Electromagnetic Logging – High–performance Computing and Effective Interpretation

The work is intended for development of a software-algorithmic base for express-simulation and inversion of high-frequency electromagnetic logging data. The effective algorithm for simulation using graphics processors of personal computers (video cards) has been developed. Comparative assessments of processing power of logs for 2D geoelectrical models of water saturated reservoirs for the central processing unit (CPU) and graphics processing unit (GPU) have been obtained. We suggest to approximate the radial and depth-dependent conductivity of formations by continuous functions and to design an algorithm for 2D inversion of induction logging data using this approach. Comparison of synthetic logs in models with piecewise constant and continuous approximations of conductivity shows that the continuous-function approximation is accurate within 5 %, and the accuracy improves as the average conductivity of the layered formation is taken into account