论文部分内容阅读
砂砾岩储层孔隙结构复杂、非均质性强,在渗透率计算方面传统的测井解释方法误差较大,目前还没有经典的计算砂砾岩渗透率的测井解释模型。以克拉玛依油田某区八道湾组砂砾岩稠油油藏为例,首先在微观层面上分析了渗透率的主控因素。其次根据本地区的实际情况建立了3套渗透率测井解释方法:一是在前人研究基础上改进了多元回归模型;二是在岩性识别的基础上分不同岩性建立了渗透率模型;三是利用BP神经网络进行了渗透率的预测。最后对传统的经验公式与文中的3种方法进行检验。结果表明,比起传统的经验公式和多元回归模型,基于不同岩性的渗透率模型与BP神经网络在实际应用中效果更好,较大幅度地提高了测井解释精度,在非均质性强的砂砾岩油藏中具有更好的应用前景。
The conglomerate reservoir has complicated pore structure and strong heterogeneity. The conventional well logging interpretation method has a large error in permeability calculation. At present, there is no typical logging interpretation model for calculating the permeability of glutenite. Taking the sand and gravel heavy oil reservoir of Badaowan Formation in a certain area of Karamay Oilfield as an example, the main controlling factors of permeability are analyzed at the micro level. Secondly, based on the actual situation in our region, three sets of permeability log interpretation methods are established: one is to improve the multiple regression model based on previous studies; the other is to establish the permeability model based on lithology identification with different lithologies Thirdly, BP neural network was used to predict the permeability. Finally, the traditional empirical formula and the text of the three methods to test. The results show that compared with the traditional empirical formula and multiple regression model, the permeability model based on different lithologies and the BP neural network are better in practical application, which greatly improve the logging interpretation accuracy. In the heterogeneity Strong conglomerate reservoir has a better application prospects.