A team of scientists has created a new AI-based tool to help block greenhouse gases like CO2 in porous rock formations faster and more accurately than ever before.
Carbon capture technology, also known as carbon sequestration, is a climate change mitigation method that redirects CO2 emitted by underground power plants. In doing so, scientists must avoid excessive pressure buildup caused by the injection of CO2 in rock, which can fracture geological formations and leak carbon into aquifers above the site and even into the atmosphere.
A new neural operator architecture named U-FNO simulates pressure levels during carbon storage in a fraction of a second while doubling the accuracy of some tasks, helping scientists find optimal injection rates and sites. It was unveiled this week in a study published in Progress in water resourceswith co-authors from Stanford University, California Institute of Technology, Purdue University, and NVIDIA.
Carbon capture and storage is one of the few methods that industries such as refining, cement and steel could use to decarbonize and meet emission reduction targets. More than a hundred carbon capture and storage facilities are under construction worldwide.
U-FNO will be used to accelerate carbon storage predictions for ExxonMobil, which funded the study.
“Reservoir simulators are computer-intensive models that engineers and scientists use to study multiphase flows and other complex physical phenomena in the Earth’s underground geology,” said James V. White, underground storage manager for the carbon at ExxonMobil. “Machine learning techniques such as those used in this work provide a robust pathway to quantify uncertainties in large-scale groundwater flow models such as carbon capture and sequestration and ultimately facilitate better decision-making. “
How carbon storage scientists are using machine learning
Scientists use carbon storage simulations to select appropriate injection sites and rates, control pressure buildup, maximize storage efficiency, and ensure injection activity does not fracture the formation rocky. For a successful storage project, it is also important to understand the carbon dioxide plume – the spread of CO2 through the ground.
Traditional carbon sequestration simulators are time consuming and computationally expensive. Machine learning models provide similar levels of accuracy while dramatically reducing the time and cost required.
Based on the U-Net neural network and Fourier neural operator architecture, known as FNO, U-FNO provides more accurate predictions of gas saturation and pressure buildup. Compared to using a state-of-the-art convolutional neural network for the task, U-FNO is twice as accurate while only requiring a third of the training data.
“Our machine learning method for scientific modeling is fundamentally different from standard neural networks, where we typically work with fixed-resolution images,” said paper co-author Anima Anandkumar, research director in machine learning at NVIDIA and Professor Bren in the Computing + Department of Mathematical Sciences at Caltech. “In scientific modeling, we have varying resolutions depending on how and where we sample. Our model can generalize well to different resolutions without the need to retrain, resulting in huge speedups. »
Trained U-FNO models are available in a web application to provide real-time predictions for carbon storage projects.
“Recent innovations in AI, with techniques such as FNOs, can speed up calculations by orders of magnitude, taking an important step in helping carbon capture and storage technologies scale,” said Ranveer Chandra, general manager of industry research at Microsoft and collaborator on the Northern Lights initiative, a large-scale carbon capture and storage project in Norway. “Our model-parallel FNO can scale to realistic 3D problem sizes using the distributed memory of many NVIDIA Tensor Core GPUs.”
New Neural Operators Accelerate CO2 Storage Predictions
U-FNO allows scientists to simulate how pressure levels will build and where CO2 will spread throughout the 30 years of injection. GPU acceleration with U-FNO allows these 30-year-old simulations to run in a hundredth of a second on a single NVIDIA A100 Tensor Core GPU, compared to 10 minutes with traditional methods.
Thanks to GPU-accelerated machine learning, researchers can now quickly simulate many injection sites. Without this tool, site selection is like a shot in the dark.
The U-FNO model focuses on modeling plume migration and pressure during the injection process – when there is the greatest risk of exceeding the amount of CO2 injected. It was developed using NVIDIA A100 GPUs in the Sherlock Computing Cluster at Stanford.
“For net zero to be achievable, we will need low-emission energy sources as well as negative-emission technologies, such as carbon capture and storage,” said U-Collaborator Farah Hariri. FNO and Technical Lead for Climate Change Mitigation Projects. for NVIDIA’s Earth-2, which will be the world’s first digital AI supercomputer. “By applying Fourier neural operators to carbon storage, we have shown how AI can help accelerate the process of climate change mitigation. Earth-2 will take advantage of these techniques.
Learn more about U-FNO on the NVIDIA Tech Blog.
Earth-2 will use FNO-like models to address challenges in climatology and contribute to global climate change mitigation efforts. Learn more about Earth-2 and the AI models used for climate science in GTC’s keynote from NVIDIA Founder and CEO Jensen Huang: