Prediction of fracture-cavity reservoirs in ultra-deep carbonate strike-slip fault zones: A multi-dimensional angle-domain imaging data attribute fusion technique
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
The strike-slip fault systems in ultra-deep carbonate formations are challenging to identify effectively due to their small fault displacements, weak vertical seismic responses, great burial depths, and strong heterogeneity. To address these challenges, this study innovatively integrates multidimensional angle-domain imaging data and proposes a three-stage prediction workflow: ① Strike-slip fault characterization: Combining drilled fault calibration data, we construct a supervised convolutional neural network (CNN) model using coherence attributes of mirror-stack volumes and full-stack fracture-related attributes (e.g., Eigenvalue, curvature). ② Subtle Fracture-Cavity Detection: By integrating azimuth-sector stacking, reflection-angle gathers, dip-sector volumes, and seismic attributes, this study employs Locally Linear Embedding (LLE) — a nonlinear dimensionality reduction technique—to enhance the identification capability of weak-amplitude scatteringtype fracture-cavity reservoirs. ③ Reservoir classification and target optimization: An integrated classification model is established by fusing full-stack energy volumes, scattering energy bodies, and dip-angle gathers. guiding the optimization of ultra-deep exploration targets.
Fields of Study & Subjects
Seismic Imaging and Inversion Techniques
Hydrocarbon exploration and reservoir analysis
Geological Studies and Exploration
Publication details
- Journal / source
- Second International Meeting for Applied Geoscience & Energy
- Volume
- 1
- Pages
- 1402–1406
- Publication date
- 10 February 2026
- Language
- EN