Natural fractures significantly influence the production performance of volcanic hydrocarbon reservoirs. However, the complex nature of fracture networks affects the elastic behavior and seismic responses of these reservoirs, creating substantial challenges for seismic-based fracture characterization. This study proposes a method for quantitatively estimating fracture parameters in volcanic formations by integrating a rock physics model with the Bayesian optimized bidirectional long short-term memory (BO-BiLSTM) network architecture. Following the established model for volcanic formations, a model-based inversion approach is introduced to calculate fracture parameters using well log data, incorporating the simulated annealing particle swarm optimization (SA-PSO) algorithm for reliable multi-parameter simultaneous estimation. Fracture density is derived from these parameters, enabling a comprehensive assessment of fracture parameters. The validity of the volcanic model and inversion framework is confirmed by the strong correlation between predicted and observed velocities during the inversion process. Subsequently, the BO-BiLSTM architecture is used to build a predictive framework that captures the intricate relationships between fracture parameters and elastic properties. This predictive model is then applied to estimate fracture parameters from seismically-derived elastic properties. The seismic-based predictions exhibit strong agreement with results obtained from well log data, demonstrating the applicability of the proposed method. The predicted fracture parameters reflect micro-scale fracture characteristics and exhibit a structure-related distribution when visualized alongside macro-scale fractures, revealing patterns of fracture networks and potential connectivity between larger fractures. These insights enhance the understanding of multi-scale fracture systems, providing critical information for volcanic reservoir characterization.

The detection of vertical fractures using seismic methods is crucial for characterizing shale gas reservoirs. When vertical fractures are aligned within a vertically transverse isotropic (VTI) shale formation, the medium behaves as an orthorhombic system. Seismic waves propagating through such fluid-saturated fractured shales exhibit anisotropic dispersion and attenuation, effects often not adequately considered by conventional fracture detection methods. This paper addresses this limitation by proposing an azimuthal frequency-dependent inversion method to compute anisotropic dispersion attributes for fracture characterization in orthorhombic shales. The method employs an azimuthal amplitude difference scheme to highlight anisotropy’s influence on the reflection coefficient by removing the VTI background effects, enhancing the robustness in anisotropic parameter estimation. A frequency-dependent inversion framework fully utilizes seismic reflection frequency information, while a frequency-scanning scheme improves mathematical rigor over traditional approaches. Numerical examples based on a viscoelastic orthorhombic shale model demonstrate the feasibility of the proposed dispersion attributes for fracture characterization. Synthetic data tests validate their effectiveness in assessing vertical fracture density, and field data applications confirm their reliability through good correlations with FMI images. Finally, a fracture indicator integrating the anisotropic dispersion attributes is developed to characterize vertical fractures in the studied shale reservoirs. This methodology holds potential for advancing the detecting of more complex fracture systems in hydrocarbon reservoirs.

Predicting microcracks is crucial for identifying tight sandstones with high permeability, which is essential for economic gas production. However, effective methods for directly detecting microcracks using prestack seismic data are currently lacking. Rock physics modeling and experiments suggest that microcracks can cause frequency-dependence of elastic properties and seismic responses, which are not fully utilized in existing seismic methods for microcrack prediction. This article addresses this gap by introducing an approach for identifying microcracks using seismic dispersion attributes derived from a new frequency-dependent amplitude variation with offset (AVO) inversion. A new AVO equation is derived, parameterized by a proposed microcrack indicator and relevant elastic properties, with its accuracy verified against exact solutions from the Zoeppritz equations. A frequency-dependent AVO inversion approach is then developed to compute the microcrack-related dispersion attribute from prestack seismic data. The effectiveness of the proposed dispersion attribute for robust microcrack prediction is validated using synthetic data. The results suggest that the proposed dispersion attribute increases with microcrack density and is unaffected by varying gas saturation. In real seismic data applications, high-value anomalies of the proposed dispersion attribute indicate tight sandstones with high permeability. Given the geological understanding that microcracks significantly enhance the permeability of tight sandstones, the reliability of the proposed dispersion attribute for detecting microcracks is further confirmed by this result. The presented method provides valuable information for the comprehensive characterization of prospective areas in tight sandstone gas reservoirs.

Prediction of fluid-saturated fractures is crucial for characterizing tight hydrocarbon reservoirs. Elastic amplitude variation versus azimuth (AVAz) methods are commonly employed for fracture prediction. Although frequency-dependent anisotropy has been explored using various experiments and numerical modeling, there is a lack of applicable methods for fracture prediction based on seismic anisotropy dispersion. To address this gap, we propose a frequency-scanning AVAz (FS-AVAz) method to predict fluid-saturated fractures by extracting anisotropy dispersion properties associated with fluid flow in fractured porous rocks from wide-azimuth seismic data. In this context, the presented FS-AVAz method offers a new perspective for fracture prediction that differs from elastic AVAz approaches. In the FS-AVAz method, the reflectivity equation formulated by azimuthal amplitude differences enables a robust estimation of the anisotropy dispersion inversion by eliminating the effect of the isotropic host rock. Meanwhile, as a critical inversion framework, the proposed frequency-scanning scheme ensures a reliable calculation of anisotropy dispersion attributes. Synthetic tests validate the capability of the FS-AVAz method in hydrocarbon-saturated fracture prediction. The effectiveness of the presented method for fracture prediction is further confirmed by field data applications using wide-azimuth seismic data. The predicted anisotropy dispersion attribute shows a good agreement with logging permeability and can serve as an indicator of fluid-saturated fractures. Meanwhile, incorporating the results of FS-AVAz and frequency-dependent amplitude variation versus offset can achieve a comprehensive characterization of fluids and fractures in tight rocks.

Natural fractures play a crucial role in shale reservoir characterization. While vertical fractures can be estimated using amplitude variation with azimuth (AVAz) inversion methods, predicting horizontal fractures remains limited due to their intricate seismic responses. This article addresses this gap by introducing a rock-physics-based method for predicting horizontal fracture parameters using well-log and seismic data. A shale model is developed using rock physics methods to quantify elastic properties associated with horizontal fractures. The sensitivity of the elastic properties to horizontal fracture parameters is analyzed to validate their potential for fracture prediction. Subsequently, a model-based inversion approach is proposed to extract horizontal fracture parameters from logging data. This method integrates a simulated annealing particle swarm optimization (SA-PSO) algorithm to ensure robustness and convergence in calculations. The results confirm the efficacy of the proposed method in estimating horizontal fracture parameters in boreholes. Based on the results obtained using the proposed method and well-log data, a prediction model is constructed to capture complex correlations between horizontal fracture density and elastic properties by employing a convolutional neural network (CNN) architecture. Following successful training and validation, the established model predicts horizontal fracture density in shales using elastic properties obtained from seismic inversion. The results align closely with estimates derived from logging data and are consistent with the geological characteristics of the studied area. This study presents a valuable method for predicting horizontal fractures using geophysical logging and seismic data, offering valuable insights into natural fracture characterization in shales.

Fluid identification using seismic data is critical for the characterization of hydrocarbon reservoirs. Poroelastic behaviors associated with wave-induced fluid flow allow for hydrocarbon detection using seismic attributes derived from frequency-dependent information. However, the conventional inversion methods usually estimate the frequency-dependent seismic reflectivity across subsurface interfaces, which may lead to ambiguities in the interpretation of the obtained results. A novel frequency-dependent elastic impedance inversion approach is proposed in the present study by extending the elastic impedance equation, as represented by the elastic modulus, to the form in the frequency domain. The proposed method has the advantage of calculating the interlayer dispersion properties, instead of the frequency-dependent interface reflectivity, for fluid identification. The estimated interlayer dispersion attributes show more apparent physical meanings than the traditional frequency-dependent interface properties. It can thereby provide intuitive interpretations for hydrocarbon identifications. Synthetic examples show that the interlayer dispersion attribute estimated by the proposed method can reliably indicate tight sandstone targets with less ambiguity than a traditional interface frequency-dependent attribute. Real data applications of tight sandstone reservoirs further validate the effectiveness of the here proposed frequency-dependent elastic impedance inversion method. For gas-bearing tight sandstones, the obtained interlayer dispersion attribute can be used as a fluid identification factor with improved accuracy. By using appropriate elastic impedance representations, the proposed method provides a valuable tool for fluid detection in various hydrocarbon resources by extending the method to the estimation of other interlayer dispersion attributes.

Seismic prediction of natural fractures is essential for the characterization of unconventional reservoirs because fractures provide seepage paths for fluid migration and storage space for hydrocarbon accumulation. However, it remains challenging to robustly estimate anisotropic parameters for fracture characterization based on seismic inversion methods. We propose an improved inversion scheme for the robust and accurate estimation of anisotropic parameters using PP-wave azimuthal amplitude differences incorporated with a hybrid optimization algorithm. Modeling analysis indicates that by removing the effect of isotropic terms, azimuthal amplitude differences are more sensitive to anisotropic parameters than the traditional azimuthal reflection coefficient; this can avoid possible instabilities in anisotropic parameter estimates that may occur in conventional methods owing to unbalanced weighting coefficients between isotropic and anisotropic terms. Meanwhile, the proposed hybrid algorithm takes advantage of the global optimization ability of the simulated annealing algorithm and the fast convergence characteristics of the particle swarm optimization algorithm. Synthetic data tests indicate that the hybrid algorithm achieves reliable and stable estimates of anisotropic parameters with higher computational accuracy and efficiency than traditional methods. The improved inversion method is applied to characterize fractures in volcanic gas reservoirs. Based on azimuthal amplitude differences obtained after estimating fracture orientation using the Fourier series method, anisotropic parameters are computed and converted to weakness properties, which is more meaningful in terms of the fracture properties. Our results indicate that the obtained fracture tangential weakness exhibits an apparent correspondence with well-log permeability, justifying the applicability of the proposed method for fracture prediction.

A proper description of the frequency-dependent seismic amplitude variation versus offset (AVO) responses should consider the effect of both the layered structure of a reservoir and the dispersive and attenuated property of the media in the reservoir. We propose an improved method to seamlessly link the rock physics modeling and the calculation for frequency-dependent reflection coefficients based on propagator matrix method. The improved AVO modeling method is implemented in frequency-wavenumber domain, and can accurately considers dispersion and attenuation that described by complex and frequency-dependent elastic properties predicted by rock physic models. Therefore, the improved method avoids errors resulting from truncating imaginary parts of the elastic properties as adopted by the conventional Zoeppritz-equation-based method. Moreover, the proposed method considers the intrinsic contribution of the layered structure to frequency-dependence of AVO responses, which has been ignored by current conventional methods. In addition, the method provides an efficient way to calculate seismograms for a dispersive and attenuated layered model. Finally, modeling results show the applicability of the improved method for the interpretation of complex frequency-dependent abnormalities, and indicate the potential for fluid detection in a layered reservoir.