
(Photo by Dilok Klaisataporn on Shutterstock)
In a nutshell
- Carbon dioxide mixes with underground brine 13% faster in real 3D conditions than previously estimated in 2D models, making geological storage more efficient than we thought.
- When CO₂ dissolves in groundwater, it creates a denser fluid that sinks downward, forming finger-like structures that enhance mixing and trap the carbon permanently.
- Researchers developed a simplified mathematical model that engineers can use to predict long-term CO₂ behavior without running complex simulations for every site.
ENSCHEDE, Netherlands — Carbon dioxide pumped underground stays there more securely than we thought, according to a new study that could boost confidence in one of our most promising climate solutions. Advanced 3D computer simulations reveal that CO2 mixes with underground brine remarkably faster than previous calculations predicted, creating a natural trapping mechanism that’s more efficient than scientists realized.
This discovery, made by researchers from the University of Twente in the Netherlands and their colleagues, contradicts earlier assumptions about how injected carbon dioxide behaves when stored deep beneath the Earth’s surface. For years, scientists relied on simplified two-dimensional models that underestimated mixing rates. Now, using advanced three-dimensional simulations, researchers found the process works significantly more efficiently.
How Carbon Storage Works
Carbon capture and storage represents one of our most promising weapons against climate change. The process captures carbon dioxide from power plants and industrial sources before it enters the atmosphere, then injects it deep underground into porous rock formations where it can remain trapped for centuries.
When first injected underground, the CO2 is lighter than the surrounding brine and floats upward, collecting beneath an impermeable rock layer. This arrangement initially seems precarious – if that caprock were to crack, the carbon dioxide could potentially escape.
“Pure CO2 has a lower density than water, but the situation changes when CO2 is dissolved in water. When the two are mixed, the total volume decreases, creating a denser liquid,” explains Marco De Paoli, head of the research project, in a statement. “Water with a high CO2 content has a higher density than water with a lower CO2 content and therefore sinks.”
This density difference creates a fascinating dynamic. As the CO2-rich water sinks, it forms distinctive finger-like structures called “plumes” that increase mixing through a process called convection.
“Because water with a higher CO2 content has a higher density than water with a lower CO2 content, the dynamics in the porous rock are highly interesting,” says De Paoli. “Where the CO2 concentration is highest, the mixture sinks faster, which in turn ensures even better mixing.” This results in a network-like pattern of areas with higher and lower CO2 concentrations.

The Discovery: Faster Mixing in 3D
The research team used powerful supercomputers to run simulations that tracked the fluid dynamics occurring when carbon dioxide dissolves into brine. Their work, now published in Geophysical Research Letters, revealed that in three-dimensional simulations, the dissolution rate was 13.5% higher than in two-dimensional models – substantially different from previous estimates suggesting a 25% difference.
From their calculations, the team was able to derive simple models that can now be used by engineers to predict CO2 flow in the ground and design injection strategies without having to carry out complex computer simulations for every situation.
The study found this process unfolds through several distinct phases. First comes a diffusion-dominated phase where carbon dioxide slowly dissolves at the interface. Next, the CO2-rich layer becomes unstable and forms the small plumes that grow downward. These fingers eventually merge with neighbors to form larger, persistent structures. Finally, as the domain becomes saturated, mixing gradually slows.
Real-World Impact
To demonstrate practical applications, they applied their model to the Sleipner site in the North Sea – one of the world’s first and largest carbon storage projects. Their calculations predicted it would take about 10 years to dissolve 100 kg/m² of carbon dioxide at this site. The model also suggested it would take less than 20 years to achieve 50% dissolution of injected CO2, but more than 100 years to reach 90% dissolution.
The geological conditions needed for this storage method aren’t rare. “You could use depleted oil reservoirs. There are also large areas called saline aquifers, located under the seabed or inland, where CO2 storage would be possible according to this scheme. At least six saline aquifers are also present in Austria,” notes De Paoli.
Once the CO2-containing water sinks downward, it remains safely stored – even in the event of geological changes. “[These changes] such as an earthquake or anthropogenic activities, would no longer affect the situation. The CO2 is safely stored in the ground,” De Paoli explains.
Governments and industries have been racing to develop carbon capture technologies. What seemed like a precarious solution – pumping CO2 underground – now appears highly secure and effective, reinforcing carbon capture’s role as a vital strategy in our climate mitigation toolkit.
Paper Summary
Methodology
The scientists performed high-resolution, large-scale simulations of solute convection in both two-dimensional and three-dimensional porous media at Rayleigh-Darcy numbers ranging from 10² to 8 × 10⁴. They used a semi-infinite domain where CO₂ concentration remains constant at the top with no flux at the bottom. The simulations employed a second-order finite-difference solver to resolve the governing equations for flow in porous media. The researchers systematically compared 2D and 3D simulations to quantify differences in mixing dynamics and CO₂ dissolution rates.
Results
The study identified distinct phases in the convection process: an initial diffusion-dominated phase, transition to convection-driven solute finger growth, and a final shutdown stage as fingers reach the bottom boundary and concentration increases. For Rayleigh-Darcy numbers ≥5 × 10³, researchers observed a constant-flux regime with dissolution flux stabilizing at 0.019 in 3D simulations, approximately 13% higher than in 2D estimates – not 25% higher as previously thought. The team also developed an accurate physical model that reproduces flow evolution across all regimes and numbers. Applied to the Sleipner site (North Sea), this model predicted less than 20 years to achieve 50% CO₂ dissolution but over 100 years to reach 90%.
Limitations
The study focused primarily on homogeneous and isotropic porous media, while real geological formations may be anisotropic, heterogeneous, or fractured. Additional factors not fully addressed include solute redistribution from mechanical dispersion, flow-induced morphology modifications due to geological formation chemistry, and other mechanisms that could influence dissolution dynamics. These complexities represent areas requiring further research.
Funding or Disclosures
The research received funding from the European Union’s Horizon Europe research and innovation programme under the Marie Sklodowska-Curie grant agreement MEDIA No. 101062123. The work was supported by the EuroHPC Joint Undertaking (project GEOCOSE number EHPC-REG-2022R03-207), which granted access to the EuroHPC supercomputer LUMI-C, hosted by the LUMI consortium in Finland. Marco De Paoli is currently working at the University of Twente in the Netherlands and at the Institute of Fluid Mechanics and Heat Transfer at TU Wien, and was awarded an ERC grant by the European Research Council in 2024.
Publication Information
The study “Simulation and Modeling of Convective Mixing of Carbon Dioxide in Geological Formations” was published in Geophysical Research Letters (2025), authored by Marco De Paoli, Francesco Zonta, Lea Enzenberger, Eliza Coliban, and Sergio Pirozzoli. The paper (DOI: 10.1029/2025GL114804) was received January 14, 2025, and accepted March 7, 2025.
About StudyFinds Staff
StudyFinds sets out to find new research that speaks to mass audiences — without all the scientific jargon. The stories we publish are digestible, summarized versions of research that are intended to inform the reader as well as stir civil, educated debate. StudyFinds Staff articles are AI assisted, but always thoroughly reviewed and edited by a Study Finds staff member. Read our AI Policy for more information.
Read More About StudyFinds
Our Editorial Process
StudyFinds publishes digestible, agenda-free, transparent research summaries that are intended to inform the reader as well as stir civil, educated debate.We do not agree nor disagree with any of the studies we post, rather, we encourage our readers to debate the veracity of the findings themselves.All articles published on StudyFinds are vetted by our editors prior to publication and include links back to the source or corresponding journal article, if possible.
Read More →
Our Editorial Team
Steve Fink
Editor-in-Chief
Sophia Naughton
Associate Editor