A numerical electromagnetic study of shallow geophysical targets

Abstract

Prediction of the response of high-frequency induction survey tools to 3-dimensional targets is needed to aid in tool and survey design, in the interpretation of data, and to analyze the interaction of the individual field components with the target of interest. To this end, two numerical algorithms (TSAR and NEC) were imported and adapted to solve geophysical electromagnetic problems. A third algorithm (EM1DSH) was used to quantitatively analyze the role of current channeling on the response of shallow targets, and to verify that the TSAR and NEC algorithms include the important effect of current channeling in their solution. TSAR (a finite difference time-domain algorithm) proved successful in modeling the ellipticity response of a vertical magnetic dipole placed over a homogenous and layered lossy dielectric as compared to published data in the 500 kHz to 30 MHz range. Cell-size versus accuracy analyses show that little accuracy gains are made with a reduction of cell-size past the one-tenth effective wavelength modeling guideline. NEC (a method-of-moments algorithm) shows substantial but limited success in modeling the response of small loop antennas to perfectly and near-perfectly conducting geophysical targets (conductivity and permeability) in the 6.4 kHz to 8 MHz range. Comparison of NEC results are made with analytic results, fields data, and other numerical algorithms. NEC shows substantial numerical error at lower frequencies due to the effective lengths (in wavelengths) of the wire segments used. Also, the Green's function look-up table used to interpolate the effect of half-space on target response is not optimized for the geophysical problem which can lead to substantial solution error at lower (kHz) frequencies. An integral equation solution (EM1DSH) analysis shows that the quantitative effect of increasing background conductivity (which affects both current channeling and target response) on the secondary field response of a buried thin-sheet can be greater than 120 percent in the geophysical induction range. Target parameter changes show current channeling to be greatest for targets that are shallow, that are horizontal, and have a large dimensional aspect ratio. Target and survey parameter sensitivity analyses help to understand the relationship of these parameters to current channeling.