Method of skin effect correction and data quality verification for a multi-frequency induction well logging instrument

Abstract

A method of correcting receiver signals in an induction well logging instrument for skin effect. The method includes determining a magnitude of signals induced in an induction receiver at each one of a plurality of different frequencies, determining a relationship of the magnitudes with respect to frequency, and determining a skin effect corrected conductivity by determining a value of the relationship which would obtain when the frequency is equal to zero. In a preferred embodiment, the step of determining the relationship includes calculating a best fit curve of the magnitudes with respect to the frequency, calculating the first derivative of the best fit curve with respect to the frequency, and calculating a correction for the signals at a selected frequency according to the relationship of the first derivative with respect to the frequency. The correction is applied to the signal at the selected frequency to calculate a skin effect corrected signal. In a specific embodiment, the step of curve fitting includes generating a weighting factor for each frequency based on the noise variance of each frequency. The specific embodiment includes constraining the best fit curve so that the first derivative of the best fit curve is generally less than zero and the second derivative of the best fit curve is generally more than zero.

Dependencies are inferred from the claim text.