Abstract:Carbonate cave reservoirs are controlled by large strike-slip fault, which are home to incredibly rich oil and gas resources and hold immense exploration potential. The effective identification of these reservoirs using geophysical methods is quite important, but often challenging due to low signal-to-noise ratio and strong background reflection shielding of raw seismic data. In order to fully exploit the information of carbonate reservoirs contained in seismic data and improve the interpretation accuracy, we involve the innovative use of wavelet reconstruction and background modeling techniques to enhance the identification accuracy of carbonate cave reservoirs based on post-stack seismic data. The wavelet reconstruction in the first step mitigate the impact of noise and background reflection, allowing us to reveal the bead-like reflections hidden underneath strong interference energy. Then, we propose a background modeling method based on the weighted logarithmic robust principal component analysis (WLRPCA) technique in the second part to further eliminate background and noise signals. Compared with traditional principal component analysis (PCA) and robust principal component analysis (RPCA), WLRPCA exhibits superior recognition accuracy and noise immunity in the background modeling process, which effectively separates the target reflection from interference reflection and provides a favorable data basis for subsequent reservoir identification. Finally, a comprehensive attribute analysis is conducted on the processed seismic data, which confirms that the sweetness is the most effective attribute in detecting cave reservoirs. The proposed method has been successfully applied to both synthetic data and real data from an oil field in northern China, demonstrating their feasibility and effectiveness in identifying carbonate cave reservoirs.

Seismic characterization of deeply buried paleocaves based on Bayesian deep learning

Paleokarst Caves Recognition from Seismic Response Simulation to Convolutional Neural Network Detection

3D seismic intelligent prediction of fault-controlled fractured-vuggy reservoirs in carbonate reservoirs based on deep learning method

Automatic Karst Cave Detection from Seismic Images Via a Convolutional Neural Network and Transfer Learning

Deep learning for predicting porosity in ultra-deep fractured vuggy reservoirs from the Shunbei oilfield in Tarim Basin, China

Identification of Carbonate Cave Reservoirs Based on Variational Bayesian Principal Component Analysis

An integrated fracture parameter prediction and characterization method in deeply-buried carbonate reservoirs based on deep neural network

Deep carbonate reservoir characterisation using multi-seismic attributes via machine learning with physical constraints

Deep carbonate reservoir characterization using multi seismic attributes:A comparison of unsupervised machine learning approaches

Identification of Carbonate Cave Reservoirs Using Wavelet Reconstruction and WLRPCA Based Background Modeling

Automatic identification of carbonate karst caves using a symmetrical convolutional neural network

Seismic Facies Recognition of Ultra-Deep Carbonate Rocks Based on Convolutional Neural Network

Deep Carbonate Reservoir Characterization Using Multiseismic Attributes: A Comparison of Unsupervised Machine-Learning Approaches

Characterization of fault-karst reservoirs based on deep learning and attribute fusion

INTELLIGENT RECOGNITION OF LAYERED GEOLOGICAL BODY BASED ON MACHINE LEARNING AND SEISMIC EXPLORATION

Shape Carving Methods of Geologic Body Interpretation from Seismic Data Based on Deep Learning

Encoder-Decoder Architecture for 3D Seismic Inversion

Seismic Stratigraphic Interpretation Based on Deep Active Learning.

Identification of Underground Artificial Cavities Based on the Bayesian Convolutional Neural Network

Reconstructing 2-D Basement Relief Using Gravity Data by Deep Neuron Network: an Application on Poyang Basin

Hybrid Deep Learning Approach for Hydraulic Fracture Morphology Diagnosis in Coalbed Methane Wells