Summary
Geological CO2 sequestration has gotten a lot of attention as a potential technical solution for decreasing human carbon emissions to the environment. Numerical reservoir simulation (NRS) has been conventionally used to model subsurface reservoirs and employed in uncertainty analysis, optimization, and decision-making. One of the challenges associated with NRS is the computational efforts required to model complex reservoir systems. Therefore, fast proxy models are suggested for quick predictions. Although several studies have been conducted about the CO2 injection & storage (CIS) process, there still exists a research gap in developing a fast tool for monitoring and performance prediction of a safe CIS project. This research project aims to develop an open-access smart tool to predict the performance of CIS with satisfactory accuracy in the mature fields of the Norwegian Continental Shelf, considering geochemical & geomechanical (GCGM) aspects. The uncertainty parameters will be the underground rock and fluid properties, operational constraints of the CO2 injection, and GCGM parameters. The target parameters include net CO2 storage, ultimate oil recovery from miscible CO2 injection, & cumulative production of gaseous CO2. The risk of CO2 leakage through induced fractures/faults should be considered by considering the effect of GCGM parameters on CIS. The effect of GCGM parameters on such processes will be useful in CO2 leakage risk quantification, making the proxy model more realistic. Machine learning tools, such as PINN, ANN, SVR, and XGBoost, will be applied to the simulated data set to digitalize the CIS process efficiently. The developed proxy model will be validated with a synthetic truth model. Lastly, the novel proxy model will be added as a separate module to MRST - an open-access software. This integration offers researchers and industry professionals an openly accessible, fast & accurate proxy model to streamline the initial evaluation of CIS projects.
Unfold all
/
Fold all
More information & hyperlinks
| Web resources: | https://cordis.europa.eu/project/id/101151733 |
| Start date: | 01-09-2025 |
| End date: | 31-08-2027 |
| Total budget - Public funding: | - 226 751,00 Euro |
Cordis data
Original description
Geological CO2 sequestration has gotten a lot of attention as a potential technical solution for decreasing human carbon emissions to the environment. Numerical reservoir simulation (NRS) has been conventionally used to model subsurface reservoirs and employed in uncertainty analysis, optimization, and decision-making. One of the challenges associated with NRS is the computational efforts required to model complex reservoir systems. Therefore, fast proxy models are suggested for quick predictions. Although several studies have been conducted about the CO2 injection & storage (CIS) process, there still exists a research gap in developing a fast tool for monitoring and performance prediction of a safe CIS project. This research project aims to develop an open-access smart tool to predict the performance of CIS with satisfactory accuracy in the mature fields of the Norwegian Continental Shelf, considering geochemical & geomechanical (GCGM) aspects. The uncertainty parameters will be the underground rock and fluid properties, operational constraints of the CO2 injection, and GCGM parameters. The target parameters include net CO2 storage, ultimate oil recovery from miscible CO2 injection, & cumulative production of gaseous CO2. The risk of CO2 leakage through induced fractures/faults should be considered by considering the effect of GCGM parameters on CIS. The effect of GCGM parameters on such processes will be useful in CO2 leakage risk quantification, making the proxy model more realistic. Machine learning tools, such as PINN, ANN, SVR, and XGBoost, will be applied to the simulated data set to digitalize the CIS process efficiently. The developed proxy model will be validated with a synthetic truth model. Lastly, the novel proxy model will be added as a separate module to MRST - an open-access software. This integration offers researchers and industry professionals an openly accessible, fast & accurate proxy model to streamline the initial evaluation of CIS projects.Status
SIGNEDCall topic
HORIZON-MSCA-2023-PF-01-01Update Date
22-11-2024
Images
No images available.
Geographical location(s)
