Zipper hydraulic fracturing in multiple wells with long horizontal sections is a primary solution means to increase the shale gas production rate and efficiency and to reduce the cost in Southern Sichuan Basin.Microse...Zipper hydraulic fracturing in multiple wells with long horizontal sections is a primary solution means to increase the shale gas production rate and efficiency and to reduce the cost in Southern Sichuan Basin.Microseismic based fracturing monitoring can be used for real-time imaging of hydraulic fractures,so it has been widely used to evaluate the fracturing effect of shale gas reservoirs and to direct the optimiza-tion and adjustment of fracturing parameters.In China,however,the microseismic fracturing monitoring on fracturing of shale gas reservoirs cannot be used to evaluate the fracturing results until the fracturing operation in the pad wells is completed according to the parameters which are designed prior to the fracturing monitoring.Its evaluation results can merely provide a guidance for the fracturing parameters of the next pad wells instead of the wells in operation.As a result,the real-time effect of microseismic fracturing monitoring is out of work.In view of this,the fractures induced by zipper hydraulic fracturing in multiple shale gas wells with long horizontal sections in the southern Sichuan Basin,was real-time imaged by using the combined technology of radially arranged microseismic surface monitoring and microseismic well monitoring on the basis of real-time positioning method.The fracturing results were assessed and used in real time for the optimization of prepad fluid parameter,perforation and temporary plugging additive releasing time,so as to effectively avoid repeated fracturing and uneven fracturing effects and improve fracturing stimulation effects.This method is applied in two well groups.It is shown that the average shale gas production rate is increased by 2e5 times.Furthermore,microseismic fracturing real-time monitoring plays a vital role in real-time evaluation of fracturing effect and real-time optimization of fracturing parameters,so it can be used as the reference and should be popularized further.展开更多
This paper introduces a deep learning workflow to predict phase distributions within complex geometries during two-phase capillary-dominated drainage.We utilize subsamples from Computerized Tomography(CT)images of roc...This paper introduces a deep learning workflow to predict phase distributions within complex geometries during two-phase capillary-dominated drainage.We utilize subsamples from Computerized Tomography(CT)images of rocks and incorporate pixel size,interfacial tension,contact angle,and pressure as inputs.First,an efficient morphology-based simulator creates a diverse dataset of phase distributions.Then,two commonly used convolutional and recurrent neural networks are explored and their deficiencies are highlighted,particularly in capturing phase connectivity.Subsequently,we develop a Higher-Dimensional Vision Transformer(HD-ViT)that drains pores solely based on their size,with phase connectivity enforced as a post-processing step.This enables inference for images of varying sizes,resolutions,and inlet-outlet setup.After training on a massive dataset of over 9.5 million instances,HD-ViT achieves excellent performance.We demonstrate the accuracy and speed advantage of the model on new and larger sandstone and carbonate images.We further evaluate HD-ViT against experimental fluid distribution images and the corresponding Lattice-Boltzmann simulations,producing similar outcomes in a matter of seconds.In the end,we train and validate a 3D version of the model.展开更多
基金Supported by the National Science and Technology Major Project“Real-time hydraulic fracturing monitoring and integrated geologic-engineering evaluation”(Grant No.2016ZX05023004).
文摘Zipper hydraulic fracturing in multiple wells with long horizontal sections is a primary solution means to increase the shale gas production rate and efficiency and to reduce the cost in Southern Sichuan Basin.Microseismic based fracturing monitoring can be used for real-time imaging of hydraulic fractures,so it has been widely used to evaluate the fracturing effect of shale gas reservoirs and to direct the optimiza-tion and adjustment of fracturing parameters.In China,however,the microseismic fracturing monitoring on fracturing of shale gas reservoirs cannot be used to evaluate the fracturing results until the fracturing operation in the pad wells is completed according to the parameters which are designed prior to the fracturing monitoring.Its evaluation results can merely provide a guidance for the fracturing parameters of the next pad wells instead of the wells in operation.As a result,the real-time effect of microseismic fracturing monitoring is out of work.In view of this,the fractures induced by zipper hydraulic fracturing in multiple shale gas wells with long horizontal sections in the southern Sichuan Basin,was real-time imaged by using the combined technology of radially arranged microseismic surface monitoring and microseismic well monitoring on the basis of real-time positioning method.The fracturing results were assessed and used in real time for the optimization of prepad fluid parameter,perforation and temporary plugging additive releasing time,so as to effectively avoid repeated fracturing and uneven fracturing effects and improve fracturing stimulation effects.This method is applied in two well groups.It is shown that the average shale gas production rate is increased by 2e5 times.Furthermore,microseismic fracturing real-time monitoring plays a vital role in real-time evaluation of fracturing effect and real-time optimization of fracturing parameters,so it can be used as the reference and should be popularized further.
基金supported by the International Cooperation Programme of Chengdu City(No.2020-GH02-00023-HZ)。
文摘This paper introduces a deep learning workflow to predict phase distributions within complex geometries during two-phase capillary-dominated drainage.We utilize subsamples from Computerized Tomography(CT)images of rocks and incorporate pixel size,interfacial tension,contact angle,and pressure as inputs.First,an efficient morphology-based simulator creates a diverse dataset of phase distributions.Then,two commonly used convolutional and recurrent neural networks are explored and their deficiencies are highlighted,particularly in capturing phase connectivity.Subsequently,we develop a Higher-Dimensional Vision Transformer(HD-ViT)that drains pores solely based on their size,with phase connectivity enforced as a post-processing step.This enables inference for images of varying sizes,resolutions,and inlet-outlet setup.After training on a massive dataset of over 9.5 million instances,HD-ViT achieves excellent performance.We demonstrate the accuracy and speed advantage of the model on new and larger sandstone and carbonate images.We further evaluate HD-ViT against experimental fluid distribution images and the corresponding Lattice-Boltzmann simulations,producing similar outcomes in a matter of seconds.In the end,we train and validate a 3D version of the model.
