Pollutant monitoring in a sand-column with Spectral Induced Polarization, Time Domain Reflectometry, and geochemistry analysis: experimentation and modeling

Geoelectrical methods have been proposed as non-invasive techniques to monitor the soil contamination. It is now well known that electrical properties of porous media depend on the composition of the electrolyte and its interaction with the surface of the minerals. So, understanding in the laboratory the influence of change in electrolyte composition on geolectrical measurements is critical to validate behaviour models. Following this path, we have been using two geolectrical methods, Time Domain Reflectometry and Spectral Induced Polarization measurements. We carried out column experiments to investigate the effect of mineralogy and electrolyte composition on the electrical signatures resulting from pollutant migration (CuSO4 and ZnSO4) in saturated flow conditions, and adsorption/desorption reaction between Cu, Zn and the surface of the minerals. Laboratory columns are filled with a 100 µm well sorted sand (nearly pure silica). These columns are saturated with different well known electrolytes (composition of the electrolyte is progressively modified during the course of the experiment). The evolution of the chemical composition in columns has been monitored to have a well understanding of transport and reaction processes. Spectral Induced Polarization measurements were performed in the frequency range 45 mHz - 12 kHz. Low-frequencies polarization characteristics in a such environment depend on two processes. There is the polarization of the mineral/electrolyte interface coating the surface of the grains (polarization of the Stern layer) and the Maxwell-Wagner polarization associated with accumulation of the electrical charges in the pore space of the medium. The model of low-frequency conductivity is connected to a double-layer model of electrochemical processes occurring at the surface of silica. Characteristics of this double-layer model (i.e., distribution of counterions between the diffuse layer and Stern layer) are computed with the code Phreeqc (USGS). Experimentation and modeling allows us to connect characteristics of the electrolyte and geoelectric responses.