O. Andersen, M. Kelley, V. Smith, S. Raziperchikolaee
In this study, we demonstrate geomechanical modeling with fully automatic parameter calibration to estimate the full geomechanical stress fields of a prospective US CO2 storage site, based on sparse measurement data. The goal is to compute full stress tensor field estimates (principal stresses and orientations) that are maximally compatible with observations within the constraints of the model assumptions, thereby extending point-wise, incomplete partial stress measurement to a simulated full formation stress field, as well as a rough assessment of the associated error. We use the Perch site, located in Otsego Country, Michigan, as our case study. Input data consists of partial stress tensor information inferred from in-situ borehole tests, geophysical well logs and processing of seismic data. A static earth model of the site was developed, and geomechanical simulation functionality of the open-source MATLAB Reservoir Simulation Toolbox (MRST) used to model the stress field. Adjoint-based nonlinear optimization was used to adjust boundary conditions and material properties to calibrate simulated results to observations. Results were interpreted through a Bayesian framework. The focus of this article is to demonstrate how the fully automatic calibration procedure works and discuss the results obtained but does not attempt a detailed analysis of the stress field in the context of the proposed CO2 storage initiatives. Our work is part of a larger effort to non-invasively determine in-situ stresses in deep formations considered for CO2 storage. Guided by previously published research on geomechanical model calibration, our work presents a novel calibration approach supporting a potentially large number of linear or nonlinear calibration parameters, in order to produce results optimally agreeing with available measurements and thus extend partial point-wise estimates to full tensor fields compatible with the physics of the site.
{"title":"Automatic Calibration of a Geomechanical Model from Sparse Data for Estimating Stress in Deep Geological Formations","authors":"O. Andersen, M. Kelley, V. Smith, S. Raziperchikolaee","doi":"10.2118/204006-ms","DOIUrl":"https://doi.org/10.2118/204006-ms","url":null,"abstract":"\u0000 In this study, we demonstrate geomechanical modeling with fully automatic parameter calibration to estimate the full geomechanical stress fields of a prospective US CO2 storage site, based on sparse measurement data. The goal is to compute full stress tensor field estimates (principal stresses and orientations) that are maximally compatible with observations within the constraints of the model assumptions, thereby extending point-wise, incomplete partial stress measurement to a simulated full formation stress field, as well as a rough assessment of the associated error. We use the Perch site, located in Otsego Country, Michigan, as our case study. Input data consists of partial stress tensor information inferred from in-situ borehole tests, geophysical well logs and processing of seismic data. A static earth model of the site was developed, and geomechanical simulation functionality of the open-source MATLAB Reservoir Simulation Toolbox (MRST) used to model the stress field. Adjoint-based nonlinear optimization was used to adjust boundary conditions and material properties to calibrate simulated results to observations. Results were interpreted through a Bayesian framework.\u0000 The focus of this article is to demonstrate how the fully automatic calibration procedure works and discuss the results obtained but does not attempt a detailed analysis of the stress field in the context of the proposed CO2 storage initiatives. Our work is part of a larger effort to non-invasively determine in-situ stresses in deep formations considered for CO2 storage. Guided by previously published research on geomechanical model calibration, our work presents a novel calibration approach supporting a potentially large number of linear or nonlinear calibration parameters, in order to produce results optimally agreeing with available measurements and thus extend partial point-wise estimates to full tensor fields compatible with the physics of the site.","PeriodicalId":11146,"journal":{"name":"Day 1 Tue, October 26, 2021","volume":"43 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76278788","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}
Recently, the elevated levels of seismicity activities in Western Canada have been demonstrated to be linked to hydraulic fracturing operations that developed unconventional resources. The underlying triggering mechanisms of hydraulic fracturing-induced seismicity are still uncertain. The interactions of well stimulation and geology-geomechanical-hydrological features need to be investigated comprehensively. The linear poroelasticity theory was utilized to guide coupled poroelastic modeling and to quantify the physical process during hydraulic fracturing. The integrated analysis is first conducted to characterize the mechanical features and fluid flow behavior. The finite-element simulation is then conducted by coupling Darcy's law and solid mechanics to quantify the perturbation of pore pressure and poroelastic stress in the seismogenic fault zone. Finally, the Mohr-coulomb failure criterion is utilized to determine the spatial-temporal faults activation and reveal the trigger mechanisms of induced earthquakes. The mitigation strategy was proposed accordingly to reduce the potential seismic hazards near this region. A case study of ML 4.18 earthquake in the East Shale Basin was utilized to demonstrate the applicability of the coupled modeling and numerical simulation. Results showed that one inferred fault cut through the Duvernay formation with the strike of NE20°. The fracture half-length of two wells owns an average value of 124 m. The brittleness index deriving from the velocity logging data was estimated to be a relatively higher value in the Duvernay formation, indicating a geomechanical bias of stimulated formation for the fault activation. The coupled poroelastic simulation was conducted, showing that the hydrologic connection between seismogenic faults and stimulated well was established by the end of the 38th stage completion for the east horizontal well. The simulated coulomb failure stress surrounding the fault reached a maximum of 4.15 MPa, exceeding the critical value to cause the fault slip. Hence the poroelastic effects on the inferred fault were responsible for the fault activation and triggered the subsequent ML 4.18 earthquake. It is essential to optimize the stimulation site selection near the existing faults to reduce risks of future seismic hazards near the East Shale Basin.
{"title":"Coupled Poroelastic Modeling to Characterize the 4.18-Magnitude Earthquake Due to Hydraulic Fracturing in the East Shale Basin of Western Canada","authors":"Gang Hui, Shengnan Chen, F. Gu","doi":"10.2118/203921-ms","DOIUrl":"https://doi.org/10.2118/203921-ms","url":null,"abstract":"\u0000 Recently, the elevated levels of seismicity activities in Western Canada have been demonstrated to be linked to hydraulic fracturing operations that developed unconventional resources. The underlying triggering mechanisms of hydraulic fracturing-induced seismicity are still uncertain. The interactions of well stimulation and geology-geomechanical-hydrological features need to be investigated comprehensively. The linear poroelasticity theory was utilized to guide coupled poroelastic modeling and to quantify the physical process during hydraulic fracturing. The integrated analysis is first conducted to characterize the mechanical features and fluid flow behavior. The finite-element simulation is then conducted by coupling Darcy's law and solid mechanics to quantify the perturbation of pore pressure and poroelastic stress in the seismogenic fault zone. Finally, the Mohr-coulomb failure criterion is utilized to determine the spatial-temporal faults activation and reveal the trigger mechanisms of induced earthquakes. The mitigation strategy was proposed accordingly to reduce the potential seismic hazards near this region. A case study of ML 4.18 earthquake in the East Shale Basin was utilized to demonstrate the applicability of the coupled modeling and numerical simulation. Results showed that one inferred fault cut through the Duvernay formation with the strike of NE20°. The fracture half-length of two wells owns an average value of 124 m. The brittleness index deriving from the velocity logging data was estimated to be a relatively higher value in the Duvernay formation, indicating a geomechanical bias of stimulated formation for the fault activation. The coupled poroelastic simulation was conducted, showing that the hydrologic connection between seismogenic faults and stimulated well was established by the end of the 38th stage completion for the east horizontal well. The simulated coulomb failure stress surrounding the fault reached a maximum of 4.15 MPa, exceeding the critical value to cause the fault slip. Hence the poroelastic effects on the inferred fault were responsible for the fault activation and triggered the subsequent ML 4.18 earthquake. It is essential to optimize the stimulation site selection near the existing faults to reduce risks of future seismic hazards near the East Shale Basin.","PeriodicalId":11146,"journal":{"name":"Day 1 Tue, October 26, 2021","volume":"114 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74851350","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}
In this paper, we introduce an efficient method to generate reservoir simulation grids and modify the fault juxtaposition on the generated grids. Both processes are based on a mapping method to displace vertices of a grid to desired locations without changing the grid topology. In the gridding process, a grid that can capture stratigraphical complexity is first generated in an unfaulted space. The vertices of the grid are then displaced back to the original faulted space to become a reservoir simulation grid. The resulting reversely mapped grid has a mapping structure that allows fast and easy fault juxtaposition modification. This feature avoids the process of updating the structural framework and regenerating the reservoir properties, which may be time-consuming. To facilitate juxtaposition updates within an assisted history matching workflow, several parameterized fault throw adjustment methods are introduced. Grid examples are given for reservoirs with Y-faults, overturned bed, and complex channel-lobe systems.
{"title":"A Fast Gridding Method for Capturing Geological Complexity and Uncertainty","authors":"Xu Yifei, Priyesh Srivastava, Xiao Ma, Karan Kaul, Hao Huang","doi":"10.2118/203902-ms","DOIUrl":"https://doi.org/10.2118/203902-ms","url":null,"abstract":"\u0000 In this paper, we introduce an efficient method to generate reservoir simulation grids and modify the fault juxtaposition on the generated grids. Both processes are based on a mapping method to displace vertices of a grid to desired locations without changing the grid topology. In the gridding process, a grid that can capture stratigraphical complexity is first generated in an unfaulted space. The vertices of the grid are then displaced back to the original faulted space to become a reservoir simulation grid. The resulting reversely mapped grid has a mapping structure that allows fast and easy fault juxtaposition modification. This feature avoids the process of updating the structural framework and regenerating the reservoir properties, which may be time-consuming. To facilitate juxtaposition updates within an assisted history matching workflow, several parameterized fault throw adjustment methods are introduced. Grid examples are given for reservoirs with Y-faults, overturned bed, and complex channel-lobe systems.","PeriodicalId":11146,"journal":{"name":"Day 1 Tue, October 26, 2021","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90336672","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}
R. Santoso, Xupeng He, M. AlSinan, H. Kwak, H. Hoteit
History matching is critical in subsurface flow modeling. It is to align the reservoir model with the measured data. However, it remains challenging since the solution is not unique and the implementation is expensive. The traditional approach relies on trial and error, which are exhaustive and labor-intensive. In this study, we propose a new workflow utilizing Bayesian Markov Chain Monte Carlo (MCMC) to automatically and accurately perform history matching. We deliver four novelties within the workflow: 1) the use of multi-resolution low-fidelity models to guarantee high-quality matching, 2) updating the ranges of priors to assure convergence, 3) the use of Long-Short Term Memory (LSTM) network as a low-fidelity model to produce continuous time-response, and 4) the use of Bayesian optimization to obtain the optimum low-fidelity model for Bayesian MCMC runs. We utilize the first SPE comparative model as the physical and high-fidelity model. It is a gas injection into an oil reservoir case, which is the gravity-dominated process. The coarse low-fidelity model manages to provide updated priors that increase the precision of Bayesian MCMC. The Bayesian-optimized LSTM has successfully captured the physics in the high-fidelity model. The Bayesian-LSTM MCMC produces an accurate prediction with narrow uncertainties. The posterior prediction through the high-fidelity model ensures the robustness and precision of the workflow. This approach provides an efficient and high-quality history matching for subsurface flow modeling.
{"title":"Bayesian Long-Short Term Memory for History Matching in Reservoir Simulations","authors":"R. Santoso, Xupeng He, M. AlSinan, H. Kwak, H. Hoteit","doi":"10.2118/203976-ms","DOIUrl":"https://doi.org/10.2118/203976-ms","url":null,"abstract":"\u0000 History matching is critical in subsurface flow modeling. It is to align the reservoir model with the measured data. However, it remains challenging since the solution is not unique and the implementation is expensive. The traditional approach relies on trial and error, which are exhaustive and labor-intensive. In this study, we propose a new workflow utilizing Bayesian Markov Chain Monte Carlo (MCMC) to automatically and accurately perform history matching. We deliver four novelties within the workflow: 1) the use of multi-resolution low-fidelity models to guarantee high-quality matching, 2) updating the ranges of priors to assure convergence, 3) the use of Long-Short Term Memory (LSTM) network as a low-fidelity model to produce continuous time-response, and 4) the use of Bayesian optimization to obtain the optimum low-fidelity model for Bayesian MCMC runs.\u0000 We utilize the first SPE comparative model as the physical and high-fidelity model. It is a gas injection into an oil reservoir case, which is the gravity-dominated process. The coarse low-fidelity model manages to provide updated priors that increase the precision of Bayesian MCMC. The Bayesian-optimized LSTM has successfully captured the physics in the high-fidelity model. The Bayesian-LSTM MCMC produces an accurate prediction with narrow uncertainties. The posterior prediction through the high-fidelity model ensures the robustness and precision of the workflow. This approach provides an efficient and high-quality history matching for subsurface flow modeling.","PeriodicalId":11146,"journal":{"name":"Day 1 Tue, October 26, 2021","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86849083","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}
Drill sequence optimization is a common challenge faced in the oil and gas industry and yet it cannot be solved efficiently by existing optimization methods due to its unique features and constraints. For many fields, the drill queue is currently designed manually based on engineering heuristics. In this paper, a heuristic priority function is combined with traditional optimizers to boost the optimization efficiency at a lower computational cost to speed up the decision-making process. The heuristic priority function is constructed to map the individual well properties such as well index and inter-well distance to the well priority values. As the name indicates, wells with higher priority values will be drilled earlier in the queue. The heuristic priority function is a comprehensive metric of inter-well communication & displacement efficiency. For example, injectors with fast support to producers or producers with a better chance to drain the unswept region tend to have high scores. It contains components that weigh the different properties of a well. These components are then optimized during the optimization process to generate the beneficial drill sequences. Embedded with reservoir engineering heuristics, the priority function helps the optimizer focus on exploring scenarios with promising outcomes. The proposed heuristic priority function, combined with the Genetic Algorithm (GA), has been tested through drill sequence optimization problems for the Brugge field and Olympus field. Optimizations that are directly performed on the drill sequence are employed as reference cases. Different continu- ous/categorical parameterization schemes and various forms of heuristic priority functions are also investigated. Our exploration reveals that the heuristic priority function including well type, constraints, well index, distance to existing wells, and adjacent oil in place yields the best outcome. The proposed approach was able to achieve a better optimization starting point (∼5-18% improvement due to more reasonable drill sequence rather than random guess), a faster convergence rate (results stabilized at 12 vs. 30 iterations), and a lower computational cost (150-250 vs. 1,300 runs to achieve the same NPV) over the reference methods. Similar performance improvement was also observed in another application to a North Sea type reservoir. This demonstrated the general applicability of the proposed method. The employment of the heuristic priority function improves the efficiency and reliability of drill sequence optimization compared to the traditional methods that directly optimize the sequence. It can be easily embedded in either commercial or research simulators as an independent module. In addition, it is also an automatic process that fits well with iterative optimization algorithms.
{"title":"Efficient Drill Sequence Optimization Using a Heuristic Priority Function","authors":"Zhenzhen Wang, Jincong He, Shusei Tanaka, X. Wen","doi":"10.2118/203986-ms","DOIUrl":"https://doi.org/10.2118/203986-ms","url":null,"abstract":"\u0000 Drill sequence optimization is a common challenge faced in the oil and gas industry and yet it cannot be solved efficiently by existing optimization methods due to its unique features and constraints. For many fields, the drill queue is currently designed manually based on engineering heuristics. In this paper, a heuristic priority function is combined with traditional optimizers to boost the optimization efficiency at a lower computational cost to speed up the decision-making process.\u0000 The heuristic priority function is constructed to map the individual well properties such as well index and inter-well distance to the well priority values. As the name indicates, wells with higher priority values will be drilled earlier in the queue. The heuristic priority function is a comprehensive metric of inter-well communication & displacement efficiency. For example, injectors with fast support to producers or producers with a better chance to drain the unswept region tend to have high scores. It contains components that weigh the different properties of a well. These components are then optimized during the optimization process to generate the beneficial drill sequences. Embedded with reservoir engineering heuristics, the priority function helps the optimizer focus on exploring scenarios with promising outcomes.\u0000 The proposed heuristic priority function, combined with the Genetic Algorithm (GA), has been tested through drill sequence optimization problems for the Brugge field and Olympus field. Optimizations that are directly performed on the drill sequence are employed as reference cases. Different continu- ous/categorical parameterization schemes and various forms of heuristic priority functions are also investigated. Our exploration reveals that the heuristic priority function including well type, constraints, well index, distance to existing wells, and adjacent oil in place yields the best outcome. The proposed approach was able to achieve a better optimization starting point (∼5-18% improvement due to more reasonable drill sequence rather than random guess), a faster convergence rate (results stabilized at 12 vs. 30 iterations), and a lower computational cost (150-250 vs. 1,300 runs to achieve the same NPV) over the reference methods. Similar performance improvement was also observed in another application to a North Sea type reservoir. This demonstrated the general applicability of the proposed method.\u0000 The employment of the heuristic priority function improves the efficiency and reliability of drill sequence optimization compared to the traditional methods that directly optimize the sequence. It can be easily embedded in either commercial or research simulators as an independent module. In addition, it is also an automatic process that fits well with iterative optimization algorithms.","PeriodicalId":11146,"journal":{"name":"Day 1 Tue, October 26, 2021","volume":"63 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82153204","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}
Hui Zhao, Wei Liu, Xiang Rao, Guanglong Sheng, H. Li, Zhenyu Guo, Deng Liu, Lin Cao
The data-driven interwell simulation model (INSIM) has been recognized as an effective tool for history matching and interwell-connectivity characterization of waterflooding reservoirs. INSIM-FT-3D (FT: front tracking) was recently developed to upgrade the applicationdimension of INSIM series data-driven models from two-dimensional (2D) to three-dimensional (3D). However, INSIM-FT-3D cannot accurately infer the dynamic change of well-connectivity and predict well's bottom-hole pressure (BHP). The main purpose of this study intends to expand the capability of INSIM-FT-3D to empower for the assimilation of BHPs, the reliable prediction of water breakthrough and the characterization of dynamic interwell-connectivities. The default setting of well index (WI) in INSIM-FT-3D based on Peaceman's equation does not yield accurate BHP estimates. We derive a WI that can honor the BHPs of a reference model composed of a set of 1D connections. When history matching BHPs of a 3D reservoir, we show that the derived WI is a better initial guess than that obtained from Peaceman's equation. We also develop a flow-path-tracking (FPT) algorithm to calculate the dynamic interwell properties (allocation factors and pore volumes (PVs)). Besides, we discuss the relationship between the INSIM-family methods and the traditional grid-based methods, which indicates that the INSIM-family methods can calculate the transmissibility of the connection between coarse-scale cells in a more accurate manner. As an improvement of INSIM-FT-3D, the newly proposed data-driven model is denoted as INSIM-FPT-3D. To verify the correctness of the derived WI, we present a 1D problem and a T-shaped synthetic reservoir simulation model as the reference models. BHPs and oil production rates are obtained as the observed data by running these two reference models with total injection/production-rate controls. An INSIM-FPT-3D model is created by specifying the transmissibilities and PVs that are the same as those in the reference model. By applying the derived WIs in INSIM-FPT-3D, the resulting BHPs and oil rates obtained agree well with the reference model without further model calibration. Applying INSIM-FPT-3D to a synthetic multi-layered reservoir shows that we obtain a reasonable match of both BHPs and oil rates with INSIM-FPT-3D. Compared with the FrontSim model, the INSIM-FPT-3D model after history matching is shown to match the dynamic PVs from FrontSim reasonably well and can correctly predict the timing of water breakthrough. By allowing for the assimilation of BHP data, we enable INSIM-FPT-3D to history match a green field with limited production history and forecast the timing of water breakthrough. The improved INSIM-FPT-3D leads to more accurate characterization of the interwell connectivities.
数据驱动井间模拟模型(INSIM)已被认为是水驱油藏历史拟合和井间连通性表征的有效工具。最近开发的INSIM-FT-3D (FT: front tracking)是为了将INSIM系列数据驱动模型的应用维度从二维(2D)提升到三维(3D)。然而，INSIM-FT-3D并不能准确推断井连通性的动态变化和预测井底压力(BHP)。本研究的主要目的是扩大INSIM-FT-3D的能力，以增强BHPs的同化，可靠的水侵预测和动态井间连通性的表征。INSIM-FT-3D中基于Peaceman公式的井指数(WI)的默认设置不能产生准确的BHP估计。我们推导了一个WI，该WI可以满足由一组1D连接组成的参考模型的BHPs。当对3D油藏的BHPs进行历史匹配时，我们发现推导出的WI比Peaceman公式得到的WI更好。我们还开发了一种流动路径跟踪(FPT)算法来计算动态井间特性(分配系数和孔隙体积(pv))。此外，我们还讨论了insim族方法与传统基于网格的方法之间的关系，表明insim族方法可以更准确地计算粗尺度单元之间连接的传递率。作为对INSIM-FT-3D的改进，新提出的数据驱动模型记为INSIM-FPT-3D。为了验证推导的WI的正确性，我们提出了一个一维问题和一个t形综合油藏模拟模型作为参考模型。通过运行这两个参考模型并控制总注入/生产速度，可以获得BHPs和产油量作为观测数据。通过指定与参考模型相同的透射率和pv来创建INSIM-FPT-3D模型。通过在INSIM-FPT-3D中应用推导出的WIs，得到的BHPs和产油速率与参考模型吻合良好，无需进一步校正模型。将INSIM-FPT-3D应用到合成多层油藏中，结果表明，使用INSIM-FPT-3D可以获得合理的BHPs和产油速率匹配。与FrontSim模型相比，经过历史拟合的INSIM-FPT-3D模型能够较好地拟合FrontSim模型的动态pv，并能正确预测见水时间。通过对BHP数据的同化，我们使INSIM-FPT-3D能够对生产历史有限的未开发油田进行历史匹配，并预测见水时间。改进的INSIM-FPT-3D可以更准确地描述井间连通性。
{"title":"INSIM-FPT-3D: A Data-Driven Model for History Matching, Water-Breakthrough Prediction and Well-Connectivity Characterization in Three-Dimensional Reservoirs","authors":"Hui Zhao, Wei Liu, Xiang Rao, Guanglong Sheng, H. Li, Zhenyu Guo, Deng Liu, Lin Cao","doi":"10.2118/203931-ms","DOIUrl":"https://doi.org/10.2118/203931-ms","url":null,"abstract":"\u0000 The data-driven interwell simulation model (INSIM) has been recognized as an effective tool for history matching and interwell-connectivity characterization of waterflooding reservoirs. INSIM-FT-3D (FT: front tracking) was recently developed to upgrade the applicationdimension of INSIM series data-driven models from two-dimensional (2D) to three-dimensional (3D). However, INSIM-FT-3D cannot accurately infer the dynamic change of well-connectivity and predict well's bottom-hole pressure (BHP). The main purpose of this study intends to expand the capability of INSIM-FT-3D to empower for the assimilation of BHPs, the reliable prediction of water breakthrough and the characterization of dynamic interwell-connectivities.\u0000 The default setting of well index (WI) in INSIM-FT-3D based on Peaceman's equation does not yield accurate BHP estimates. We derive a WI that can honor the BHPs of a reference model composed of a set of 1D connections. When history matching BHPs of a 3D reservoir, we show that the derived WI is a better initial guess than that obtained from Peaceman's equation. We also develop a flow-path-tracking (FPT) algorithm to calculate the dynamic interwell properties (allocation factors and pore volumes (PVs)). Besides, we discuss the relationship between the INSIM-family methods and the traditional grid-based methods, which indicates that the INSIM-family methods can calculate the transmissibility of the connection between coarse-scale cells in a more accurate manner. As an improvement of INSIM-FT-3D, the newly proposed data-driven model is denoted as INSIM-FPT-3D.\u0000 To verify the correctness of the derived WI, we present a 1D problem and a T-shaped synthetic reservoir simulation model as the reference models. BHPs and oil production rates are obtained as the observed data by running these two reference models with total injection/production-rate controls. An INSIM-FPT-3D model is created by specifying the transmissibilities and PVs that are the same as those in the reference model. By applying the derived WIs in INSIM-FPT-3D, the resulting BHPs and oil rates obtained agree well with the reference model without further model calibration. Applying INSIM-FPT-3D to a synthetic multi-layered reservoir shows that we obtain a reasonable match of both BHPs and oil rates with INSIM-FPT-3D. Compared with the FrontSim model, the INSIM-FPT-3D model after history matching is shown to match the dynamic PVs from FrontSim reasonably well and can correctly predict the timing of water breakthrough.\u0000 By allowing for the assimilation of BHP data, we enable INSIM-FPT-3D to history match a green field with limited production history and forecast the timing of water breakthrough. The improved INSIM-FPT-3D leads to more accurate characterization of the interwell connectivities.","PeriodicalId":11146,"journal":{"name":"Day 1 Tue, October 26, 2021","volume":"86 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83747618","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}
Supersonic gas injection can help deliver gas uniformly to a reservoir, regardless of reservoir conditions. This technology has played a key role in enhanced oil recovery (EOR) and in particular, thermal enhanced oil recovery operations. Most previous studies have focused on single phase gas injection whereas in most field applications, multiphase and multicomponent situations occur. In the research documented in this paper, we report on results of evaluations of compressible multiphase supersonic gas flows in which gas is the continuous phase is seeded with dispersed liquid droplets or solid particles. Theoretical derivation and numerical simulations with and without relative motions between continuous and disperse phases are examined first. The results illustrate that the shock wave structures and flow properties associated with the multiphase gas flows are different than that of single-phase isentropic flows. The existence and importance of relaxation zones after the normal shock wave in multiphase flow is described. Numerical computational fluid dynamics (CFD) simulations are conducted to show how the multiphase multicomponent flow affects gas phase injection under different conditions. The impact of solid/liquid mass loading on flow performance is discussed. Finally, the practical application of the findings is discussed.
{"title":"Theoretical and Numerical Investigation of Supersonic Multiphase Gas Injection","authors":"Da Zhu, M. Sivagnanam, I. Gates","doi":"10.2118/203911-ms","DOIUrl":"https://doi.org/10.2118/203911-ms","url":null,"abstract":"\u0000 Supersonic gas injection can help deliver gas uniformly to a reservoir, regardless of reservoir conditions. This technology has played a key role in enhanced oil recovery (EOR) and in particular, thermal enhanced oil recovery operations. Most previous studies have focused on single phase gas injection whereas in most field applications, multiphase and multicomponent situations occur.\u0000 In the research documented in this paper, we report on results of evaluations of compressible multiphase supersonic gas flows in which gas is the continuous phase is seeded with dispersed liquid droplets or solid particles. Theoretical derivation and numerical simulations with and without relative motions between continuous and disperse phases are examined first. The results illustrate that the shock wave structures and flow properties associated with the multiphase gas flows are different than that of single-phase isentropic flows. The existence and importance of relaxation zones after the normal shock wave in multiphase flow is described. Numerical computational fluid dynamics (CFD) simulations are conducted to show how the multiphase multicomponent flow affects gas phase injection under different conditions. The impact of solid/liquid mass loading on flow performance is discussed. Finally, the practical application of the findings is discussed.","PeriodicalId":11146,"journal":{"name":"Day 1 Tue, October 26, 2021","volume":"78 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83744455","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}
Soham Sheth, François McKee, K. Neylon, Ghazala Fazil
We present a novel reservoir simulator time-step selection approach which uses machine-learning (ML) techniques to analyze the mathematical and physical state of the system and predict time-step sizes which are large while still being efficient to solve, thus making the simulation faster. An optimal time-step choice avoids wasted non-linear and linear equation set-up work when the time-step is too small and avoids highly non-linear systems that take many iterations to solve. Typical time-step selectors use a limited set of features to heuristically predict the size of the next time-step. While they have been effective for simple simulation models, as model complexity increases, there is an increasing need for robust data-driven time-step selection algorithms. We propose two workflows – static and dynamic – that use a diverse set of physical (e.g., well data) and mathematical (e.g., CFL) features to build a predictive ML model. This can be pre-trained or dynamically trained to generate an inference model. The trained model can also be reinforced as new data becomes available and efficiently used for transfer learning. We present the application of these workflows in a commercial reservoir simulator using distinct types of simulation model including black oil, compositional and thermal steam-assisted gravity drainage (SAGD). We have found that history-match and uncertainty/optimization studies benefit most from the static approach while the dynamic approach produces optimum step-sizes for prediction studies. We use a confidence monitor to manage the ML time-step selector at runtime. If the confidence level falls below a threshold, we switch to traditional heuristic method for that time-step. This avoids any degradation in the performance when the model features are outside the training space. Application to several complex cases, including a large field study, shows a significant speedup for single simulations and even better results for multiple simulations. We demonstrate that any simulation can take advantage of the stored state of the trained model and even augment it when new situations are encountered, so the system becomes more effective as it is exposed to more data.
{"title":"Intelligent Time-Stepping for Practical Numerical Simulation","authors":"Soham Sheth, François McKee, K. Neylon, Ghazala Fazil","doi":"10.2118/204002-ms","DOIUrl":"https://doi.org/10.2118/204002-ms","url":null,"abstract":"\u0000 We present a novel reservoir simulator time-step selection approach which uses machine-learning (ML) techniques to analyze the mathematical and physical state of the system and predict time-step sizes which are large while still being efficient to solve, thus making the simulation faster. An optimal time-step choice avoids wasted non-linear and linear equation set-up work when the time-step is too small and avoids highly non-linear systems that take many iterations to solve.\u0000 Typical time-step selectors use a limited set of features to heuristically predict the size of the next time-step. While they have been effective for simple simulation models, as model complexity increases, there is an increasing need for robust data-driven time-step selection algorithms. We propose two workflows – static and dynamic – that use a diverse set of physical (e.g., well data) and mathematical (e.g., CFL) features to build a predictive ML model. This can be pre-trained or dynamically trained to generate an inference model. The trained model can also be reinforced as new data becomes available and efficiently used for transfer learning.\u0000 We present the application of these workflows in a commercial reservoir simulator using distinct types of simulation model including black oil, compositional and thermal steam-assisted gravity drainage (SAGD). We have found that history-match and uncertainty/optimization studies benefit most from the static approach while the dynamic approach produces optimum step-sizes for prediction studies. We use a confidence monitor to manage the ML time-step selector at runtime. If the confidence level falls below a threshold, we switch to traditional heuristic method for that time-step. This avoids any degradation in the performance when the model features are outside the training space. Application to several complex cases, including a large field study, shows a significant speedup for single simulations and even better results for multiple simulations. We demonstrate that any simulation can take advantage of the stored state of the trained model and even augment it when new situations are encountered, so the system becomes more effective as it is exposed to more data.","PeriodicalId":11146,"journal":{"name":"Day 1 Tue, October 26, 2021","volume":"23 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82812519","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}
Evaluating large basin-scale formations for CO2 sequestration is one of the most important challenges for our industry. The technical complexity and the quantification of risks associated with these operations call for new reservoir engineering and reservoir simulation tools. The impact of multiple coupled physical phenomena, the century timescale, and basin-sized models in these operations force us to completely take apart and revisit the numerical backbone of existing simulation tools. We need a reservoir simulation tool designed for scalability and portability on high-performance computing architectures. To achieve this, we are proposing a new, open-source, multiphysics, and multilevel physics simulation tool called GEOSX. This tool is jointly created by Lawrence Livermore National Laboratory, Stanford University, and Total. It is designed for scalability on multiple CPUs and multiple GPUs and offers a suite of physical solvers that can be extended easily while achieving a balance between performance and portability. GEOSX is initially targeting multiphysics simulations with coupled geomechanics, flow, and transport mechanics but with its open architecture, it allows access to high-performance physical solvers as building blocks of other multiphysics problems and provides users with a suite of tools for numerical optimization across platforms. In this paper, we introduce GEOSX, expose its fundamental architecture principles, and show an example of geological sequestration of CO2 modeling on real data. We demonstrate our ability to simulate fluid and rock poromechanical interactions over long periods and basin-scale dimensions. GEOSX demonstrates its usefulness for such complex and large problems and proves to be scalable and portable across multiple high-performance systems.
{"title":"GEOSX: A Multiphysics, Multilevel Simulator Designed for Exascale Computing","authors":"H. Gross, A. Mazuyer","doi":"10.2118/203932-ms","DOIUrl":"https://doi.org/10.2118/203932-ms","url":null,"abstract":"\u0000 Evaluating large basin-scale formations for CO2 sequestration is one of the most important challenges for our industry. The technical complexity and the quantification of risks associated with these operations call for new reservoir engineering and reservoir simulation tools. The impact of multiple coupled physical phenomena, the century timescale, and basin-sized models in these operations force us to completely take apart and revisit the numerical backbone of existing simulation tools. We need a reservoir simulation tool designed for scalability and portability on high-performance computing architectures.\u0000 To achieve this, we are proposing a new, open-source, multiphysics, and multilevel physics simulation tool called GEOSX. This tool is jointly created by Lawrence Livermore National Laboratory, Stanford University, and Total. It is designed for scalability on multiple CPUs and multiple GPUs and offers a suite of physical solvers that can be extended easily while achieving a balance between performance and portability. GEOSX is initially targeting multiphysics simulations with coupled geomechanics, flow, and transport mechanics but with its open architecture, it allows access to high-performance physical solvers as building blocks of other multiphysics problems and provides users with a suite of tools for numerical optimization across platforms.\u0000 In this paper, we introduce GEOSX, expose its fundamental architecture principles, and show an example of geological sequestration of CO2 modeling on real data. We demonstrate our ability to simulate fluid and rock poromechanical interactions over long periods and basin-scale dimensions. GEOSX demonstrates its usefulness for such complex and large problems and proves to be scalable and portable across multiple high-performance systems.","PeriodicalId":11146,"journal":{"name":"Day 1 Tue, October 26, 2021","volume":"206 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80418444","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}
Solving the equations governing multiphase flow in geological formations involves the generation of a mesh that faithfully represents the structure of the porous medium. This challenging mesh generation task can be greatly simplified by the use of unstructured (tetrahedral) grids that conform to the complex geometric features present in the subsurface. However, running a million-cell simulation problem using an unstructured grid on a real, faulted field case remains a challenge for two main reasons. First, the workflow typically used to construct and run the simulation problems has been developed for structured grids and needs to be adapted to the unstructured case. Second, the use of unstructured grids that do not satisfy the K-orthogonality property may require advanced numerical schemes that preserve the accuracy of the results and reduce potential grid orientation effects. These two challenges are at the center of the present paper. We describe in detail the steps of our workflow to prepare and run a large-scale unstructured simulation of a real field case with faults. We perform the simulation using four different discretization schemes, including the cell-centered Two-Point and Multi-Point Flux Approximation (respectively, TPFA and MPFA) schemes, the cell- and vertex-centered Vertex Approximate Gradient (VAG) scheme, and the cell- and face-centered hybrid Mimetic Finite Difference (MFD) scheme. We compare the results in terms of accuracy, robustness, and computational cost to determine which scheme offers the best compromise for the test case considered here.
{"title":"Comparison of Various Discretization Schemes for Simulation of Large Field Case Reservoirs Using Unstructured Grids","authors":"Samier Pierre, Raguenel Margaux, Darche Gilles","doi":"10.2118/203949-ms","DOIUrl":"https://doi.org/10.2118/203949-ms","url":null,"abstract":"\u0000 Solving the equations governing multiphase flow in geological formations involves the generation of a mesh that faithfully represents the structure of the porous medium. This challenging mesh generation task can be greatly simplified by the use of unstructured (tetrahedral) grids that conform to the complex geometric features present in the subsurface. However, running a million-cell simulation problem using an unstructured grid on a real, faulted field case remains a challenge for two main reasons. First, the workflow typically used to construct and run the simulation problems has been developed for structured grids and needs to be adapted to the unstructured case. Second, the use of unstructured grids that do not satisfy the K-orthogonality property may require advanced numerical schemes that preserve the accuracy of the results and reduce potential grid orientation effects. These two challenges are at the center of the present paper. We describe in detail the steps of our workflow to prepare and run a large-scale unstructured simulation of a real field case with faults. We perform the simulation using four different discretization schemes, including the cell-centered Two-Point and Multi-Point Flux Approximation (respectively, TPFA and MPFA) schemes, the cell- and vertex-centered Vertex Approximate Gradient (VAG) scheme, and the cell- and face-centered hybrid Mimetic Finite Difference (MFD) scheme. We compare the results in terms of accuracy, robustness, and computational cost to determine which scheme offers the best compromise for the test case considered here.","PeriodicalId":11146,"journal":{"name":"Day 1 Tue, October 26, 2021","volume":"8 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87880389","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：