基于地震属性堆叠泛化集成学习的辫状河储层构型表征——以渤海湾盆地C-2油田为例

doi:10.13809/j.cnki.cn32-1825/te.2025.01.008

摘要/Abstract

摘要：

C-2油田是位于渤海湾盆地的河流相海上油田，主要采用水平井开发，储层厚度薄、纵向多期砂体叠置、横向相变快，储层内部结构与连通关系复杂，受复杂的储层结构与井震资料的双重影响，稀疏井网条件下储层描述难度大，制约了油田精细挖潜效果，常规地震反演难以满足薄储层高分辨率预测与储层内部结构精细解剖的需要。针对研究油田辫状河储层结构特征，采用基于地震属性堆叠泛化集成学习的方法完成河流相复杂结构储层的预测，相较于单一的机器学习模型提高了预测精度。综合地质、地球物理与油藏动态等多维信息进行迭代优化，进一步降低地下储层预测与结构认识的不确定性，实现了研究区辫状河复杂结构储层的精细表征，为油田剩余油与潜力砂体精细挖潜提供了依据。研究表明：基于地震属性堆叠泛化集成学习的储层预测方法，不仅能有效提高地震纵向分辨率，同时具有较强的横向“相控”指示能力，预测结果包括砂体叠置关系与储层内部结构特征，更加适用于相变快、储层空间建筑结构复杂的陆相河流沉积体系储层预测与精细刻画，可为稀疏井网海上油田开发中后期河流相沉积储层构型精细表征提供借鉴。

关键词: 机器学习, 储层预测, 辫状河, 储层构型, 水平井

Abstract:

The C-2 oilfield, located in the Bohai Bay Basin, is a fluvial-facies offshore oilfield primarily developed with horizontal wells. It is characterized by thin reservoir layers, vertically stacked multi-phase sandbodies, and rapid lateral facies transitions, leading to complex internal reservoir structures and connectivity. The combined effects of complex reservoir structures and well-seismic data make reservoir characterization challenging under sparse well patterns, hindering refined development. Conventional seismic inversion fails to meet the requirements for high-resolution prediction of thin reservoirs and detailed characterization of internal reservoir structures. To study the structural characteristics of braided river reservoirs in the oilfield, a stacking ensemble learning method based on seismic attributes was applied to predict the complex fluvial-facies reservoir structures. This approach significantly improved prediction accuracy compared to a single machine learning model. By integrating multi-dimensional information such as geology, geophysics, and reservoir dynamics, iterative optimization was conducted to further reduce the uncertainty in subsurface reservoir prediction and structural understanding. This enabled the precise characterization of the complex braided river reservoir structures in the study area, providing a basis for refined exploitation of remaining oil and potential sandbodies in the oilfield. The study demonstrates that the reservoir prediction method based on stacking ensemble learning not only enhances seismic vertical resolution, but also exhibits strong horizontal phase-control capabilities. The prediction results include sandbody stacking relationships and internal reservoir structures, making it more suitable for the prediction and fine characterization of continental fluvial sedimentary systems with rapid facies transitions and complex spatial architectural structures. This method can serve as a reference for the detailed characterization of fluvial-facies reservoir configurations during the middle and late development stages of offshore oilfields with sparse well patterns.

Key words: machine learning, reservoir prediction, braided river, reservoir architecture, horizontal well

中图分类号:

TE122

张章,孟鹏,杨威等. 基于地震属性堆叠泛化集成学习的辫状河储层构型表征——以渤海湾盆地C-2油田为例[J]. 油气藏评价与开发, 2025, 15(1): 64-72.

ZHANG Zhang,MENG Peng,YANG Wei, et al. Characterization of braided river reservoir architecture based on seismic attribute stacking ensemble learning: A case study of the C-2 oilfield in the Bohai Bay Basin[J]. Petroleum Reservoir Evaluation and Development, 2025, 15(1): 64-72.

图/表 9

图1

图2

表1

表2

图3

图4

图5

图6

图7

参考文献 22

[1]	李洪辉, 岳大力, 李伟, 等. 基于分频智能反演的曲流河点坝与废弃河道识别[J]. 石油地球物理勘探, 2023, 58(2): 358-368.
	LI Honghui, YUE Dali, LI Wei, et al. Identification of point bar and abandoned channel of meandering river by spectral decomposition inversion based on machine learning[J]. Oil Geophysical Prospecting, 2023, 58(2): 358-368.
[2]	赵鹏飞, 刘财, 冯晅, 等. 基于神经网络的随机地震反演方法[J]. 地球物理学报, 2019, 62(3): 1172-1180.
	ZHAO Pengfei, LIU Cai, FENG Xuan, et al. Stochastic seismic inversion based on neural network[J]. Chinese Journal of Geophysics, 2019, 62(3): 1172-1180.
[3]	程冰洁, 徐天吉, 罗诗艺, 等. 基于机器学习的深层页岩有利储集层预测方法及实践[J]. 石油勘探与开发, 2022, 49(5): 1-11.
	CHENG Bingjie, XU Tianji, LUO Shiyi, et al. Method and practice of deep favorable shale reservoir prediction based on machine learning[J]. Petroleum Exploration and Development, 2022, 49(5): 1-11.
[4]	安鹏, 曹丹平, 赵宝银, 等. 基于LSTM循环神经网络的储层物性参数预测方法研究[J]. 地球物理学进展, 2019, 34(5): 1849-1858.
	AN Peng, CAO Danping, ZHAO Baoyin, et al. Reservoir physical parameters prediction based on LSTM recurrent neural network[J]. Progress in Geophysics, 2019, 34(5): 1849-1858.
[5]	张国印, 王志章, 林承焰, 等. 基于小波变换和卷积神经网络的地震储层预测方法及应用[J]. 中国石油大学学报(自然科学版), 2020, 44(4): 83-93.
	ZHANG Guoyin, WANG Zhizhang, LIN Chengyan, et al. Seismic reservoir prediction method based on wavelet transform and convolutional neural network and its application[J]. Journal of China University of Petroleum(Edition of Natural Science), 2020, 44(4): 83-93.
[6]	LI W, YUE D L, WANG W F, et al. Fusing multiple frequency-decomposed seismic attributes with machine learning for thickness prediction and sedimentary facies interpretation in fluvial reservoirs[J]. Journal of Petroleum Science and Engineering, 2019, 177: 1087-1102.
[7]	张宇航, 时保宏, 张曰静, 等. 机器学习方法在浅层滩坝相薄储层孔隙度预测中的应用: 以准噶尔盆地车排子地区白垩系为例[J]. 沉积学报, 2023, 41(5): 1559-1567.
	ZHANG Yuhang, SHI Baohong, ZHANG Yuejing, et al. Application of Machine Learning for Porosity Estimation of Beach and Bar Sand Bodies in a Lacustrine Basin: A case study of the Lower Cretaceous strata in Chepaizi area, Junggar Basin, NW China[J]. Acta Sedimentologica Sinica, 2023, 41(5): 1559-1567.
[8]	蔡义峰, 熊婷, 姚卫江, 等. 地震多属性分析技术在薄层砂体预测中的应用[J]. 石油地球物理勘探, 2017, 52(增刊2): 140-145.
	CAI Yifeng, XIONG Ting, YAO Weijiang, et al. Thin sandstone prediction with seismic multi-attribute analysis[J]. Oil Geophysical Prospecting, 2017, 52(Suppl. 2): 140-145.
[9]	曲志鹏, 王芳芳, 张云银, 等. 基于关联规则与随机森林的地震多属性砂体厚度预测[J]. 地质科技通报, 2021, 40(3): 211-218.
	QU Zhipeng, WANG Fangfang, ZHANG Yunyin, et al. Thickness prediction of seismic multi-attributes sand based on association rules and random forests[J]. Bulletin of Geological Science and Technology, 2021, 40(3): 211-218.
[10]	史长林, 魏莉, 张剑, 等. 基于机器学习的储层预测方法[J]. 油气地质与采收率, 2022, 29(1): 90-97.
	SHI Changlin, WEI Li, ZHANG Jian, et al. Reservoir prediction method based on machine learning[J]. Petroleum Geology and Recovery Efficiency, 2022, 29(1): 90-97.
[11]	YANG R X, SUN C Y, XU L. Prediction of photovoltaic power generation based on stacking model fusion[J]. Computer System Application, 2020, 29(5): 36-45.
[12]	曹志民, 丁璐, 韩建. 基于Stacking集成学习的声波时差测井曲线复原研究[J]. 化工自动化及仪表, 2024, 51(3): 470-476.
	CAO Zhimin, DING Lu, HAN Jian. Research on acoustic moveout logging curves restoration based on stacking ensemble learning[J]. Control and Instruments in Chemical Industry, 2024, 51(3): 470-476.
[13]	WOLPERT D H. Stacked generalization[J]. Neural Networks, 1992, 5(2): 241-259.
[14]	钱玉贵. 机器深度学习技术在致密砂岩储层预测中的应用: 以川西坳陷新场须家河组为例[J]. 油气藏评价与开发, 2023, 13(5): 600-607.
	QIAN Yugui. Application of machine deep learning technology in tight sandstones reservoir prediction: A case study of Xujiahe Formation in Xinchang, western Sichuan Depression[J]. Petroleum Reservoir Evaluation and Development, 2023, 13(5): 600-607.
[15]	吴胜和, 岳大力, 刘建民, 等. 地下古河道储层构型的层次建模研究[J]. 中国科学(地球科学), 2008, 38(增刊1): 111-121.
	WU Shenghe, YUE Dali, LIU Jianmin, et al. Hierarchy modeling of subsurface palaeochannel reservoir architecture[J]. Scientia Sinica(Terrae): Earth Sciences, 2008, 38(Suppl. 1): 111-121.
[16]	MIALL A D. Architectural-element analysis: A new method of facies analysis applied to fluvial deposits[J]. Earth-Science Reviews, 1985, 22(4): 261-308.
[17]	MIALL A D. The geology of fluvial deposits: sedimentary facies, basin analysis and petroleum geology[M]. Berlin:Springer, 1996.
[18]	侯东梅, 赵秀娟, 汪巍, 等. 地下曲流河点坝砂体规模定量表征研究[J]. 油气藏评价与开发, 2018, 8(3): 7-11.
	HOU Dongmei, ZHAO Xiujuan, WANG Wei, et al. Quantitative characterization research for point bar sand body of subsurface meandering river environment: Taking Minghua Formation of Bohai C Oilfield as an Instance[J]. Petroleum Reservoir Evaluation and Development, 2018, 8(3): 7-11.
[19]	KELLY S. Scaling and hierarchy in braided rivers and their deposits: Examples and implications for reservoir modeling[M]. UK: Blackwell Publishing, 2006.
[20]	陈薪凯, 刘景彦, 陈程, 等. 主要构型要素细分下的曲流河单砂体识别[J]. 沉积学报, 2019, 38(1): 205-217.
	CHEN Xinkai, LIU Jingyan, CHEN Cheng, et al. The identification of single sand body in meandering river deposits based on the subdivision of main architecture[J]. Acta Sedimentologica Sinica, 2019, 38(1): 205-217.
[21]	徐中波, 汪利兵, 申春生, 等. 渤海蓬莱19-3 油田新近系明下段曲流河储层构型表征[J]. 岩性油气藏, 2023, 35(5): 100-107.
	XU Zhongbo, WANG Libing, SHEN Chunsheng, et al. Architecture characterization of meandering river reservoirs of lower Minghuazhen Formation of Neogene in Penglai 19-3 oilfield, Bohai Sea[J]. Lithologic Reservoirs, 2023, 35(5): 100-107.
[22]	权勃, 侯东梅, 牟松茹, 等. 基于水平井信息的辫状河储层构型单元空间展布研究[J]. 中国海上油气, 2020, 32(4): 96-103.
	QUAN Bo, HOU Dongmei, MOU Songru, et al. Study on configuration unit spatial distribution of braided river reservoirs based on horizontal well information[J]. China Offshore Oil and Gas, 2020, 32(4): 96-103.

参数	90°相移	RMS 振幅	甜点属性	振幅包络	瞬时频率	衰减	纵波阻抗
90°相移	1.00	0.87	0.88	0.75	0.24	0.08	0.77
RMS振幅	0.87	1.00	0.97	0.96	0.13	0.08	0.70
甜点属性	0.88	0.97	1.00	0.92	0.35	0.10	0.68
振幅包络	0.75	0.96	0.92	1.00	0.12	0.08	0.58
瞬时频率	0.24	0.13	0.35	0.12	1.00	0.10	0.12
衰减	0.08	0.08	0.1	0.08	0.1	1.00	0.07
纵波阻抗	0.77	0.70	0.68	0.58	0.12	0.07	1.00

模型层次	模型类别	模型名称	预测均方根误差
基础模型	神经网络	多层感知机	5.91
	神经网络	FastAI网络	9.29
	梯度提升决策树	LightGBM	3.16
		XGBoost	3.76
		CatBoost	7.52
	随机决策树	随机森林	3.35
	随机决策树	极度随机树	3.13
	支持向量机	多项式核函数	8.77
	支持向量机	径向基函数	8.65
混合模型	神经网络	多层感知机	2.31
	随机决策树	随机森林	2.17
	多项式拟合	线性回归	2.89