Flexural-dipole sonic logging has been widely used as the primary method to measure formation shear slowness because it can be applied in both fast and slow formations and can resolve azimuthal anisotropy. The flexuraldipole waveforms are dispersive borehole-guided waves that are sensitive to borehole geometry, mud and formation properties, and therefore the processing techniques need to honor the physical dispersive signatures to obtain an accurate estimation of shear slowness. Traditional processing techniques are based on either a modeldependent method, in which an isotropic model is used as a reference to compensate for the dispersion effect, or a model-independent method, which optimizes nonphysical parameters to fit a simplified model to the field dispersion data extracted in the slowness-frequency domain. Many methods often require inputs such as mud slowness, frequency bandpass filter, or an initial guess of formation shear. Consequently, these methods often fail to interpret the dispersion signature properly when those inputs are inaccurate or when the waveform quality is poor due to downhole logging noises. The users must manually tune the processing parameters and/or choose different methods as a workaround, which causes extra time and effort to obtain the result hence imposes a significant challenge for automating sonic shear interpretation. We develop a physics-driven machine learning-based method for enhancing the interpretation of borehole sonic dipole data for both wireline logging and loggingwhile- drilling. Extensive synthetic databases (i.e., lookup tables) are generated from an anisotropic root-finding mode-search routine and used to train neural network models as accurate and efficient proxies. Those neural network proxies can be used for real-time sensitivity analysis and performing inversion to the measured sonic dipole dispersion data to estimate relevant model parameters with associated uncertainties. Alternatively, various machine learning methods can also be developed based on the generated training dataset and that can be used for inferring relevant model parameters with uncertainties from the field data directly. We introduce how these trained models can be used to enhance the labeling and extraction of different dispersion modes. We developed a new method that outperforms previous modeldependent and model-independent approaches because the new method introduces a mechanism to constrain the solution with physics that also has the capability to incorporate more complicated physical dispersion signatures. This new method needs neither prior information such as mud slowness and formation shear slowness, nor any tuning parameter to be played by the user. It also paves a way to automatically identify different anisotropy mechanisms such as intrinsic, layering, stress, or fractures. This leads to significant progress toward automated sonic interpretation. The algorithm and workflow have been tested on field data for challenging borehole and geological conditions and compared with traditional flexural-dipole processing techniques with great success.

Abstract Petrophysical analysis for determining porosity and water saturation in complex mineralogy can be challenging owing to the strong effect of light hydrocarbons on sensors such as density-neutron, nuclear magnetic resonance, and sonic sensors. In the new frontier areas of offshore Brazil and East Africa, complex mineralogy that results from the presence of feldspars, heavy minerals, and varying clay types further complicates traditional log interpretation methods. When drilling uses oil-base mud, additional uncertainty arises from the varying amount of invasion of the filtrate and its effect on log response. In such situations, using acoustic dipole shear and flexural wave dispersion data coupled with induction logs in the petrophysical workflow provides an alternative method for evaluating porosity and water saturation. Shear and flexural wave data are less influenced by fluid, and coupling the radial response of the flexural wave data with array induction logs by using a pixel-based joint inversion can provide porosity and water saturation profiles from the near wellbore to the far field. These results compare favorably with routine core analysis for porosity and with special core analysis for saturation height modeling, which indicates a significant improvement over traditional log analysis for key intervals. Effective medium models suitable for complex mineralogy and porosity types found in these frontier areas were studied and applied within the multisensor inversion workflow in conjunction with Gassmann fluid substitution to evaluate light hydrocarbon effects on the elastic properties of the formation. The results demonstrate an efficient and accurate inversion workflow that can complement traditional formation evaluation in challenging conditions.

ABSTRACT: A novel acoustic logging data processing and interpretation technique is developed to assess the radial property change of a low-velocity formation. Such a formation is frequently encountered in shallow marine environment, where the sediment is often unconsolidated and mechanically weak. Determining the rock property change around the wellbore is important for assessing well stability and reducing potential hazards. Compared to dipole-shear waves, the compressional wave (i.e. P wave) measured in this environment is more advantageous for determining the radial change. Besides its low P-wave velocity, the formation usually has a high Poisson's ratio, resulting in strong attenuation and dispersion of the P waves along the borehole. The dispersive borehole P-wave is therefore called leaky-P wave. Sensitivity analyses show that the wave's dispersion characteristics are very sensitive to the near-borehole velocity change. This allows for developing an inversion method to estimate the radial change from the dispersion data. We also discuss the nonuniqueness issue of the inversion and ways to solve it. The dispersion data and the inversion results clearly demonstrate that formation alteration is a common characteristic for unconsolidated soft sediments, which occurs even while the well is being drilled. INTRODUCTION For drilling and exploration of offshore reservoirs, the unconsolidated shallow sediments often cause well stability problems because the formations surrounding the borehole are often acoustically very slow and mechanically weak. Determining the acoustic property around the wellbore can help for accessing the well stability and reducing potential hazards. Acoustic logging in such formations often measures a compressional (or P) wave that is both dispersive and attenuative. This P wave is therefore called leaky-P wave. The dispersion characteristics have been discussed by Hornby and Pastemark (2000) for wireline situations and by Goldenberg et al (2003) and Tang et al (2005) for LWD situations.

With the world-wide application of LWD acoustic technology in recent years, various formation types and drilling environments are encountered in the practice. A challenge for the LWD acoustic measurement is the shallow, unconsolidated, and acoustically very slow formations that are frequently encountered in drilling a deep-water reservoir. In this environment, the borehole is usually very large, varying in the range of 12 ? 18 inches. Consequently, LWD acoustic devices that are designed for high-frequency (~ 10 kHz or higher measurement often fail to excite the desirable acoustic propagation and to correctly measure formation velocity. By analyzing the acoustic wave phenomena in this environment, we have designated a procedure that incorporates the low-frequency measurement in the 2–4 kHz frequency range. The acoustic compressional waves in this frequency range have a good signal-to-noise ratio and the data quality is good even in the presence of drilling noise. The waves, however, are usually attenuative (or leaky, from radiating energy into formation and dispersive. The degree of dispersion depends on the softness (or more specifically, Poisson?s ratio of the formation and the possible radial alteration of the formation velocity. A dispersion analysis is then applied to the measured data to estimate and to correct for the dispersion effect. The measured dispersion can also be compared with theoretical modeling to assess the possible formation alteration. Field data examples demonstrate that the above procedure yields reliable compressional wave slowness for the unconsolidated slow formations.