2.5-D Controlled-source EM Modeling by Finite Element Method and Arbitrary Sources

Modeling controlled-source EM data using the finite-element method is preferred because of its ability to model complex structures and to accommodate both small and large structures in the same grid. Most of current 2.5-D FE codes are designed for point dipole sources oriented parallel or perpendicular to the strike direction. Additionally, most codes use a primary / secondary separation field approach to simplify and speed-up calculations. We have developed a 2.5-D controlled-source EM algorithm that uses the finite element method, models the sources directly (no primary/secondary approach) and applies an alternative scheme for the transformation of the EM fields from the strike-parallel wavenumber domain into the spatial domain. This approximation does not assume any parity of the transformed EM fields and can be used for any orientation of the sources. The accuracy of this scheme is shown against a semi-analytical solution for a 1D canonical reservoir model. A second test uses a 2D model, in which a small conductive block is modeled below the seafloor. The anomaly distorts the fields at all frequencies studied, demonstrating the utility of the direct source modeling, in contrast to the usual primary/secondary field separation technique, which cannot handle this type of structures.