Experimental and Numerical Modelling of Self-Potential Response to Saltwater Intrusion

Abstract

<p>Saltwater intrusion (SWI) in coastal aquifers poses a significant hazard to freshwater security for many of the world’s population centers. SWI is challenging to monitor and model due to the physical complexity of real aquifers. Self-Potential (SP) has been an important method for monitoring the subsurface for many years. Previous studies have suggested that borehole measurements of SP could be used to identify saline interface movement and provide advance warning of imminent saline breakthrough at an abstraction borehole. SP produced during SWI comprises the combined effects of electro-kinetic potential, arising from transport of excess charge in response to water potential (head) gradients, and exclusion-diffusion potential, arising from transport of excess charge in response to ion (salt) concentration gradients. SP can have advantages over other geophysical methods, such as electrical resistivity tomography and borehole fluid electrical conductivity measurements, because the effect of  moving saltwater fronts can be determined using a relatively small number of localized probes.</p><p>We quantitatively investigate the relationship between SP and SWI using experimental and numerical modelling with the aim of reproducing experimentally measured SP response via simulation. Building on well-established methods, a novel laboratory setup has been developed to optically monitor SWI in a thin homogenous aquifer while simultaneously recording SP data at multiple probe points. A Matlab solver is used to calculate SP data from simulated hydrodynamic SWI data computed by the fixed-grid finite element software SUTRA. Similarly, finite element SWI simulations using adaptive meshing are carried out using the IC-FERST software, which directly computes hydrodynamic and SP solutions. We compare these numerical results with experimental data and show similarity in SP signal trends as functions of brine movement near probe locations. We conclude with a discussion of the merits of SP modelling and its suitability for interpreting SP signals for monitoring and characterization of saltwater intrusion in coastal aquifers.</p>