Sonic $Q_{\mathrm{P}}/Q_{\mathrm{S}}$ ratio obtained from full-waveform acoustic logs in conventional sandstone reservoirs is known to be sensitive to the presence of gas, and it is regarded as a potential diagnostic tool for saturation discrimination. However, it is not known if such a saturation diagnostic tool will be applicable in unconventional reservoirs, such as in gas-saturated shales. We have analyzed the monopole and dipole waveform logs acquired from a shale gas exploration well in the Cooper Basin, South Australia. The depth interval of interest is 300 m thick, and it intersects three shale units in which the two underlying formations contain gas saturation of more than 30% and are identified as the primary exploration targets. We use the statistical average method to extract the $\mathrm{P}$- and the $\mathrm{S}$-wave attenuation profiles and obtain an average $\mathrm{P}$-wave quality factor of $Q_{\mathrm{P}}=33$ and $\mathrm{S}$-wave quality factor of $Q_{\mathrm{S}}=46$. The gas saturation of the lithological layers having $Q_{\mathrm{P}}/Q_{\mathrm{S}}<1$ is appreciably larger than the gas saturation of the others having $Q_{\mathrm{P}}/Q_{\mathrm{S}}>1$. The net difference indicates that the saturation is a dominant factor in controlling the $Q_{\mathrm{P}}/Q_{\mathrm{S}}$ ratio in these shale formations. Based on the criterion $Q_{\mathrm{P}}/Q_{\mathrm{S}}<1$, we identify the intervals with high gas potential. This result is in good agreement with the prediction from an independently obtained saturation log based on petrophysical analysis. Furthermore, we found that the $Q_{\mathrm{P}}/Q_{\mathrm{S}}$ ratio can be jointly interpreted with the $V_{\mathrm{P}}/V_{\mathrm{S}}$ ratio to differentiate between the saturation and the lithology effects for a shale reservoir interbedded with sandstone layers. Our results underpin the concept of using the $Q_{\mathrm{P}}/Q_{\mathrm{S}}$ ratio as a hydrocarbon saturation indicator and provide insights into application of this technique for shale gas detection.

Seismic attenuation and dispersion in layered sedimentary structures are often interpreted in terms of the classical White model for wave-induced pressure diffusion across the layers. However, this interlayer flow is severely dependent on the properties of the interface separating two layers. This interface behaviour can be described by a pressure jump boundary condition involving a non-vanishing interfacial impedance. In this paper, we incorporate the interfacial impedance into the White model by solving a boundary value problem in the framework of quasi-static poroelasticity. We show that the White model predictions for attenuation and dispersion substantially change. These changes can be attributed to petrophysically plausible scenarios such as imperfect hydraulic contacts or the presence of capillarity.

Quantitative interpretation of waveform attenuation for determining petrophysical properties remains one of the most challenging problems associated with rock physics. In this study, we extract $P$...

In field seismic data, abnormal environmental noise from the acquisition environment often is unavoidable, and its characteristics are complex. Abnormal environmental noise has a strong amplitude (tens of thousands times the average seismic amplitudes) and masks the effective seismic signal. The traditional method for removing such abnormal environmental noise relies on energy scanning to identify the noise location. However, energy scanning cannot always accurately identify such noise due to the sudden energy change at the locations of first-breaks and surface waves. A deep learning method for removing abnormal environmental noise has been developed. The presence of environmental noise with large amplitude also results in an uneven energy distribution. The existing denoising networks rarely consider this situation. Based on the characteristics of environmental noise received by the nodal seismic acquisition system, a workflow for automatically generating the training data sets has been established. The network is built based on the Unet with the incorporation of a residual block. The Batchnorm layer is specifically omitted to adapt to the uneven energy distribution. Synthetic and field data testing have confirmed the effectiveness and applicability of the proposed method. The new method indicates better denoising performance and greater flexibility than the traditional method. Comparison between different networks also indicates that the new Residual Block Connected Unet is much more effective in denoising seismic data that contain abnormal environmental noise.

Formation S-wave attenuation, when combined with compressional attenuation, serves as a potential hydrocarbon indicator for seismic reservoir characterization. Sonic flexural wave measurements provide a direct means for obtaining the in situ S-wave attenuation at log scale. The key characteristic of the flexural wave is that it propagates at the formation shear slowness and experiences shear attenuation at low frequency. However, in a fast formation, the dipole log consists of refracted P- and S-waves in addition to the flexural wave. The refracted P-wave arrives early and can be removed from the dipole waveforms through time windowing. However, the refracted S-wave, which is often embedded in the flexural wave packet, is difficult to separate from the dipole waveforms. The additional energy loss associated with the refracted S-wave results in the estimated dipole attenuation being higher than the shear attenuation at low frequency. To address this issue, we have developed a new method for accurately determining the formation shear attenuation from the dipole sonic log data. The method uses a multifrequency inversion of the frequency-dependent flexural wave attenuation based on energy partitioning. We first developed our method using synthetic data. Application to field data results in a shear attenuation log that is consistent with lithologic interpretation of other available logs.

The quality of wavefield separation on vertical seismic profiling (VSP) data directly affects subsequent imaging and inversion processes. However, the traditional methods have various defects in separating upgoing and downgoing waves. The application of deep learning brings another train of thought to solve related problems. The traditional methods, such as f-k filtering and Radon transform, produce results with spatial aliasing and inaccurate amplitudes. The median filtering method relies on accurate first-break picking and waveform consistency. Moreover, the results of traditional methods are subject to manual intervention. To overcome these problems, a deep-learning method is developed for the automatic wavefield separation of VSP data. First, the traditional Radon transform method is used to produce training data sets. Then, to ensure amplitude preservation and suppress spatial aliasing, new input data is formed by recombining the upgoing and downgoing waves extracted by Radon transform. The developed deep-learning network is highly efficient, and its accuracy exceeds that of the traditional methods only after one epoch training. A practical workflow of wavefield separation based on deep learning is established, and it is applied to synthetic data and field distributed acoustic sensing VSP data. The results indicate that our method is superior in terms of amplitude preservation and spatial aliasing suppression. The time consumption of our method is very acceptable and can be further minimized by training the network using downsampling data.