A Data-Driven Method for Permeability Estimation of Rough Fractures

C. A.S. Ferreira, T. Kadeethum, N. Bouklas, H. M. Nick

    Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

    Abstract

    The work presented here is novel in the application of the image-to-image translation concept for scaling-up highly heterogeneous aperture distributions of fractures in a porous medium. We show that this framework can not only parameterize complex input fields, but more importantly realistic inputs based in fractal distributions. The heterogeneous aperture distributions are replaced by the upscaled tensors and the requirement of locally fine grids is removed thus reducing the computational cost. The model provides sufficient accuracy for capturing the anisotropy of the fracture with a speed-up of at-least 800-folds for highly heterogeneous aperture.

    Original languageEnglish
    Title of host publicationProceedings of the 83rd EAGE Conference and Exhibition 
    Volume4
    PublisherEuropean Association of Geoscientists and Engineers
    Publication date2022
    Pages2964-2968
    ISBN (Electronic)978-171385931-4
    DOIs
    Publication statusPublished - 2022
    Event83rd EAGE Annual Conference & Exhibition - IFEMA, Madrid, Spain
    Duration: 6 Jun 20229 Jun 2022
    Conference number: 83
    https://eage.eventsair.com/eageannual2022/

    Conference

    Conference83rd EAGE Annual Conference & Exhibition
    Number83
    LocationIFEMA
    Country/TerritorySpain
    CityMadrid
    Period06/06/202209/06/2022
    Internet address

    Fingerprint

    Dive into the research topics of 'A Data-Driven Method for Permeability Estimation of Rough Fractures'. Together they form a unique fingerprint.

    Cite this