Implications of heterogeneous embankment conditions for geoelectrical investigations on dams: A case study at Mactaquac Dam, Canada

Abstract Electrical resistivity tomography (ERT) has been shown to be effective for surveying and monitoring dams, due to the method's sensitivity to moisture content and relevant physical properties (e.g., porosity). Automated ERT systems, capable of time‐lapse monitoring, can be used to detect variations in ground conditions. However, dam environments are often structurally heterogenous due to, for example, zoned embankments, supporting bedrock or concrete structures and adjoining headponds. If such factors are not accounted for, off‐grid effects may obscure and distort features of interest (e.g., leakage zones) in resulting ERT images. Synthetic modelling simulating conditions at Mactaquac Dam, Canada, was carried out to evaluate whether the abutting concrete structure and properties of the adjacent headpond (water resistivity and level variations) need to be accounted for in an inversion of ERT data. This was achieved through a synthetic numerical model of the dam, including headpond, concrete abutment, core and dry and wet rockfill components. The results show that internal features and dynamic changes through time (e.g., headpond level and resistivity variation) can induce 3D effects in the inversions, which have the potential to be misinterpreted. The modelling revealed that leakage zones could be resolved, showing that features of interest in dam monitoring can still be identified despite potential 3D effects. Overall, these results show that 3D effects from internal structure and a water body are likely to distort modelled resistivity distributions in dam settings. This research sheds light on how ERT can be impacted by structural complexity in dams, using synthetic modelling to understand and quantify the nature of expected artefacts resulting from heterogeneities outside the footprint of the survey area.