1887

Abstract

Summary

The paper presents a deep learning-based approach to improve the alignment of 4D seismic volumes, addressing inconsistencies caused by surveying or data processing. These inconsistencies can lead to misinterpretations that affect production efficiency. The proposed 3D deep learning algorithm offers better results by accurately aligning monitor and base seismic data. The method uses the VoxelMorph framework, incorporating a modified U-Net CNN architecture to generate 3D deformation fields. This unsupervised technique as a result produces deformation and corresponding uncertainty cubes to enhance alignment accuracy. By comparing conventional 1D methods with the proposed 3D deep learning approach results show improved volume alignment, reducing mismatches and revealing finer details of reservoir dynamics. This method enhances the vertical resolution of 4D differences, optimizing the extraction of 4D responses from hydrocarbon-producing intervals, leading to better seismic interpretation and more effective reservoir management.

© EAGE Publications BV
Loading

Article metrics loading...

/content/papers/10.3997/2214-4609.202539087
2025-03-24
2025-12-06

  • From This Site

    /content/papers/10.3997/2214-4609.202539087
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Journal -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

References

  1. BalakrishnanG., ZhaoA., SabuncuM.R., GuttagJ. and DalcaA.V. [2019]. VoxelMorph: A Learning Framework for Deformable Medical Image Registration. IEEE Transactions on Medical Imaging, vol. 38, iss. 8, pp. 1788–1800
     [Google Scholar]
  2. Dramsch, J. S., Christensen, Anders, N., MacbethC. and Lüthje, M. [2021]. Deep Unsupervised 4-D Seismic 3-D Time-Shift Estimation with Convolutional Neural Networks. IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1–16, Art no. 5903116
     [Google Scholar]
  3. FomelS. and JinL. [2009], Time-lapse image registration using the local similarity attribute. Geophysics, vol. 74, no. 2, pp. A7–A11
     [Google Scholar]
  4. HaleD. [2013]. Dynamic warping of seismic images. Geophysics, vol. 78, no. 2, pp. S105–S115
     [Google Scholar]
  5. HatchellP. and BourneS. [2005]. Rocks under strain: Strain-induced timelapse time shifts are observed for depleting reservoirs. Lead. Edge, vol. 24, no. 12, pp. 1222–1225
     [Google Scholar]
  6. WongL. J., AminiH. and MacBethC. [2020]. Noisy 4D seismic data interpretation: Case study of a Brazilian carbonate reservoir. The Leading Edge, vol. 39, iss. 7, pp. 488–496
     [Google Scholar]
/content/papers/10.3997/2214-4609.202539087
Loading
Deep Learning Based 4D Volume Matching for Improved Alignment Compensation
Conference Proceedings 2025, 1 (2025); https://doi.org/10.3997/2214-4609.202539087
/content/papers/10.3997/2214-4609.202539087
/content/papers/10.3997/2214-4609.202539087
Loading

Data & Media loading...

Most Cited This Month Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error