Pub Date : 2026-10-01Epub Date: 2026-10-12DOI: 10.1016/j.ijggc.2026.104806
Tuvshinjargal Otgonbayar, Marco Mazzotti
Carbon dioxide removal (CDR) is increasingly recognized as a key component of climate strategies to offset unabatable residual emissions on the path to net zero. In recent years, corporations in particular have shown growing interest in using CDR to offset their historical or “legacy” emissions. However, the prevailing assumption that one tonne of CO2 removed today cancels a tonne emitted in the past neglects the temporal mismatch between emissions and removals, during which CO2 exerts a warming effect. This paper considers specific entities that emit a very small fraction of global emissions, have a net-zero carbon goal and pledge to compensate for their legacy emissions. It examines the effectiveness of such compensation through climate consequences of delayed CDR, arguing that a simple mass-balance approach may not be sufficient. Using a simplified linear climate model, we analyze the impact of the timing and rate of CDR deployment on two complementary climate metrics: peak surface temperature and abyssal ocean temperature anomaly, reflecting direct short-term and long-term cumulative impacts, respectively. Our results show that earlier CDR deployment, even at slower rates, yields a more favorable climate outcome than delayed, faster deployment. The analysis also reveals that removing more CO2 than was originally emitted can help mitigate the effects of delayed action, particularly for long-term climate impacts. These findings highlight the importance of accounting for temporal dynamics in climate mitigation strategies. Our work aims to inform corporate climate policies and encourage further scientific discourse on the dynamic effects of legacy emissions and legacy removals.
{"title":"Climate impact of delayed carbon dioxide removal by a single emitting entity to compensate for their legacy emissions","authors":"Tuvshinjargal Otgonbayar, Marco Mazzotti","doi":"10.1016/j.ijggc.2026.104806","DOIUrl":"10.1016/j.ijggc.2026.104806","url":null,"abstract":"<div><div>Carbon dioxide removal (CDR) is increasingly recognized as a key component of climate strategies to offset unabatable residual emissions on the path to net zero. In recent years, corporations in particular have shown growing interest in using CDR to offset their historical or “legacy” emissions. However, the prevailing assumption that one tonne of CO<sub>2</sub> removed today cancels a tonne emitted in the past neglects the temporal mismatch between emissions and removals, during which CO<sub>2</sub> exerts a warming effect. This paper considers specific entities that emit a very small fraction of global emissions, have a net-zero carbon goal and pledge to compensate for their legacy emissions. It examines the effectiveness of such compensation through climate consequences of delayed CDR, arguing that a simple mass-balance approach may not be sufficient. Using a simplified linear climate model, we analyze the impact of the timing and rate of CDR deployment on two complementary climate metrics: peak surface temperature and abyssal ocean temperature anomaly, reflecting direct short-term and long-term cumulative impacts, respectively. Our results show that earlier CDR deployment, even at slower rates, yields a more favorable climate outcome than delayed, faster deployment. The analysis also reveals that removing more CO<sub>2</sub> than was originally emitted can help mitigate the effects of delayed action, particularly for long-term climate impacts. These findings highlight the importance of accounting for temporal dynamics in climate mitigation strategies. Our work aims to inform corporate climate policies and encourage further scientific discourse on the dynamic effects of legacy emissions and legacy removals.</div></div>","PeriodicalId":334,"journal":{"name":"International Journal of Greenhouse Gas Control","volume":"156 ","pages":"Article 104806"},"PeriodicalIF":5.9,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149200973","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-10-01Epub Date: 2026-10-06DOI: 10.1016/j.ijggc.2026.104805
Onni Linjala, Lauri Kujanpää
This study examines the availability, infrastructure demand and supply costs of biogenic CO2 for carbon capture, utilization and storage (CCUS) in Finland. A heuristic geospatial allocation model that aims to minimize CO2 transport demand is developed to evaluate potential CO2 supply and hub configurations under policy-based CCUS deployment scenarios. The examined large industrial facilities in Finland emit 29 Mt of biogenic CO2 annually. Due to the lack of domestic geological storage capacity and the need for CO2 shipping, the model consistently forms CCS hubs near seaports adjacent to large emission sources. Within the scenarios and system boundary examined, centralized CCU hubs can reduce investments, but the cost benefit is largely offset by increased transport demand. Seasonally variable sources have higher specific investment costs when dimensioned for peak load but excluding such sources from a hub-based CCUS system appears to yield little or no economic advantage because the hub areas expand and transport demand increases. Across the scenarios examined, the levelized cost of CO2 supply, including capture, compression and transport to CCS hub ports or utilization sites, ranges from 79 to 95 €/tCO2. With a CO2 capture volume of 14.8 Mtpa, investments in CO2 supply infrastructure range from EUR 3.5 to 5.0 billion. For CO2 utilization capacities ranging from 3.7 to 11.1 Mtpa, investments in CO2 utilization and hydrogen production infrastructure range from EUR 3.6 to 9.5 billion. The findings provide insights into CCUS infrastructure planning in regions with abundant CO2 sources but limited domestic geological storage capacity.
{"title":"Geospatial allocation of biogenic CO2 supply and demand for CCUS in Finland: A scenario analysis","authors":"Onni Linjala, Lauri Kujanpää","doi":"10.1016/j.ijggc.2026.104805","DOIUrl":"10.1016/j.ijggc.2026.104805","url":null,"abstract":"<div><div>This study examines the availability, infrastructure demand and supply costs of biogenic CO2 for carbon capture, utilization and storage (CCUS) in Finland. A heuristic geospatial allocation model that aims to minimize CO2 transport demand is developed to evaluate potential CO2 supply and hub configurations under policy-based CCUS deployment scenarios. The examined large industrial facilities in Finland emit 29 Mt of biogenic CO2 annually. Due to the lack of domestic geological storage capacity and the need for CO2 shipping, the model consistently forms CCS hubs near seaports adjacent to large emission sources. Within the scenarios and system boundary examined, centralized CCU hubs can reduce investments, but the cost benefit is largely offset by increased transport demand. Seasonally variable sources have higher specific investment costs when dimensioned for peak load but excluding such sources from a hub-based CCUS system appears to yield little or no economic advantage because the hub areas expand and transport demand increases. Across the scenarios examined, the levelized cost of CO2 supply, including capture, compression and transport to CCS hub ports or utilization sites, ranges from 79 to 95 €/tCO2. With a CO2 capture volume of 14.8 Mtpa, investments in CO2 supply infrastructure range from EUR 3.5 to 5.0 billion. For CO2 utilization capacities ranging from 3.7 to 11.1 Mtpa, investments in CO2 utilization and hydrogen production infrastructure range from EUR 3.6 to 9.5 billion. The findings provide insights into CCUS infrastructure planning in regions with abundant CO2 sources but limited domestic geological storage capacity.</div></div>","PeriodicalId":334,"journal":{"name":"International Journal of Greenhouse Gas Control","volume":"156 ","pages":"Article 104805"},"PeriodicalIF":5.9,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149160740","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-10-01Epub Date: 2026-10-07DOI: 10.1016/j.ijggc.2026.104808
Amir Hossein Nazari, Mohammadali Ahmadi
Selecting where to permanently store CO₂ across the offshore Gulf of Mexico (GoM) is a high-stakes decision that precedes every well, requiring thousands of candidate grid blocks to be ranked from sparse, spatially correlated evidence. The prevailing approach, expert-weighted multi-criteria decision analysis (MCDA), encodes this choice in subjective, scenario-dependent weights that cannot be validated, quantify no uncertainty, and predict none of the physical reservoir properties, storage capacity and injectivity, that ultimately govern where CO₂ can be stored. Data-driven models promise objectivity, but naïve machine learning (ML) on spatial geoscience data silently leaks information across neighboring blocks, inflates reported accuracy, recycles the very score it should predict (circularity), and returns point estimates with no error bars. Using the 2,559-block NETL GoM screening dataset as the sole data source, we build a leakage-controlled pipeline: an anti-circularity raw-feature whitelist, physics-based interaction features, per-target leakage whitelists, and a single spatial-block split tuned to match the train/test distribution of every target. Six complementary spatial learners, a physics-motivated gradient-boosting stack, a graph neural network, a physics-informed network, a geospatial transformer, a tabular foundation model, and a Gaussian-process kriging model, predict storage capacity, injectivity, and a multiplicative storage-deliverability index, each wrapped in distribution-free conformal uncertainty and explained with a common SHAP protocol. The best learner reaches a test R2 of 0.98 on the composite suitability target under the original predictor set, and 0.89 under a stricter predictor set that additionally excludes two ordinal variables encoding part of that target's own containment-risk term; nominal 90% conformal intervals achieve empirical coverage of 0.83 to 0.95, and the data-derived suitability index reproduces the experts' geologic-storage rankings (quadratic-weighted κ up to 0.65). Because the composite target is constructed rather than observed, this figure quantifies learnability under leakage-safe spatial validation rather than validated storage performance. Where our ranking diverges from the expert screen, we show the divergence is driven by the weight carried by hydrocarbon-derived criteria rather than by superior storage quality, and we report it as a criterion-weighting diagnostic. The result is a reproducible, uncertainty-aware suitability surface for regional high-grading, intended to prioritize blocks for subsequent characterization rather than to initialize dynamic reservoir simulation.
{"title":"From expert weighting to data-derived evidence: A leakage-safe, physics-motivated, uncertainty-aware machine-learning framework for CO₂ storage site screening in the Gulf of Mexico","authors":"Amir Hossein Nazari, Mohammadali Ahmadi","doi":"10.1016/j.ijggc.2026.104808","DOIUrl":"10.1016/j.ijggc.2026.104808","url":null,"abstract":"<div><div>Selecting where to permanently store CO₂ across the offshore Gulf of Mexico (GoM) is a high-stakes decision that precedes every well, requiring thousands of candidate grid blocks to be ranked from sparse, spatially correlated evidence. The prevailing approach, expert-weighted multi-criteria decision analysis (MCDA), encodes this choice in subjective, scenario-dependent weights that cannot be validated, quantify no uncertainty, and predict none of the physical reservoir properties, storage capacity and injectivity, that ultimately govern where CO₂ can be stored. Data-driven models promise objectivity, but naïve machine learning (ML) on spatial geoscience data silently leaks information across neighboring blocks, inflates reported accuracy, recycles the very score it should predict (circularity), and returns point estimates with no error bars. Using the 2,559-block NETL GoM screening dataset as the sole data source, we build a leakage-controlled pipeline: an anti-circularity raw-feature whitelist, physics-based interaction features, per-target leakage whitelists, and a single spatial-block split tuned to match the train/test distribution of every target. Six complementary spatial learners, a physics-motivated gradient-boosting stack, a graph neural network, a physics-informed network, a geospatial transformer, a tabular foundation model, and a Gaussian-process kriging model, predict storage capacity, injectivity, and a multiplicative storage-deliverability index, each wrapped in distribution-free conformal uncertainty and explained with a common SHAP protocol. The best learner reaches a test R<sup>2</sup> of 0.98 on the composite suitability target under the original predictor set, and 0.89 under a stricter predictor set that additionally excludes two ordinal variables encoding part of that target's own containment-risk term; nominal 90% conformal intervals achieve empirical coverage of 0.83 to 0.95, and the data-derived suitability index reproduces the experts' geologic-storage rankings (quadratic-weighted κ up to 0.65). Because the composite target is constructed rather than observed, this figure quantifies learnability under leakage-safe spatial validation rather than validated storage performance. Where our ranking diverges from the expert screen, we show the divergence is driven by the weight carried by hydrocarbon-derived criteria rather than by superior storage quality, and we report it as a criterion-weighting diagnostic. The result is a reproducible, uncertainty-aware suitability surface for regional high-grading, intended to prioritize blocks for subsequent characterization rather than to initialize dynamic reservoir simulation.</div></div>","PeriodicalId":334,"journal":{"name":"International Journal of Greenhouse Gas Control","volume":"156 ","pages":"Article 104808"},"PeriodicalIF":5.9,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149160742","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}
Pub Date : 2026-10-01Epub Date: 2026-10-14DOI: 10.1016/j.ijggc.2026.104811
Aowang Wang, Tingting Luo, Liwei Zhang, Xiang Sun, Tao Han, Zhen Zhang, Yanghui Li, Yongchen Song, Ye Zhang, Madhusudhan B N, Weihao Yang
Carbon Capture, Utilization, and Storage (CCUS) is a critical pathway for mitigating the global greenhouse effect. Sub-seabed CO2 sequestration, particularly through hydrate-based methods, offers immense potential due to the vast storage capacity and favorable thermodynamic conditions of marine sediments. However, the complex phase transitions and associated geomechanical deformations pose significant challenges for in-situ monitoring of CO2 evolution and reservoir stability. This study utilizes a triaxial-resistivity apparatus to investigate the electrical response of silty-clayey sediments during CO2 hydrate formation and subsequent deformation under varying saturations and effective stresses. Key findings indicate that resistivity is highly sensitive to temperature and pore fluid content, with a 10 % difference in initial water saturation leading to an 88 % variation in resistivity. Furthermore, hydrate morphology and residual water distribution significantly influence electrical signals, resulting in resistivity fluctuations of up to 853 Ω·m at identical saturation levels. Beyond phase identification, resistivity effectively reflects sediment consolidation and pore structure evolution during the shearing stage, exhibiting a non-linear correlation with applied stress. These results demonstrate the feasibility of resistivity-based monitoring for tracking CO2 dynamics and assessing the mechanical integrity of sub-seabed reservoirs, providing essential theoretical support for large-scale sequestration in the future.
{"title":"Experimental study on the electrical resistivity characteristics of CO2 hydrate-bearing silty-clayey sediments under formation stress","authors":"Aowang Wang, Tingting Luo, Liwei Zhang, Xiang Sun, Tao Han, Zhen Zhang, Yanghui Li, Yongchen Song, Ye Zhang, Madhusudhan B N, Weihao Yang","doi":"10.1016/j.ijggc.2026.104811","DOIUrl":"10.1016/j.ijggc.2026.104811","url":null,"abstract":"<div><div>Carbon Capture, Utilization, and Storage (CCUS) is a critical pathway for mitigating the global greenhouse effect. Sub-seabed CO<sub>2</sub> sequestration, particularly through hydrate-based methods, offers immense potential due to the vast storage capacity and favorable thermodynamic conditions of marine sediments. However, the complex phase transitions and associated geomechanical deformations pose significant challenges for in-situ monitoring of CO<sub>2</sub> evolution and reservoir stability. This study utilizes a triaxial-resistivity apparatus to investigate the electrical response of silty-clayey sediments during CO<sub>2</sub> hydrate formation and subsequent deformation under varying saturations and effective stresses. Key findings indicate that resistivity is highly sensitive to temperature and pore fluid content, with a 10 % difference in initial water saturation leading to an 88 % variation in resistivity. Furthermore, hydrate morphology and residual water distribution significantly influence electrical signals, resulting in resistivity fluctuations of up to 853 Ω·m at identical saturation levels. Beyond phase identification, resistivity effectively reflects sediment consolidation and pore structure evolution during the shearing stage, exhibiting a non-linear correlation with applied stress. These results demonstrate the feasibility of resistivity-based monitoring for tracking CO<sub>2</sub> dynamics and assessing the mechanical integrity of sub-seabed reservoirs, providing essential theoretical support for large-scale sequestration in the future.</div></div>","PeriodicalId":334,"journal":{"name":"International Journal of Greenhouse Gas Control","volume":"156 ","pages":"Article 104811"},"PeriodicalIF":5.9,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149200976","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}
Pub Date : 2026-09-01Epub Date: 2026-08-22DOI: 10.1016/j.ijggc.2026.104759
Iwona Galeczka, Martin Voigt, Daniel Andres Duque Carrillo, Kjartan Marteinsson, Thomas Ratouis, Sandra Ósk Snæbjörnsdóttir, Fríða Jónsdóttir, Paula Fernández Acosta, Gunnar Hrafn Gunnarsson, Bergur Sigfússon, Jóhann Gunnarsson Robin, Sif Pétursdóttir, Salka Kolbeinsdóttir, Magnús Á. Sigurgeirsson, Helga M. Helgadóttir, Gunnlaug Ásgeirsdóttir, Ólafur Elínarson
{"title":"Corrigendum to “Geochemical framework for CO2 mineralization in coastal aquifers: Lessons from the Coda Terminal project, Iceland” [International Journal of Greenhouse Gas Control 149 (2026) 104565]","authors":"Iwona Galeczka, Martin Voigt, Daniel Andres Duque Carrillo, Kjartan Marteinsson, Thomas Ratouis, Sandra Ósk Snæbjörnsdóttir, Fríða Jónsdóttir, Paula Fernández Acosta, Gunnar Hrafn Gunnarsson, Bergur Sigfússon, Jóhann Gunnarsson Robin, Sif Pétursdóttir, Salka Kolbeinsdóttir, Magnús Á. Sigurgeirsson, Helga M. Helgadóttir, Gunnlaug Ásgeirsdóttir, Ólafur Elínarson","doi":"10.1016/j.ijggc.2026.104759","DOIUrl":"10.1016/j.ijggc.2026.104759","url":null,"abstract":"","PeriodicalId":334,"journal":{"name":"International Journal of Greenhouse Gas Control","volume":"155 ","pages":"Article 104759"},"PeriodicalIF":5.9,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148958534","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-09-01Epub Date: 2026-08-28DOI: 10.1016/j.ijggc.2026.104763
Mikael Erlström, Peter Dahlqvist, Jan-Erik Rosberg
Two cored boreholes were drilled in 2023 through the Cambro–Silurian succession on the southernmost part of the island of Gotland as part of a project coordinated by the Geological Survey of Sweden to assess the CO2-storage potential in the Cambrian sandstone reservoirs in the south Baltic Sea. The investigations were performed in cooperation with the Swedish national research infrastructure for scientific drilling “Riksriggen”. We describe and assess the drilling, geophysical wire-line logging, and stratigraphical, lithological, and petrophysical analysis of the Cambro–Silurian succession. The results are discussed with respect to the reservoir and sealing properties for CO2-storage. The results indicate that the Faludden sandstone, along with the overlying Ordovician and Silurian caprock interval meets the necessary requirements for CO2-storage. In contrast the Cambrian Viklau and När sandstone intervals display less favourable reservoir conditions due to high degree of cementation contributing to very low permeability. Formation integrity tests and laboratory analyses verify an Ordovician and Silurian caprock interval with high compressive strength and low permeability. This new information strengthens the judgement that CO2 storage is feasible in the Swedish sector of the Baltic Sea.
{"title":"Characterization of the lower Palaeozoic regarding its CO2-storage potential in the Swedish sector of the Baltic Sea","authors":"Mikael Erlström, Peter Dahlqvist, Jan-Erik Rosberg","doi":"10.1016/j.ijggc.2026.104763","DOIUrl":"10.1016/j.ijggc.2026.104763","url":null,"abstract":"<div><div>Two cored boreholes were drilled in 2023 through the Cambro–Silurian succession on the southernmost part of the island of Gotland as part of a project coordinated by the Geological Survey of Sweden to assess the CO<sub>2</sub>-storage potential in the Cambrian sandstone reservoirs in the south Baltic Sea. The investigations were performed in cooperation with the Swedish national research infrastructure for scientific drilling “Riksriggen”. We describe and assess the drilling, geophysical wire-line logging, and stratigraphical, lithological, and petrophysical analysis of the Cambro–Silurian succession. The results are discussed with respect to the reservoir and sealing properties for CO<sub>2</sub>-storage. The results indicate that the Faludden sandstone, along with the overlying Ordovician and Silurian caprock interval meets the necessary requirements for CO<sub>2</sub>-storage. In contrast the Cambrian Viklau and När sandstone intervals display less favourable reservoir conditions due to high degree of cementation contributing to very low permeability. Formation integrity tests and laboratory analyses verify an Ordovician and Silurian caprock interval with high compressive strength and low permeability. This new information strengthens the judgement that CO<sub>2</sub> storage is feasible in the Swedish sector of the Baltic Sea.</div></div>","PeriodicalId":334,"journal":{"name":"International Journal of Greenhouse Gas Control","volume":"155 ","pages":"Article 104763"},"PeriodicalIF":5.9,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148853964","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-09-01Epub Date: 2026-08-30DOI: 10.1016/j.ijggc.2026.104766
Muhammad Ali, Ning Wei, Ren Jiang, Wendong Wang, Keyao Lin, Chang Jia
Hybrid reservoir property modeling is important for reliable CO2 storage assessment, particularly in areas with limited well control and strong reservoir heterogeneity. This study presents a Hybrid Geostatistics and Machine Learning workflow, termed HyGEM, to generate high-resolution reservoir property models for CO2 geological storage applications. The workflow integrates upscaled well-log data, multi-attribute seismic volumes, geostatistical modeling, and supervised machine learning to predict petrophysical properties within a 3D reservoir grid. Arithmetic averaging is first applied to upscale well-log data to the simulation-grid scale, helping preserve volumetric consistency and reduce high-frequency noise. Several seismic attributes, including Instantaneous Bandwidth, RMS Amplitude, Sweetness, RGB, Semblance, and thin-bed indicators, are then used to describe stratigraphic and structural variations around the injection zone. Based on these inputs, the workflow generates 3D models of porosity, permeability, water saturation, volume of shale, bulk density, and pore pressure while preserving vertical layering and lateral heterogeneity. Validation results show strong agreement with original and upscaled well log data, with improved geological continuity compared to conventional stochastic and deterministic approaches. The predicted property grids are further applied in dynamic CO2 injection simulation, where the results indicate controlled plume migration within high permeability zones, effective caprock containment, stable geomechanical behavior, and no clear indication of caprock failure during injection. These results demonstrate that the HyGEM workflow provides a practical and reliable approach for reservoir characterization, CO2 plume prediction, and long-term storage safety assessment.
{"title":"Integrated geostatistical and machine learning reservoir modeling for CO2 plume prediction and storage safety assessment","authors":"Muhammad Ali, Ning Wei, Ren Jiang, Wendong Wang, Keyao Lin, Chang Jia","doi":"10.1016/j.ijggc.2026.104766","DOIUrl":"10.1016/j.ijggc.2026.104766","url":null,"abstract":"<div><div>Hybrid reservoir property modeling is important for reliable CO<sub>2</sub> storage assessment, particularly in areas with limited well control and strong reservoir heterogeneity. This study presents a Hybrid Geostatistics and Machine Learning workflow, termed HyGEM, to generate high-resolution reservoir property models for CO2 geological storage applications. The workflow integrates upscaled well-log data, multi-attribute seismic volumes, geostatistical modeling, and supervised machine learning to predict petrophysical properties within a 3D reservoir grid. Arithmetic averaging is first applied to upscale well-log data to the simulation-grid scale, helping preserve volumetric consistency and reduce high-frequency noise. Several seismic attributes, including Instantaneous Bandwidth, RMS Amplitude, Sweetness, RGB, Semblance, and thin-bed indicators, are then used to describe stratigraphic and structural variations around the injection zone. Based on these inputs, the workflow generates 3D models of porosity, permeability, water saturation, volume of shale, bulk density, and pore pressure while preserving vertical layering and lateral heterogeneity. Validation results show strong agreement with original and upscaled well log data, with improved geological continuity compared to conventional stochastic and deterministic approaches. The predicted property grids are further applied in dynamic CO<sub>2</sub> injection simulation, where the results indicate controlled plume migration within high permeability zones, effective caprock containment, stable geomechanical behavior, and no clear indication of caprock failure during injection. These results demonstrate that the HyGEM workflow provides a practical and reliable approach for reservoir characterization, CO<sub>2</sub> plume prediction, and long-term storage safety assessment.</div></div>","PeriodicalId":334,"journal":{"name":"International Journal of Greenhouse Gas Control","volume":"155 ","pages":"Article 104766"},"PeriodicalIF":5.9,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148958531","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}
Pub Date : 2026-09-01Epub Date: 2026-08-28DOI: 10.1016/j.ijggc.2026.104764
Hongjin Yu, Yixuan Fang, Shiwei Wu, Ziyou Zhu, Huanxiang Gao, Quan Xie, Ranjith Pathegama Gamage, Zhaoyang Ma
Balancing mineral trapping with pore-network preservation remains a key challenge for CO2 storage in deep saline sandstones. Here, we propose a precipitation-provenance framework to distinguish carbonate recycling from silicate- and clay-fed net carbonate formation. Two tight sandstones, a calcite-cemented red sandstone (RS1) and a plagioclase/chlorite-rich grey sandstone (GS1), were reacted with scCO2-brine at 120°C and 30 MPa. Image-registered ESEM-MapsMin automated mineralogy, CT-resolved 3D pore-network analysis, and ICP-OES brine chemistry were integrated to track coupled dissolution-precipitation from grain to core scale. RS1 followed a carbonate-buffered recycling pathway, where calcite dissolution-reprecipitation and feldspar corrosion increased CT porosity from ∼2.0% to ∼5.5% but produced limited net mineral trapping. In contrast, GS1 showed silicate-fed carbonate formation: plagioclase and chlorite supplied Ca-Mg-Fe, increasing MapsMin-derived calcite abundance from near-zero to ∼13 wt%, while pore-wall coatings and local throat restriction moderated pore growth. We further distinguish likely in situ pore-wall carbonates from depressurization-related carbonates and drying-derived halite to avoid overestimating reservoir-condition trapping. Together, these results provide mechanistic criteria for screening sandstone storage formations and for designing pressure-management strategies that maximize mineral trapping while limiting adverse pore-network feedbacks.
{"title":"Mineral-controlled precipitation provenance dictates the mineral trapping-injectivity trade-off in sandstone CO2 storage","authors":"Hongjin Yu, Yixuan Fang, Shiwei Wu, Ziyou Zhu, Huanxiang Gao, Quan Xie, Ranjith Pathegama Gamage, Zhaoyang Ma","doi":"10.1016/j.ijggc.2026.104764","DOIUrl":"10.1016/j.ijggc.2026.104764","url":null,"abstract":"<div><div>Balancing mineral trapping with pore-network preservation remains a key challenge for CO<sub>2</sub> storage in deep saline sandstones. Here, we propose a precipitation-provenance framework to distinguish carbonate recycling from silicate- and clay-fed net carbonate formation. Two tight sandstones, a calcite-cemented red sandstone (RS<sub>1</sub>) and a plagioclase/chlorite-rich grey sandstone (GS<sub>1</sub>), were reacted with scCO<sub>2</sub>-brine at 120°C and 30 MPa. Image-registered ESEM-MapsMin automated mineralogy, CT-resolved 3D pore-network analysis, and ICP-OES brine chemistry were integrated to track coupled dissolution-precipitation from grain to core scale. RS<sub>1</sub> followed a carbonate-buffered recycling pathway, where calcite dissolution-reprecipitation and feldspar corrosion increased CT porosity from ∼2.0% to ∼5.5% but produced limited net mineral trapping. In contrast, GS<sub>1</sub> showed silicate-fed carbonate formation: plagioclase and chlorite supplied Ca-Mg-Fe, increasing MapsMin-derived calcite abundance from near-zero to ∼13 wt%, while pore-wall coatings and local throat restriction moderated pore growth. We further distinguish likely in situ pore-wall carbonates from depressurization-related carbonates and drying-derived halite to avoid overestimating reservoir-condition trapping. Together, these results provide mechanistic criteria for screening sandstone storage formations and for designing pressure-management strategies that maximize mineral trapping while limiting adverse pore-network feedbacks.</div></div>","PeriodicalId":334,"journal":{"name":"International Journal of Greenhouse Gas Control","volume":"155 ","pages":"Article 104764"},"PeriodicalIF":5.9,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148853965","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}
Pub Date : 2026-09-01Epub Date: 2026-08-31DOI: 10.1016/j.ijggc.2026.104765
Yun-Ting Chou, Hailun Ni, Hannah Lu
As global CO2 emissions rise, geological carbon storage remains a critical mitigation strategy. While traditional storage focuses on reservoirs with continuous caprocks, composite confining systems, comprised of discontinuous barriers, have recently emerged as promising alternatives due to their ability to inhibit leakage through heterogeneity. Existing assessments using physical sand tanks or commercial simulators are often computationally expensive or lack flexibility. In this work, we develop a full-field RU-Net-based surrogate model trained on simulation data to predict flow paths and retention capacity from various capillary entry pressure fields and injection locations. We demonstrate how the resulting computationally efficient and accurate surrogate model enables global sensitivity analysis of barrier geometry and injection optimization that would otherwise be expensive using physical experiments or full-physics simulations. Our results indicate that retention capacity cannot be attributed to any single geometric parameter and instead arise from interaction effects among barrier length, wing angle, and dip angle. Together, these results show that the full-field machine learning surrogate model can provide an efficient and practical decision-support tool for screening, sensitivity analysis, and injection design in heterogeneous geological carbon storage systems.
{"title":"A RU-Net-based surrogate study of sand tank experiments for CO2 storage capacity in composite confining systems","authors":"Yun-Ting Chou, Hailun Ni, Hannah Lu","doi":"10.1016/j.ijggc.2026.104765","DOIUrl":"10.1016/j.ijggc.2026.104765","url":null,"abstract":"<div><div>As global CO<sub>2</sub> emissions rise, geological carbon storage remains a critical mitigation strategy. While traditional storage focuses on reservoirs with continuous caprocks, composite confining systems, comprised of discontinuous barriers, have recently emerged as promising alternatives due to their ability to inhibit leakage through heterogeneity. Existing assessments using physical sand tanks or commercial simulators are often computationally expensive or lack flexibility. In this work, we develop a full-field RU-Net-based surrogate model trained on simulation data to predict flow paths and retention capacity from various capillary entry pressure fields and injection locations. We demonstrate how the resulting computationally efficient and accurate surrogate model enables global sensitivity analysis of barrier geometry and injection optimization that would otherwise be expensive using physical experiments or full-physics simulations. Our results indicate that retention capacity cannot be attributed to any single geometric parameter and instead arise from interaction effects among barrier length, wing angle, and dip angle. Together, these results show that the full-field machine learning surrogate model can provide an efficient and practical decision-support tool for screening, sensitivity analysis, and injection design in heterogeneous geological carbon storage systems.</div></div>","PeriodicalId":334,"journal":{"name":"International Journal of Greenhouse Gas Control","volume":"155 ","pages":"Article 104765"},"PeriodicalIF":5.9,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148958530","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}
Pub Date : 2026-09-01Epub Date: 2026-08-28DOI: 10.1016/j.ijggc.2026.104754
Jakob Kulich, Holger Ott
The carbon capture and storage (CCS) potential in Austria remains to be elucidated as the country plans to lift its CO2 storage ban and permit geologic storage for hard-to-abate emitters. This study presents a static screening and assessment of Austria’s CCS and underground hydrogen storage (UHS) potential in hydrocarbon reservoirs. Using reported ultimate recoveries from oil and gas fields, CO2 storage capacities were estimated across 73 reservoirs. The results indicate a total static CCS capacity of 263 Mt (183–357 Mt, depending on efficiency factor variation) for pure CO2, which decreases to 151–293 Mt when high levels of impurities are considered. The five largest gas fields, with individual storage capacities exceeding 10 Mt, account for approximately 100 Mt of CCS capacity. Site-specific challenges in Austria’s largest hydrocarbon fields may include numerous abandoned boreholes, strong aquifer support, and competition with UHS. Storage efficiency factors and CO2 density variations significantly impact capacity estimates, particularly in reservoirs near the supercritical threshold. For UHS, the total theoretical storage capacity across suitable fields was calculated to be 55 TWh, with 21 TWh identified in operational underground gas storage sites. Competition between CCS and UHS technologies is anticipated in relatively large gas fields with overlapping suitability. While individual CO2 storage capacities in hydrocarbon fields are small compared to offshore saline aquifers, they offer advantages such as existing infrastructure, detailed reservoir characterization, and proximity to major emission sources. The identified CCS capacity could bridge crucial periods before CO2 export or saline aquifer storage becomes viable.
{"title":"Screening of Austrian geological carbon and hydrogen storage options. Part I: Hydrocarbon fields","authors":"Jakob Kulich, Holger Ott","doi":"10.1016/j.ijggc.2026.104754","DOIUrl":"10.1016/j.ijggc.2026.104754","url":null,"abstract":"<div><div>The carbon capture and storage (CCS) potential in Austria remains to be elucidated as the country plans to lift its CO<sub>2</sub> storage ban and permit geologic storage for hard-to-abate emitters. This study presents a static screening and assessment of Austria’s CCS and underground hydrogen storage (UHS) potential in hydrocarbon reservoirs. Using reported ultimate recoveries from oil and gas fields, CO<sub>2</sub> storage capacities were estimated across 73 reservoirs. The results indicate a total static CCS capacity of 263 Mt (183–357 Mt, depending on efficiency factor variation) for pure CO<sub>2</sub>, which decreases to 151–293 Mt when high levels of impurities are considered. The five largest gas fields, with individual storage capacities exceeding 10 Mt, account for approximately 100 Mt of CCS capacity. Site-specific challenges in Austria’s largest hydrocarbon fields may include numerous abandoned boreholes, strong aquifer support, and competition with UHS. Storage efficiency factors and CO<sub>2</sub> density variations significantly impact capacity estimates, particularly in reservoirs near the supercritical threshold. For UHS, the total theoretical storage capacity across suitable fields was calculated to be 55 TWh, with 21 TWh identified in operational underground gas storage sites. Competition between CCS and UHS technologies is anticipated in relatively large gas fields with overlapping suitability. While individual CO<sub>2</sub> storage capacities in hydrocarbon fields are small compared to offshore saline aquifers, they offer advantages such as existing infrastructure, detailed reservoir characterization, and proximity to major emission sources. The identified CCS capacity could bridge crucial periods before CO<sub>2</sub> export or saline aquifer storage becomes viable.</div></div>","PeriodicalId":334,"journal":{"name":"International Journal of Greenhouse Gas Control","volume":"155 ","pages":"Article 104754"},"PeriodicalIF":5.9,"publicationDate":"2026-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148853966","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","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