Although high resolution can be provided by electrical logging, the measured electrical log range is narrow and is limited to near the well. Borehole-surface electric potential measurements are able to detect a wide e...Although high resolution can be provided by electrical logging, the measured electrical log range is narrow and is limited to near the well. Borehole-surface electric potential measurements are able to detect a wide enough range but its resolution is limited, particularly for reservoirs with complex oil and water distribution or complicated structure. In this study, we attempt to accurately locate the 3-D reservoir water and oil distribution by combining borehole-surface and crosswell electric potentials. First, the distributions of oil and water in both vertical and horizontal directions are detected by the borehole-surface and erosswell electric potential methods, respectively, and then the measured crosswell potential result is used to calibrate the measured borehole-surface electric potential data to improve vertical resolution so that the residual oil distribution is determined in a lower half-space with three dimensions. The evaluation of residual oil distribution is obtained by investigation of differences between the simulation results of the reservoir with and without water flooding. The finite difference numerical simulation results prove that the spatial residual oil distribution can be effectively determined by combining the crosswell and borehole-surface electric potentials.展开更多
3D inversion of borehole-surface electrical data for complex geo-electrical models is still a challenging problem in geophysical exploration. We have developed a program for 3D inversion to borehole-surface electrical...3D inversion of borehole-surface electrical data for complex geo-electrical models is still a challenging problem in geophysical exploration. We have developed a program for 3D inversion to borehole-surface electrical data based on the quasi-analytical approximation （QA） and re-weighted regularized conjugate gradient method （RRCG） algorithms using Visual Fortran 6.5. Application of the QA approximation to forward modeling and Frechet derivative computations speeds up the calculation dramatically. The trial calculation for synthetic data of theoretical model showed that the program is fast and highly precise.展开更多
文摘Although high resolution can be provided by electrical logging, the measured electrical log range is narrow and is limited to near the well. Borehole-surface electric potential measurements are able to detect a wide enough range but its resolution is limited, particularly for reservoirs with complex oil and water distribution or complicated structure. In this study, we attempt to accurately locate the 3-D reservoir water and oil distribution by combining borehole-surface and crosswell electric potentials. First, the distributions of oil and water in both vertical and horizontal directions are detected by the borehole-surface and erosswell electric potential methods, respectively, and then the measured crosswell potential result is used to calibrate the measured borehole-surface electric potential data to improve vertical resolution so that the residual oil distribution is determined in a lower half-space with three dimensions. The evaluation of residual oil distribution is obtained by investigation of differences between the simulation results of the reservoir with and without water flooding. The finite difference numerical simulation results prove that the spatial residual oil distribution can be effectively determined by combining the crosswell and borehole-surface electric potentials.
文摘3D inversion of borehole-surface electrical data for complex geo-electrical models is still a challenging problem in geophysical exploration. We have developed a program for 3D inversion to borehole-surface electrical data based on the quasi-analytical approximation （QA） and re-weighted regularized conjugate gradient method （RRCG） algorithms using Visual Fortran 6.5. Application of the QA approximation to forward modeling and Frechet derivative computations speeds up the calculation dramatically. The trial calculation for synthetic data of theoretical model showed that the program is fast and highly precise.
