Redox potential distribution inferred from self-potential measurements associated with the corrosion of a burden metallic body
J. Castermant, C.A. Mendonça, A. Revil, F. Trolard, G. Bourrié and N. Linde
Journal name: Geophysical Prospecting
Issue: Vol 56, No 2, March 2008 pp. 269 - 282
Info: Article, PDF ( 547.97Kb )
Negative self-potential anomalies can be generated at the ground surface by ore bodies and ground water contaminated with organic compounds. These anomalies are connected to the distribution of the redox potential of the ground water. To study the relationship between redox and self-potential anomalies, a controlled sandbox experiment was performed. We used a metallic iron bar inserted in the left-hand side of a thin Plexiglas sandbox filled with a calibrated sand infiltrated by an electrolyte. The self-potential signals were measured at the surface of the tank (at different time lapses) using a pair of non-polarizing electrodes. The self-potential, the redox potential, and the pH were also measured inside the tank on a regular grid at the end of the experiment. The self-potential distribution sampled after six weeks presents a strong negative anomaly in the vicinity of the top part of the iron bar with a peak amplitude of −82 mV. The resulting distributions of the pH, redox, and self-potentials were interpreted in terms of a geobattery model combined with a description of the electrochemical mechanisms and reactions occurring at the surface of the iron bar. The corrosion of iron yields the formation of a resistive crust of fougerite at the surface of the bar. The corrosion modifies both the pH and the redox potential in the vicinity of the iron bar. The distribution of the self-potential is solved with Poisson's equation with a source term given by the divergence of a source current density at the surface of the bar. In turn, this current density is related to the distribution of the redox potential and electrical resistivity in the vicinity of the iron bar. A least-squares inversion method of the self-potential data, using a 2D finite difference simulation of the forward problem, was developed to retrieve the distribution of the redox potential.