The Study On The Simulation Of The Borehole Acoustic Reflection Imaging Logging Tool And It's Waveform Processing Method

The borehole acoustic reflection imaging logging is a newly developed acoustic logging method that detects the waves reflected from the objects within 10 meters nearby borehole. Geometrical acoustic methods are so far the most commonly used methods to analysis such a wave field because they are simple to use and easy to be understood. However these methods are not able to reveal the most important characteristics related to the wave energy and frequency. To gain an insight into the full wave field and fully understand the tool responses, the 3D finite difference (3DFD) method for the simulation of the borehole acoustic reflection imaging logging has been studied and implemented for a number of cases, including an orthogonal anisotropic formation model. Considering the characteristics of the wave fields under study, i.e., higher frequencies (up to 10 kHz), longer travel distance (up to 10m) and wide dynamic range (60db), we have applied PML absorbing boundary, parallelization and non-uniform staggered grids algorithm to improve the numerical simulations. All these techniques have been implemented for different cases and these results have been tested and compared with the results by conventional algorithm. These comparisons show that the simulation method developed in this thesis can simulate the borehole acoustic reflection imaging logging correctly and much more efficiently. From these wave field simulations, it has been discovered that the amplitude of the P-P reflected wave decreases when the source frequency increases; different investigation depth would be achieved at the different offset if the dip of a reflective interface is varying; the major wave energy is distributed within a narrow offset range; the dipole source-receiver may be used to identify the azimuth of the reflective interface because of the directional wave energy radiation; the maximum amplitude would be received when the dipole source is perpendicular to the reflective interface. Due to the reflected waves are mixed with the borehole guided waves and have relatively smaller amplitudes, the signal processing techniques for effectively extracting reflected waves are also a very important issue for the new logging method. Two new methods, namely, the multi-scale semblance method and the modified ridgelet transform., are introduced in the processing of the reflected waves. The multi-scale semblance method is the combination of the dual tree complex wavelets transform and the slowness-time semblance method. By such a combination, we can get STC information at different scales. And this information can be used to eliminate the interference noise with a scale. When applied to the simulated and the field array signals, this method can appropriately separate the reflected waves while suppress all noises. The modified ridgelet transform (MRT) is a ridgelet transform incorporated with the high resolution radon transform. By such a modification, this transform is more apt to process array signals. In this thesis, the MRT is employed to separate the reflected waves directly from the common midpoint gathers and the common shot gathers. The application of MRT to the simulated signals demonstrates the advantages of direct identification of the reflected waves in the array acoustic wave fields.