AccScience Publishing / JSE / Online First / DOI: 10.36922/JSE026210087
Cite this article
4
Download
36
Views
Related Info Links
More by Authors Links
Journal Browser
Volume | Year
Issue
Search
News and Announcements
View All
ARTICLE

Multi-level facies-constrained reservoir prediction and application in a sparse-well area of the Qiongdongnan Basin

Wanyuan Sun1* ,  Shiyou Liu1 ,  Rui Wang1 ,  Shuxian Liu1
Show Less
1 Research Institute, China National Offshore Oil Corporation Hainan branch, Haikou, Hainan , China
Received: 19 May 2026 | Revised: 30 July 2026 | Accepted: 5 August 2026 | Published online: 16 November 2026
© 2026 by the Author(s). This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution 4.0 International License ( https://creativecommons.org/licenses/by/4.0/ )
Abstract

The deepwater area of the Qiongdongnan Basin is characterized by sparse well control, low-porosity, and low-permeability reservoirs, weak seismic responses, and small elastic-parameter contrasts. These conditions increase the non-uniqueness of conventional reservoir prediction. To reduce this uncertainty, this study proposes a multi-level facies-constrained reservoir prediction workflow that couples sedimentary facies, seismic facies, and logging lithofacies. The workflow first constructs a facies prior from seismic waveform classification, well-log interpretation, and sedimentary constraints. It then performs Bayesian pre-stack inversion based on the exact Zoeppritz equation, followed by semi-supervised porosity prediction and frequency-dependent amplitude variation with offset flowability analysis. Finally, the P/S velocity ratio, porosity, and water-saturation-related attributes are integrated using a competitive neural network to delineate favorable reservoir zones. Synthetic tests and a Stanford reservoir model test show that the proposed method improves facies agreement from 78.57% to 92.86% relative to a conventional facies-constrained workflow. Because the field area contains only one drilled well, the field results are interpreted as probabilistic exploration targets rather than blind-well proof of deterministic accuracy. The workflow provides a reproducible and geologically constrained approach for reservoir prediction in sparse-well deepwater settings.

Keywords
Reservoir prediction
Facies-constrained inversion
Competitive neural network
Sparse-well area
Qiongdongnan Basin
Funding
None.
Conflict of interest
The authors declare no potential conflict of interest.
References
  1. Chen P, Lu Y, Wang Z, Chen L. Seismic image prediction of Cenozoic reef-banks in the northwestern South China Sea. J Earth Sci. 2012;23(4):597-611. doi: 10.1007/s12583-012-0277-y
  2. Feng Y, Ren Y, Li Z, Jin L. Geological interpretation and hydrocarbon exploration potential of three types of mound-shaped reflectors in the Meishan Formation, southern Qiongdongnan Basin. Acta Geol Sin (Engl). 2021;95(1):167-176. doi: 10.1111/1755-6724.14636
  3. Pacheco JMÚ, Bayón PV, Correia U, Kneller E, Peiro M. Unlocking the properties of a presalt carbonate reservoir offshore Brazil with facies-constrained geostatistical inversion. Lead Edge. 2022;41(12):832-839. doi: 10.1190/tle41120832.1
  4. Xie Q, Wu Y, Huang Q, et al. Prediction of marine thin shale gas reservoir with seismic phase-controlled nonlinear stochastic inversion. Processes. 2023;11(8):2301. doi: 10.3390/pr11082301
  5. Zheng C, Zhang B, Li R, et al. Facies-controlled prediction of dolomite reservoirs in the Middle Permian Qixia Formation in Shuangyushi, northwestern Sichuan Basin. Energy Geosci. 2024;5(2):100230. doi: 10.1016/j.engeos.2023.100230
  6. Su Q, Xu X, Chen T, Zong J, Wang H. A facies-constrained geostatistical seismic inversion method based on multi-scale sparse representation. Geophys Prospect. 2024;72(5):1657-1671. doi: 10.1111/1365-2478.13476
  7. Fernandes FJD, Teixeira L, Freire AFM, Lupinacci WM. Stochastic seismic inversion and Bayesian facies classification applied to porosity modeling and igneous rock identification. Pet Sci. 2024;21(2):918-935. doi: 10.1016/j.petsci.2023.11.020
  8. Zhang H, Chen L, Zhu H, Xin Y, Wang Y, Sun X. Prediction of marl reservoir distribution based on facies-constrained reflectivity inversion method. Front Earth Sci. 2024;12:1495720. doi: 10.3389/feart.2024.1495720
  9. Caers J, Hoffman T. The probability perturbation method: a new look at Bayesian inverse modeling. Math Geol. 2006;38(1):81-100. doi: 10.1007/s11004-005-9005-9
  10. Huang Y, Xu LH, Yang HD, He QL, He YH, Yang QJ. Fanyan yueshu de duodian dizhi tongjixue jianmo—yi Daqing changyuan luxiang youtian weili [Multi-point geostatistical modeling with inversion constraints: A case study of continental oil fields in Daqing placanticline]. Oil Geophys Prospect. 2022;57(6):1445-1452. [In Chinese]. doi: 10.13810/j.cnki.issn.1000-7210.2022.06.020
  11. Bosch M, Mukerji T, Gonzalez EF. Seismic inversion for reservoir properties combining statistical rock physics and geostatistics: a review. Geophysics. 2010;75(5):75A165-75A176. doi: 10.1190/1.3478209
  12. Buland A, Omre H. Bayesian linearized AVO inversion. Geophysics. 2003;68(1):185-198. doi: 10.1190/1.1543206
  13. Chen H, Han J, Li K. Facies-constrained simultaneous inversion for elastic parameters and fracture weaknesses using azimuthal partially incidence-angle-stacked seismic data. Geophys Prospect. 2024;73(2):562-574. doi: 10.1111/1365-2478.13659
  14. Larsen AL, Ulvmoen M, Omre H, Buland A. Bayesian lithology/fluid prediction and simulation on the basis of a Markov-chain prior model. Geophysics. 2006;71(5):R69-R78. doi: 10.1190/1.2245469
  15. Liu M, Jervis M, Li W, Nivlet P. Seismic facies classification using supervised convolutional neural networks and semisupervised generative adversarial networks. Geophysics. 2020;85(4):O47-O58. doi: 10.1190/geo2019-0627.1
  16. Cai H, Wu Q, Ren H, et al. Pre-stack texture-based semi-supervised seismic facies analysis using global optimization. J Seism Explor. 2019;28(6):513-532.
  17. Zhao M, Wang YQ, Peng ZM, et al. Seismic multi-attribute fusion using fast independent component analysis and its application. J Seism Explor. 2019;28(1):89-101.
  18. Song Z, Li S, He S, Yuan S, Wang S. Gas-bearing prediction of tight sandstone reservoir using semi-supervised learning and transfer learning. IEEE Geosci Remote Sens Lett. 2022;19:1-5. doi: 10.1109/LGRS.2022.3177314
  19. Dhara A, Sen MK, Biswas R. Facies-constrained transdimensional amplitude versus angle inversion using machine learning assisted priors. Geophys Prospect. 2023;71(4):590-613. doi: 10.1111/1365-2478.13339
  20. Xie Y, Zhang G, Tang W, Zhao Z. Nanhai beibu shenshuiqu youqi chengcang lilun jishu chuangxin yu kantan zhongda tupo [Theoretical and technological innovation of oil and gas accumulation and major exploration breakthroughs in deep-water areas, northern South China Sea]. Nat Gas Ind. 2020;40(12):1-11. [In Chinese]. doi: 10.3787/j.issn.1000-0976.2020.12.001
  21. Xie Y, Gao Y. Zhongguo haiyou jinqi guonei kantan jinzhan yu kantan fangxiang [Recent domestic exploration progress and direction of CNOOC]. China Pet Explor. 2020;25(1):20-30. [In Chinese]. doi: 10.3969/j.issn.1672-7703.2020.01.003
  22. Xie YH. Jiyu dizhen gouxiang de xiangkong chuceng yuce fangfa [Facies-controlled reservoir prediction based on the seismic configuration]. Geophys Prospect Pet. 2021;60(5):784-793. [In Chinese]. doi: 10.3969/j.issn.1000-1441.2021.05.009
  23. Fu C, Xie Y, Song L, et al.  Jiyu chenji moni de xuni jing jianli yu shenshui xiagu xiangkong jianmo celue—yi Qiongdongnan pendi zhongyang xiagu Lingshui duan weili [Pseudo-well setting and strategy of facies-controlled reservoir modeling of deep water canyon based on sedimentary simulation: Taking the Lingshui section of the Central Canyon in Qiongdongnan Basin as an example]. Mar Orig Pet Geol. 2023;28(4):413-424. [In Chinese]. Accessed November 3, 2026. https://www.hxyqdz.com/CN/Y2023/V28/I4/413
  24. Yin X, Liu C, Wang B. Jiyu hunhe yichuan suanfa de dieqian suiji fanyan fangfa [Pre-stack stochastic inversion based on hybrid genetic algorithm]. J China Univ Pet (Ed Nat Sci). 2017;41(4):57-64. [In Chinese]. doi: 10.3969/j.issn.1673-5005.2017.04.008
  25. Yin X, Kong S, Zhang F, Zhang X. Jiyu chayi yanhua suanfa de dieqian AVO fanyan [Pre-stack AVO inversion based on the differential evolution algorithm]. Oil Geophys Prospect. 2013;48(6):948-953. [In Chinese]. doi: 10.13810/j.cnki.issn.1000-7210.2013.04.013
  26. Liang J, Wang Q, Hao L, Tang J, Liao P. Chengyanxiang fenxi fangfa zai Nanhai beibu shenshuiqu chuceng yuce de yingyong—yi Zhujiangkou pendi Baiyun aoaxian weili [Application of diagenetic facies analysis to reservoir prediction in deep water area of the northern South China Sea: A case study from Baiyun Sag, Zhujiangkou Basin]. Acta Sedimentol Sin. 2011;29(3):503-512. [In Chinese]. doi: 10.14027/j.cnki.cjxb.2011.03.002
  27. Yin X, Zhang S, Zhang F, et al. Liyong jiyu Russell jinsi de tanxing bozu kang fanyan jinxing chuceng miaoshu he liuti shibie [Utilizing Russell approximation-based elastic wave impedance inversion to conduct reservoir description and fluid identification]. Oil Geophys Prospect. 2010;45(3):373-380. [In Chinese]. Accessed November 3, 2026. http://www.ogp-cn.com.cn/EN/Y2010/V45/I3/373
  28. Gu W, Yin X, Wu F, Li K, Zhai H. Boxing qudong xia duocanshu yueshu gao fenbianlv fanyan fangfa—yi Sichuan pendi Yuxi diqu Longmaxi zu yeyanqi weili [Multi-parameter constrained high-resolution inversion method driven by waveform: A case study of Longmaxi Formation shale gas in western Chongqing, Sichuan Basin]. Oil Geophys Prospect. 2021;56(6):1311-1321. [In Chinese]. doi: 10.13810/j.cnki.issn.1000-7210.2021.06.014
  29. Ling T, You L, Lu Y, Luo W, Ye Q. Qiongdongnan pendi Baodao aoaxian nan xiepo haidishan chenji tezheng yu youli chuceng yuce [Sedimentary characteristics and high-quality reservoir prediction of submarine fan in the southern continental slope of Baodao Sag, Qiongdongnan Basin]. Complex Hydrocarbon Reserv. 2025;18(3):311-317. [In Chinese]. doi: 10.16181/j.cnki.fzyqc.2025.03.008
  30. Xu X, Zhou W, Zhang C, et al. Shenshui shenceng di shen youqi cang chuceng tezheng ji wuxing kongzhi yinsu—yi Qiongdongnan pendi Baodao aoaxian nan po Lingsan duan weili [Reservoir characteristics and physical property control factors in deep water and low-permeability oil and gas reservoirs: A case study of the 3rd member of Lingshui Formation on the southern slope of Baodao Sag in Qiongdongnan Basin]. Mar Orig Pet Geol. 2026;31(1):48-60. [In Chinese]. doi: 10.3969/j.issn.1672-9854.2026.01.004
  31. Xu C, You L. North slope transition zone of Songnan-Baodao sag in Qiongdongnan Basin and its control on medium and large gas fields, South China Sea. Pet Explor Dev. 2022;49(6):1229-1242. doi: 10.1016/S1876-3804(23)60345-3
  32. Park DC. Centroid neural network for unsupervised competitive learning. IEEE Trans Neural Netw. 2000;11(2):520-528. doi: 10.1109/72.839021
  33. Park DC, Kwon OH, Jeon BN. Centroid neural network with a divergence measure for GPDF data clustering. IEEE Trans Neural Netw. 2008;19(6):948-957. doi: 10.1109/TNN.2007.2000051
Share
Back to top
Journal of Seismic Exploration, Print ISSN: 0963-0651, Published by AccScience Publishing