Pub Date : 2021-10-18DOI: 10.3997/2214-4609.202113194
X. Gao, T. Gong, Z. Wang, R. Zhang, D. Hu
Summary Higher seismic imaging resolution is in need as the targets of exploration and development are transformed from structural reservoirs to lithologic reservoirs, which are becoming more and more complex. To meet the demands of lithologic exploration, seismic data results should be broadband and amplitude preserved. A full-layer Q compensation technology which consists of surface absorption compensation, Q tomography and Q pre-stack depth migration imaging technology is raised. It takes into account the near-surface anomalies with strong absorption and the influence of viscoelastic absorption and attenuation factors during the propagation of seismic waves. Different absorption and attenuation problems from shallow to deep are treated separately. The amplitude attenuation and phase change caused by the absorption of stratum in the travel path from shots to detectors are compensated. The resolution and amplitude attributes are effectively improved in the study area. The seismic result supports the prediction of continental shale oil sweet spots and the deployment need of horizontal wells in Songliao Basin.
{"title":"Application of Full-layer Q Compensation Technology in Shale Oil Sweet Spot Prediction in Songliao Basin","authors":"X. Gao, T. Gong, Z. Wang, R. Zhang, D. Hu","doi":"10.3997/2214-4609.202113194","DOIUrl":"https://doi.org/10.3997/2214-4609.202113194","url":null,"abstract":"Summary Higher seismic imaging resolution is in need as the targets of exploration and development are transformed from structural reservoirs to lithologic reservoirs, which are becoming more and more complex. To meet the demands of lithologic exploration, seismic data results should be broadband and amplitude preserved. A full-layer Q compensation technology which consists of surface absorption compensation, Q tomography and Q pre-stack depth migration imaging technology is raised. It takes into account the near-surface anomalies with strong absorption and the influence of viscoelastic absorption and attenuation factors during the propagation of seismic waves. Different absorption and attenuation problems from shallow to deep are treated separately. The amplitude attenuation and phase change caused by the absorption of stratum in the travel path from shots to detectors are compensated. The resolution and amplitude attributes are effectively improved in the study area. The seismic result supports the prediction of continental shale oil sweet spots and the deployment need of horizontal wells in Songliao Basin.","PeriodicalId":265130,"journal":{"name":"82nd EAGE Annual Conference & Exhibition","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130469134","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-18DOI: 10.3997/2214-4609.202113181
W. Sang, S. Yuan, S. Li, J. Cheng, S. Wang
Summary Traditional porosity prediction methods usually adopt two consecutive steps including seismic inversion and petrophysical modelling to convert seismic data into porosity. Machine learning can take full advantage of available geophysical information to directly build the nonlinear mapping for predicting porosity from seismic data. To realize the one-step reservoir porosity estimation, we propose the semi-supervised recurrent neural networks (SSRNNs) based porosity modelling method. SSRNNs include an encoder subnet and a decoder subnet. The encoder simulates the generalized seismic inversion to convert the input post-stack seismic data into the predicted porosity, and the decoder acts as a forward model to make the predicted porosity can return to the generated seismic data and reduce resolution space. In addition, seismic data at the non-well positions are randomly selected in each iteration of SSRNNs to boost the lateral continuity of the predicted porosity result. Without the demand of some approximate assumptions and accurate elastic parameters, well logs and seismic data at well locations and non-well locations are integrated into SSRNNs to directly predict high-precision porosity from seismic data. A numerical model example and a real data example are used to verify the effectiveness and accuracy of the SSRNNs based reservoir lateral porosity prediction method.
{"title":"Semi-supervised seismic data and well logs integration for reservoir lateral porosity prediction","authors":"W. Sang, S. Yuan, S. Li, J. Cheng, S. Wang","doi":"10.3997/2214-4609.202113181","DOIUrl":"https://doi.org/10.3997/2214-4609.202113181","url":null,"abstract":"Summary Traditional porosity prediction methods usually adopt two consecutive steps including seismic inversion and petrophysical modelling to convert seismic data into porosity. Machine learning can take full advantage of available geophysical information to directly build the nonlinear mapping for predicting porosity from seismic data. To realize the one-step reservoir porosity estimation, we propose the semi-supervised recurrent neural networks (SSRNNs) based porosity modelling method. SSRNNs include an encoder subnet and a decoder subnet. The encoder simulates the generalized seismic inversion to convert the input post-stack seismic data into the predicted porosity, and the decoder acts as a forward model to make the predicted porosity can return to the generated seismic data and reduce resolution space. In addition, seismic data at the non-well positions are randomly selected in each iteration of SSRNNs to boost the lateral continuity of the predicted porosity result. Without the demand of some approximate assumptions and accurate elastic parameters, well logs and seismic data at well locations and non-well locations are integrated into SSRNNs to directly predict high-precision porosity from seismic data. A numerical model example and a real data example are used to verify the effectiveness and accuracy of the SSRNNs based reservoir lateral porosity prediction method.","PeriodicalId":265130,"journal":{"name":"82nd EAGE Annual Conference & Exhibition","volume":"215 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134539732","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-18DOI: 10.3997/2214-4609.202113155
A. Paprouschi, M. Aliahmadi
Summary Implementing conformance treatments has been essential for enhanced oil recovery in recent years. Reducing water production becomes an issue for the oil industries when it competes directly with oil rates. Enhanced Preformed Particle Gels (PPGs) are one of the superabsorbent materials to shut off high permeable zones and promote flood sweep improvement. This research aims to evaluate the behaviour of in-house synthesized silicate preformed particle gels (SPPGs) through Hele-Shaw cell as a dynamic test. The experimental results were validated with the model and implemented into a reservoir simulator called UTCHEM. Various injection rates were injected in the silicate gel model of the modified simulator. We obtain that the simulation results contain assumption for the silicate gel injection are compatible with the experimental values obtained from injections of SPPG through Hele-Shaw Cell. The results indicate that gel injection pressure increases with flow rate. After Gel placement, results showed that water flood pressure decreases as the fracture width increases from 0.5 mm to 1 mm. Simulation results indicated a 21% pressure increase while using silicate gel.
{"title":"Dynamic Investigation of Silicate Preformed Particle GELS (SPPGS) Using Transparent Fracture Model (Hele-Shaw Cell) and Simulator","authors":"A. Paprouschi, M. Aliahmadi","doi":"10.3997/2214-4609.202113155","DOIUrl":"https://doi.org/10.3997/2214-4609.202113155","url":null,"abstract":"Summary Implementing conformance treatments has been essential for enhanced oil recovery in recent years. Reducing water production becomes an issue for the oil industries when it competes directly with oil rates. Enhanced Preformed Particle Gels (PPGs) are one of the superabsorbent materials to shut off high permeable zones and promote flood sweep improvement. This research aims to evaluate the behaviour of in-house synthesized silicate preformed particle gels (SPPGs) through Hele-Shaw cell as a dynamic test. The experimental results were validated with the model and implemented into a reservoir simulator called UTCHEM. Various injection rates were injected in the silicate gel model of the modified simulator. We obtain that the simulation results contain assumption for the silicate gel injection are compatible with the experimental values obtained from injections of SPPG through Hele-Shaw Cell. The results indicate that gel injection pressure increases with flow rate. After Gel placement, results showed that water flood pressure decreases as the fracture width increases from 0.5 mm to 1 mm. Simulation results indicated a 21% pressure increase while using silicate gel.","PeriodicalId":265130,"journal":{"name":"82nd EAGE Annual Conference & Exhibition","volume":"192 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132291701","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-18DOI: 10.3997/2214-4609.202113294
C. Jin, D. Cao, B. Zhou
Summary Quality factor (Q-factor) evaluates the attenuation of seismic wave propagation, playing a fundamental role of reservoir characterization, which can be obtained accurately from Vertical Seismic Profile (VSP). The common methods usually use the downgoing wavefields in VSP data. However, the downgoing wavefields consist of more than 90% energy of the spectrum of the VSP data due to the energy fraction of the upgoing and downgoing wavefields, which makes difficult to estimate the viscoacoustic parameters accurately. Thus, a joint viscoacoustic waveform inversion of velocity and Q-factor is proposed to measure the difference between the separated upgoing and downgoing wavefields in VSP data based on the multi-objective functions. A simple separating step is accomplished by the reflectivity method to obtain the pure individual wavefields in VSP data, and then a joint inversion step is carried out to make full use of the information of the individual wavefields and improve the convergence of viscoacoustic waveform inversion. The sensitivity analysis about the velocity and Q-factor shows that the upgoing and downgoing wavefields contribute differently to the viscoacoustic parameters. Numerical examples and a field test indicate the accuracy and efficiency of the proposed method.
{"title":"Velocity and Q estimation from the separated upgoing and downgoing wavefields in VSP data","authors":"C. Jin, D. Cao, B. Zhou","doi":"10.3997/2214-4609.202113294","DOIUrl":"https://doi.org/10.3997/2214-4609.202113294","url":null,"abstract":"Summary Quality factor (Q-factor) evaluates the attenuation of seismic wave propagation, playing a fundamental role of reservoir characterization, which can be obtained accurately from Vertical Seismic Profile (VSP). The common methods usually use the downgoing wavefields in VSP data. However, the downgoing wavefields consist of more than 90% energy of the spectrum of the VSP data due to the energy fraction of the upgoing and downgoing wavefields, which makes difficult to estimate the viscoacoustic parameters accurately. Thus, a joint viscoacoustic waveform inversion of velocity and Q-factor is proposed to measure the difference between the separated upgoing and downgoing wavefields in VSP data based on the multi-objective functions. A simple separating step is accomplished by the reflectivity method to obtain the pure individual wavefields in VSP data, and then a joint inversion step is carried out to make full use of the information of the individual wavefields and improve the convergence of viscoacoustic waveform inversion. The sensitivity analysis about the velocity and Q-factor shows that the upgoing and downgoing wavefields contribute differently to the viscoacoustic parameters. Numerical examples and a field test indicate the accuracy and efficiency of the proposed method.","PeriodicalId":265130,"journal":{"name":"82nd EAGE Annual Conference & Exhibition","volume":"78 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114918999","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-18DOI: 10.3997/2214-4609.202112438
X. Li, J. Shen, J. Wei, J. Li, X. Yang, K. Wang
Summary In order to the fine evaluation on pore structure in fractured-vuggy reservoirs, the mathematical morphology method combined with singular spectrum interpolation is proposed to automatically extract fractures and vugs based on the electric imaging logging data with the high coverage rate and resolution. The singular spectrum interpolation is applied to reconstruct the full borehole electric logging images in spatial domain by calculating the low-rank conductivity matrix. For implementing the edge detection of the conductivity anomalies and constructing the fractured-vuggy pore structure spectrum, the structural elements on different scales and configurations are selected to perform various kinds of morphological filtering operators. The fusion technology combining mathematical morphology and singular spectrum interpolation is utilized to quantitatively characterize fractures and vugs in carbonate reservoirs in the east of the right bank of Amu Darya Basin. The results show that the novel fusion method can not only recognize fractures and vugs automatically, but also help to improve the efficiency of electric imaging logging data processing.
{"title":"Electric Imaging Fusion Technique Combining Singular Spectrum Interpolation with Mathematical Morphology in Amu Darya Basin","authors":"X. Li, J. Shen, J. Wei, J. Li, X. Yang, K. Wang","doi":"10.3997/2214-4609.202112438","DOIUrl":"https://doi.org/10.3997/2214-4609.202112438","url":null,"abstract":"Summary In order to the fine evaluation on pore structure in fractured-vuggy reservoirs, the mathematical morphology method combined with singular spectrum interpolation is proposed to automatically extract fractures and vugs based on the electric imaging logging data with the high coverage rate and resolution. The singular spectrum interpolation is applied to reconstruct the full borehole electric logging images in spatial domain by calculating the low-rank conductivity matrix. For implementing the edge detection of the conductivity anomalies and constructing the fractured-vuggy pore structure spectrum, the structural elements on different scales and configurations are selected to perform various kinds of morphological filtering operators. The fusion technology combining mathematical morphology and singular spectrum interpolation is utilized to quantitatively characterize fractures and vugs in carbonate reservoirs in the east of the right bank of Amu Darya Basin. The results show that the novel fusion method can not only recognize fractures and vugs automatically, but also help to improve the efficiency of electric imaging logging data processing.","PeriodicalId":265130,"journal":{"name":"82nd EAGE Annual Conference & Exhibition","volume":"125 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114643207","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-18DOI: 10.3997/2214-4609.202011567
M. Nami, M. Ahmadi
Summary Microbial enhanced oil recovery (MEOR) relies on the microbial metabolisms yielding surfactant and polymer. Surfactant production decreases the interfacial tension and mobilizes the remained oil. On the other hand, polymer production increases the water viscosity and consequently transports the injected fluid into unswept zones as it plugs the swept zones of reservoir. Several phenomena are involved in a MEOR process, including growth and decay of bacteria, consumption of substrate, metabolite production and adsorption/desorption of bacteria on/from rock surface. Lots of complexities are introduced in any attempt to create a mathematical model of MEOR, especially when multiphase flow is modelled in multiple dimensions. Our multiphase multidimensional model for MEOR can handle these complex phenomena. Equations of MEOR multi-physics model include convection / diffusion equations, black oil model, interfacial tension reduction equations, relative permeability alteration model, absolute permeability reduction equations and viscosity enhancement equations. System of multi-physics equations have been discretized using control volume finite difference method and then solved with a fully implicit approach. Implemented model can properly represent the transportation and metabolism of microbe in porous media and provides reliable predictions of improvement in oil recovery due to microbial activities.
{"title":"Multiphase, Multidimensional and Multiphysics (M3) Modeling and Simulation of Microbial Enhanced Oil Recovery Process","authors":"M. Nami, M. Ahmadi","doi":"10.3997/2214-4609.202011567","DOIUrl":"https://doi.org/10.3997/2214-4609.202011567","url":null,"abstract":"Summary Microbial enhanced oil recovery (MEOR) relies on the microbial metabolisms yielding surfactant and polymer. Surfactant production decreases the interfacial tension and mobilizes the remained oil. On the other hand, polymer production increases the water viscosity and consequently transports the injected fluid into unswept zones as it plugs the swept zones of reservoir. Several phenomena are involved in a MEOR process, including growth and decay of bacteria, consumption of substrate, metabolite production and adsorption/desorption of bacteria on/from rock surface. Lots of complexities are introduced in any attempt to create a mathematical model of MEOR, especially when multiphase flow is modelled in multiple dimensions. Our multiphase multidimensional model for MEOR can handle these complex phenomena. Equations of MEOR multi-physics model include convection / diffusion equations, black oil model, interfacial tension reduction equations, relative permeability alteration model, absolute permeability reduction equations and viscosity enhancement equations. System of multi-physics equations have been discretized using control volume finite difference method and then solved with a fully implicit approach. Implemented model can properly represent the transportation and metabolism of microbe in porous media and provides reliable predictions of improvement in oil recovery due to microbial activities.","PeriodicalId":265130,"journal":{"name":"82nd EAGE Annual Conference & Exhibition","volume":"105 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134204163","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-18DOI: 10.3997/2214-4609.202112544
G. Dam, M. Soenderholm, A. Mathiesen, U. Gregersen, J. Bojesen‐Koefoed, L. Kristensen, E. Willerslev, F. Moerk, T. Varming, M. Brandt, T. Kristensen, P. Ventris
Summary A play-based Yet-to-Find resource assessment of conventional hydrocarbons has been carried out for the West Greenland continental shelf that constitutes one of the last huge frontier areas of the World. The basin fill is divided into six main tectono-stratigraphic phases and eight play intervals. Source rock intervals include Ordovician, Albian, Cenomanian-Turonian, Campanian and Paleocene-Eocene. Reservoir rocks are present at virtually all stratigraphic levels. High-quality regional seals are well documented from all play intervals. Volume estimates for more than 152 structural leads have been integrated into the play analysis and the identified prospectivity has been calculated. The Yet-to-Find analysis is based on a feature (lead) density calculation approach for each of the identified play intervals calibrated with data from the most extensively explored areas (analogue areas). Based on these analogue areas the unidentified prospectivity has been calculated for the underexplored areas. Having calculated both identified and unidentified prospectivity, the roll-up of all play intervals provide the Total Mean Case Risked Recoverable MMBOE. The total Mean risked recoverable for AU1 is 5500 MMBOE, for AU2 9100 MMBOE and for AU3 2800 MMBOE. A final portfolio analysis shows which areas of the West Greenland continental margin are the most prospective for future exploration.
{"title":"Play-Based Yet-to-Find Resource Assessment of the West Greenland Continental Shelf","authors":"G. Dam, M. Soenderholm, A. Mathiesen, U. Gregersen, J. Bojesen‐Koefoed, L. Kristensen, E. Willerslev, F. Moerk, T. Varming, M. Brandt, T. Kristensen, P. Ventris","doi":"10.3997/2214-4609.202112544","DOIUrl":"https://doi.org/10.3997/2214-4609.202112544","url":null,"abstract":"Summary A play-based Yet-to-Find resource assessment of conventional hydrocarbons has been carried out for the West Greenland continental shelf that constitutes one of the last huge frontier areas of the World. The basin fill is divided into six main tectono-stratigraphic phases and eight play intervals. Source rock intervals include Ordovician, Albian, Cenomanian-Turonian, Campanian and Paleocene-Eocene. Reservoir rocks are present at virtually all stratigraphic levels. High-quality regional seals are well documented from all play intervals. Volume estimates for more than 152 structural leads have been integrated into the play analysis and the identified prospectivity has been calculated. The Yet-to-Find analysis is based on a feature (lead) density calculation approach for each of the identified play intervals calibrated with data from the most extensively explored areas (analogue areas). Based on these analogue areas the unidentified prospectivity has been calculated for the underexplored areas. Having calculated both identified and unidentified prospectivity, the roll-up of all play intervals provide the Total Mean Case Risked Recoverable MMBOE. The total Mean risked recoverable for AU1 is 5500 MMBOE, for AU2 9100 MMBOE and for AU3 2800 MMBOE. A final portfolio analysis shows which areas of the West Greenland continental margin are the most prospective for future exploration.","PeriodicalId":265130,"journal":{"name":"82nd EAGE Annual Conference & Exhibition","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134210430","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-18DOI: 10.3997/2214-4609.202113247
S. Namie, Dongmei Wang, Z. Yin, G. Radu, N. Samson, A. Dandekar, D. Cercone, J. Ciferno, W. Xindan
Summary History Match Relative Permeability Polymer Flood EOR Milne Point Low salinity water flooding Field-scale simulation model
拟合相对渗透率聚合物驱提高采收率Milne Point低矿化度水驱现场规模模拟模型
{"title":"Challenge Solutions for a Significant Water Cut Reduction History Match on Heavy Oil Polymer EOR","authors":"S. Namie, Dongmei Wang, Z. Yin, G. Radu, N. Samson, A. Dandekar, D. Cercone, J. Ciferno, W. Xindan","doi":"10.3997/2214-4609.202113247","DOIUrl":"https://doi.org/10.3997/2214-4609.202113247","url":null,"abstract":"Summary History Match Relative Permeability Polymer Flood EOR Milne Point Low salinity water flooding Field-scale simulation model","PeriodicalId":265130,"journal":{"name":"82nd EAGE Annual Conference & Exhibition","volume":"297 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134501118","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-18DOI: 10.3997/2214-4609.202010574
Q. Li, P. Duan, G. Wu
Summary Reverse time migration has to store both forward- and backward-propagated wavefield which cost a large of memory, especially in elastic world. Such problems can be solved by wavefield reconstruction method. To reconstruct source wavefield with high spacing accuracy, tradition methods still cost more computer storage. In this study, we propose a 3D elastic wavefield reconstruction method based on optimal operator boundary storage strategy, which cost less storage and reconstruct source wavefield with high spacing accuracy same as previous methods. Our algorithm is success in accurately reconstruct elastic wavefield and reducing 80% of the storage.
{"title":"3D Elastic Wavefield Reconstruction Method Based on Optimal Operator Boundary Storage Strategy","authors":"Q. Li, P. Duan, G. Wu","doi":"10.3997/2214-4609.202010574","DOIUrl":"https://doi.org/10.3997/2214-4609.202010574","url":null,"abstract":"Summary Reverse time migration has to store both forward- and backward-propagated wavefield which cost a large of memory, especially in elastic world. Such problems can be solved by wavefield reconstruction method. To reconstruct source wavefield with high spacing accuracy, tradition methods still cost more computer storage. In this study, we propose a 3D elastic wavefield reconstruction method based on optimal operator boundary storage strategy, which cost less storage and reconstruct source wavefield with high spacing accuracy same as previous methods. Our algorithm is success in accurately reconstruct elastic wavefield and reducing 80% of the storage.","PeriodicalId":265130,"journal":{"name":"82nd EAGE Annual Conference & Exhibition","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129387827","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-10-18DOI: 10.3997/2214-4609.202113304
T. Konuk, J. Shragge
Summary Many real-world seismic modeling and imaging applications require computing frequency-domain numerical solutions of acoustic wave equation (AWE). However, obtaining such solutions in media characterized by strong parameter contrasts and anisotropy poses significant practical challenges to existing numerical solvers, especially for 3D scenarios. Physics-informed neural networks (PINN) provide a computationally efficient alternative approach for AWE solutions. However, PINNs solve only a single instance of AWE and need to be re-trained for each different subsurface models and frequencies. Fourier neural operators, on the other hand, can solve AWE for a wide range of models and frequencies with a single set of network configuration and parameters. This method, though, requires a tremendous amount of data, which can be difficult and expensive to obtain. Here, we propose a methodology that combines PINNs with Fourier neural operators to learn AWE solution operators that are valid for a wide range of frequencies without requiring any training data. We present two numerical examples that demonstrate the capabilities of the proposed method in modeling the acoustic wavefield accurately and efficiently in the frequency domain.
{"title":"Physics-guided deep learning using Fourier neural operators for solving the acoustic VTI wave equation","authors":"T. Konuk, J. Shragge","doi":"10.3997/2214-4609.202113304","DOIUrl":"https://doi.org/10.3997/2214-4609.202113304","url":null,"abstract":"Summary Many real-world seismic modeling and imaging applications require computing frequency-domain numerical solutions of acoustic wave equation (AWE). However, obtaining such solutions in media characterized by strong parameter contrasts and anisotropy poses significant practical challenges to existing numerical solvers, especially for 3D scenarios. Physics-informed neural networks (PINN) provide a computationally efficient alternative approach for AWE solutions. However, PINNs solve only a single instance of AWE and need to be re-trained for each different subsurface models and frequencies. Fourier neural operators, on the other hand, can solve AWE for a wide range of models and frequencies with a single set of network configuration and parameters. This method, though, requires a tremendous amount of data, which can be difficult and expensive to obtain. Here, we propose a methodology that combines PINNs with Fourier neural operators to learn AWE solution operators that are valid for a wide range of frequencies without requiring any training data. We present two numerical examples that demonstrate the capabilities of the proposed method in modeling the acoustic wavefield accurately and efficiently in the frequency domain.","PeriodicalId":265130,"journal":{"name":"82nd EAGE Annual Conference & Exhibition","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134305054","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：