Identifying Fracture by Time-Frequency Analysis of Well Logs

摘要

It is important to precisely identify the fracture in the fractured reservoirs for exploiting gas and oil. The common method, identifying the fracture by the size and difference of dual laterolog, is affected by many factors, so the resolution of the fracture identification is not high. To solve these problems, this paper studies the principles of time-frequency analysis and Short Time Fourier transform, and then discusses how to extract the implicit information of the fracture in the dual laterolog by using the time-frequency analysis. Through theoretical modeling and practical applications, the results show that using Hamming window function and selecting 27 samples of window function we can calculate the amplitude and spectrum of the resistivity curves, which can greatly improve the ability of distinguishing fracture.