Investigation of Lotsane earth fill dam for internal defects using time-lapse electrical resistivity imaging and frequency domain electromagnetic methods

Internal erosion and development of seepage pathways is an ever present concern in earth fill dams, and are a major reason for many dam failures. Internal erosional progresses that are responsible for many dam failures are usually difficult to detect at their onset using the currently in-built monitoring systems that are installed during dam construction. Importantly, increased seepage in dams can only be observed after manifesting at the surface, at which time the integrity of the dam may have already been compromised. Therefore, there is need for improved monitoring in dams using methods that are able to detect defects at an earlier stage. The utility of the electrical resistivity imaging (ERI) and frequency domain electromagnetic (FDEM) methods in characterizing dam embankment conditions has been demonstrated and can be used to elucidate dam defects before they become disasters. Additionally, these methods can be employed on a monitoring basis to study changes that occur in dam embankments across time scales.In this thesis, the ERI was used in conjunction with the FDEM profiling to investigate and locate possible zones of seepage and internal erosion within the Lotsane Dam. Firstly, data enhancement filters were applied to the aeromagnetic data to enhance shallow subsurface anomalies and highlight geological features such as dykes and faults in the study area. A two dimensional (2D) electrical resistivity profile of 1180 m long was established along the dam embankment and five time-lapse measurements were taken at approximately one and a half months interval over a one year period. The commercially available Geotomo Res2dinv software was used to analyze the resistivity data sets and to produce the pseudo sections. The 2D ERI sections were appraised by forward modeling and synthetic data inversion in order to gain confidence and avoid over/under interpretation of resistivity results. On the basis of interpretation of the resistivity data sets, two major zones were identified. The first is an upper conductive layer varying in thickness from 10 to 25 m with a resistivity range of 26-92 Ωm. Underlying this layer is a more resistive medium of 300-6000 Ωm, interpreted to be rocks of the crystalline basement. Additional to these two layers, there are two conductive zones that cut across the resistivity subsurface model and have been interpreted as fractures. The same anomalous zones were observed in the FDEM profiling data. A fault is also identified in both the resistivity and the aeromagnetic data. From this study it wasconcluded that the Lotsane Dam may be at a risk of potential embankment failure due to the identified structural discontinuities, hence further work is needed to monitor the defective zones.