Deep learning inversion of gravity data for detection of CO2 plumes in overlying aquifers
Abstract
In this work, we developed an effective U-Net based deep learning (DL) model for inversion of surface gravity data on a rectangular grid to predict 2-D high-resolution subsurface CO2 distribution along a vertical cross-section due to CO2 leakage through a wellbore within a deep CO2 storage reservoir. We used synthetic data to model two types of CO2 leakage scenarios: one CO2 plume in a shallow aquifer (single plume case), and two plumes present at different depths (double plume case). The 3-D synthetic plume samples were created by sampling among predetermined CO2 plume depths, saturations, and volumes. The corresponding surface gravity data on a rectangular grid were generated by a 3-D forward model. The U-Net model detected 72% of single-plume samples, and one or both plumes in 75% of double-plume samples. Most of the undetected single plumes have small gravity field strengths below the typical noise level of 5 μGal. This model generated reproducible, reliable predictions with acceptable errors and demonstrated improved spatial resolution over the conventional least-squares inversion. In contrast to the conventional least-squares inversion, which often overestimates the size of its target and underestimates its density, this U-Net model accurately delineated the boundary of a target. Furthermore, this DLmore »
- Authors:
-
- Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
- Publication Date:
- Research Org.:
- Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)
- Sponsoring Org.:
- USDOE Office of Fossil Energy (FE); USDOE Office of Fossil Energy and Carbon Management (FECM)
- OSTI Identifier:
- 1834488
- Alternate Identifier(s):
- OSTI ID: 1862482
- Report Number(s):
- LLNL-JRNL-820456
Journal ID: ISSN 0926-9851; 1031719
- Grant/Contract Number:
- AC52-07NA27344
- Resource Type:
- Accepted Manuscript
- Journal Name:
- Journal of Applied Geophysics
- Additional Journal Information:
- Journal Volume: 196; Journal Issue: na; Journal ID: ISSN 0926-9851
- Publisher:
- Elsevier
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 58 GEOSCIENCES; Deep learning; Inversion; Gravity; U-Net; CO2; Detection
Citation Formats
Yang, Xianjin, Chen, Xiao, and Smith, Megan M. Deep learning inversion of gravity data for detection of CO2 plumes in overlying aquifers. United States: N. p., 2021.
Web. doi:10.1016/j.jappgeo.2021.104507.
Yang, Xianjin, Chen, Xiao, & Smith, Megan M. Deep learning inversion of gravity data for detection of CO2 plumes in overlying aquifers. United States. https://doi.org/10.1016/j.jappgeo.2021.104507
Yang, Xianjin, Chen, Xiao, and Smith, Megan M. Tue .
"Deep learning inversion of gravity data for detection of CO2 plumes in overlying aquifers". United States. https://doi.org/10.1016/j.jappgeo.2021.104507. https://www.osti.gov/servlets/purl/1834488.
@article{osti_1834488,
title = {Deep learning inversion of gravity data for detection of CO2 plumes in overlying aquifers},
author = {Yang, Xianjin and Chen, Xiao and Smith, Megan M.},
abstractNote = {In this work, we developed an effective U-Net based deep learning (DL) model for inversion of surface gravity data on a rectangular grid to predict 2-D high-resolution subsurface CO2 distribution along a vertical cross-section due to CO2 leakage through a wellbore within a deep CO2 storage reservoir. We used synthetic data to model two types of CO2 leakage scenarios: one CO2 plume in a shallow aquifer (single plume case), and two plumes present at different depths (double plume case). The 3-D synthetic plume samples were created by sampling among predetermined CO2 plume depths, saturations, and volumes. The corresponding surface gravity data on a rectangular grid were generated by a 3-D forward model. The U-Net model detected 72% of single-plume samples, and one or both plumes in 75% of double-plume samples. Most of the undetected single plumes have small gravity field strengths below the typical noise level of 5 μGal. This model generated reproducible, reliable predictions with acceptable errors and demonstrated improved spatial resolution over the conventional least-squares inversion. In contrast to the conventional least-squares inversion, which often overestimates the size of its target and underestimates its density, this U-Net model accurately delineated the boundary of a target. Furthermore, this DL inversion detected deep, small, or low saturation CO2 plumes that are often more difficult to resolve with conventional gravity inversion methods. We note the limitations of this feasibility study, including the use of synthetic data with regular CO2 plume shapes, and the prediction of a 2-D plume cross-section rather than the full 3-D plume, as well, we recognize the lower detection fraction for double-plume scenarios. Nevertheless, this study demonstrates that DL gravity inversion is a promising and potentially superior method to conventional least-squares inversion. Our U-Net based deep learning inversion approach may be adapted for inversion of other types of geophysical data. DL inversion can facilitate near real-time monitoring of geologic carbon sequestration to provide site operators with prompt information about subsurface CO2 distribution for risk management and mitigation.},
doi = {10.1016/j.jappgeo.2021.104507},
journal = {Journal of Applied Geophysics},
number = na,
volume = 196,
place = {United States},
year = {Tue Nov 23 00:00:00 EST 2021},
month = {Tue Nov 23 00:00:00 EST 2021}
}
Works referenced in this record:
3D non-smooth inversion of gravity data by zero order minimum entropy stabilizing functional
journal, September 2019
- Rezaie, Mohammad
- Physics of the Earth and Planetary Interiors, Vol. 294
Evaluating time-lapse borehole gravity for CO2 plume detection at SECARB Cranfield
journal, October 2013
- Dodds, Kevin; Krahenbuhl, Richard; Reitz, Anya
- International Journal of Greenhouse Gas Control, Vol. 18
Historical development of the gravity method in exploration
journal, November 2005
- Nabighian, M. N.; Ander, M. E.; Grauch, V. J. S.
- GEOPHYSICS, Vol. 70, Issue 6
Sensitivity thresholds of groundwater parameters for detecting CO2 leakage at a geologic carbon sequestration site
journal, October 2019
- Berger, Peter M.; Wimmer, Bracken; Iranmanesh, Abbas
- Environmental Monitoring and Assessment, Vol. 191, Issue 11
20 Years of Monitoring CO2-injection at Sleipner
journal, July 2017
- Furre, Anne-Kari; Eiken, Ola; Alnes, Håvard
- Energy Procedia, Vol. 114
Deep-learning tomography
journal, January 2018
- Araya-Polo, Mauricio; Jennings, Joseph; Adler, Amir
- The Leading Edge, Vol. 37, Issue 1
Physics-driven deep-learning inversion with application to transient electromagnetics
journal, April 2021
- Colombo, Daniele; Turkoglu, Ersan; Li, Weichang
- GEOPHYSICS, Vol. 86, Issue 3
Deep-learning electromagnetic monitoring coupled to fluid flow simulators
journal, January 2020
- Colombo, Daniele; Li, Weichang; Sandoval-Curiel, Ernesto
- GEOPHYSICS, Vol. 85, Issue 4
Constraints in 3D gravity inversion: 3D gravity inversion
journal, March 2001
- Boulanger, Olivier; Chouteau, Michel
- Geophysical Prospecting, Vol. 49, Issue 2
Deep-Learning Inversion of Seismic Data
journal, March 2020
- Li, Shucai; Liu, Bin; Ren, Yuxiao
- IEEE Transactions on Geoscience and Remote Sensing, Vol. 58, Issue 3
Downhole pressure and chemical monitoring for CO2 and brine leak detection in aquifers above a CO2 storage reservoir
journal, December 2019
- Buscheck, Thomas A.; Mansoor, Kayyum; Yang, Xianjin
- International Journal of Greenhouse Gas Control, Vol. 91
Managing well leakage risks at a geologic carbon storage site with many wells
journal, September 2019
- Lackey, Greg; Vasylkivska, Veronika S.; Huerta, Nicolas J.
- International Journal of Greenhouse Gas Control, Vol. 88
Deep learning electromagnetic inversion with convolutional neural networks
journal, May 2019
- Puzyrev, Vladimir
- Geophysical Journal International, Vol. 218, Issue 2
Time-lapse gravity surveying as a monitoring tool for CO 2 storage
journal, May 2017
- Wilkinson, M.; Mouli-Castillo, J.; Morgan, P.
- International Journal of Greenhouse Gas Control, Vol. 60
3-D linear inversion of gravity data: method and application to Basse-Terre volcanic island, Guadeloupe, Lesser Antilles
journal, February 2016
- Barnoud, Anne; Coutant, Olivier; Bouligand, Claire
- Geophysical Journal International, Vol. 205, Issue 1
Ü-Net: Deep-Learning Schemes for Ground Penetrating Radar Data Inversion
journal, June 2020
- Xie, Longhao; Zhao, Qing; Ma, Chunguang
- Journal of Environmental and Engineering Geophysics, Vol. 25, Issue 2
Risk-based post injection site care and monitoring for commercial-scale carbon storage: Reevaluation of the FutureGen 2.0 site using NRAP-Open-IAM and DREAM
journal, November 2019
- Bacon, Diana H.; Yonkofski, Catherine M. R.; Brown, Christopher F.
- International Journal of Greenhouse Gas Control, Vol. 90
A deep learning approach to the inversion of borehole resistivity measurements
journal, April 2020
- Shahriari, M.; Pardo, D.; Picon, A.
- Computational Geosciences, Vol. 24, Issue 3
Cooperative deep learning inversion of controlled-source electromagnetic data for salt delineation
journal, June 2020
- Oh, Seokmin; Noh, Kyubo; Seol, Soon Jee
- GEOPHYSICS, Vol. 85, Issue 4
Program to calculate the gravitational field and gravity gradient tensor resulting from a system of right rectangular prisms
journal, June 1992
- Montana, Carlos J.; Mickus, Kevin L.; Peeples, Wayne J.
- Computers & Geosciences, Vol. 18, Issue 5
Monitoring a large-volume injection at Cranfield, Mississippi—Project design and recommendations
journal, October 2013
- Hovorka, Susan D.; Meckel, Timothy A.; Treviño, Ramón H.
- International Journal of Greenhouse Gas Control, Vol. 18
Monitoring gas production and C O2 injection at the Sleipner field using time-lapse gravimetry
journal, November 2008
- Alnes, Håvard; Eiken, Ola; Stenvold, Torkjell
- GEOPHYSICS, Vol. 73, Issue 6
Geophysical inversion versus machine learning in inverse problems
journal, December 2018
- Kim, Yuji; Nakata, Nori
- The Leading Edge, Vol. 37, Issue 12
Time-lapse gravity monitoring of CO2 migration based on numerical modeling of a faulted storage complex
journal, April 2020
- Appriou, Delphine; Bonneville, Alain; Zhou, Quanlin
- International Journal of Greenhouse Gas Control, Vol. 95
Hyper-parameter optimization of deep learning model for prediction of Parkinson’s disease
journal, May 2020
- Kaur, Sukhpal; Aggarwal, Himanshu; Rani, Rinkle
- Machine Vision and Applications, Vol. 31, Issue 5
Occam’s inversion to generate smooth, two‐dimensional models from magnetotelluric data
journal, December 1990
- deGroot‐Hedlin, C.; Constable, S.
- GEOPHYSICS, Vol. 55, Issue 12
Deep-learning inversion: A next-generation seismic velocity model building method
journal, July 2019
- Yang, Fangshu; Ma, Jianwei
- GEOPHYSICS, Vol. 84, Issue 4
Pure and Pseudo-pure Fluid Thermophysical Property Evaluation and the Open-Source Thermophysical Property Library CoolProp
journal, January 2014
- Bell, Ian H.; Wronski, Jorrit; Quoilin, Sylvain
- Industrial & Engineering Chemistry Research, Vol. 53, Issue 6
One-dimensional deep learning inversion of electromagnetic induction data using convolutional neural network
journal, April 2020
- Moghadas, Davood
- Geophysical Journal International, Vol. 222, Issue 1
Deep Convolutional Neural Network for Inverse Problems in Imaging
journal, September 2017
- Jin, Kyong Hwan; McCann, Michael T.; Froustey, Emmanuel
- IEEE Transactions on Image Processing, Vol. 26, Issue 9
Deep Learning Inversion of Electrical Resistivity Data
journal, August 2020
- Liu, Bin; Guo, Qian; Li, Shucai
- IEEE Transactions on Geoscience and Remote Sensing, Vol. 58, Issue 8
3-D inversion of gravity data
journal, January 1998
- Li, Yaoguo; Oldenburg, Douglas W.
- GEOPHYSICS, Vol. 63, Issue 1
Inversion of 1D frequency- and time-domain electromagnetic data with convolutional neural networks
journal, April 2021
- Puzyrev, Vladimir; Swidinsky, Andrei
- Computers & Geosciences, Vol. 149
Machine learning and geophysical inversion — A numerical study
journal, July 2019
- Russell, Brian
- The Leading Edge, Vol. 38, Issue 7
Physics-guided deep learning for seismic inversion with hybrid training and uncertainty analysis
journal, March 2021
- Sun, Jian; Innanen, Kristopher A.; Huang, Chao
- GEOPHYSICS, Vol. 86, Issue 3
Seismic inversion by Newtonian machine learning
journal, June 2020
- Chen, Yuqing; Schuster, Gerard T.
- GEOPHYSICS, Vol. 85, Issue 4
Fast 3D inversion of gravity data using solution space priorconditioned lanczos bidiagonalization
journal, January 2017
- Rezaie, Mohammad; Moradzadeh, Ali; Kalateh, Ali Nejati
- Journal of Applied Geophysics, Vol. 136
Assessment of geophysical monitoring methods for detection of brine and CO2 leakage in drinking water aquifers
journal, November 2019
- Yang, Xianjin; Buscheck, Thomas A.; Mansoor, Kayyum
- International Journal of Greenhouse Gas Control, Vol. 90
Resolution, stability and efficiency of resistivity tomography estimated from a generalized inverse approach
journal, May 2003
- Friedel, Sven
- Geophysical Journal International, Vol. 153, Issue 2
The state of the art in monitoring and verification—Ten years on
journal, September 2015
- Jenkins, Charles; Chadwick, Andy; Hovorka, Susan D.
- International Journal of Greenhouse Gas Control, Vol. 40
Progress in monitoring strategies for risk reduction in geologic CO2 storage
journal, August 2016
- Harbert, William; Daley, Thomas M.; Bromhal, Grant
- International Journal of Greenhouse Gas Control, Vol. 51