Study Of Theory And Application Of Array Sonic Logging

Sonic logging is important for exploration and development of oil industry, especially the mechanical parameters of rock in great depth provided by it, can be used in oil recovery and safety drilling. Based on the rock physics, the time lapsed seismic can be served for the determination of residual oil’s distribution in the final stage of development. And the velocity provided by array sonic logging can guide the time lapsed seismic processing and interpretation. The anisotropy detected by cross-dipole sonic logging can be used to detect fracture, to evaluate fracturing effect and to determine in situ stress. It can be used to guide the seismic data inversion too. In one word, the study of array sonic logging is very necessary and important in science and application. The study also can be used to design and optimize new tools.The dissertation emphasizes on the theory and the numerical modeling of the array sonic logging. The data processing or the parameters inversion of the array sonic data is also studied intensively. The following is the major contents of the paper.Firstly, the real axis integration (RAI) and the finite difference method is studied and used for forward modeling of the multipole array sonic logging. The perfectly matched layer absorbing boundary condition is also used for the finite difference to eliminate the reflection wave of the outer boundary. The RAI can analyse acoustic field of multi-radial layer formation. For the complex borehole-information condition, the finite difference method or other numerical method is necessary. The acoustic fields of three typical formations are computed by RAI. While the acoustic wave propagation in anisotropic formations and the vertical layered formation is modeled by finite difference.Secondly, methods to get the slowness from the array data are studied widely. The time domain semblance method, the frequency domain Prony’s method and the phase velocity processing in frequency-time domain are all used to extract the slowness from the array waveforms. It is validated that the methods of ours have the same resolution to the business software. It shows the methods developed in the dissertation have higher resolution and can take the place of the software imported. The different method can satisfy different application area. To get more accurate slowness from dispersive flexural wave, the dispersion correction method for it is developed. Thirdly, two different simulated annealing methods are studied, and one complex problem is used to test and compare their properties. Then based on their properties and the local continuous characteristic of sonic logging, the hybrid simulated annealing is used to anisotropy inversion. The actual data test shows that the method is faster and more stable than the method only using one of the two simulated annealing.Finally, the application of the array sonic logging is studied, including the hydraulic fracture evaluation, reservoir fracture detection and formation stress determination. Based on the Gassman equation, the fluid prediction is studied. This can be useful for the seismic interpretation. The two seismic attributes, such as passion ratio and compressional wave impedance, or Lamda-Rho and compressional wave impedance, is easier to determine the fluid type when cross plotted. And the discrimination of gas and water is affected by porosity.