Evaluation of 2D Electrical Resistivity Tomography to detect leakages within highly 3D structures (earth dykes)

a numerical approach based on realistic mechanical models

Titre : Evaluation of 2D Electrical Resistivity Tomography to detect leakages within highly 3D structures (earth dykes) : a numerical approach based on realistic mechanical models
Laboratoire de rattachement : ISTerre

Encadrant : Grégory Bièvre (gregory.bievre univ-grenoble-alpes.fr)
Téléphone : +33 (0)4 76 63 51 73

Mots clés : 2D electrical resistivity tomography, earth dyke, leakage, numerical modeling

Contexte et objectifs de la mission de stage :
Earth dykes are complex 3D structures which are prone to internal erosion, piping and leakages. ERT has been widely used in the past decades to try to detect leakages within such structures, mainly in 2D along the dyke crest for cost-effective purposes. Results show that it is mainly a time-lapse approach which allows to detect the physico-mechanical evolution within the ground where leakages occur. However, many questions stay unanswered regarding the sensitivity of such a method in a strongly 3D context. At which stage of development can the piping be detected by ERT ? Does the piping have to be located below the electrode line to be detected ?
Modern simulation codes allow to compute easily the electrical response of the ground to the development of a piping phenomena. However, realistic mechanical models, encompassing the development of a realistic internal erosion phenomena, have not been used so far to test the sensitivity of the method (electrode spacing, protocol, etc.). Noticeably, the progressive development of the piping across the dyke still (i.e. from one side to the other) needs to be taken into account to evaluate the sensitivity of ERT. Furthermore, ERT coverage recorded at surface varies laterally across the dyke as a function of the electrode configuration (i.e. varying conditions provide varying lighting of the target ; Bièvre et al., in prep).
The aim of this project is to set up numerical experiments based on realistic mechanical and geophysical 3D models. Simulated data will be inverted in 2D using the BERT algorithm (resistivity.net) developed by Günther et al (2006). These experiments will simulate an actual piping development within an earth dyke, from its initiation to failure. The main objective is to test the hability of the method to detect piping with constrained mechanical conditions, and to evaluate the experimental conditions necessary to succeed in this operation.
Simulations will be conducted using the F3DM code developed at IRSTEA in France (Clément et al. 2016) which allows to compute multichannel and multielectrodes acquisitions via the Comsol Multiphysics software. Visualization of the (and mathematical operations on the) results will have to be developed using the free Paraview software (paraview.org).

References
Bièvre G., Oxarango L., Günther T., Goutaland D. & Massardi M. (in prep.) Geophysical assessment of an earth-filled dyke with a permanent hydraulic head. 4D effects on 2D Electrical Resistivity Tomography monitoring. To be submitted to Journal of Applied Geophysics.
Clement R. & Moreau S. (2016) How should an electrical resistivity tomography laboratory test cell be designed ? Numerical investigation of error on electrical resistivity measurement. Journal of Applied Geophysics, 127, 45 – 55.
Günther T., Rücker C. & Spitzer K. (2006) Three-dimensional modelling and inversion of DC resistivity data incorporating topography - II. Inversion. Geophysical Journal International, 166, 506-517.

Prérequis : physics of ERT, measurements, inversion