{"title":"Spatio-temporal neural networks for monitoring and prediction of CO2 plume migration from measurable field data","authors":"Yingxiang Liu, Zhen Qin, Fangning Zheng, Behnam Jafarpour","doi":"10.1016/j.jclepro.2024.144080","DOIUrl":null,"url":null,"abstract":"Carbon capture, utilization, and storage (CCUS) technologies are crucial for mitigating greenhouse gas emissions. The success of CCUS projects hinges on accurate prediction and monitoring of the CO<span><span style=\"\"></span><span data-mathml='<math xmlns=\"http://www.w3.org/1998/Math/MathML\"><msub is=\"true\"><mrow is=\"true\" /><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math>' role=\"presentation\" style=\"font-size: 90%; display: inline-block; position: relative;\" tabindex=\"0\"><svg aria-hidden=\"true\" focusable=\"false\" height=\"1.509ex\" role=\"img\" style=\"vertical-align: -0.582ex;\" viewbox=\"0 -399.4 453.9 649.8\" width=\"1.054ex\" xmlns:xlink=\"http://www.w3.org/1999/xlink\"><g fill=\"currentColor\" stroke=\"currentColor\" stroke-width=\"0\" transform=\"matrix(1 0 0 -1 0 0)\"><g is=\"true\"><g is=\"true\"></g><g is=\"true\" transform=\"translate(0,-150)\"><g is=\"true\"><use transform=\"scale(0.707)\" xlink:href=\"#MJMAIN-32\"></use></g></g></g></g></svg><span role=\"presentation\"><math xmlns=\"http://www.w3.org/1998/Math/MathML\"><msub is=\"true\"><mrow is=\"true\"></mrow><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math></span></span><script type=\"math/mml\"><math><msub is=\"true\"><mrow is=\"true\"></mrow><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math></script></span> plume migration during and after injection. To address the computational burden of traditional numerical simulation methods, previous studies have successfully used neural networks as proxy models to expedite the prediction of the CO<span><span style=\"\"></span><span data-mathml='<math xmlns=\"http://www.w3.org/1998/Math/MathML\"><msub is=\"true\"><mrow is=\"true\" /><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math>' role=\"presentation\" style=\"font-size: 90%; display: inline-block; position: relative;\" tabindex=\"0\"><svg aria-hidden=\"true\" focusable=\"false\" height=\"1.509ex\" role=\"img\" style=\"vertical-align: -0.582ex;\" viewbox=\"0 -399.4 453.9 649.8\" width=\"1.054ex\" xmlns:xlink=\"http://www.w3.org/1999/xlink\"><g fill=\"currentColor\" stroke=\"currentColor\" stroke-width=\"0\" transform=\"matrix(1 0 0 -1 0 0)\"><g is=\"true\"><g is=\"true\"></g><g is=\"true\" transform=\"translate(0,-150)\"><g is=\"true\"><use transform=\"scale(0.707)\" xlink:href=\"#MJMAIN-32\"></use></g></g></g></g></svg><span role=\"presentation\"><math xmlns=\"http://www.w3.org/1998/Math/MathML\"><msub is=\"true\"><mrow is=\"true\"></mrow><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math></span></span><script type=\"math/mml\"><math><msub is=\"true\"><mrow is=\"true\"></mrow><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math></script></span> plume migration. However, these models rely on uncertain inputs, such as the distribution of heterogeneous permeability and porosity maps, which can lead to erroneous predictions and pose a significant hurdle for their adoption in real-world applications. To address this issue, this study introduces a framework for reconstruction and short-term prediction of CO<span><span style=\"\"></span><span data-mathml='<math xmlns=\"http://www.w3.org/1998/Math/MathML\"><msub is=\"true\"><mrow is=\"true\" /><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math>' role=\"presentation\" style=\"font-size: 90%; display: inline-block; position: relative;\" tabindex=\"0\"><svg aria-hidden=\"true\" focusable=\"false\" height=\"1.509ex\" role=\"img\" style=\"vertical-align: -0.582ex;\" viewbox=\"0 -399.4 453.9 649.8\" width=\"1.054ex\" xmlns:xlink=\"http://www.w3.org/1999/xlink\"><g fill=\"currentColor\" stroke=\"currentColor\" stroke-width=\"0\" transform=\"matrix(1 0 0 -1 0 0)\"><g is=\"true\"><g is=\"true\"></g><g is=\"true\" transform=\"translate(0,-150)\"><g is=\"true\"><use transform=\"scale(0.707)\" xlink:href=\"#MJMAIN-32\"></use></g></g></g></g></svg><span role=\"presentation\"><math xmlns=\"http://www.w3.org/1998/Math/MathML\"><msub is=\"true\"><mrow is=\"true\"></mrow><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math></span></span><script type=\"math/mml\"><math><msub is=\"true\"><mrow is=\"true\"></mrow><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math></script></span> plume migration based solely on field measurements, which can directly provide information on CO<span><span style=\"\"></span><span data-mathml='<math xmlns=\"http://www.w3.org/1998/Math/MathML\"><msub is=\"true\"><mrow is=\"true\" /><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math>' role=\"presentation\" style=\"font-size: 90%; display: inline-block; position: relative;\" tabindex=\"0\"><svg aria-hidden=\"true\" focusable=\"false\" height=\"1.509ex\" role=\"img\" style=\"vertical-align: -0.582ex;\" viewbox=\"0 -399.4 453.9 649.8\" width=\"1.054ex\" xmlns:xlink=\"http://www.w3.org/1999/xlink\"><g fill=\"currentColor\" stroke=\"currentColor\" stroke-width=\"0\" transform=\"matrix(1 0 0 -1 0 0)\"><g is=\"true\"><g is=\"true\"></g><g is=\"true\" transform=\"translate(0,-150)\"><g is=\"true\"><use transform=\"scale(0.707)\" xlink:href=\"#MJMAIN-32\"></use></g></g></g></g></svg><span role=\"presentation\"><math xmlns=\"http://www.w3.org/1998/Math/MathML\"><msub is=\"true\"><mrow is=\"true\"></mrow><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math></span></span><script type=\"math/mml\"><math><msub is=\"true\"><mrow is=\"true\"></mrow><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math></script></span> plumes, thus eliminating the dependency on uncertain geological information as model input. The framework trains a spatio-temporal neural network model using simulated data under various geologic scenarios to capture the plume evolution dynamics without constraining it to specific geological scenarios. Once trained, the model integrates global and local field measurements from multiple sources to reconstruct the CO<span><span style=\"\"></span><span data-mathml='<math xmlns=\"http://www.w3.org/1998/Math/MathML\"><msub is=\"true\"><mrow is=\"true\" /><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math>' role=\"presentation\" style=\"font-size: 90%; display: inline-block; position: relative;\" tabindex=\"0\"><svg aria-hidden=\"true\" focusable=\"false\" height=\"1.509ex\" role=\"img\" style=\"vertical-align: -0.582ex;\" viewbox=\"0 -399.4 453.9 649.8\" width=\"1.054ex\" xmlns:xlink=\"http://www.w3.org/1999/xlink\"><g fill=\"currentColor\" stroke=\"currentColor\" stroke-width=\"0\" transform=\"matrix(1 0 0 -1 0 0)\"><g is=\"true\"><g is=\"true\"></g><g is=\"true\" transform=\"translate(0,-150)\"><g is=\"true\"><use transform=\"scale(0.707)\" xlink:href=\"#MJMAIN-32\"></use></g></g></g></g></svg><span role=\"presentation\"><math xmlns=\"http://www.w3.org/1998/Math/MathML\"><msub is=\"true\"><mrow is=\"true\"></mrow><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math></span></span><script type=\"math/mml\"><math><msub is=\"true\"><mrow is=\"true\"></mrow><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math></script></span> plume and predict its spatio-temporal evolution. The effectiveness of the proposed framework is tested using two case studies: one using a synthetic dataset and another with data generated from a model of a real field in the Southern San Joaquin Basin. The results show that the proposed framework can accurately reconstruct and perform short-term predictions of the CO<span><span style=\"\"></span><span data-mathml='<math xmlns=\"http://www.w3.org/1998/Math/MathML\"><msub is=\"true\"><mrow is=\"true\" /><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math>' role=\"presentation\" style=\"font-size: 90%; display: inline-block; position: relative;\" tabindex=\"0\"><svg aria-hidden=\"true\" focusable=\"false\" height=\"1.509ex\" role=\"img\" style=\"vertical-align: -0.582ex;\" viewbox=\"0 -399.4 453.9 649.8\" width=\"1.054ex\" xmlns:xlink=\"http://www.w3.org/1999/xlink\"><g fill=\"currentColor\" stroke=\"currentColor\" stroke-width=\"0\" transform=\"matrix(1 0 0 -1 0 0)\"><g is=\"true\"><g is=\"true\"></g><g is=\"true\" transform=\"translate(0,-150)\"><g is=\"true\"><use transform=\"scale(0.707)\" xlink:href=\"#MJMAIN-32\"></use></g></g></g></g></svg><span role=\"presentation\"><math xmlns=\"http://www.w3.org/1998/Math/MathML\"><msub is=\"true\"><mrow is=\"true\"></mrow><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math></span></span><script type=\"math/mml\"><math><msub is=\"true\"><mrow is=\"true\"></mrow><mrow is=\"true\"><mn is=\"true\">2</mn></mrow></msub></math></script></span> plume migration by integrating various forms of field measurements.","PeriodicalId":349,"journal":{"name":"Journal of Cleaner Production","volume":null,"pages":null},"PeriodicalIF":9.7000,"publicationDate":"2024-10-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Cleaner Production","FirstCategoryId":"93","ListUrlMain":"https://doi.org/10.1016/j.jclepro.2024.144080","RegionNum":1,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, ENVIRONMENTAL","Score":null,"Total":0}
引用次数: 0
Abstract
Carbon capture, utilization, and storage (CCUS) technologies are crucial for mitigating greenhouse gas emissions. The success of CCUS projects hinges on accurate prediction and monitoring of the CO plume migration during and after injection. To address the computational burden of traditional numerical simulation methods, previous studies have successfully used neural networks as proxy models to expedite the prediction of the CO plume migration. However, these models rely on uncertain inputs, such as the distribution of heterogeneous permeability and porosity maps, which can lead to erroneous predictions and pose a significant hurdle for their adoption in real-world applications. To address this issue, this study introduces a framework for reconstruction and short-term prediction of CO plume migration based solely on field measurements, which can directly provide information on CO plumes, thus eliminating the dependency on uncertain geological information as model input. The framework trains a spatio-temporal neural network model using simulated data under various geologic scenarios to capture the plume evolution dynamics without constraining it to specific geological scenarios. Once trained, the model integrates global and local field measurements from multiple sources to reconstruct the CO plume and predict its spatio-temporal evolution. The effectiveness of the proposed framework is tested using two case studies: one using a synthetic dataset and another with data generated from a model of a real field in the Southern San Joaquin Basin. The results show that the proposed framework can accurately reconstruct and perform short-term predictions of the CO plume migration by integrating various forms of field measurements.
期刊介绍:
The Journal of Cleaner Production is an international, transdisciplinary journal that addresses and discusses theoretical and practical Cleaner Production, Environmental, and Sustainability issues. It aims to help societies become more sustainable by focusing on the concept of 'Cleaner Production', which aims at preventing waste production and increasing efficiencies in energy, water, resources, and human capital use. The journal serves as a platform for corporations, governments, education institutions, regions, and societies to engage in discussions and research related to Cleaner Production, environmental, and sustainability practices.