Abstract
Volcanic reservoirs contain highly heterogeneous and extremely tight pore and fracture systems that exert a primary control on natural gas storage and migration. Accurate identification of micro fractures in µCT images remains challenging because fracture apertures are small, morphologies are complex, and grayscale contrast is low. To address this issue, this study develops an integrated framework that combines ensemble learning with a 2.5D deep learning model to achieve efficient and precise segmentation of volcanic micro fractures. A Random Forest model is first used to provide rapid pre segmentation, followed by a U Net plus plus model that captures through plane continuity and improves fracture boundary recognition. A semi-automatic label as you train strategy reduces annotation requirements while maintaining high accuracy. The model achieves a Dice coefficient of 0.902 within ten epochs. Based on the segmentation results, three dimensional reconstruction, pore throat network modeling and gas flow simulations were conducted to quantify fracture connectivity and reveal gas migration behavior. The results show that highly connected fractures create axial flow channels and lateral micro pathways that enhance seepage efficiency and govern the overall migration process of natural gas. Samples with larger pore throat radii and stronger multiscale connectivity present significantly higher permeabilities and more continuous flow paths. This study provides a practical and accurate fracture segmentation workflow and quantitatively demonstrates how micro fracture connectivity regulates gas transport in tight volcanic reservoirs. The findings offer theoretical and methodological support for digital rock analysis, reservoir evaluation and multiscale modeling in geomechanics.
Data availability
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to being collected and sorted by authors from some paid databases.
Abbreviations
- LPA:
-
Low-pressure gas adsorption
- MIP:
-
Mercury intrusion porosimetry
- NMR:
-
Nuclear magnetic resonance
- SANS:
-
Small-angle neutron scattering
- µCT:
-
Micro-computed tomography
- PNM:
-
Pore network modeling
- EIA:
-
Environmental impact assessment
- FOV:
-
Field of view
- REV:
-
Representative fundamental volume
References
Tao, S. Exploration and development of unconventional oil and gas resources: latest advances and prospects. Energies 18, 3933. https://doi.org/10.3390/en18153933 (2025).
Zhang, N. et al. Fracture conductivity and rock appearance in volcanic reservoirs treated by various stimulation techniques. Energy 295, 131045. https://doi.org/10.1016/j.energy.2024.131045 (2024).
Liao, Q. et al. Reservoir stimulation for unconventional oil and gas resources: recent advances and future perspectives. Adv. Geo-Energy Res. 13, 7–9. https://doi.org/10.46690/ager.2024.07.02 (2024).
Li, Y. et al. Quantitative prediction of fracture parameters in volcanic reservoirs using integrated rock physics and BO-BiLSTM network approach. IEEE Trans. Geosci. Remote Sens. 63, 1–14. https://doi.org/10.1109/TGRS.2025.3581090 (2025).
Tahir, M. U. & Guo, S. Preliminary investigation of fracture behavior during carbon dioxide fracturing of natural hydrogen reservoir with Hard-Core imperfections. Reserv. Sci. 2, 34–51. https://doi.org/10.62762/RS.2025.759326 (2026).
Gao, X. et al. Micro–nano 3D CT scanning to assess the impact of microparameters of volcanic reservoirs on gas migration. Processes 12, 2000. https://doi.org/10.3390/pr12092000 (2024).
Da Silva, L. C. V. et al. Entropy-based measure of rock sample heterogeneity derived from micro-CT images. Transp. Porous Media. 152, 44. https://doi.org/10.1007/s11242-025-02183-3 (2025).
Lee, D. et al. Detecting micro fractures: A comprehensive comparison of conventional and machine-learning-based segmentation methods. Solid Earth. 13, 1475–1494. https://doi.org/10.5194/se-13-1475-2022 (2022).
Zhang, L. et al. The characterization of bituminous coal microstructure and permeability by liquid nitrogen fracturing based on µCT technology. Fuel 262, 116635. https://doi.org/10.1016/j.fuel.2019.116635 (2020).
Chen, S. et al. Evolution of coal microfracture by Cyclic fracturing of liquid nitrogen based on µCT and convolutional neural networks. Rock. Mech. Rock. Eng. 57, 2103–2124. https://doi.org/10.1007/s00603-023-03649-w (2024).
Xu, J. et al. Basin evolution and oil shale deposition during upper cretaceous in the Songliao basin (NE China): implications from sequence stratigraphy and geochemistry. Int. J. Coal Geol. 149, 9–23. https://doi.org/10.1016/j.coal.2015.07.005 (2015).
Chi, H. et al. Geophysical characteristics of internal weathered volcanic crust in the Songliao basin. Acta Geophys. 73, 221–233. https://doi.org/10.1007/s11600-024-01356-w (2024).
Wu, H. C. et al. Integrated chronostratigraphic framework for cretaceous strata in the Songliao basin. Earth Sci. Front. 31, 431–445. https://doi.org/10.13745/j.esf.sf.2024.1.22 (2024).
Meng, S-W. et al. Mechanical characteristics and reservoir stimulation mechanisms of the Gulong shale oil reservoirs, the Northern Songliao basin. Pet. Sci. 21, 2023–2036. https://doi.org/10.1016/j.petsci.2023.11.002 (2024).
Liu, Y-K. et al. Rare Earth elemental and Sr isotopic evidence for seawater intrusion event of the Songliao basin 91 million years ago. Pet. Sci. 20, 1347–1362. https://doi.org/10.1016/j.petsci.2022.11.015 (2023).
Zhu, J. et al. Multi-scale CT imaging technique characterization of volcanic rock reservoirs: A study case from the Changling fault depression of the Southern Songliao basin. Front. Earth Sci. 13, 1456257. https://doi.org/10.3389/feart.2025.1456257 (2025).
Sun, Y. et al. Structural characteristics of faults in Wangfu fault depression and their control on coal-rock gas enrichment, Songliao Basin, NE China. Pet. Explor. Dev. 52, 649–662. https://doi.org/10.1016/S1876-3804(25)60594-5 (2025).
Liu, L. et al. The relationship between physical property evolution and reservoir formation in volcanic rock reservoirs: the huoshiling Formation, Dehui fault Depression, Southern Songliao basin. Mar. Pet. Geol. 179, 107446. https://doi.org/10.1016/j.marpetgeo.2025.107446 (2025).
Wang, T. et al. High-precision geochronology of the early cretaceous Yingcheng formation and its stratigraphic implications for Songliao Basin, China. Geosci. Front. 13, 101386. https://doi.org/10.1016/j.gsf.2022.101386 (2022).
Withers, P. J. et al. X-ray computed tomography. Nat. Rev. Methods Primer. 1, 18. https://doi.org/10.1038/s43586-021-00015-4 (2021).
Miyazaki, T. et al. Measurement of the particle density of small amounts of pharmaceutical powders using high-contrast micro X-ray computed tomography. Powder Technol. 457, 120929. https://doi.org/10.1016/j.powtec.2025.120929 (2025).
Wu, H. et al. Imaged based fractal characterization of micro-fracture structure in coal. Fuel 239, 53–62. https://doi.org/10.1016/j.fuel.2018.10.117 (2019).
Fan, Y. et al. Reconstructing data representation for multi-label feature selection. Pattern Recognit. 169, 111941. https://doi.org/10.1016/j.patcog.2025.111941 (2026).
Jin, C. et al. Segmentation of ore and waste rocks in borehole images using the multi-module densely connected U-net. Comput. Geosci. 159, 105018. https://doi.org/10.1016/j.cageo.2021.105018 (2022).
Shaker, M. H. & Hüllermeier, E. Random forest calibration. Knowl-Based Syst. 328, 114143. https://doi.org/10.1016/j.knosys.2025.114143 (2025).
He, C. et al. 3D analysis of mineralogical and fracture features in granite and impact of biotite distribution on uniaxial compressive strength. Eng. Geol. 356, 108279. https://doi.org/10.1016/j.enggeo.2025.108279 (2025).
Xing, S. et al. 2.5D-UNet-HC: 2.5D-UNet based on hybrid Convolution for prostate ultrasound image segmentation. IEEE Access. 13, 178652–178662. https://doi.org/10.1109/ACCESS.2025.3620348 (2025).
Singh, A. et al. On representative elementary volumes of grayscale micro-CT images of porous media. Geophys. Res. Lett. 47, e2020GL088594. https://doi.org/10.1029/2020GL088594 (2020).
Zhang, P. et al. Unet-based image segmentation and binarization for water level detection. Vis. Comput. 41, 7367–7377. https://doi.org/10.1007/s00371-025-03810-2 (2025).
Wei, S. et al. Classification of ground information during wildfires: using an improved UNet model. Int. J. Disaster Risk Reduct. 116, 105157. https://doi.org/10.1016/j.ijdrr.2024.105157 (2025).
Hou, Z. et al. UAV target segmentation based on depse Unet + + modeling. Drones 9, 166. https://doi.org/10.3390/drones9030166 (2025).
Angermann, C. & Haltmeier, M. Random 2.5D U-net for fully 3D segmentation. In: (eds Liao, H., Balocco, S., Wang, G., Zhang, F., Liu, Y., Ding, Z. et al.) Mach. Learn. Med. Eng. Cardiovasc. Health Intravasc. Imaging Comput. Assist. Stenting, Shenzhen, China: Springer International Publishing; 158–166. https://doi.org/10.1007/978-3-030-33327-0_19. (2019).
Zhou, Z. et al. UNet++: A nested U-Net architecture for medical image segmentation. In: (eds Stoyanov, D., Taylor, Z., Carneiro, G., Syeda-Mahmood, T., Martel, A. & Maier-Hein, L.) Deep Learn. Med. Image Anal. Multimodal Learn. Clin. Decis. Support, Granada, Spain: Springer International Publishing; 3–11. https://doi.org/10.1007/978-3-030-00889-5_1. (2018).
Tan, K. J. et al. Research on effectiveness of effective pore aspect ratio based on pore-throat characteristics of digital core. Chin. J. Geophys. 65, 4433–4447. https://doi.org/10.6038/cjg2022P0393 (2022).
Zhao, X. et al. Modeling of permeability and formation factor of carbonate digital rocks: Dual-pore-network and pore-network-continuum models. Transp. Porous Media. 152, 37. https://doi.org/10.1007/s11242-025-02177-1 (2025).
Bataineh, B. An iterative thinning algorithm for binary images based on sequential and parallel approaches. Pattern Recognit. Image Anal. 28, 34–43. https://doi.org/10.1134/S1054661818010030 (2018).
Alcántara-López, F. et al. Spatial fractional darcy’s law on the diffusion equation with a fractional time derivative in Single-Porosity naturally fractured reservoirs. Energies 15, 4837. https://doi.org/10.3390/en15134837 (2022).
Zhang, H. et al. Pore-scale characterization and PNM simulations of multiphase flow in carbonate rocks. Energies 14, 6897. https://doi.org/10.3390/en14216897 (2021).
Keogh, R. R. et al. Active darcy’s law. Phys. Rev. Lett. 132, 188301. https://doi.org/10.1103/PhysRevLett.132.188301 (2024).
Liu, Y. et al. Quantitative characterization of tight rock microstructure of digital core. Geofluids 2022, 1–18. https://doi.org/10.1155/2022/3554563 (2022).
Ge, L. et al. Label correlation for partial label learning. J. Syst. Eng. Electron. 33, 1043–1051. https://doi.org/10.23919/JSEE.2022.000102 (2022).
Zhang, G. et al. Automatic lung tumor segmentation from CT images using improved 3D densely connected UNet. Med. Biol. Eng. Comput. 60, 3311–3323. https://doi.org/10.1007/s11517-022-02667-0 (2022).
Song, M. S. et al. Study on the fracture distribution pattern of volcanicrock in thrust fault developed zone. Nat. Gas Geosci. 28, 989–999. https://doi.org/10.11764/j.issn.1672-1926.2017.06.011 (2017).
Bao, D. et al. Pore structure and seepage characteristics of hydrate-bearing sediments. Energy Fuels. 39, 3087–3096. https://doi.org/10.1021/acs.energyfuels.4c05720 (2025).
Wang, C. Y. Fluid seepage into a rotating channel filled with a porous medium. J. Eng. Mech. 151, 4024107. https://doi.org/10.1061/JENMDT.EMENG-8037 (2025).
Li, W. & Liao, J. Microscopic analysis of flow resistance of oil displacement fluid in reservoir fractures. Reserv. Sci. 2, 16–33. https://doi.org/10.62762/RS.2025.837826 (2026).
Xu, N. & Wang, Y. Effect of nanomaterials on improving the apparent viscosity of heavy oil and the environmental evaluation of reservoir environment. Reserv. Sci. 2, 1–15. https://doi.org/10.62762/RS.2025.277961 (2026).
Acknowledgements
We thank Sanying Precision Instruments Co., Ltd. for the technical support and equipment provided for this research. We appreciate the assistance of the Exploration and Development Research Institute of Jilin Oilfield Branch, China National Petroleum Corporation (CNPC), for sample preparation., and are grateful for their support.
Funding
This research was funded by the National Natural Science Foundation of China (No. 41472101).
Author information
Authors and Affiliations
Contributions
Conceptualization, Jiacheng Zhang and Yunliang Yu; methodology, Jiacheng Zhang; software, Jiacheng Zhang; validation, formal analysis, Jiacheng Zhang and Yunliang Yu; investigation, Jiacheng Zhang; resources, Yunliang Yu; data curation, Yunliang Yu; writing—original draft preparation, Jiacheng Zhang; writing—review and editing, Jiacheng Zhang, Yunliang Yu, Hongchen Cai and Mengyu Li; visualization, Jiacheng Zhang; supervision, Yunliang Yu and Yingchun Liu; project administration, Yunliang Yu; funding acquisition, Yunliang Yu. All authors have read and agreed to the published version of the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Zhang, J., Yu, Y., Cai, H. et al. Multiscale characterization of micro fracture connectivity and gas migration in volcanic reservoirs using µCT and hybrid learning segmentation. Sci Rep (2026). https://doi.org/10.1038/s41598-026-39657-3
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-026-39657-3
