Combining electrical resistivity, self-potential and induced polarization to image ground water flow in 3D: Theory and applications

We demonstrate that electrical resistivity imaging can be combined with self-potential and induced polarization to produce 3D images of the Darcy-velocity field in the subsurface of the Earth. We first review the basic concept behind this new appraoch and then, we apply it to two case studies associated with leakage in dams and enbankments. The dam of Lampy (Black Mountain, Aude, France) is considered as one of the oldest dams in France. A geophysical survey is performed to better understand the pattern of groundwater flow downstream of this dam in the granitic substratum. Induced polarization is first used to image both electrical conductivity and normalized chargeability. 8 core samples of granite from this site are measured and analyzed in the laboratory. Their electrical conductivity and normalized chargeability are expressed as a function of the porosity and Cation Exchange Capacity (CEC). The field data and the petrophysical results are used to image the water content, the CEC, and the permeability distribution of the substratum. Then, self-potential is used as a complementary passive geophysical technique, which, in absence of metallic bodies, is directly sensitive to groundwater flow through the so-called streaming potential effect. Indeed, the excess of electrical charges in the vicinity of the solid grains, in the so-called double layer, is dragged by the ground water flow generating in turn an electrical (streaming) current and therefore an electrical field. A map of the resulting self-potential signals is done over the area covered by the induced polarization profiles. This map shows a large positive anomaly with an amplitude of ~80 mV possibly associated with upwelling groundwater in an area where the soil is water-saturated. A groundwater flow simulation is performed to model this anomaly. This is done in two steps. A preliminary groundwater flow model is built using the permeability and water content distributions obtained from the induced polarization data. Then, this groundwater flow model is updated using the information contained in the self-potential data including the electrical conductivity distribution obtained through resistivity tomography. The algorithm for the inversion of the self-potential data is validated through a 2D numerical test. This analysis yields a groundwater flow model with the flow being focused through a high permeability zone. A similar appraoch is then apply to a leakage through a small dam in Easter France. We also provide a synthetic case study to demonstrate the value of our approach. This study shows how three geoelectrical methods (self-potential, induced polarization and electrical resistivity) can be efficiently combined to image groundwater flow in the vicinity of a dam or an embankment. We are currebtly working on using this approach on landslides.