| 摘要: |
| 以大芦湖油田樊29块沙三中亚段储层为研究对象,运用孔喉半径法并结合沉积特征、流动特征、物性特征及生产动态特征开展取芯井储层流动单元研究。在此基础上,选取与浊积扇低渗储层渗流特征相关的13个储层特征参数,采用支持向量机(SVM)算法开展未取芯井流动单元定量评价。基于52组测试样本对SVM预测模型性能进行检验,并利用生产动态资料对流动单元定量评价结果进行合理性验证。结果表明:研究区取芯井储层可划分为Ⅰ、Ⅱ、Ⅲ、Ⅳ 4类流动单元,Ⅰ类、Ⅱ类流动单元储集物性和渗流能力最好,Ⅲ类流动单元次之,Ⅳ类流动单元最差;基于SVM的流动单元预测结果与岩心分析结果吻合,模型的正判率达90.38％;流动单元预测结果与油井初期产能、吸水特征及水淹特征存在较强的相关性,SVM预测模型为储层精细解释提供有效途径;不同类型流动单元内剩余油可动用储量及采出程度不同,当前剩余油主要集中于Ⅱ类、Ⅲ类流动单元中,是下一步调整挖潜的重点。 |
| 关键词: 浊积扇 低渗透 流动单元 支持向量机 剩余油 |
| DOI：10.3969/j.issn.1673-5005.2019.06.001 |
| 分类号::TE 122.2 |
| 文献标识码:A |
| 基金项目:国家科技重大专项(2017ZX05009001)；国家自然科学基金面上项目(41772138) |
|
| Quantitative characterization of reservoir flow units of basin floor-fan turbidite |
|
LU Yan1,2, XU Shouyu1,2, LIU Keyu1,2, WANG Ya1,2
|
|
(1.School of Geosciences in China University of Petroleum(East China), Qingdao 266580, China;2.Key Laboratory of Deep Oil and Gas in China University of Petroleum(East China),Qingdao 266580, China)
|
| Abstract: |
| This paper selected the middle Es3 reservoir in Block Fan29, Daluhu Oilfield as the research area. Flow units were divided by using the pore throat method, also considering sedimentary characteristics, flow characteristics, petrophysical properties characteristics, and production dynamic characteristics. Based on these studies, 13 parameters related to permeability characteristics of low-permeability reservoir were selected, and a quantitative identification of flow units in uncored wells was carried out using the support vector machine(SVM) algorithm. The established model was tested by 52 groups of predicting samples, and verified using the production dynamic data. The results show that the reservoir can be divided into four types of flow units, including Type Ⅰ, Type Ⅱ, Type Ⅲ and Type Ⅳ, respectively Type Ⅰ and Type Ⅱ flow units have the best reservoir petrophysical properties and permeability capacity, while the permeability capacity of Type Ⅳ flow unit is the lowest. The prediction of flow unit based on SVM is consistent with the core analysis data, with an accuracy rate of 90.38%. The prediction also has a high correspondence with the initial oil production, characteristics of water absorption and initial water cut. The SVM prediction model provides an effective way for fine reservoir interpretation. In addition, the reserves and recovery factors of remaining oil in different types of flow units are different. The remaining oil is mainly concentrated in Type Ⅱ and Type Ⅲ units, which are the target area for remaining oil development of the next step. |
| Key words: basin floor-fan turbidite low-permeability flow units support vector machine remaining oil |
