... Recent advances in deep learning and computer vision have resulted in giant leaps in automating some of the cumbersome oil and gas exploration and production operations. Deep convolutional neural networks have been widely used for seismic interpretation tasks including detection, classification...

... We develop a method for automatic rock classification (ARC) from core data to well-log scale using unsupervised machine learning. Our ARC method estimates rock classes along a well to accelerate the petrophysical interpretation of well logs using core data for rapid formation evaluation...

... cepstral coefficient (MFCC) feature. Then, the 2D MFCC feature is the input pattern for CNN classification training. Over 99% accuracies are obtained for training and testing after 300 epochs. Consequently, only 1/2500 misclassification is recorded in the confusion matrix. resource in place estimate...

... This study investigates explainable AI (XAI) methods to understand the inner workings of a machine learning (ML) architecture trained for seismic facies classification in the Canterbury Basin, New Zealand. Specifically, we analyze the global and local importance of a suite seismic attributes...

... logs—the backbones of every cement evaluation procedure—into visual, time-frequency representations. Building upon this, we introduce a transfer learning strategy for the classification of the acquired images, focusing on assessing the cement isolation potential. The cornerstone of transfer learning...

... developed to make PS monitoring time efficient and cost effective. geologist complex reservoir steam-assisted gravity drainage artificial intelligence heavy oil play unconventional play sagd enhanced recovery thermal method machine learning heavy oil field integrity classification cold...

... geophysicist case study test area classification identification lithology Regionally consistent depositional models using data analytics and machine learning: a Gulf of Mexico case study. Arpana Sarkar*, Dan Ferdinand Fernandez, and Daniel Smith, SLB Summary In this study we present an effective...

... resolution to find patterns on the volume which can be associated with seismic facies. The classification of the seismic facies in 3D volumes provides insights into depositional setting and potential reservoir distribution. Furthermore, ML-assisted facies classification is a crucial first step for regional...

... networks exceeds 92%, and CNN presents the best performance among the three networks, suggesting the potential of seismic data in vehicle classification. The results are promising for the future application in real traffic management. upstream oil & gas ground transportation deep learning...

... permeability porosity hydraulic tortuosity diameter classification permeability prediction relation american association discriminator society sedimentary rock clastic rock tortuosity prediction modeling exploration geophysicist rocklog consultancy summary applied geoscience international...

... neural network anomaly reservoir characterization gradient artificial intelligence segmentation geology north america government machine learning ferromagnetic pipeline international applied geoscience national energy technology laboratory prediction american association classification...

... Many classification schemes of sedimentary basins have been developed but an effective classification system remains elusive, rendering communication and comparisons difficult. Here we propose a simple yet robust classification scheme based on driving mechanisms and lithosphere mechanical...

... reservoir characterization biolithite international energy 10 matrix society realization classification pore magnitude carbonate extraction resolution Porosity and permeability calculations in a Biolithite using x-ray tomography images Adetomiwa Aderemi*, Elli-Maria Charalampidou, Zeyun Jiang...

... atlas american association expanded abstract exploration geophysicist disagreement applied geoscience conference alregib learning algorithm selection classification What Samples must Seismic Interpreters Label for Ef cient Machine Learning? Ryan Benkert, Mohit Prabhushankar, and Ghassan...

... We describe the second half of a new workflow, Bayesian Integrated Reservoir Characterization, or BIRCh (Paramo et al. 2023, Biswas et. al 2023, Walker et. al., 2023). The second half consists of Bayesian classification of Vp, Vs and density estimates from a Bayesian seismic inversion scheme...

.... Reservoir evaluation via well logging requires manual classification based on logging data. Machine learning can solve this issue by discovering new rules, presenting tremendous potential for growth. Our study aims to categorize reservoirs by the reservoir quality index (RQI) and the pore geometry...

... Innovative interpretation methods based on efficient measurements are the key to establishing accurate “geological-geophysical” mappings. The multi-solvability of petrophysical responses leads to the complex non-linear relationship between multi-scale geophysical data and the classification...

... in the full resolution image. Finally, neural networks trained on Dunham rock types show better interpretability than the pore type dataset. upstream oil & gas machine learning texture representation artificial intelligence baechle interpretation deep learning classification...

..., subjective and non-repeatable. Thus, we propose to train an end-to-end deep neural network (DNN) to identify the facies of the carbonate rocks from acoustic image logs and gamma ray logs. The well-trained DNN generalizes well to the test and uncertain sections. Compared with manual classification, the DNN...

... oil & gas machine learning explainable machine learning logistic regression classifier classification prospect alregib datapoint interpretation american association applied geoscience neural network explanation dataset international exploration geophysicist alfarraj Explainable...