Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Results
The numerical solution of wave equations plays a crucial role in computational geophysics problems, which forms the foundation of inverse problems and directly impact the high-precision imaging results of earth models. However, common numerical methods often lead to signifcant computational and storage requirements. Due to the heavy reliance on forward modeling methods in inversion techniques, particularly full waveform inversion, enhancing the computational efficiency and reducing storage demands of traditional numerical methods becomes a key issue in computational geophysics. In this paper, we present the deep Lax-Wendroff correction method (DeepLWC), a deep learning-based numerical format for solving two#xD; dimensional (2D) hyperbolic wave equations. DeepLWC combines the advantages of the traditional numerical schemes with a deep neural network. We provide a detailed comparison of this method with representative traditional Lax-Wendroff correction (LWC) method. Our numerical results indicate that the DeepLWC signifcantly improves calculation speed (by more than ten times) and reduces storage space by over 10000 times compared to traditional numerical methods. In contrast to the more popular Physics Informed Neural Network (PINN) method, DeepLWC maximizes the advantages of traditional mathematical methods in solving PDEs and employs a new sampling approach, leading to improved accuracy and faster computations. It is particularly worth pointing that, DeepLWC introduces a novel research paradigm for numerical equation-solving, which can be combined with various traditional numerical methods, enabling acceleration and reduction in storage requirements of conventional approaches.
- Reservoir Description and Dynamics > Reservoir Characterization > Seismic processing and interpretation (1.00)
- Data Science & Engineering Analytics > Information Management and Systems > Neural networks (1.00)
During geophysical exploration, inpainting defective logging images caused by mismatches between logging tools and borehole sizes can affect fracture and hole identification, petrographic analysis and stratigraphic studies. However, existing methods do not describe stratigraphic continuity enough. Also, they ignore the completeness of characterization in terms of fractures, gravel structures, and fine-grained textures in the logging images. To address these issues, we propose a deep learning method for inpainting stratigraphic features. First, to enhance the continuity of image inpainting, we build a generative adversarial network (GAN) and train it on numerous natural images to extract relevant features that guide the recovery of continuity characteristics. Second, to ensure complete structural and textural features are found in geological formations, we introduce a feature-extraction-fusion module with a co-occurrence mechanism consisting of channel attention(CA) and self-attention(SA). CA improves texture effects by adaptively adjusting control parameters based on highly correlated prior features from electrical logging images. SA captures long-range contextual associations across pre-inpainted gaps to improve completeness in fractures and gravels structure representation. The proposed method has been tested on various borehole images demonstrating its reliability and robustness.
- Geology > Geological Subdiscipline > Stratigraphy (0.74)
- Geology > Rock Type > Sedimentary Rock > Clastic Rock (0.46)
- Geophysics > Seismic Surveying > Borehole Seismic Surveying (1.00)
- Geophysics > Borehole Geophysics (1.00)
- Reservoir Description and Dynamics > Reservoir Characterization > Exploration, development, structural geology (1.00)
- Reservoir Description and Dynamics > Formation Evaluation & Management > Open hole/cased hole log analysis (1.00)
- Production and Well Operations > Well & Reservoir Surveillance and Monitoring > Borehole imaging and wellbore seismic (1.00)
- (2 more...)
The use of virtual reality (VR), a technology that immerses the user in a realistic 3D experience, is becoming increasingly affordable and accessible across different industries. Previously, its use has been mainly associated with entertainment and gaming, but the technology has also seen immense success in health care for training in complex procedures such as surgery, in the mining industry to simulate emergency situations and to explore difficult terrain, and in the automotive industry to improve driving and reduce accidents. This paper describes how the use of VR has been transformative in the approach taken to training in major-hazard industries including oil and gas. The use of VR for training has seen immense growth because it provides an interactive learning environment that is both engaging and fun. For most people, it is something they do not use at home, so the experience is a major boon for engagement of a workforce that would have previously experienced many different, conventional training programs.
- Energy > Oil & Gas (0.40)
- Education > Educational Setting (0.38)
- Health, Safety, Environment & Sustainability (1.00)
- Data Science & Engineering Analytics > Information Management and Systems (0.94)
- Management > Professionalism, Training, and Education (0.60)
The goal of the talk is to start you on the way to becoming a data connoisseur, instead of merely an indiscriminate consumer. The talk will be example-driven for a broad audience, however I will also have tutorial sections that I can include for academic audiences wanting to dig deeper with a longer talk.
- Reservoir Description and Dynamics > Reservoir Characterization > Seismic processing and interpretation (1.00)
- Data Science & Engineering Analytics > Information Management and Systems (1.00)
- Information Technology > Knowledge Management (0.76)
- Information Technology > Communications > Collaboration (0.76)
Arnab Dhara discusses his paper, "Physics-guided deep autoencoder to overcome the need for a starting model in full-waveform inversion," in the June issue of The Leading Edge. Arnab proposes employing deep learning as a regularization in full-waveform inversion. He explains why physics-based solutions with machine learning are challenging to develop, how he made it possible to train the network without known answers, and why he tested his approach with the Marmousi and SEAM models. Arnab also shares why this research took over 20 years to build on the initial idea and how he used full-waveform inversion without a starting model. This is a cutting-edge conversation that may represent the future of FWI.
- Reservoir Description and Dynamics > Reservoir Characterization (1.00)
- Data Science & Engineering Analytics > Information Management and Systems > Neural networks (0.70)
- Information Technology > Artificial Intelligence > Representation & Reasoning (0.70)
- Information Technology > Artificial Intelligence > Machine Learning > Neural Networks > Deep Learning (0.32)
Aria Abubakar was born in Bandung, Indonesia. He received an MSc degree in Electrical Engineering in 1997 and a PhD in Technical Sciences in 2000, both from the Delft University of Technology, The Netherlands. He joined Schlumberger-Doll Research in Ridgefield, CT, USA in 2003, where he remained for 10 years, ending his tenure as a Scientific Advisor and the Manager of the Multi-Physics Modeling and Inversion Program. From 2013 until mid-2017, he was the Interpretation Engineering Manager at Schlumberger Houston Formation Evaluation in Sugar Land, TX. From mid-2017 until mid-2020, he was Data Analytics Program Manager for Software Technology and then Head of Data Science for the Schlumberger Exploration and Field Development Platform based in Houston, TX. Aria is currently the Head of Data Science for the Digital Subsurface Solutions.
- Europe > Netherlands > South Holland > Delft (0.25)
- North America > United States > Texas > Harris County > Houston (0.25)
- North America > United States > Texas > Fort Bend County > Sugar Land (0.25)
- (2 more...)
- Reservoir Description and Dynamics > Reservoir Characterization > Seismic processing and interpretation (1.00)
- Reservoir Description and Dynamics > Formation Evaluation & Management (1.00)
- Data Science & Engineering Analytics > Information Management and Systems > Artificial intelligence (1.00)
- Information Technology > Data Science (1.00)
- Information Technology > Knowledge Management (0.76)
- Information Technology > Communications > Collaboration (0.76)
- Information Technology > Artificial Intelligence > Natural Language (0.48)
Fabien Allo highlights his award-winning article, "Characterization of a carbonate geothermal reservoir using rock-physics-guided deep neural networks." Fabien shares the potential of deep neural networks (DNNs) in integrating seismic data for reservoir characterization. He explains why DNNs have yet to be widely utilized in the energy industry and why utilizing a training set was key to this study. Fabien also details why they did not include any original wells in the final training set and the advantages of neural networks over seismic inversion. This episode is an exciting opportunity to hear directly from an award-winning author on some of today's most cutting-edge geophysics tools.
- Reservoir Description and Dynamics > Reservoir Characterization (1.00)
- Data Science & Engineering Analytics > Information Management and Systems > Neural networks (1.00)
Seismic Characterization of the Individual Geological Factor with Disentangled Features
Fei, Yifeng (University of Electronic Science and Technology of China (UESTC)) | Cai, Hanpeng (University of Electronic Science and Technology of China (UESTC)) | Zhou, Cheng (University of Electronic Science and Technology of China (UESTC)) | He, Xin (University of Electronic Science and Technology of China (UESTC)) | Liang, Jiandong (University of Electronic Science and Technology of China (UESTC)) | Su, Mingjun (PetroChina) | Hu, Guangmin (University of Electronic Science and Technology of China (UESTC))
Seismic attributes are critical in understanding geological factors, such as sand body configuration, lithology, and porosity. However, existing attributes typically reflect a combined response of multiple geological factors. The interplay between these factors can obscure the features of the target factor, posing a challenge to its direct seismic characterization, particularly when the factor is subtle. To address this, we develop an innovative neural network designed to disentangle and characterize the individual geological factor within seismic data. Our approach divides the geological information in the seismic data into two categories: the single geological factor of interest and an aggregate of all other information. A novel feature-swapping mechanism within our network facilitates the disentanglement of these two categories, providing an interpretable representation. We employ a triplet loss function to differentiate data samples with similar waveforms but varying subtle geological details, thus enhancing the extraction of distinct features. Additionally, our network employs a co-training strategy to integrate synthetic and actual field data during the training process. This strategy helps mitigate potential performance degradation arising from discrepancies between simulated and actual field data. We apply our method to synthetic data experiments and field data from two geologically distinct areas. Current results indicate that our method surpasses traditional approaches such as a deep autoencoder and a convolutional neural network classifier in extracting seismic attributes with more explicit geophysical implications.
- Geology > Rock Type > Sedimentary Rock (0.46)
- Geology > Geological Subdiscipline > Geomechanics (0.45)
- Reservoir Description and Dynamics > Reservoir Characterization > Seismic processing and interpretation (1.00)
- Data Science & Engineering Analytics > Information Management and Systems > Neural networks (1.00)
Petroleum Engineering, University of Houston, 2. Metarock Laboratories, 3. Department of Earth and Atmospheric Sciences, University of Houston) 16:00-16:30 Break and Walk to Bizzell Museum 16:30-17:30 Tour: History of Science Collections, Bizzell Memorial Library, The University of Oklahoma 17:30-19:00 Networking Reception: Thurman J. White Forum Building
- Research Report > New Finding (0.93)
- Overview (0.68)
- Geology > Geological Subdiscipline > Geomechanics (1.00)
- Geology > Mineral (0.72)
- Geology > Rock Type > Sedimentary Rock > Carbonate Rock (0.68)
- (2 more...)
- Geophysics > Borehole Geophysics (1.00)
- Geophysics > Seismic Surveying > Seismic Modeling > Velocity Modeling (0.93)
During the last few years, there has been an increasing focus on automation solutions for drilling, with many of these solutions already in the market. Artificial Intelligence (AI) techniques are usually part of these solutions. During this SPE Live, we focus on AI systems that are targeting higher levels of automation, some reaching all the way to autonomy. One requirement for such systems to be taken into use is to be trustworthy. But what does this mean and how to achieve it? At the same time, drilling involves multiple providers, which reflects on the data and information flow, hence adding a new level of complexity for any automation solution.