Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Facies classification of image logs plays a vital role in reservoir characterization, especially in the heterogeneous and anisotropic carbonate formations of the Brazilian pre-salt region. Although manual classification remains the industry standard for handling the complexity and diversity of image logs, it has notable disadvantages of being time-consuming, labor-intensive, subjective, and non-repeatable. Recent advancements in machine learning offer promising solutions for automation and acceleration. However, previous attempts to train deep neural networks for facies identification have struggled to generalize to new data due to insufficient labeled data and the inherent intricacy of image logs. Additionally, human errors in manual labels further hinder the performance of trained models. To overcome these challenges, we propose adopting the state-of-the-art SwinV2-Unet to provide depthwise facies classification for Brazilian pre-salt acoustic image logs. The training process incorporates transfer learning to mitigate overfitting and confident learning to address label errors. Through a k-fold cross-validation experiment, with each fold spanning over 350 meters, we achieve an impressive macro F1 score of 0.90 for out-of-sample predictions. This significantly surpasses the previous model modified from the widely recognized U-Net, which provides a macro F1 score of 0.68. These findings highlight the effectiveness of the employed enhancements, including the adoption of an improved neural network and an enhanced training strategy. Moreover, our SwinV2-Unet enables highly efficient and accurate facies analysis of the complex yet informative image logs, significantly advancing our understanding of hydrocarbon reservoirs, saving human effort, and improving productivity.
- Geology > Structural Geology > Tectonics > Salt Tectonics (1.00)
- Geology > Geological Subdiscipline (1.00)
- Geology > Rock Type > Sedimentary Rock > Carbonate Rock (0.67)
- 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...)
Adaptive laterally constrained inversion of time-domain electromagnetic data using Hierarchical Bayes
Li, Hai (Chinese Academy of Sciences, Chinese Academy of Sciences) | Di, Qingyun (Chinese Academy of Sciences, Chinese Academy of Sciences) | Li, Keying (Chinese Academy of Sciences, Chinese Academy of Sciences, University of Chinese Academy of Sciences)
Laterally constrained inversion (LCI) of time-domain electromagnetic (TEM) data is effective in recovering quasi-layered models, particularly in sedimentary environments. By incorporating lateral constraints, LCI enhances the stability of the inverse problem and improves the resolution of stratified interfaces. However, a limitation of the LCI is the recovery of laterally smooth transitions, even in regions unsupported by the available datasets. Therefore, we have developed an adaptive LCI scheme within a Bayesian framework. Our approach introduces user-defined constraints through a multivariate Gaussian prior, where the variances serve as hyperparameters in a Hierarchical Bayes algorithm. By simultaneously sampling the model parameters and hyperparameters, our scheme allows for varying constraints throughout the model space, selectively preserving lateral constraints that align with the available datasets. We demonstrated the effectiveness of our adaptive LCI scheme through a synthetic example. The inversion results showcase the self-adaptive nature of the strength of constraints, yielding models with smooth lateral transitions while accurately retaining sharp lateral interfaces. An application to field TEM data collected in Laizhou, China, supports the findings from the synthetic example. The adaptive LCI scheme successfully images quasi-layered environments and formations with well-defined lateral interfaces. Moreover, the Bayesian inversion provides a measure of uncertainty, allowing for a comprehensive illustration of the confidence in the inversion results.
- Geology > Mineral (0.93)
- Geology > Sedimentary Geology > Depositional Environment (0.34)
- Oceania > Australia > Western Australia > North West Shelf > Carnarvon Basin > Exmouth Plateau > WA-1-R > Scarborough Field (0.99)
- Europe > Norway (0.91)
- Reservoir Description and Dynamics > Reservoir Characterization > Exploration, development, structural geology (1.00)
- Reservoir Description and Dynamics > Reservoir Simulation > Evaluation of uncertainties (0.93)
- Reservoir Description and Dynamics > Reservoir Characterization > Seismic processing and interpretation (0.79)
- (2 more...)
Power-law frequency-dependent Q simulations in viscoacoustic media using decoupled fractional Laplacians
Zhang, Yabing (China University of Mining and Technology) | Zhu, Hejun (The University of Texas at Dallas) | Liu, Yang (China University of Petroleum (Beijing)) | Chen, Tongjun (China University of Mining and Technology)
Quantifying seismic attenuation of wave propagation in the Earthยs interior is essential for studying subsurface structures. Previous approaches for attenuation simulations (e.g., the standard linear solid and the fractional derivative model) are mainly based on the frequency-independent quality factor Q assumption. However, seismic attenuation in high-temperature and high-pressure regions usually exhibits power-law frequency-dependent Q characteristics. To simulate this Q effect in attenuative media, we derive a new viscoacoustic wave equation with decoupled fractional Laplacians in the time domain. Unlike the existing methods using relaxation functions to fit the power-law relationship in a specific frequency band, our proposed equation is directly derived from the approximated complex modulus, which explicitly involves the reference quality factor and fractional exponent parameters. Furthermore, this equation contains two fractional Laplacians, which can easily simulate decoupled amplitude dissipation and phase distortion effects, making it amenable to Q-compensated reverse-time migration. In the implementation, a Taylor-series expansion and a pseudo-spectral method are introduced to solve the fractional Laplacians with variable fractional exponents. Numerical experiments demonstrate the effectiveness of the proposed method for power-law frequency-dependent Q simulations. As a forward-modeling engine, our derived viscoacoustic wave equation is a good supplement to the current Q simulation methods and it could be applied in many seismic applications, including Q-compensated reverse time migration and full-waveform inversion.
Distributed acoustic sensing for seismic surface wave data acquisition in an intertidal environment
Trafford, Andrew (University College Dublin) | Ellwood, Robert (Optasense Limited, QinetiQ) | Godfrey, Alastair (Optasense Limited, Indeximate Limited) | Minto, Christopher (Optasense Limited, Indeximate Limited) | Donohue, Shane (University College Dublin)
This paper assesses the use of Distributed Acoustic Sensing (DAS) for shallow marine seismic investigations, in particular the collection of seismic surface wave data, in an intertidal setting. The paper considers appropriate selection and directional sensitivity of fiber optic cables and validates the measured data with respect to conventional seismic data acquisition approaches ,using geophones and hydrophones, along with independent borehole and Seismic Cone Penetration Test (SCPT) data. In terms of cable selection, a reduction of amplitude and frequency response of an armored cable is observed, when compared to an unarmored cable. For seismic surface wave surveys in an offshore environment where the cable would need to withstand significant stresses, the use of the armored variant with limited loss in frequency response may be acceptable, from a practical perspective. The DAS approach has also shown good consistency with conventional means of surface wave data acquisition, and the inverted Vs is also very consistent with downhole SCPT data. Observed differences in phase velocity between high tide (Scholte wave propagation) and low tide (Rayleigh wave propagation) are not thought to be related to the particular type of interface wave due to shallow water depth. These differences are more likely to be related to the development of capillary forces in the partially saturated granular medium at low tide. Overall, this study demonstrates that the proposed novel approach of DAS using seabed fiber-optic cables in the intertidal environment is capable of rapidly providing near-surface shear wave velocity data across considerable spatial scales (multi-km) at high resolution, beneficial for the design of subsea cables routes and landfall locations. The associated reduction in deployment and survey duration, when compared to conventional approaches, is particularly important when working in the marine environment due to potentially short weather windows and expensive downtime.
- Europe (1.00)
- North America > United States > Illinois > Madison County (0.24)
- Research Report > New Finding (0.68)
- Research Report > Experimental Study (0.54)
PINNslope: seismic data interpolation and local slope estimation with physics#xD;informed neural networks
Brandolin, F. (King Abdullah University of Science and Technology (KAUST)) | Ravasi, M. (King Abdullah University of Science and Technology (KAUST)) | Alkhalifah, T. (King Abdullah University of Science and Technology (KAUST))
Interpolation of aliased seismic data constitutes a key step in a seismic processing workflow to obtain high-quality velocity models and seismic images. Building on the idea of describing seismic wavefields as a superposition of local plane waves, we propose to interpolate seismic data by using a physics informed neural network (PINN). In the proposed framework, two feed-forward neural networks are jointly trained using the local plane wave differential equation as well as the available data as two terms in the objective function: a primary network assisted by positional encoding is tasked with reconstructing the seismic data, while an auxiliary, smaller network estimates the associated local slopes. Results on synthetic and field data validate the effectiveness of the proposed method in handling aliased (coarsely sampled) data and data with large gaps. Our method compares favorably against a classic least-squares inversion approach regularized by the local plane-wave equation as well as a PINN-based approach with a single network and precomputed local slopes. We find that introducing a second network to estimate the local slopes while at the same time interpolating the aliased data enhances the overall reconstruction capabilities and convergence behavior of the primary network. Moreover, an additional positional encoding layer embedded as the first layer of the wavefield network confers to the network the ability to converge faster, improving the accuracy of the data term.
- Geophysics > Seismic Surveying > Seismic Processing (1.00)
- Geophysics > Seismic Surveying > Seismic Modeling > Velocity Modeling (0.34)
Introduction to special section: Borehole image data applications in reservoir characterization ? Case studies and updates on new developments
Li, Bingjian (Blackriver Geoscience LLC) | Egorov, Vsevolod (GeoExpera) | Perona, Ricardo (Repsol) | Haddad, David (Apache Corporation) | Sementelli, Katy (Woodside Energy) | Xu, Chicheng (Aramco Americas Company) | Mardi, Chrystianto (BPX Energy)
Borehole image data have played an important role in the oil and gas industry for decades, providing invaluable insights into hydrocarbon exploration, reservoir appraisal, and development. Recent advancements in borehole image technologies, encompassing data acquisition, processing, and interpretation, have ushered in a new era of possibilities. Geoscientists have expanded the applications of image data, progressing from basic natural fracture detection to comprehensive reservoir characterization. This special section explores significant advances in sedimentological and structural interpretation, full-scale fracture and fault analysis, wellbore geomechanics, reservoir heterogeneity evaluation, and 3-D reservoir modeling. Applications of borehole image log data have transcended reservoir types, spanning clastics, carbonates, naturally fractured basements, and unconventional shales. With these developments in mind, we have invited submissions that showcase studies utilizing borehole image log data for the successful characterization of any reservoir type, along with related case studies of interest to the exploration and development community. The overwhelming response to our call-for-papers resulted in the selection of seven high-quality contributions for inclusion in this special publication. Mohebian et al. revolutionize fracture identification by employing the YOLOv5 deep learning algorithm on borehole image logs, shifting from manual to automated processes.
- Geology > Geological Subdiscipline > Geomechanics (1.00)
- Geology > Rock Type > Sedimentary Rock > Clastic Rock (0.56)
- Geophysics > Seismic Surveying > Borehole Seismic Surveying (1.00)
- Geophysics > Borehole Geophysics (1.00)
- Asia > Kazakhstan > West Kazakhstan > Uralsk Region > Precaspian Basin > Karachaganak Field (0.99)
- Asia > China > Bohai Basin (0.99)
Introduction to special section: South China Sea deep structures and tectonics
Zhang, Ruwei (Guangzhou Marine Geological Survey) | Zhang, Baojin (Guangzhou Marine Geological Survey) | Zhu, Hongtao (China University of Geosciences) | Sibuet, Jean-Claude (Ifremer Centre de Brest) | Briais, Anne (Centre National de la Recherche Scientifique) | Wu, Jonny (University of Houston) | Susilohadi, Susilohadi (National Research and Innovation Agency) | Zeng, Hongliu (The University of Texas at Austin) | Chen, Jianxiong (Anadarko Petroleum Corporation) | Zhong, Guangfa (TongJi University)
E-mail: susi021@brin.go.id 8.The University of Texas at Austin, USA. The South China Sea (SCS) is one of the largest Cenozoic marginal seas in the Western Pacific region. This oceanic basin was opened from the southeastern edge of the Asia continent under the interaction of the Eurasian, Indo-Australian, and Philippine Sea-Pacific plates. Therefore, it provides an exceptional natural laboratory to investigate the genesis of marginal seas and to explore plate-tectonic interactions. It is suggested that the deep structural and tectonic characteristics in the SCS reflect the conditions of the formation and geodynamic evolution of the basin.
- Asia > China (1.00)
- North America > United States > Texas > Travis County > Austin (0.26)
More than 1,000 mound structures have been mapped in shallow marine sediments at the Cretaceous ย Paleogene boundary in the Rubย Al-Khali of Saudi Arabia. Mapping utilized 3D reflection seismic data in a 37,000 square kilometer study area. No wells penetrate the mounds themselves. The mounds are at a present-day subsurface depth of approximately 1 km and are convex-up with diameters of 200 ย 400 m and elevation of 10 ย 15 m. The mounds display spatial self-organization with a mean separation of approximately 3.75 km. Comparison with mound populations in other study areas with known spatial distribution statistics and modes of origin indicates that the mound population in this study has the characteristics of fluid escape structures, and they are interpreted here as mud volcanoes. The observation that the mounds occur at the Cretaceous ย Paleogene boundary demands a singular trigger at that moment in time. We develop a model of seismic energy ย related mud volcanism mechanism including the Chicxulub asteroid impact as the energy source that accounts for the timing of the mound structures, and a drainage cell model based on producing water wells that provides a mechanism for spatial self-organization into a regular pattern.
- Europe (1.00)
- Asia > Middle East > Saudi Arabia (1.00)
- Africa (1.00)
- Phanerozoic > Cenozoic > Paleogene > Paleocene (0.67)
- Phanerozoic > Mesozoic > Cretaceous > Upper Cretaceous (0.46)
- Geology > Sedimentary Geology > Depositional Environment > Marine Environment (1.00)
- Geology > Rock Type > Sedimentary Rock (1.00)
- Geology > Geological Subdiscipline > Volcanology (1.00)
- (2 more...)
- Geophysics > Seismic Surveying > Seismic Processing (1.00)
- Geophysics > Seismic Surveying > Seismic Interpretation > Well Tie (0.46)
- Oceania > New Zealand > South Island > South Pacific Ocean > Great South Basin (0.99)
- North America > Canada > Saskatchewan > Prairie Evaporite Basin (0.99)
- Europe > Norway > North Sea > Central North Sea > Norwegian-Danish Basin (0.99)
- (6 more...)
- Reservoir Description and Dynamics > Reservoir Characterization > Seismic processing and interpretation (1.00)
- Reservoir Description and Dynamics > Reservoir Characterization > Exploration, development, structural geology (1.00)
- Data Science & Engineering Analytics > Information Management and Systems > Artificial intelligence (1.00)
Unsupervised Deep Learning Framework for 5D Seismic Denoising and Interpolation
Saad, Omar M. (National Research Institute of Astronomy and Geophysics (NRIAG), King Abdullah University of Science and Technology (KAUST)) | Helmy, Islam (National Research Institute of Astronomy and Geophysics (NRIAG)) | Chen, Yangkang (The University of Texas at Austin)
We propose an unsupervised framework to reconstruct the missing data from the noisy and incomplete five-dimensional (5D) seismic data. The proposed method comprises two main components: a deep learning network and a projection onto convex sets (POCS) method. The model works iteratively, passing the data between the two components and splitting the data into a group of patches using a patching scheme. Specifically, the patching scheme breaks the input data into small segments which are then reshaped to a vector of one dimension feeding the deep learning model. Afterward, POCS is utilized to optimize the output data from the deep learning model, which is proposed to denoise and interpolate the extracted patches. The proposed deep learning model consists of several blocks, that are, fully connected layers, attention block, and several skip connections. Following this, the output of the POCS algorithm is considered as the input of the deep learning model for the following iteration. The proposed model iteratively works in an unsupervised scheme where labeled data is not required. A performance comparison with benchmark methods using several synthetic and field examples shows that the proposed method outperforms the traditional methods.
- North America > United States (0.28)
- Asia (0.28)
Elliptical anisotropy is convenient to use as the reference medium in perturbation methods designed to study P-wave propagation for transverse isotropy (TI). We make the elliptically anisotropic TI model attenuative and discuss the corresponding P-wave dispersion relation and the wave equation. Our analysis leads to two conditions in terms of the Thomsen type parameters, which guarantee that the P-wave slowness surface and the dispersion relation satisfy elliptical equations. We also obtain the viscoacoustic wave equation for such elliptically anisotropic media and solve it for point-source radiation using the correspondence principle. For the constant-Q TI model, we use the weighting function method to derive the viscoacoustic wave equation in differential form. Numerical examples validate the proposed elliptical conditions and illustrate the behavior of the P-wavefield in attenuative elliptical TI models.