Detection of Boulder Using GEM-2 EM System Based on Three-Dimensional Inversion Algorithm

Rong Liu,Xiaowu Shen,Chunfei Chen,Jianxin Liu,Jianping Xiao,Mengyu Sun,Rongwen Guo
DOI: https://doi.org/10.1016/j.jappgeo.2021.104516
IF: 1.845
2022-01-01
Journal of Applied Geophysics
Abstract:The granite body, or referred to as boulder, that shows high resistivity in the near-surface is a common geohazard in pipeline construction by trenchless technology. The GEM-2 system using small coils as the transmitter and receiver (Tx-Rx) has been widely used to delineate the low resistivity target with 1D results in the near-surface due to its efficiency and non-invasive characteristics. However, the 3D inversion might be preferred since the boulder is highly resistive and has a three-dimensional (3D) structure. To study the effectiveness of boulder detection using GEM-2 system and 3D inversion, we use the footprint-guided compact finite element method (CFEM) for 3D numerical modeling, and Gauss-Newton optimization is applied to delineate the 3D structure of the boulder. Because of the recorded response is dominated by the induced current within a limited volume directly below the Tx, the modeling volume of each Tx-Rx pair is defined as subareas of the entire survey area, and the sensitivity is explicitly calculated along with the modeling process. We setup an experiment with granite boulder scenario and conduct a GEM-2 test survey to show the performance of our 3D inversion method in practice. The inversion results illustrate that our inversion algorithm has a higher horizontal resolution than the previous 1D inversion, and the three-dimensional contour of the boulder is indicated by relatively high resistivity.
What problem does this paper attempt to address?