Estimation of the electrical potential distribution along metallic casing from surface self-potential profile

Abstract

Corroding casings of wells generate negative self-potential (SP) anomalies, increasing from about -10 to -500 mV in the vicinity of the well to 0 mV at large distances. As reported in previous laboratory experiment, SP can be used to retrieve the distribution of electrical potential along the casing, which is somehow a proxy for the corrosion state of the casing. These studies used 3D (whole space) or surface 2D (whole surface) measurements of SP distribution; here we reported a field example, for which only a 1D surface SP profile is available. In order to retrieve the most probable associated potential distribution (defined by a spline) along the 11.1-m long metallic casing, we develop a direct model based on geometrical and geoelectrical properties of the medium, which was then used in a (non-deterministic) optimization procedure by simulated annealing, including some physical constrains. Tests carried out on a synthetic case allowed the initial source to be correctly retrieved, provided that the number of nodes used for the spline defining the potential distribution along the casing is large enough. The inversion of real field data provided a dipolar anomaly, with minimal negative amplitude of around -600 mV at 5 m, and maximal positive amplitude of about 1100 mV at 9 m (close to the level of the water table), this shape being in agreement with the results of previous laboratory studies.

Spectral Induced Polarization Characterization of Non‐Consolidated Clays for Varying Salinities—An Experimental Study

“…where, A mes represents the measured amplitude vector, 〈 A mes 〉 represents the mean of the measured amplitude vector, A mod , the modeled or calculated amplitude vector, via the double‐Pelton model, N a is the number of amplitude data points that have been preserved, P mes is the measured phase vector, 〈 P mes 〉 is the mean of the measured phase vector, P mod is the modeled or calculated phase vector, and N p is the number of phase data points that have been kept. The strategy we used was to first optimize with a simulated annealing approach, that has been explained in detail in Maineult (2016). For the parameters m 1 , m 2 , c 1 , and c 2 , we let them vary between [0–1], for ρ 0 we usually use , for τ 1 we usually use [10 −3 − 10 6 ]s, and finally for τ 2 we use [10 −10 − 10 1 ]s. Here, is the arithmetic mean electrical resistivity for all frequencies.…”

Saturation Dependence of the Streaming Potential Coefficient

“…Self-potentials rely on electrokinetic coupling and proved to be efficient for characterizing groundwater resources and mapping contamination plumes, since those signals are directly sensitive to hydrological fluxes [Fournier, 1989;Titov et al, 2002;Pinettes et al, 2002;Maineult et al, 2004;Jouniaux et al, 2009;Desroches and Butler, 2016] and solute transport [Maineult et al, 2005;Kabir et al, 2015;Giampaolo et al, 2016]. Self-potentials appeared especially useful to estimate fluid flow [Jouniaux et al, 1999;Sailhac et al, 2004;Rangarajan et al, 2014], to detect propagating vapor-front during oil-extraction processes [e.g., Saunders et al, 2008], to detect hydrothermal flows [Mauri et al, 2010;Brothelande et al, 2014;Gonzales et al, 2014], to map advected reacting redox fronts [Maineult et al, 2006a,b], and to provide a tool to evaluate the evolution of the corrosion of metallic casings [Maineult, 2016]. Self-potentials are also often associated with natural hydrothermal circulation [Ishido and Pritchett, 1999;Matsushima et al, 2000;Jouniaux and Ishido, 2012;Monetti et al, 2014].…”

Characterizing Stream‐Aquifer Exchanges with Self‐Potential Measurements

“…Metallic well casings can also induce very large redox potentials (e.g. Maineult ). Nevertheless, in the context of high redox activity in the aquifer, Arora et al () noticed a good fit between observed SP and the modeled electrokinetic contribution.…”