SPE Journal最新文献

{"title":"Application of Physics-Informed Neural Networks for Estimation of Saturation Functions from Countercurrent Spontaneous Imbibition Tests","authors":"J. Abbasi, P. Andersen","doi":"10.2118/218402-pa","DOIUrl":"https://doi.org/10.2118/218402-pa","url":null,"abstract":"\u0000 In this work, physics-informed neural networks (PINNs) are used for history matching data from core-scale countercurrent spontaneous imbibition (COUCSI) tests. To our knowledge, this is the first work exploring the variation in saturation function solutions from COUCSI tests. 1D flow was considered, in which two phases flow in opposite directions driven by capillary forces with one boundary open to flow.\u0000 The partial differential equation (PDE) depends only on a saturation-dependent capillary diffusion coefficient (CDC). Static properties such as porosity, permeability, interfacial tension, and fluid viscosities are considered known. In contrast, the CDC or its components [relative permeability (RP) and capillary pressure (PC)], are considered unknown. We investigate the range of functions (CDCs or RP/PC combinations) that explain different (synthetic or real) experimental COUCSI data: recovery from varying extents of early-time and late-time periods, pressure transducers, and in-situ saturation profiles.\u0000 History matching was performed by training a PINN to minimize a loss function based on observational data and terms related to the PDE, boundary, and initial conditions. The PINN model was generated with feedforward neural networks, Fourier/inverse-Fourier transformation, and an adaptive tanh activation function, and trained using full batching. The trainable parameters of both the neural networks and saturation functions (parameters in RP and PC correlations) were initialized randomly.\u0000 The PINN method successfully matched the observed data and returned a range of possible saturation function solutions. When a full observed recovery curve was provided (recovery data reaching close to its final value), unique and correct CDC functions and correct spatial saturation profiles were obtained. However, different RP/PC combinations composing the CDC were calculated. For limited amounts of recovery data, different CDCs matched the observations equally well but predicted different recovery behavior beyond the collected data period. With limited recovery data, when all points were still following a square root of time trend, a CDC with a low magnitude and peak shifted to high saturations gave the same match as a CDC with a high magnitude and peak shifted to low saturations. Recovery data with sufficient points not being proportional to the square root of time strongly constrained how future recovery would behave and thus which CDCs could explain the results. Limited recovery data combined with an observed in-situ profile of saturations allowed for accurate determination of CDC and prediction of future recovery, suggesting in-situ data allowed for shortened experiments. With full recovery data, in-situ PC data calibrated the PC toward unique solutions matching the input. The RPs were determined, where their phase had much lower mobility than the others. The CDC is virtually independent of the highest fluid mobility, and RPs could not be matched at thei","PeriodicalId":22252,"journal":{"name":"SPE Journal","volume":"2 5","pages":""},"PeriodicalIF":3.6,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139457019","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Experimental Study on Permeability and Gas Production Characteristics of Montmorillonite Hydrate Sediments Considering the Effective Stress and Gas Slippage Effect","authors":"Zhaoran Wu, Qing Gu, Lei Wang, Guijing Li, Cheng Shi, Yufa He, Qingping Li, Yanghui Li","doi":"10.2118/218385-pa","DOIUrl":"https://doi.org/10.2118/218385-pa","url":null,"abstract":"\u0000 Gas permeability in hydrate reservoirs is the decisive parameter in determining the gas production efficiency and gas production of hydrate. In the South China Sea (SCS), the gas flow in tight natural gas hydrate (NGH) silty clay reservoirs is significantly affected by the gas slippage effect and the effective stress (ES) of overlying rock. To improve the effectiveness of hydrate exploitation, it is necessary to understand the influence of gas slippage in hydrate reservoirs on the permeability evolution law. For this paper, the gas permeability characteristics and methane production of hydrate montmorillonite sediments were studied at different pore pressures and ESs. Experimental data revealed that the gas permeability of montmorillonite samples before methane hydrate (MH) formation is seriously affected by the Klinkenberg effect. The gas permeability of montmorillonite sediments before hydrate formation is generally smaller than that after hydrate formation, and the gas slippage effect in the sediments after hydrate formation is weaker than that before hydrate formation. With the change in ES, the intrinsic permeability of sediment has a power law relationship with the simple ES. The ES law coefficient n was determined using the response surface method to eliminate the influence of gas slip on gas permeability. As pore pressure decreases and MH decomposes, montmorillonite swelling seriously affects gas permeability. However, the gas slippage effect has a good compensation effect on the permeability of montmorillonite sediments after MH decomposition under low pore pressure. The multistage depressurization-producing process of MH in montmorillonite sediments is mainly 3 MPa depressurization-producing stage and 2 MPa depressurization-producing stage. In this paper, the influence mechanism of gas slippage effect of hydrate reservoir is studied, which is conducive to improving the prediction accuracy of gas content in the process of hydrate exploitation and exploring the best pressure reduction method to increase the gas production of hydrate in the process of exploitation.","PeriodicalId":22252,"journal":{"name":"SPE Journal","volume":"80 10","pages":""},"PeriodicalIF":3.6,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139458262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Feature Extraction in Time-Lapse Seismic Using Deep Learning for Data Assimilation","authors":"Rodrigo Exterkoetter, Gustavo R. Dutra, Leandro P. de Figueiredo, Fernando Bordignon, Gilson M. S. Neto, Alexandre A. Emerick","doi":"10.2118/212196-pa","DOIUrl":"https://doi.org/10.2118/212196-pa","url":null,"abstract":"<p>Assimilation of time-lapse (4D) seismic data with ensemble-based methods is challenging because of the massive number of data points. This situation requires excessive computational time and memory usage during the model updating step. We addressed this problem using a deep convolutional autoencoder to extract the relevant features of 4D images and generate a reduced representation of the data. The architecture of the autoencoder is based on the VGG-19 network, a deep convolutional architecture with 19 layers well-known for its effectiveness in image classification and object recognition. Some advantages of VGG-19 are the possibility of using some pretrained convolutional layers to create a feature extractor and taking advantage of the transfer learning technique to address other related problem domains. Using a pretrained model bypasses the need for large training data sets and avoids the high computational demand to train a deep network. For further improvements in the reconstruction of the seismic images, we apply a fine-tuning of the weights of the latent convolutional layer. We propose to use a fully convolutional architecture, which allows the application of distance-based localization during data assimilation with the ensemble smoother with multiple data assimilation (ES-MDA). The performance of the proposed method is investigated in a synthetic benchmark problem with realistic settings. We evaluate the methodology with three variants of the autoencoder, each one with a different level of data reduction. The experiments indicate that it is possible to use latent representations with major data reductions without impairing the quality of the data assimilation. Additionally, we compare central processing unit (CPU) and graphics processing unit (GPU) implementations of the ES-MDA update step and show in another synthetic problem that the reduction in the number of data points obtained with the application of the deep autoencoder may provide a substantial improvement in the overall computation cost of the data assimilation for large reservoir models.</p>","PeriodicalId":22252,"journal":{"name":"SPE Journal","volume":"35 1","pages":""},"PeriodicalIF":3.6,"publicationDate":"2023-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140581288","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluation of Flow Units and Capillary Pressures of the Giant Chicontepec Tight Oil Paleochannel in Mexico and a Fresh Look at Drilling and Completions","authors":"Alejandra Gutierrez Oseguera, Roberto Aguilera","doi":"10.2118/212745-pa","DOIUrl":"https://doi.org/10.2118/212745-pa","url":null,"abstract":"&lt;p&gt;The Chicontepec Paleochannel in Mexico is a giant shaly sandstone reservoir with volumes of original oil in place (OOIP) ranging between 137 and 59 billion STB (Guzmán 2022). However, the oil recoveries are very small, ranging between 0.32% and 0.75% of the OOIP. Under these conditions, consistent interpretation of flow units and mercury injection capillary pressures up to 55,000 psi provide useful information that helps in deciphering the rock quality and pore sizes at levels that might not be reached by thin-section petrography. This is important because the Chicontepec Paleochannel (Misantla-Tampico Basin) has been recently equated to the Permian Basin in the United States and has been termed by Guzman (2022) “a premier super basin in waiting.”&lt;/p&gt;&lt;p&gt;The current cumulative oil production of Chicontepec is 440 million STB. Although it is a significant volume, it represents a very small percentage of recovery from the reservoir (0.32–0.75% of the OOIP). To help improve recovery, a method is developed for characterizing the tight Chicontepec Paleochannel using flow units and capillary pressures. Like in the case of many tight unconventional reservoirs, the capillary pressures can go to very high values, reaching 55,000 psi in the Chicontepec case. Therefore, a special procedure is developed to generate a consistent interpretation of all the available capillary pressure curves for the entire range of pressures.&lt;/p&gt;&lt;p&gt;The results highlight the important oil recovery potential. The assessment is supported by quantitative formation evaluation work performed by Gutierrez Oseguera and Aguilera (2023). Although natural fractures are present, most wells must be hydraulically fractured to achieve commercial success.&lt;/p&gt;&lt;p&gt;Process or delivery speed (the ratio of permeability and porosity) for the Chicontepec samples used in the capillary pressure experimental work range between 159.1 md and 0.17 md (porosity in the denominator is a fraction). Flow units show pore throat radii (&lt;em&gt;r&lt;sub&gt;p&lt;/sub&gt;&lt;/em&gt;&lt;sub&gt;35&lt;/sub&gt;) range from less than 0.1 µm to about 4.5 µm. These values and flow units compare well with data available for prolific unconventional reservoirs such as the Cardium sandstone in Canada and the giant Permian Basin in the United States. The radius &lt;em&gt;r&lt;sub&gt;p&lt;/sub&gt;&lt;/em&gt;&lt;sub&gt;35&lt;/sub&gt; refers to the pore throat radius at 35% cumulative pore volume (PV) of injected mercury. This is different from &lt;em&gt;r&lt;sub&gt;p&lt;/sub&gt;&lt;/em&gt; also discussed in this paper, which is the pore throat radius at any water saturation (for example, at 40% water saturation). Thus, in the case where water saturation is 65%, &lt;em&gt;r&lt;sub&gt;p&lt;/sub&gt;&lt;/em&gt; is equal to &lt;em&gt;r&lt;sub&gt;p&lt;/sub&gt;&lt;/em&gt;&lt;sub&gt;35&lt;/sub&gt;.&lt;/p&gt;&lt;p&gt;The novelty of this study is the development of a consistent procedure for interpreting the entire range of pressures measured during mercury injection capillary pressures. Such pressures go up to 55,000 psi for the core samples considered in this study. The integration with flow units a","PeriodicalId":22252,"journal":{"name":"SPE Journal","volume":"159 1","pages":""},"PeriodicalIF":3.6,"publicationDate":"2023-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140581206","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Bi-Objective Optimization of Subsurface CO2 Storage with Nonlinear Constraints Using Sequential Quadratic Programming with Stochastic Gradients","authors":"Quang Minh Nguyen, Mustafa Onur, Faruk Omer Alpak","doi":"10.2118/214363-pa","DOIUrl":"https://doi.org/10.2118/214363-pa","url":null,"abstract":"<p>This study focuses on carbon capture, utilization, and sequestration (CCUS) via the means of nonlinearly constrained production optimization workflow for a CO₂-enhanced oil recovery (EOR) process, in which both the net present value (NPV) and the net present carbon tax credits (NPCTC) are bi-objectively maximized, with the emphasis on the consideration of injection bottomhole pressure (IBHP) constraints on the injectors, in addition to field liquid production rate (FLPR) and field water production rate (FWPR), to ensure the integrity of the formation and to prevent any potential damage during the life cycle injection/production process. The main optimization framework used in this work is a lexicographic method based on the line-search sequential quadratic programming (LS-SQP) coupled with stochastic simplex approximate gradients (StoSAG). We demonstrate the performance of the optimization algorithm and results in a field-scale realistic problem, simulated using a commercial compositional reservoir simulator. Results show that the workflow can solve the single-objective and bi-objective optimization problems computationally efficiently and effectively, especially in handling and honoring nonlinear state constraints imposed onto the problem. Various numerical settings have been experimented with to estimate the Pareto front for the bi-objective optimization problem, showing the trade-off between the two objectives of NPV and NPCTC. We also perform a single-objective optimization on the total life cycle cash flow, which is the aggregated quantity of NPV and NPCTC, and quantify the results to further emphasize the necessity of performing bi-objective production optimization, especially when used in conjunction with commercial flow simulators that lack the capability of computing adjoint-based gradients.</p>","PeriodicalId":22252,"journal":{"name":"SPE Journal","volume":"47 1","pages":""},"PeriodicalIF":3.6,"publicationDate":"2023-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140581285","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Relationship Between Residual Saturations and Wettability Using Pore-Network Modeling","authors":"Prakash Purswani, Russell T. Johns, Zuleima T. Karpyn","doi":"10.2118/206379-pa","DOIUrl":"https://doi.org/10.2118/206379-pa","url":null,"abstract":"<p>The relationship between residual saturation and wettability is critical for modeling multiphase processes like enhanced oil recovery, CO₂ sequestration, and geologic storage of hydrogen. The wetting state of a core is often quantified through Amott indices, which are estimated from the ratio of the saturation fraction that flows spontaneously to the total saturation change that occurs due to spontaneous flow and forced injection. Observations from traditional coreflooding experiments show a minimum in the trends of residual oil saturation (<em>S_or</em>) around mixed-wet conditions. Amott indices, however, provide an average measure of wettability because of their intrinsic dependence on a variety of factors such as the initial oil saturation, aging conditions, rock heterogeneity, etc. Thus, the use of Amott indices could potentially cloud the observed trends of residual saturation with wettability.</p><p>Using pore-network modeling (PNM), we show that <em>S_or</em> varies monotonically with the contact angle, which is a direct measure of wettability. That is, for fixed initial oil saturation, the <em>S_or</em> decreases monotonically as the reservoir becomes more water-wet (decreasing contact angle). Further, the calculation of Amott indices for the PNM data sets shows that a plot of the <em>S_or</em> vs. Amott indices also shows this monotonic trend, but only if the initial oil saturation is kept fixed. Thus, for the cases presented here, we show that there is no minimum residual saturation at mixed-wet conditions as wettability changes.</p><p>In this research, we employ a numerical approach to quantify trends of <em>S_or</em> against the traditional definition of wettability. Through the analysis of our numerical work and literature experiments, we find that under isolated conditions (constant initial saturation), linear trends exist between <em>S_or</em> and wettability. This can have important implications for low salinity waterflooding, water-alternating-gas enhanced oil recovery, or CO₂ sequestration where the effects of wettability are critical to understand phase trapping.</p>","PeriodicalId":22252,"journal":{"name":"SPE Journal","volume":"32 1","pages":""},"PeriodicalIF":3.6,"publicationDate":"2023-12-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140581375","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Accelerated Design of Sidetrack and Deepening Well Trajectories","authors":"Umesh Singh, Rizwan Pathan, Amol Dilip Joshi, Adrien Cavé, Cédric Fouchard, Antonin Baume","doi":"10.2118/218395-pa","DOIUrl":"https://doi.org/10.2118/218395-pa","url":null,"abstract":"\u0000 Sidetrack and deepening wells play a pivotal role in enhancing oil and gas recovery while simultaneously reducing drilling costs, particularly in cluster well scenarios. These wells leverage existing wellbores effectively, resulting in substantial reductions in development expenses. Deepening wells maximize cost savings by utilizing the entire length of preexisting wellbores. These wells strategically access low-permeability layers, thin pay zones, wedge zones, and marginal reserves while also serving as rapid response solutions during emergencies to expedite risk mitigation in accidents. There is a pressing need for expedient, safer, and cost-effective well designs to achieve economic efficiency, which necessitates the development of advanced design methodologies. However, designing optimized 3D sidetrack and deepening well trajectories for oil and gas reservoir access while mitigating collision risks is a complex and time-consuming task that demands meticulous planning and exhaustive well path analysis, often involving multiple iterations to ensure cost-effective solutions meeting drillability and safety constraints. In this study, we develop an integrated framework for the accelerated design of sidetrack and deepening well trajectories, complemented by a trajectory optimization algorithm to generate safer and cost-effective well trajectories. The developed framework is rigorously tested in a live Nigerian oil and gas field. The case study involves the design of a sidetrack and a deepening well trajectory in a crowded brownfield consisting of 21 legacy wells. The results of the case study exhibit the significance of the established framework on streamlining the well design process, leading to expedited creation of efficient and safe sidetrack and deepening well trajectories.","PeriodicalId":22252,"journal":{"name":"SPE Journal","volume":"282 2","pages":""},"PeriodicalIF":3.6,"publicationDate":"2023-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138985912","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Generalized Analytical Solutions of Vertically Fractured Wells in Commingled Reservoirs: Field Case Study","authors":"Cao Wei, Haitao Li, Hongwen Luo, Ying Li, Shiqing Cheng","doi":"10.2118/218391-pa","DOIUrl":"https://doi.org/10.2118/218391-pa","url":null,"abstract":"\u0000 Accurate identification of the individual-layer parameters for vertically fractured wells in commingled reservoirs is essential for development plan design, reservoir numerical simulation, and stimulation measure selection. Different semi-analytical and numerical models are generally applied in multilayer transient testing (MLT) analysis to determine the properties of individual layer. However, these approaches require numerous computations and are complicated to program due to the fracture and reservoir discretization. This work thus presents the generalized analytical solutions of vertically fractured wells in infinite, closed, or constant-pressure commingled reservoirs with both computational and functional simplicity. The fully analytical solutions are derived based on the early-time approximate solutions of infinite-conductivity fracture and trilinear flow models, infinite-conductivity fracture solutions, pressure superposition principle, and Duhamel principle. A systematic verification by employing a standardized well testing software and trilinear flow model is conducted to ensure the general application accuracy of the presented solutions. The results show that the developed analytical solutions are valid when the dimensionless fracture conductivity is more than 2 (FcD > 2) with an average absolute percent deviation (AAD) of ~2% for pressure and that is ~4% for pressure derivative. The developed analytical solutions also exhibit improvements in early-time pressure and derivative calculation. Finally, a field case of a four-layer fractured well is interpreted by the developed solutions and well testing software to illustrate the feasibility. The interpretation results of two methods are nearly identical, with only a minor difference. The developed analytical solutions are computationally accurate while maintaining functional simplicity and can be considered as an alternative to the current semi-analytical and numerical approaches in MLT analysis.","PeriodicalId":22252,"journal":{"name":"SPE Journal","volume":"534 ","pages":""},"PeriodicalIF":3.6,"publicationDate":"2023-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139023564","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Oil-Water Flowing Experiments and Water-Cut Range Classification Approach Using Distributed Acoustic Sensing","authors":"Junrong Liu, Yanhui Han, Qingsheng Jia, Lei Zhang, Ming Liu, Zhigang Li","doi":"10.2118/218389-pa","DOIUrl":"https://doi.org/10.2118/218389-pa","url":null,"abstract":"\u0000 The accurate measurement of dynamic water cut is of great interest for analyzing reservoir performance and optimizing oilwell production. Downhole water-cut measurement is a very challenging work. Moreover, the surface-measured water cut is a comprehensive indicator of commingled producing well and it is difficult to use this parameter to deduce the downhole water cut of each contributing layer. In this paper, we propose to use distributed fiber-optic acoustic sensing (DAS) technology for the classification of water-cut range. DAS can dynamically monitor the entire wellbore by “listening” to the acoustic signals during flow. A large number of laboratory experimental data from DAS have been collected and analyzed using wavelet time scattering transform and short-time Fourier transform (STFT). The extracted low-variance scattering feature, short time-frequency feature, and fusion feature (combination of two extracted features) were learned with backpropagation (BP) neural network, decision tree (DT), and random forest (RF) algorithm. Then, a classification method of water-cut range in oil-water flow was established with machine learning. Field DAS data were collected from two oil wells to verify the effectiveness of the proposed method. The classification accuracies for the vertical well (Well A) are 92.4% and 87.4% by DT and RF model, respectively. For the horizontal well (Well B), the average classification accuracy exceeds 90% for all three methods. Water shutoff measure was conducted in Well B, and an obvious water decrease was realized. The result shows that the fusion feature overweighs single feature in machine learning with DAS data. This study provides a novel way to identify downhole water-cut range and detect water entry location in horizontal, vertical, and deviated oil-producing wells.","PeriodicalId":22252,"journal":{"name":"SPE Journal","volume":"118 46","pages":""},"PeriodicalIF":3.6,"publicationDate":"2023-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138608278","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Deep Regression Method for Gas Well Liquid Loading Prediction","authors":"Yan Chen, Bo Miao, Yang Wang, Yunan Huang, YuQiang Jiang, Xiangchao Shi","doi":"10.2118/218387-pa","DOIUrl":"https://doi.org/10.2118/218387-pa","url":null,"abstract":"\u0000 Liquid loading occurs when gas production falls below the critical liquid-carrying flow rate of the gas well, resulting in the inability to remove the condensate or water in the gas well. Liquid loading can lead to a sharp reduction in production, which affects the gas well ultimate recovery. Accurate prediction of the timing of liquid loading is important for implementing mitigations that reduce liquid accumulation in the production tubing and prevent gas production impairment, as well as for the stability of production. Existing liquid-loading forecasting methods have a time offset in the determination of liquid loading, and there is great variation in the results for different gas wells. Currently, supervisory control and data acquisition (SCADA) systems are widely used for gas well production data acquisition, but the data are not effectively utilized. Deep machine learning techniques are applied to the field data from gas wells and have achieved significant effectiveness. In this study, a bidirectional long short-term memory network (Bi-LSTM) was used to conduct feature extraction on the production data, and the extracted feature was spliced together with the geological and engineering parameter feature. These features were combined with self-attention mechanisms to predict the time of the next liquid loading. Because the modeling results fit the actual liquid loading in production scenarios better, our method also customizes the loss functions. Experimental verification was conducted using actual production data from 13 gas wells. The recall was 1 and F1 was 0.87 for the experimental data in the model, and the customized loss function led to a 6.5% improvement in F1. The experimental results verify that our method can accurately forecast liquid-loading onset in gas wells in a timely manner, which can help reduce costs and increase efficiency in shale gas production.","PeriodicalId":22252,"journal":{"name":"SPE Journal","volume":" 1","pages":""},"PeriodicalIF":3.6,"publicationDate":"2023-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138613358","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}