Non-conventional arrays for self-potential surveys

The exponential growth of electrical resistivity tomography (ERT) methods for exploring the subsurface at large depths widened the applicability of the self-potential (SP) method, a passive geoelectrical technique suitable for a variety of purposes like mapping ore bodies or inferring fluid flow in the subsurface. Several new-generation resistivity meters have been designed to continuously log the electric potentials thus allowing for the identification of weak amplitude signals and resulting in deeper inversion models. In such approaches, long SP time-series are collected but are totally ignored as only marginal intervals are retained and analysed in the ERT procedure. The discarded SP records could be valuable although not collected using the traditional methodology, based on a reference electrode. We present an SP forward modelling feasibility study of different array techniques, based on numerical finite-element methods. The SP has been modelled in a variety of electrical settings to assess the imaging potentials of non-conventional (i.e. sparse gradient and full sparse gradient) arrays in comparison to traditional (i.e. fixed-base and the leapfrog) arrays. The analytic signal amplitude (ASA) algorithm was employed to compare numerical modelling results obtained from the different type of arrays, highlighting the great potentials of non-conventional arrays for the recognition of several sources of SP anomalies. The ASA maps, presenting a single peak centred over the targets, can significantly help in identifying the source anomalies for all the analysed array techniques. The cost-effectiveness along with the imaging capability of these non-conventional arrays constitute important benefits that could be exploited resulting in a systematic inclusion of SP analysis when collecting deep ERT data using distributed systems.