Multi-level facies-constrained reservoir prediction and application in a sparse-well area of the Qiongdongnan Basin
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.
- 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
- 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
- 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
- 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
- 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
- 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
- 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
- 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
- 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
- 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
- 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
- Buland A, Omre H. Bayesian linearized AVO inversion. Geophysics. 2003;68(1):185-198. doi: 10.1190/1.1543206
- 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
- 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
- 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
- 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.
- 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.
- 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
- 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
- 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
- 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
- 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
- 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
- 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
- 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
- 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
- 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
- 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
- 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
- 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
- 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
- Park DC. Centroid neural network for unsupervised competitive learning. IEEE Trans Neural Netw. 2000;11(2):520-528. doi: 10.1109/72.839021
- 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
