Abstract Numerical simulation of underground CO2 storage requires accurate, fast and field-scale calculations of saturation and reservoir pressures, yet high-fidelity physics-based simulators are computationally expensive to do a rapid analysis. Recent advances in deep learning, especially Graph Neural Networks (GNNs), offer promising alternatives by approximating simulation outputs while dramatically reducing computational costs by minimizing the total reliance on traditional simulators. However, training GNN models on large-scale, spatiotemporal datasets remains a significant bottleneck when constrained to a single GPU. To address the computational bottleneck of training GNN on large-scale and high-resolution datasets, we propose to use multi-GPU scalable training of GNNs. The U-Net Enhanced Graph Neural Network (U-GCN) specifically was adopted to predict the spatiotemporal evolution of gas saturation and pressure buildup in CO2 storage in saline aquifers. The UGCN model was trained across multiple GPUs, allowing for concurrent gradient computations and parameter updates on separate GPU processes. The model operates natively on regular reservoir meshes to learn spatiotemporal evolution of pressure and saturation from thousands of simulation cases, and the training pipeline couples process-per-GPU data parallelism with a node-local caching strategy that deamplifies the source data file (in Hierarchical Data Format HDF5 format) Input-Output (I/O) by performing a single deserialization per node and broadcasting preprocessed graphs to sibling ranks. This eliminates multi-node startup contention and renders runs compute bound. Scaling experiments from 1 to 16 GPUs demonstrate substantial wall-clock reductions while preserving validation behavior. With per-GPU batch size 100, total training time decreased from 955.46 s (1 GPU) to 145.17 s (16 GPUs), a 6.58× speedup (≈41% parallel efficiency). With per-GPU batch size 8, time decreased from 5519.00 s to 1047.17 s, a 5.27× speedup (≈33% efficiency). The speedups are reported per epoch. Cross-node overheads relative to ideal doubling were ∼11–35%, more pronounced at small per-GPU batches where collective latency dominates. Training and validation losses rose modestly as GPU count increased under weak scaling—consistent with the larger global batch and fewer optimizer updates per epoch—yet validation tracked training across all settings, indicating stable generalization and no data leakage. These effects are mitigated by standard large-batch remedies (e.g., linear learning-rate scaling with warmup, longer schedules, optional gradient accumulation). By integrating domain-aware batching, multi-GPU data parallelism, mixed precision, and node-local caching with a GNN tailored to structured meshes, the proposed approach delivers a reproducible, high-throughput storage surrogate training regime that scales to multi-node clusters without compromising model validity. The resulting surrogates enable rapid scenario screening and outer-loop workflows (optimization and UQ) at orders-of-magnitude lower training cost compared to conventional GNN training on single a GPU baseline.
27CO2 storage in saline aquifers: A simulation on quantifying the impact of permeability heterogeneity
CO2 storage in saline aquifers: A simulation on quantifying the impact of permeability heterogeneity
77Controlled CO2 injection into heterogeneous geologic formations for improved solubility and residual trapping
Geologic CO2 storage (GCS) has been proposed as a potentially viable climate change mitigation option. Among the trapping mechanisms known for permanent CO2 storage in saline aquifers, solubility and residual trapping are important for safe short‐term entrapment. These storage mechanisms are hampered by density‐driven upward CO2 movement that inhibits the lateral migration of the plume, leaving a large portion of the aquifer volume unexposed to CO2 and unavailable for storage. Heterogeneity of aquifer hydraulic properties and high horizontal/vertical hydraulic conductivity ratio of geologic formations are two competing mechanisms that can hinder upward CO2 migration. Postinjection displacement of free‐phase CO2 is somewhat controlled by the heterogeneity in rock permeability and porosity distributions. In particular, low permeability shale layers that act as vertical flow barrier and high‐permeability horizontal channels that form flow conduits can spread the CO2 plume laterally in the aquifer. In this paper, we consider CO2 storage in heterogeneous saline aquifers and propose controlled CO2 injection, based on existing knowledge of heterogeneity, to increase CO2 contact with brine and improve the solubility and residual trapping and the overall aquifer storage potential. We examine two optimization methods: directly maximizing the total stored gas in the aquifer, and maximizing the sweep efficiency of the CO2 flood to promote uniform displacement in all directions. We consider the effect of geologic uncertainty on the performance of the controlled injection schemes by using an ensemble of model realizations to represent the uncertainty in aquifer property distribution. We also show how a controlled injection can be used to mitigate the risk of leakage from potential pathways, such as an abandoned well, by restricting CO2 movement toward the leakage zone. Our results suggest that controlled injection can lead to substantial improvements in residual and dissolution trapping and can be used to divert a CO2 plume from approaching sensitive zones or leakage pathways.
Read more59Risk mitigation through the optimization of residual gas and solubility trapping for CO2 storage in saline aquifers
Risk mitigation through the optimization of residual gas and solubility trapping for CO2 storage in saline aquifers
1Numerical Modeling for Assessing the Effects of Thermodynamic Properties on CO2 Storage in Saline Aquifers
Summary According to the Intergovernmental Panel on Climate Change (IPCC), global CO2 emissions must be reduced by 50 to 80 percent by 2050 to avoid dramatic consequences of global warming. Geological storage of CO2 in saline aquifers is a promising method for reducing atmospheric CO2 concentration. For this aim, accurate modeling of CO2 sequestration into underground formations (saline aquifers) is required. In petroleum industry, normally this is achieved by using compositional reservoir simulators which is computationally expensive and time consuming. To overcome this, an accurate fluid model was coupled to a flow simulator to model CO2 sequestration in saline aquifers. Next, sensitivity analyses of thermodynamic properties (pressure, temperature and salinity) were done on some saline aquifers of Alberta basin, in Canada to investigate the effects of thermodynamics properties on CO2 dissolution in these aquifers. Results show that salinity has the strongest effect on CO2 dissolution in our studied aquifers compared to temperature and pressure effects. Results of this study enable us to assess the potential of each saline aquifer for CO2 storage and therefore help us in selecting suitable injection sites for CO2 sequestration.
Read more4Key Considerations for Screening and Selection of Saline Aquifers for CO2 Storage
Carbon capture and storage (CCS) presents a key solution to reducing emissions especially from stationary power and industrial sites. A major component of CCS is CO2 storage in geologic formations, including saline aquifers, presents a great opportunity for the oil and gas industry to capitalize on their subsurface expertise to ensure that CO2 is stored safely underground over the long term. Saline aquifers are particularly relevant in the Middle East, where hydrocarbon reservoirs may not be at high depletion stages and so may not be available for CO2 storage yet. This paper looks at key geologic formation considerations related to CO2 storage in saline aquifers, highlighting the static and dynamic characteristics of the rock and the fluids that determine how much CO2 can be stored and at what injection rate. It also highlights how these characteristics affect the economics of the subsurface component of a CCS project. This paper shall provide an overview of the key aspects that guide subsurface storage evaluations: containment ability, trapping mechanisms, storage capacity, and injectivity. It will highlight the different storage capacity assessment methods. Then it shall discuss resource classification and categorization, and determination of commerciality as per the SPE CO2 Storage Resources Management System (SRMS). Finally, the paper goes into further considerations that need to be taken into account beyond the initial screening, including the data acquisition program and the impact of CO2 stream impurities on storage potential, geologic formation properties and cap rock integrity. The work illustrates the importance of understanding the physics of CO2 injection into a water-bearing system and lays out considerations for screening potential saline aquifer sites from a reservoir engineering perspective. The screening process should evaluate the anticipated CO2 trapping mechanism, assure its ability to contain the injected fluids over the long term, and estimate its storage capacity and injectivity.
Read more2Influence of geochemical processes on the geomechanical response of the overburden due to CO2 storage in saline aquifers
Saline aquifers have been identified as desirable geologic formations, suitable for the storage of carbon dioxide (CO2) in large amounts. While the impermeable caprock layer(s) in the overburden provide(s) the primary trap for injected carbon dioxide, other trapping mechanisms, such as solubility, residual, ionic or mineral trapping, help contribute to CO2 storage. Geochemical reactions alter the petrophysical properties such as porosity in the target reservoir, and may have an influence on the reservoir storage capacity. When CO2 is injected for long periods of time, it changes the fluid pressures and the overburden geomechanical response. The geomechanical response associated with time-dependent geochemical reactions may also influence the integrity of the caprock layer and the long-term fate of injected CO 2.;In the current study, coupled multiphase fluid flow and geomechanical modeling with geochemical reactions was performed to simulate the long-term (1,000 years), large-scale injection of CO2 (up to10 million metric tons per year) into a deep saline aquifer. The primary objective of the study is to investigate the influence of geochemical reactions on the geomechanical response of the overburden during long-term CO2 injection. The geochemical modeling results show that geochemical reactions, such as mineral dissolution and precipitation, do not have a significant influence on reservoir rock porosity (about 2% reduction). However, an increase in the mineral reaction rate constants resulted in a reduction of about 35% in rock porosity. Modeling results show that the inclusion of geochemical reactions in the geomechanical models does not have a significant influence on the computed ground displacements during the injection and post-injection periods. However, brine salinity and CO2 solubility have an influence on the computed pressure buildup and ground displacements during CO2 injection. In this study, the influence of a vertical permeable zone (damage zone) around the wellbore on CO2 leakage during long-term CO2 injection was investigated. Modeling results show that the damage zone permeability and the vertical extent of the damage zone have a significant influence on the amount of CO2 leakage. Modeling results also show that the vertical extent of the damage zone does not have a significant influence on the computed ground displacements. Therefore, the computed ground displacements may not be a good indicator of the extent of damage zone around the wellbore.
Read more8Assessing roles of geochemical reactions on CO2 plume, injectivity and residual trapping
Assessing roles of geochemical reactions on CO2 plume, injectivity and residual trapping
35Analysis of CO2 Migration during Nanofluid-Based Supercritical CO2 Geological Storage in Saline Aquifers
Carbon dioxide (CO_2) geological storage in deep saline aquifers is a key measure to mitigate global warming. However, it still faces a variety of technical challenges such as enhancing CO_2 effective storage capacities. In this paper, a preliminary model is developed to simulate CO_2 migration during nanofluid-based supercritical CO_2 geological storage in saline aquifers. The main mechanisms, including Brownian motion, thermophoresis, thermal energy transfer, and interfacial tension, are included in the proposed conceptual model. Based on the high-resolution space-time conservation element and solution element (CE/SE) method, the model is used to simulate CO_2 migration and distribution in the in-situ heterogeneous saline aquifer. It can be inferred that the involvement of nanoparticles decreases shear stresses opposing flow and enhances CO_2 mobility in the flow boundary layer. In addition, nanoparticles increase shear stresses outside the boundary layer and retard CO_2 velocity. These competitive mechanisms result in homogeneous migration of CO_2 in the saline formation. One preliminary suggestion is that nanofluids enhance homogeneous CO_2 transport in the reservoir and mitigate the negative effects of stratigraphic heterogeneity on migration and accumulation of the CO_2 plume. CO_2 effective storage capacity may be greatly elevated by means of nanofluid-based CO_2 geological sequestration. The concept of nanofluid-based CO_2 geological storage may be potentially conducive to large-scale commercial CO_2 geological storage and useful for exploration of geothermal resources in deep-seated hot rocks. The effects of CO_2 solubility and geochemical reactions on nanofluid flows may be considered in a future study.
Read more6GRAPHITE — Generative Reasoning and Analysis for Predictive Handling in Traffic Efficiency
Traffic forecasting is a crucial aspect of modern Intelligent Transportation Systems (ITS) and the Internet of Vehicles (IoV), playing a vital role in improving the safety and efficiency of daily transportation activities. Despite the valuable contributions of traditional machine learning (ML) models and advanced deep learning (DL) techniques, there persist challenges in capturing the intricate spatial and temporal dependencies inherent in traffic flow. In response to these challenges, we present GRAPHITE, an innovative framework that combines Graph Neural Networks (GNNs) and Generative Adversarial Networks (GANs) to leverage generative reasoning for efficient traffic management. Our model seamlessly integrates historical traffic volume data collected by road sensors with local spatial information encoded through knowledge graphs (KGs) associated with each sensor. These KGs offer a structured representation of relationships between traffic sensors and points of interest (POIs) in their neighborhood, thereby enhancing the comprehension of the urban context and leading to more accurate traffic predictions. Extensive experiments conducted on diverse datasets underscore the efficacy of GRAPHITE. Notably, we achieved a maximum decrease in RMSE of 31.05% compared to GAN-GRU and a maximum increase in R2 of 8.15% compared to GAN-RNN, positioning GRAPHITE as a standout solution among the current state-of-the-art approaches. Our code is available at: https://github.com/MODAL-UNINA/GRAPHITE.
Read moreField-scale hydrodynamics and optimization of CO2 storage in saline aquifers using a hybrid Darcy scale flow–RSM approach
Field-scale hydrodynamics and optimization of CO2 storage in saline aquifers using a hybrid Darcy scale flow–RSM approach
Read more7Semi-analytical model to determine perforation interval for secure CO2 storage in saline aquifers
Semi-analytical model to determine perforation interval for secure CO2 storage in saline aquifers
Why It is Important to Perform a Fully Integrated Basin-Scale Modeling of CO2 Storage in Saline Aquifers? Insigths from the Paris Basin (France)
SummaryCarbon capture, utilization, and storage (CCUS) is an emerging strategy for managing and reducing CO2 emissions into the atmosphere. For the last two decades, such technique has been investigated as an effective measure for mitigating climate change—by storing CO2 in depleted oil and gas reservoirs, and deep saline aquifers.In this paper, we present an efficient workflow for constructing an accurate geological basin model for the evaluation of CO2 storage capacity and simulate the injection of CO2 and its long-term effects on the pressure field. Our workflow has been applied to the Paris Basin of France. This Basin is one of the most mature areas in the Western Europe in terms of the oil and gas exploration, with approximately 100 oil fields producing since the 1950’s.The results of our integrated workflow show the relevance of an integrated basin modeling workflow for CO2 screening purposes at the basin scale including geomechanical impacts of CO2 injection.
Read more18Simultaneous Prediction of Equilibrium, Interfacial, and Transport Properties of CO2-Brine Systems Using Molecular Dynamics Simulation: Applications to CO2 Storage
This study aims to provide accurate simultaneous predictions of crucial properties affecting CO2 storage in saline aquifers with deep microstructural insights using comprehensive molecular dynamics (MD) simulations under actual operating conditions ranges. We investigated the effects of pressure, temperature, and salinity on mutual solubility, interfacial tension (IFT), density, and viscosity of a CO2-brine system over the temperature range of 323.15–393.15 K, pressures up to 30 MPa, and salinity range of 0.98–2.97 mol/kg. Brine was resembled by employing different brine systems like actual reservoirs, including mixed electrolyte solutions containing monovalent and divalent ions. MD model configurations and force fields were validated by experimental data of a CO2-water/NaCl aqueous solution. The maximum average absolute deviations for the solubility, density, and IFT were 5.50, 6.95, and 7.136%, respectively. The simulation results indicated that mutual solubilities of CO2 and water in the CO2-NaCl + KCl solution system are higher than those of CO2-NaCl + CaCl2, while for IFT, the opposite trend is observed. Unlike pressure, salt concentration variations result in considerable density and viscosity changes for all systems. The maximum density and viscosity values are obtained for the brine-containing divalent cations. This study gives insight into the storage process and relates the microstructure effects to solubility, IFT, and viscosity of CO2-brine systems using a single MD model.
Read more12Flow Simulation of CO2 Storage in Saline Aquifers Using Black Oil Simulator
Flow Simulation of CO2 Storage in Saline Aquifers Using Black Oil Simulator
Evaluating CO2 Storage Potential in an Under-Explored Basin: Stord Basin, Northern North Sea
Summary Regional studies have shown that it may be possible to store a large amount of CO2 on the Norwegian Continental Shelf. In order to meet ambitious climate targets we now need detailed geological characterisation to rank and de-risk potential new storage sites. The Stord Basin is located in the northern North Sea, south of the Horda Platform where current CCS infrastructure is under development, and may provide viable opportunities for CO2 storage in saline aquifers. In this study we generate a model for the tectono-stratigraphic evolution of the Stord Basin, and use this to assess the basin potential for structural trapping, stratigraphic trapping, and residual trapping. We then perform a detailed evaluation of the identified potential storage sites in the Stord Basin using currently available data. The Stord Basin provides good stratigraphic and residual trapping opportunities for CO2 storage, but there are no structural traps at depths suitable for CO2 storage observed in the present data.
Read more
Copyright 2026 Cactus Communications. All rights reserved.
