| 摘要 |
Fracture detection in logging signals comprises use of a calculated absolute mean gradient of a wavelet transform characteristic value to define an analysis window. Detecting fractures in logging signals concerning an ambient zone and comprising logs of different types recorded for the zone as a function of depth, comprises: (a) selecting a portion in each log such that the selected portions have a common depth range, one of the portions being taken as a reference portion; (b) determining an spatial analysis frequency sequence; (c) using a selected parent wavelet function to construct a family of analysis wavelet functions dependent on the spatial frequency and the depth; (d) calculating a wavelet transform of the selected portion of each log for each spatial analysis frequency; (e) using a chosen characteristic value of the wavelet transform as a representation of the wavelet transform; (f) calculating, for each portion of the selected log and for each depth, the absolute value of the mean gradient of the wavelet transform characteristic value for different spatial analysis frequencies; (g) selecting, for each portion of the processed log, the peaks of the absolute value of the absolute mean gradient value, each peak corresponding to a fracture; (h) determining corresponding fractures on the reference log portion; (i) defining an analysis window centered on each fracture of the reference log portion; and (j) selecting the fractures of the other log portions which are located within the analysis window. An Independent claim is also included for a process for depth correlation of logs of different types recorded for the zone as a function of depth, the process comprising carrying out the above method and recalculating the values of the other log portions such that their fractures, associated with each of the fractures of the reference log portion, appear at the same depth as the reference log portion fracture. Preferred Features: The result of wavelet transformation is a complex number and the characteristic value of the wavelet transform is the real part of this complex number. The log providing the reference portion is a gamma radiography log. The spatial analysis frequencies, selected for calculating the wavelet transform, have frequency limits corresponding to wavelengths of 4 and 200 (especially 10 and 100) m. The spatial analysis frequency sequence is a geometric progression. The parent wavelet function is a function of the type f(z) = (1 - z<2>) x exp(-z<2>/2), where z = depth.
|
