this contribution, we propose an innovative 3D ERT acquisition procedure to reduce the field efforts and duration of the 3D acquisition procedure. The method is based on standard 2D parallel lines, but, in contrast with previous methodologies, we also acquire cross-line measurements in several directions to increase the ability of ERT to image 3D structures. To ensure a fast acquisition of large area, we limit the cross-line measurements to pre-defined line spacing and implement a roll-along technique moving previously acquired 2D lines in the perpendicular direction. The data can then be acquired with a standard 64 electrodes equipment. We first demonstrate the increased imaging capacities of our technique compared to standard acquisition methods with a numerical benchmark. Then, we validate it through a field application to detect the 3D geometry of karstic features and unaltered limestone formations. We analyze the minimum amount of cross-line measurements required for a proper imaging of the 3D structures. The proposed 3D survey induces extra costs of about 50% compared to a traditional 2D survey, but this extra cost is compensated by a largely better imaging of the subsurface. This cost will be reduced in the future by optimization of the survey to reduce acquisition time.