CO2 sequestration into deep saline aquifers is a promising solution in addressing the global warming effects arising out of CO2 emissions into the atmosphere. The performance of CO2 sequestration requires knowledge of the coupled fluid flow, deformation, and heat transport phenomenon. Nevertheless, numerical analysis has been identified as an economical means of analysing these coupled processes. Hence, CO2 sequestration process is modelled using three models viz., Hydraulic-H, Hydro-Mechanical-HM and Thermo-Hydro-Mechanical-THM, which were initially validated with 2-D two-phase flow in a saline aquifer reported in literature, followed by a comparative analysis. The analysis projected that H model to underestimate the storage capacity factor due to overestimation of pressure change with a delayed migration of the CO2 saturation front to about 2718 m along the top of the aquifer. Further, THM model highlighted the impact of thermal strain arising from non-isothermal conditions. Hence, a sensitivity analysis on the initial aquifer temperature and CO2 fluid injection temperatures, projected maximum storage efficiencies at a higher initial aquifer temperature of 323.15 K due to minimal pore pressure buildup of about 0.779 MPa, promoting maximum CO2 gas saturation plume migrating distances of about 4486 m. Though the lower CO2 injection temperatures of 303.15 K resulted in a slightly higher storage efficiency factor of about 0.29 and mass based storage efficiency of about 0.23, it is crucial to consider the significant pore pressure buildup occurring at these conditions, owing to the risk of failure of the rock.
{"title":"Relative performance of hydraulic, hydro-mechanical and thermo-hydro-mechanical models on the geological sequestration of CO2 in deep saline aquifers","authors":"Khumujam Jeffry Singh , Tummuri Naga Venkata Pavan , Srinivasa Reddy Devarapu , Suresh Kumar Govindarajan","doi":"10.1016/j.geoen.2026.214410","DOIUrl":"10.1016/j.geoen.2026.214410","url":null,"abstract":"<div><div>CO<sub>2</sub> sequestration into deep saline aquifers is a promising solution in addressing the global warming effects arising out of CO<sub>2</sub> emissions into the atmosphere. The performance of CO<sub>2</sub> sequestration requires knowledge of the coupled fluid flow, deformation, and heat transport phenomenon. Nevertheless, numerical analysis has been identified as an economical means of analysing these coupled processes. Hence, CO<sub>2</sub> sequestration process is modelled using three models viz., Hydraulic-H, Hydro-Mechanical-HM and Thermo-Hydro-Mechanical-THM, which were initially validated with 2-D two-phase flow in a saline aquifer reported in literature, followed by a comparative analysis. The analysis projected that H model to underestimate the storage capacity factor due to overestimation of pressure change with a delayed migration of the CO<sub>2</sub> saturation front to about 2718 m along the top of the aquifer. Further, THM model highlighted the impact of thermal strain arising from non-isothermal conditions. Hence, a sensitivity analysis on the initial aquifer temperature and CO<sub>2</sub> fluid injection temperatures, projected maximum storage efficiencies at a higher initial aquifer temperature of 323.15 K due to minimal pore pressure buildup of about 0.779 MPa, promoting maximum CO<sub>2</sub> gas saturation plume migrating distances of about 4486 m. Though the lower CO<sub>2</sub> injection temperatures of 303.15 K resulted in a slightly higher storage efficiency factor of about 0.29 and mass based storage efficiency of about 0.23, it is crucial to consider the significant pore pressure buildup occurring at these conditions, owing to the risk of failure of the rock.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"260 ","pages":"Article 214410"},"PeriodicalIF":4.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190688","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 : 2026-05-01Epub Date: 2026-01-30DOI: 10.1016/j.geoen.2026.214388
Dong Yang , Shang Xu , Bingchang Liu , Yunxuan Guo , Yufan Wang , Kang Wen , Yuerui Jia , Qiqi Li , Jingheng Nie
Pore pressure–stress coupling (PSC) is a fundamental mechanism in subsurface geomechanics that links changes in pore pressure with the evolution of in-situ stresses. It is central to the performance of geoenergy systems such as sustainable hydrocarbon production, geological CO2 storage, and enhanced geothermal operations. PSC arises from poroelastic interactions, in which pore pressure disturbances redistribute stresses through mechanical constraints on rock deformation. This article provides a critical review of the mechanisms, modeling approaches, and engineering implications of PSC. It synthesizes evidence from basin, reservoir, and laboratory scales to clarify the mechanisms, controlling factors, and engineering implications of PSC. Recent modeling advances have progressed from one-dimensional poroelastic formulations to fully time-dependent, multi-dimensional frameworks. These approaches incorporate the main controlling factors of PSC, namely geological heterogeneity, elastic contrasts, fluid properties, and the spatiotemporal extent of pressure perturbations. The review further integrates analytical and numerical formulations that link PSC with fault slip tendency and critical reactivation pressure, highlighting their conceptual connections across different spatial scales. These insights demonstrate that PSC governs fracture evolution and stress redistribution in subsurface systems and provide a basis for improving reservoir management, evaluating storage security, and ensuring the safe and sustainable utilization of geoenergy resources.
{"title":"A critical review of pore pressure–stress coupling: Mechanisms, modeling, and implications for subsurface energy systems","authors":"Dong Yang , Shang Xu , Bingchang Liu , Yunxuan Guo , Yufan Wang , Kang Wen , Yuerui Jia , Qiqi Li , Jingheng Nie","doi":"10.1016/j.geoen.2026.214388","DOIUrl":"10.1016/j.geoen.2026.214388","url":null,"abstract":"<div><div>Pore pressure–stress coupling (PSC) is a fundamental mechanism in subsurface geomechanics that links changes in pore pressure with the evolution of in-situ stresses. It is central to the performance of geoenergy systems such as sustainable hydrocarbon production, geological CO<sub>2</sub> storage, and enhanced geothermal operations. PSC arises from poroelastic interactions, in which pore pressure disturbances redistribute stresses through mechanical constraints on rock deformation. This article provides a critical review of the mechanisms, modeling approaches, and engineering implications of PSC. It synthesizes evidence from basin, reservoir, and laboratory scales to clarify the mechanisms, controlling factors, and engineering implications of PSC. Recent modeling advances have progressed from one-dimensional poroelastic formulations to fully time-dependent, multi-dimensional frameworks. These approaches incorporate the main controlling factors of PSC, namely geological heterogeneity, elastic contrasts, fluid properties, and the spatiotemporal extent of pressure perturbations. The review further integrates analytical and numerical formulations that link PSC with fault slip tendency and critical reactivation pressure, highlighting their conceptual connections across different spatial scales. These insights demonstrate that PSC governs fracture evolution and stress redistribution in subsurface systems and provide a basis for improving reservoir management, evaluating storage security, and ensuring the safe and sustainable utilization of geoenergy resources.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"260 ","pages":"Article 214388"},"PeriodicalIF":4.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190220","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 : 2026-05-01Epub Date: 2026-01-27DOI: 10.1016/j.geoen.2026.214383
Fantong Kong , Yongxiang Liu , Biqi Zhang , Chengming Luo , Xihao Gu
Dispersion processing of dipole waveforms is essential for determining the shear slowness of formations, especially in slow formations where flexural modes dominate. Traditional methods often struggle to achieve accurate estimations due to noise contamination, especially in the low-frequency range. In this study, we propose a physics-informed deep learning method for enhancing dispersion curve processing and shear slowness estimation. A synthetic dataset, parameterized by four key physical properties, is used to generate clean dispersion curves, while a noise simulation method, designed to capture the statistical characteristics of dispersion curves, creates realistic scatter conditions for training an Attention-MLP network. This network leverages residual connections and attention mechanisms to improve feature extraction and robustness against noise. The designed physics-informed loss function ensures accurate parameter predictions and physically consistent dispersion curves. Experimental results show that the proposed method achieves a high SNR of 43.75 and accurately determines formation shear slowness, highlighting its potential as a reliable and efficient tool for borehole acoustic applications.
{"title":"A physics-informed deep learning method for dispersive processing of borehole dipole wave data using synthetic dataset","authors":"Fantong Kong , Yongxiang Liu , Biqi Zhang , Chengming Luo , Xihao Gu","doi":"10.1016/j.geoen.2026.214383","DOIUrl":"10.1016/j.geoen.2026.214383","url":null,"abstract":"<div><div>Dispersion processing of dipole waveforms is essential for determining the shear slowness of formations, especially in slow formations where flexural modes dominate. Traditional methods often struggle to achieve accurate estimations due to noise contamination, especially in the low-frequency range. In this study, we propose a physics-informed deep learning method for enhancing dispersion curve processing and shear slowness estimation. A synthetic dataset, parameterized by four key physical properties, is used to generate clean dispersion curves, while a noise simulation method, designed to capture the statistical characteristics of dispersion curves, creates realistic scatter conditions for training an Attention-MLP network. This network leverages residual connections and attention mechanisms to improve feature extraction and robustness against noise. The designed physics-informed loss function ensures accurate parameter predictions and physically consistent dispersion curves. Experimental results show that the proposed method achieves a high SNR of 43.75 and accurately determines formation shear slowness, highlighting its potential as a reliable and efficient tool for borehole acoustic applications.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"260 ","pages":"Article 214383"},"PeriodicalIF":4.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190225","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 : 2026-04-01Epub Date: 2026-01-20DOI: 10.1016/j.geoen.2026.214372
Samira Khani Rasmussen, Behzad Rostami, Wael Fadi Al-Masri, Nikolai Andrianov
Intermittent CO2 injection, alternating with brine, is a realistic scenario in CO2 storage operations, often resulting from injection strategies or maintenance activities. This cyclic injection can influence reservoir reactivity and injectivity. To investigate the risk of injectivity impairment due to cyclic injection, a laboratory flooding experiment is conducted on core samples from the Nini-4 well (Frigg sand, Horda Formation, Danish North Sea) under reservoir conditions of 60 °C and 200 bar, simulating supercritical CO2 (scCO2) injection and brine backflow. The experiment includes two drainage and one imbibition periods. Produced water is measured during both scCO2 and brine injection using an acoustic separator, enabling differentiation between water displaced and dissolved in the pore space. A compositional numerical modeling approach using the CMG GEM model is employed for history-matching the experimental data to determine the relative permeability (Kr) and capillary pressure (Pc) curves, besides accounting for water evaporation to predict saturation data. Primarily, the results indicate that injectivity is significantly enhanced over extended periods up to 200 pore volumes and restored to initial brine permeability due to a dominant drying-out effect in both drainage periods, in which differential pressure and saturation data show strong agreement. Secondarily, identical Kr and Pc functions are derived from both drainage periods, differing from the imbibition period, highlighting hysteresis effects. These findings provide insights into near and far wellbore flow behavior under cyclic injection, which are directly relevant for the Nini West site’s maturation and certification process, supporting its viability for long-term CO2 storage.
{"title":"Injectivity analysis and determination of saturation functions during cyclic injection of supercritical CO2 in Nini West sandstone","authors":"Samira Khani Rasmussen, Behzad Rostami, Wael Fadi Al-Masri, Nikolai Andrianov","doi":"10.1016/j.geoen.2026.214372","DOIUrl":"10.1016/j.geoen.2026.214372","url":null,"abstract":"<div><div>Intermittent CO<sub>2</sub> injection, alternating with brine, is a realistic scenario in CO<sub>2</sub> storage operations, often resulting from injection strategies or maintenance activities. This cyclic injection can influence reservoir reactivity and injectivity. To investigate the risk of injectivity impairment due to cyclic injection, a laboratory flooding experiment is conducted on core samples from the Nini-4 well (Frigg sand, Horda Formation, Danish North Sea) under reservoir conditions of 60 °C and 200 bar, simulating supercritical CO<sub>2</sub> (scCO<sub>2</sub>) injection and brine backflow. The experiment includes two drainage and one imbibition periods. Produced water is measured during both scCO<sub>2</sub> and brine injection using an acoustic separator, enabling differentiation between water displaced and dissolved in the pore space. A compositional numerical modeling approach using the CMG GEM model is employed for history-matching the experimental data to determine the relative permeability (Kr) and capillary pressure (Pc) curves, besides accounting for water evaporation to predict saturation data. Primarily, the results indicate that injectivity is significantly enhanced over extended periods up to 200 pore volumes and restored to initial brine permeability due to a dominant drying-out effect in both drainage periods, in which differential pressure and saturation data show strong agreement. Secondarily, identical Kr and Pc functions are derived from both drainage periods, differing from the imbibition period, highlighting hysteresis effects. These findings provide insights into near and far wellbore flow behavior under cyclic injection, which are directly relevant for the Nini West site’s maturation and certification process, supporting its viability for long-term CO<sub>2</sub> storage.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"259 ","pages":"Article 214372"},"PeriodicalIF":4.6,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146039131","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 : 2026-04-01Epub Date: 2025-12-15DOI: 10.1016/j.geoen.2025.214327
Tobias Orlander , Frederik Peter Ditlevsen , Hanne Dahl Holmslykke , Leonardo Teixeira Pinto Meireles , Amirhossein Shamsolhodaei , Ida Lykke Fabricius
Denmark is currently exploring options to reuse depleted North Sea chalk oil and gas reservoirs for CO2 storage. Monitoring CO2 saturation, well-bore CO2 leakage, and plume propagation are among the challenges, with well-bore logging and seismic time-lapse techniques as likely instruments. Such instruments rely on an understanding of how elastic waves respond to changes in the saturating fluids. For chalk reservoirs, mineral dissolution and/or precipitation caused by CO2 injection may cause geomechanical and seismic property changes. To identify changes in properties due to CO2 injection, we designed four geomechanical and one geochemical test with supercritical CO2 (SC.CO2) injection into brine-saturated North Sea chalk. Two mechanical tests only involved SC.CO2 injection and two tests also included creep and brine injection phases. For the geochemical test, the brine was first injected through a dummy chalk sample to mimic reservoir conditions in the geochemical experiment. Mechanical tests involved injection at either constant stress below or above pore collapse. Ultrasonic wave velocities were measured throughout, allowing quantification of potential softening due to injection, as well as modelling of SC.CO2 saturation. The modelled SC.CO2 saturation shows that 1.5 pore volumes injected SC.CO2 cannot fully replace the saturating brine. The achieved SC.CO2 saturation is higher with injection below pore collapse stress than above, and brine injection appears to replace the injected SC.CO2 to a level below detection. Softening from SC.CO2 and brine injections (calcite dissolution) is indicated by increasing Biot's coefficient, yet counteracted by time-dependent compaction.
{"title":"Laboratory measurement of ultrasonic wave velocities in chalk - effect of supercritical CO2 and brine injection. Modelling stiffness and saturation by using rock physics","authors":"Tobias Orlander , Frederik Peter Ditlevsen , Hanne Dahl Holmslykke , Leonardo Teixeira Pinto Meireles , Amirhossein Shamsolhodaei , Ida Lykke Fabricius","doi":"10.1016/j.geoen.2025.214327","DOIUrl":"10.1016/j.geoen.2025.214327","url":null,"abstract":"<div><div>Denmark is currently exploring options to reuse depleted North Sea chalk oil and gas reservoirs for CO<sub>2</sub> storage. Monitoring CO<sub>2</sub> saturation, well-bore CO<sub>2</sub> leakage, and plume propagation are among the challenges, with well-bore logging and seismic time-lapse techniques as likely instruments. Such instruments rely on an understanding of how elastic waves respond to changes in the saturating fluids. For chalk reservoirs, mineral dissolution and/or precipitation caused by CO<sub>2</sub> injection may cause geomechanical and seismic property changes. To identify changes in properties due to CO<sub>2</sub> injection, we designed four geomechanical and one geochemical test with supercritical CO<sub>2</sub> (SC.CO<sub>2</sub>) injection into brine-saturated North Sea chalk. Two mechanical tests only involved SC.CO<sub>2</sub> injection and two tests also included creep and brine injection phases. For the geochemical test, the brine was first injected through a dummy chalk sample to mimic reservoir conditions in the geochemical experiment. Mechanical tests involved injection at either constant stress below or above pore collapse. Ultrasonic wave velocities were measured throughout, allowing quantification of potential softening due to injection, as well as modelling of SC.CO<sub>2</sub> saturation. The modelled SC.CO<sub>2</sub> saturation shows that 1.5 pore volumes injected SC.CO<sub>2</sub> cannot fully replace the saturating brine. The achieved SC.CO<sub>2</sub> saturation is higher with injection below pore collapse stress than above, and brine injection appears to replace the injected SC.CO<sub>2</sub> to a level below detection. Softening from SC.CO<sub>2</sub> and brine injections (calcite dissolution) is indicated by increasing Biot's coefficient, yet counteracted by time-dependent compaction.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"259 ","pages":"Article 214327"},"PeriodicalIF":4.6,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145979800","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 : 2026-04-01Epub Date: 2026-01-02DOI: 10.1016/j.geoen.2025.214357
Masahiro Nagao , Akhil Datta-Gupta
Monitoring the CO2 plume movement in the subsurface is essential for safety and storage integrity during CO2 storage and utilization. During field-scale CO2 EOR, routine well-wise injection/production data contain significant information which can be used for closed-loop reservoir management. Traditional physics-based numerical reservoir simulation can be computationally prohibitive for short-term decision cycles, and it requires detailed geologic models. As an alternative, reduced physics models provide an efficient simulator-free workflow but often have a limited range of applicability. Pure machine learning models lack physical interpretability and fail to provide process-based insights, resulting in limited predictive power. To address these challenges, we propose hybrid models, combining machine learning and physics-based approach, for rapid production forecasting and reservoir connectivity characterization using routine injection/production and pressure data collected during CO2 EOR operation.
We combine reduced physics models into a neural network architecture by utilizing two different approaches. In the first approach, the reduced physics model is used for pre-processing to obtain approximate solutions that feed into a neural network as input. This physics-based input feature can reduce the model complexity and provide significant improvement in prediction performance. In the second approach, physics-informed neural network (PINN) is applied. The residual terms are augmented in the neural network loss function using a physics-based regularization that relies on the governing partial differential equations (PDE). Reduced physics models are used for the governing PDE to enable efficient neural network training.
Our proposed hybrid models are first validated using a benchmark reservoir simulation case and then applied to a field case to show the robustness and efficacy of the method. The hybrid models are shown to provide superior prediction performance than pure machine learning models in terms of multiphase production rates. Specifically, the trained PINN model satisfies the reduced physics system, providing inter-well connectivity in terms of well-flux allocation. The flux allocation estimated from the hybrid model was compared with streamline-based flux allocation, and reasonable agreement was obtained for both the benchmark case and the field case.
{"title":"Physics-informed machine learning for identification of preferential flow paths and performance forecasting of subsurface carbon storage and utilization","authors":"Masahiro Nagao , Akhil Datta-Gupta","doi":"10.1016/j.geoen.2025.214357","DOIUrl":"10.1016/j.geoen.2025.214357","url":null,"abstract":"<div><div>Monitoring the CO2 plume movement in the subsurface is essential for safety and storage integrity during CO2 storage and utilization. During field-scale CO2 EOR, routine well-wise injection/production data contain significant information which can be used for closed-loop reservoir management. Traditional physics-based numerical reservoir simulation can be computationally prohibitive for short-term decision cycles, and it requires detailed geologic models. As an alternative, reduced physics models provide an efficient simulator-free workflow but often have a limited range of applicability. Pure machine learning models lack physical interpretability and fail to provide process-based insights, resulting in limited predictive power. To address these challenges, we propose hybrid models, combining machine learning and physics-based approach, for rapid production forecasting and reservoir connectivity characterization using routine injection/production and pressure data collected during CO2 EOR operation.</div><div>We combine reduced physics models into a neural network architecture by utilizing two different approaches. In the first approach, the reduced physics model is used for pre-processing to obtain approximate solutions that feed into a neural network as input. This physics-based input feature can reduce the model complexity and provide significant improvement in prediction performance. In the second approach, physics-informed neural network (PINN) is applied. The residual terms are augmented in the neural network loss function using a physics-based regularization that relies on the governing partial differential equations (PDE). Reduced physics models are used for the governing PDE to enable efficient neural network training.</div><div>Our proposed hybrid models are first validated using a benchmark reservoir simulation case and then applied to a field case to show the robustness and efficacy of the method. The hybrid models are shown to provide superior prediction performance than pure machine learning models in terms of multiphase production rates. Specifically, the trained PINN model satisfies the reduced physics system, providing inter-well connectivity in terms of well-flux allocation. The flux allocation estimated from the hybrid model was compared with streamline-based flux allocation, and reasonable agreement was obtained for both the benchmark case and the field case.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"259 ","pages":"Article 214357"},"PeriodicalIF":4.6,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145979735","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 : 2026-04-01Epub Date: 2026-01-18DOI: 10.1016/j.geoen.2026.214370
Tao Wan , Yan Dong , Longxin Wang , Jixiang He , Bing Hou
In CO2 energized fracturing, injected CO2 not only supplies additional energy and promotes complex fracture networks, but also induces coupled physical–chemical effects that may significantly modify reservoir properties. In order to examine how CO2 energized fracturing influences rock properties and fracture morphology in the Jimsar shale oil reservoir, laboratory experiments were conducted on outcrop core samples under in-situ temperature and pressure. Integrated characterization techniques, including X-ray diffraction (XRD), computed tomography (CT) scanning, contact-angle measurements, porosity–permeability tests and pressure monitoring, were employed to reveal fracture propagation mechanisms and quantify CO2-induced alterations in rock properties. Key findings reveal that: (1) CO2 exposure significantly alters shale wettability, shifting it from strongly water-/oil-wet to a near-neutral state. As the pressure increases to 16 MPa, the water-wet contact angle on the rock surface decreases from 148° to 79°, while the oil-wet contact angle increases from 45° to 82°. (2) The CO2 aqueous solution significantly dissolves carbonate minerals, preferentially attacking calcite over dolomite. After 14 days, permeability increases by roughly 2.5 times and porosity increases by about 45 % compared to the raw state. (3) Bedding planes are readily activated during CO2 fracturing, diminishing the critical role of horizontal stress difference in complex fracture network formation. (4) The fracture complexity induced by different fracturing media differs significantly. The fracture complexity ranking as SC-CO2 > CO2 energization > slickwater > Gel. Consequently, SC-CO2 energized fracturing demonstrates significant potential for field application.
{"title":"Impact of CO2 energized fracturing on fracture morphology and rock properties applied in Jimsar shale oil","authors":"Tao Wan , Yan Dong , Longxin Wang , Jixiang He , Bing Hou","doi":"10.1016/j.geoen.2026.214370","DOIUrl":"10.1016/j.geoen.2026.214370","url":null,"abstract":"<div><div>In CO<sub>2</sub> energized fracturing, injected CO<sub>2</sub> not only supplies additional energy and promotes complex fracture networks, but also induces coupled physical–chemical effects that may significantly modify reservoir properties. In order to examine how CO<sub>2</sub> energized fracturing influences rock properties and fracture morphology in the Jimsar shale oil reservoir, laboratory experiments were conducted on outcrop core samples under in-situ temperature and pressure. Integrated characterization techniques, including X-ray diffraction (XRD), computed tomography (CT) scanning, contact-angle measurements, porosity–permeability tests and pressure monitoring, were employed to reveal fracture propagation mechanisms and quantify CO<sub>2</sub>-induced alterations in rock properties. Key findings reveal that: (1) CO<sub>2</sub> exposure significantly alters shale wettability, shifting it from strongly water-/oil-wet to a near-neutral state. As the pressure increases to 16 MPa, the water-wet contact angle on the rock surface decreases from 148° to 79°, while the oil-wet contact angle increases from 45° to 82°. (2) The CO<sub>2</sub> aqueous solution significantly dissolves carbonate minerals, preferentially attacking calcite over dolomite. After 14 days, permeability increases by roughly 2.5 times and porosity increases by about 45 % compared to the raw state. (3) Bedding planes are readily activated during CO<sub>2</sub> fracturing, diminishing the critical role of horizontal stress difference in complex fracture network formation. (4) The fracture complexity induced by different fracturing media differs significantly. The fracture complexity ranking as SC-CO<sub>2</sub> > CO<sub>2</sub> energization > slickwater > Gel. Consequently, SC-CO<sub>2</sub> energized fracturing demonstrates significant potential for field application.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"259 ","pages":"Article 214370"},"PeriodicalIF":4.6,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146039132","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 : 2026-04-01Epub Date: 2026-01-16DOI: 10.1016/j.geoen.2026.214371
Hongwei Yang , Biao Wang , Jun Li , Geng Zhang , Jiahao Zhan , Gonghui Liu , Zhenyu Long , Wenbo Zhang
To address the challenges in predicting leakage rate and location during drilling in deep and complex formations, this study proposes a quantitative interpretation method for leakage conditions based on Physics-Informed Neural Networks (PINN). By utilizing automatic differentiation techniques, the method employs neural networks to solve the wellbore hydraulics model under leakage conditions, enabling intelligent and accurate predictions of both leakage rate and location. The neural network takes time and well depth as inputs, while its outputs include annulus flow velocity, pressure, and leakage rate, with the leakage location treated as a trainable variable. The total loss function is composed of several components: the residuals of the mass and momentum conservation equations, prediction errors of dual-point pressure data obtained from a downhole dual-measurement tool, prediction errors of wellhead pressure data, and errors derived from expert knowledge constraints. The Gradnorm algorithm is employed to assign weights to each loss term adaptively, and the neural network is trained by minimizing the total loss. Test results demonstrate that the neural network model trained with this approach can efficiently and reliably solve the wellbore hydraulics model under leakage conditions. Physical constraints are satisfied throughout the input–output process, achieving a mean relative error (MRE) of less than 10 % for leakage rate predictions and an absolute error (AE) within 20 m for leakage location predictions. Compared with methods such as the Unscented Kalman Filter (UKF) and Genetic Algorithm (GA), this approach, which leverages a global optimization strategy and intrinsic physical constraints, exhibits superior stability and accuracy across different noise levels. When integrated with the downhole dual-measurement tool, this approach provides critical guidance for leakage mitigation operations during drilling processes in deep and ultra-deep wells.
{"title":"A quantitative interpretation method for leakage conditions based on Physics-Informed Neural Networks","authors":"Hongwei Yang , Biao Wang , Jun Li , Geng Zhang , Jiahao Zhan , Gonghui Liu , Zhenyu Long , Wenbo Zhang","doi":"10.1016/j.geoen.2026.214371","DOIUrl":"10.1016/j.geoen.2026.214371","url":null,"abstract":"<div><div>To address the challenges in predicting leakage rate and location during drilling in deep and complex formations, this study proposes a quantitative interpretation method for leakage conditions based on Physics-Informed Neural Networks (PINN). By utilizing automatic differentiation techniques, the method employs neural networks to solve the wellbore hydraulics model under leakage conditions, enabling intelligent and accurate predictions of both leakage rate and location. The neural network takes time and well depth as inputs, while its outputs include annulus flow velocity, pressure, and leakage rate, with the leakage location treated as a trainable variable. The total loss function is composed of several components: the residuals of the mass and momentum conservation equations, prediction errors of dual-point pressure data obtained from a downhole dual-measurement tool, prediction errors of wellhead pressure data, and errors derived from expert knowledge constraints. The Gradnorm algorithm is employed to assign weights to each loss term adaptively, and the neural network is trained by minimizing the total loss. Test results demonstrate that the neural network model trained with this approach can efficiently and reliably solve the wellbore hydraulics model under leakage conditions. Physical constraints are satisfied throughout the input–output process, achieving a mean relative error (MRE) of less than 10 % for leakage rate predictions and an absolute error (AE) within 20 m for leakage location predictions. Compared with methods such as the Unscented Kalman Filter (UKF) and Genetic Algorithm (GA), this approach, which leverages a global optimization strategy and intrinsic physical constraints, exhibits superior stability and accuracy across different noise levels. When integrated with the downhole dual-measurement tool, this approach provides critical guidance for leakage mitigation operations during drilling processes in deep and ultra-deep wells.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"259 ","pages":"Article 214371"},"PeriodicalIF":4.6,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146038648","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 : 2026-04-01Epub Date: 2025-12-23DOI: 10.1016/j.geoen.2025.214352
Jorge Rodrigo Lliguizaca-Davila , P.E. Valverde-Armas , Hilde Halsøy , Jorge Segundo Mendoza Sanz , Arne Graue , Jacquelin E. Cobos , Bergit Brattekås , Zachary Paul Alcorn
Foam technologies reduce CO2 mobility, thereby enhancing sweep efficiency and potentially benefiting CO2 storage and enhanced oil recovery (EOR). However, maintaining foam strength and stability under reservoir conditions remains challenging. This study examined whether adding hydrolyzed polyacrylamide (HPAM) polymers to a nonionic surfactant solution to create CO2 polymer enhanced foam (PEF) improves mobility control. Dynamic CO2 mobility tests were conducted on four 9 cm Bentheimer short cores (BSC) and one 83 cm Berea long core (BLC), both initially saturated with 100 % brine at 90 bar and 40 °C. The tests included coinjections at 70 % foam quality and CO2 chase injections at flow rates of 2 and 4 ft/D to assess foam characteristics, including foam generation, strength, stability, residual mobility reduction, and propagation, by analyzing the DP measurements, mobility reduction factor (MRF), and apparent viscosity (μapp). The DP, MRF, and μapp values during and after CO2 PEF were higher than those of CO2 foam, indicating that PEF achieved a greater reduction in CO2 mobility. For example, the maximum MRF during CO2 PEF was 11.03 and 5.59 versus CO2 foam maxima of 1.91 and 2.07 (BSC1 and BSC2 at 2 ft/D). Moreover, pure CO2 injection post PEF reduced the MRFs to 6.50 and 2.33 (BSC1 and BSC2 at 2 ft/D), whereas the MRF post CO2 foam decreased below the baseline (MRF = 1). Similarly, the maximum MRF value at 4 ft/D (BSC3) for CO2 foam was 3.19, whereas CO2 PEF flooding reached a maximum MRF of 22.79 (mean 12.48). In addition, during CO2 injection after PEF, the MRFs ranged from 8.91 to 11.68, indicating sustained mobility control without continuous chemical injection. The improved flow resistance is attributed to increased foam viscosity and lamellar elasticity, as well as to blockage of flow paths by polymer retention. The μapp comparison from the tests at 2 and 4 ft/D demonstrated the shear-thinning behaviors of foam and PEF. At 2 ft/D in BSC1, the μapp values were 10.36 cP and 78.84 cP, whereas at 4 ft/D in BSC3, they were 9.78 cP and 67.74 cP for CO2 foam and CO2 PEF, respectively. This result may imply an advantage in mitigating injectivity challenges. The sectional pressure analysis of the 83 cm BLC showed greater upstream resistance and delayed downstream propagation under PEF, consistent with diversion by foam and with polymer-induced channel-blocking effects. This study shows that HPAM PEF markedly increases CO2 mobility control compared to CO2 foam during and after CO2 PEF flooding.
{"title":"HPAM polymer enhanced foam for CO2 mobility control: A coreflooding experimental study","authors":"Jorge Rodrigo Lliguizaca-Davila , P.E. Valverde-Armas , Hilde Halsøy , Jorge Segundo Mendoza Sanz , Arne Graue , Jacquelin E. Cobos , Bergit Brattekås , Zachary Paul Alcorn","doi":"10.1016/j.geoen.2025.214352","DOIUrl":"10.1016/j.geoen.2025.214352","url":null,"abstract":"<div><div>Foam technologies reduce CO<sub>2</sub> mobility, thereby enhancing sweep efficiency and potentially benefiting CO<sub>2</sub> storage and enhanced oil recovery (EOR). However, maintaining foam strength and stability under reservoir conditions remains challenging. This study examined whether adding hydrolyzed polyacrylamide (HPAM) polymers to a nonionic surfactant solution to create CO<sub>2</sub> polymer enhanced foam (PEF) improves mobility control. Dynamic CO<sub>2</sub> mobility tests were conducted on four 9 cm Bentheimer short cores (BSC) and one 83 cm Berea long core (BLC), both initially saturated with 100 % brine at 90 bar and 40 °C. The tests included coinjections at 70 % foam quality and CO<sub>2</sub> chase injections at flow rates of 2 and 4 ft/D to assess foam characteristics, including foam generation, strength, stability, residual mobility reduction, and propagation, by analyzing the DP measurements, mobility reduction factor (MRF), and apparent viscosity (μ<sub>app</sub>). The DP, MRF, and μ<sub>app</sub> values during and after CO<sub>2</sub> PEF were higher than those of CO<sub>2</sub> foam, indicating that PEF achieved a greater reduction in CO<sub>2</sub> mobility. For example, the maximum MRF during CO<sub>2</sub> PEF was 11.03 and 5.59 versus CO<sub>2</sub> foam maxima of 1.91 and 2.07 (BSC1 and BSC2 at 2 ft/D). Moreover, pure CO<sub>2</sub> injection post PEF reduced the MRFs to 6.50 and 2.33 (BSC1 and BSC2 at 2 ft/D), whereas the MRF post CO<sub>2</sub> foam decreased below the baseline (MRF = 1). Similarly, the maximum MRF value at 4 ft/D (BSC3) for CO<sub>2</sub> foam was 3.19, whereas CO<sub>2</sub> PEF flooding reached a maximum MRF of 22.79 (mean 12.48). In addition, during CO<sub>2</sub> injection after PEF, the MRFs ranged from 8.91 to 11.68, indicating sustained mobility control without continuous chemical injection. The improved flow resistance is attributed to increased foam viscosity and lamellar elasticity, as well as to blockage of flow paths by polymer retention. The μ<sub>app</sub> comparison from the tests at 2 and 4 ft/D demonstrated the shear-thinning behaviors of foam and PEF. At 2 ft/D in BSC1, the μ<sub>app</sub> values were 10.36 cP and 78.84 cP, whereas at 4 ft/D in BSC3, they were 9.78 cP and 67.74 cP for CO<sub>2</sub> foam and CO<sub>2</sub> PEF, respectively. This result may imply an advantage in mitigating injectivity challenges. The sectional pressure analysis of the 83 cm BLC showed greater upstream resistance and delayed downstream propagation under PEF, consistent with diversion by foam and with polymer-induced channel-blocking effects. This study shows that HPAM PEF markedly increases CO<sub>2</sub> mobility control compared to CO<sub>2</sub> foam during and after CO<sub>2</sub> PEF flooding.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"259 ","pages":"Article 214352"},"PeriodicalIF":4.6,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145903919","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 : 2026-04-01Epub Date: 2026-01-17DOI: 10.1016/j.geoen.2026.214374
Peng Lu , Nikolaos Michael , Abrar Alabbad , Rainer Zühlke , Christopher Ellis , Aaron Ketchmark , Chris Paola , Hussain Al-Qatari
We conducted a year-long flume-tank experiment to investigate early marine cementation along a supratidal-to-subtidal carbonate profile. Artificial seawater and cyclic water-level oscillations were used to generate supratidal, intertidal, and subtidal diagenetic environments, while periodic sediment addition simulated subtidal deposition of reworked carbonate sands. Monthly sampling was conducted, with four samples collected across the entire profile to characterize each depositional environment.
Petrographic observations show that the amount of aragonite cement formed over the one-year period ranges from ∼0 to 4 % of total sediment (grain + porosity), with an estimated average cement fraction of ∼1.47 % based on diagenetic modeling. The calculated cementation rate is 1.96 × 10−9 mol/m2/s, which is comparable to the aragonite growth rate measured in a recent mixed-flow reactor experiment reported in the literature.
The experimentally derived rate constant was then implemented in reactive-transport models of an upscaled carbonate ramp to estimate the timescales required to develop a significant cementation front (10 % cement) under Phanerozoic seawater chemistry and atmospheric CO2 levels. Modeled cementation times range from 8.9 to 14.3 kyr and are strongly facies-dependent, with faster cementation in calcite seas than in aragonite seas. Cementation times are inversely correlated with carbonate saturation state. By explicitly coupling hydrodynamics, cementation kinetics, and pore-space evolution, this framework improves predictions of porosity and permeability evolution in carbonate reservoirs and subsurface CO2 storage systems.
{"title":"A diagenetic flume-tank experiment of carbonate marine cementation along a carbonate profile","authors":"Peng Lu , Nikolaos Michael , Abrar Alabbad , Rainer Zühlke , Christopher Ellis , Aaron Ketchmark , Chris Paola , Hussain Al-Qatari","doi":"10.1016/j.geoen.2026.214374","DOIUrl":"10.1016/j.geoen.2026.214374","url":null,"abstract":"<div><div>We conducted a year-long flume-tank experiment to investigate early marine cementation along a supratidal-to-subtidal carbonate profile. Artificial seawater and cyclic water-level oscillations were used to generate supratidal, intertidal, and subtidal diagenetic environments, while periodic sediment addition simulated subtidal deposition of reworked carbonate sands. Monthly sampling was conducted, with four samples collected across the entire profile to characterize each depositional environment.</div><div>Petrographic observations show that the amount of aragonite cement formed over the one-year period ranges from ∼0 to 4 % of total sediment (grain + porosity), with an estimated average cement fraction of ∼1.47 % based on diagenetic modeling. The calculated cementation rate is 1.96 × 10<sup>−9</sup> mol/m<sup>2</sup>/s, which is comparable to the aragonite growth rate measured in a recent mixed-flow reactor experiment reported in the literature.</div><div>The experimentally derived rate constant was then implemented in reactive-transport models of an upscaled carbonate ramp to estimate the timescales required to develop a significant cementation front (10 % cement) under Phanerozoic seawater chemistry and atmospheric CO<sub>2</sub> levels. Modeled cementation times range from 8.9 to 14.3 kyr and are strongly facies-dependent, with faster cementation in calcite seas than in aragonite seas. Cementation times are inversely correlated with carbonate saturation state. By explicitly coupling hydrodynamics, cementation kinetics, and pore-space evolution, this framework improves predictions of porosity and permeability evolution in carbonate reservoirs and subsurface CO<sub>2</sub> storage systems.</div></div>","PeriodicalId":100578,"journal":{"name":"Geoenergy Science and Engineering","volume":"259 ","pages":"Article 214374"},"PeriodicalIF":4.6,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146039135","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}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号