Abstract
SUMMARY: Efficient simulation of controlled-source electromagnetic methods in complex geological settings is critical for accurate subsurface imaging and resource evaluation. This study presents an advanced adaptive finite element algorithm leveraging non-conforming arbitrary hexahedral meshes and combining the flexibility of adaptive refinement with the accuracy of high-order basis functions to solve challenges in electromagnetic field modelling. We present a high-order finite element algorithm for electromagnetic geophysics that supports adaptive refinement on arbitrary hexahedral meshes, a capability not commonly available in existing implementations. The proposed algorithm resolves sign conflict issues inherent in non-conforming meshes by employing global vertex indices and adjusting constraint matrices, ensuring the tangential continuity of the electromagnetic field across the computational domain. Adaptive refinement is guided by goal-oriented error estimation, enabling precise solving for the electromagnetic response while maintaining computational efficiency. Numerical examples validate the algorithm’s accuracy and performance. In a steel-cased well model, the computational accuracy of our algorithm is verified by comparison with the 3-D cylindrically symmetric algorithm of the open-source software SimPEG. Furthermore, it is demonstrated that arbitrary hexahedral meshes can achieve accurate modelling of the steel casing with a reduced mesh count. The application to the Vallès Basin model demonstrates substantial computational savings using arbitrary hexahedral meshes compared to tetrahedral meshes. Finally, the algorithm successfully models a large-scale oil-gas reservoir, illustrating its robustness in addressing complex geological scenarios and its potential for monitoring dynamic processes such as water intrusion during resource extraction. These results establish the proposed method as a reliable and efficient forward modelling tool for controlled-source electromagnetic applications, providing significant improvement in flexibility, accuracy and computational efficiency.
Funding Information
  • National Natural Science Foundation of China (42474115)
  • China National Petroleum Corporation (2023ZZ05)
  • Henan Polytechnic University (GCCYJ202407, GCCRC202502, XJBS202512)

This publication has 49 references indexed in Scilit: