Abstract: The tight sandstone reservoirs in the Member 9 of Triassic Yanchang Formation in Ordos Basin are characterized by low porosity, low permeability, low oil saturation, and strong heterogeneity, making reservoir classification and productivity prediction particularly challenging. In this study, oil test results from 31 reservoir intervals of the Member 9 of Yanchang Formation in the Jiyuan area, combined with corresponding logging curve responses and interpretation results, are utilized to establish the correlations between multiple logging parameters and productivity. The study results indicate that reservoirs characterized by smooth box-shaped or bell-shaped sandbody structures generally exhibit high productivity. Additionally, hydrocarbon productivity shows an overall upward trend with increases in the effective thickness, porosity and oil saturation of reservoirs. The correlation between productivity and reservoir physical properties is stronger than that with electrical properties, and the correlation between productivity and directly measured original logging parameters is higher than that with indirectly calculated parameters based on models. Based on these findings, a productivity prediction method for tight sandstone reservoirs is proposed, integrating multiple logging parameters. Firstly, reservoirs with smooth box-shaped or bell-shaped sandbody structures are selected using attributes such as the relative center of gravity of the natural gamma curve and homogeneity. Moreover, by incorporating key influencing factors, including effective thickness, deep resistivity, and density logging data, both comprehensive evaluation index and productivity prediction model for tight sandstone reservoirs has been developed. Finally, reservoir classification criteria based on productivity capacity indicators are established. The proposed method was applied to classify and predict the productivity of 10 tight sandstone reservoir intervals in the Member 9 of Yanchang Formation of Jiyuan area. The results show that the accuracy of reservoir classification reaches 90 %, while the oil production prediction accuracy rate is 80 %, thus validating the effectiveness and practicality of the model. This method highlights the comprehensive influence of factors such as sandbody structure, physical properties, and effective thickness on the productivity of tight sandstone reservoirs. It effectively enhances the productivity prediction capability of logging curves, providing valuable insights for the classification and productivity research of tight sandstone reservoirs.

Key words: tight sandstone, conventional logging, attribution fusion, reservoir classification, productivity prediction, Triassic, Jiyuan area, Ordos Basin

摘要： 鄂尔多斯盆地三叠系延长组9段(长9段)致密砂岩储层具有低孔、低渗、低含油饱和度与强非均质性等特点,其储层分类及产能预测难度较大。利用姬塬地区长9段31个储层段的试油结果以及对应的测井数据响应特征与测井解释结果,分别建立了多个测井属性与产能之间的配置关系。研究结果表明:具有平滑箱型/钟型砂体结构的储层的产能整体较高,且产能随着储层有效厚度、孔隙度以及含油饱和度的增大呈整体上升的趋势;其中,产能与储层物性的相关性高于储层电性,产能与测井数据原始属性的相关性高于间接计算属性。基于此,提出了多种测井属性相融合的致密砂岩储层产能预测方法:首先,利用自然伽马曲线的相对重心和均质度属性,筛选出具有平滑箱型/钟型砂体结构的储层;然后结合储层有效厚度、深侧向电阻率测井和密度测井数据等关键影响因素,构建致密砂岩储层的产能潜力综合评价指标和产能预测模型,并据此建立储层分类标准。利用该方法对姬塬地区长9段10个致密砂岩层段的产能进行了分类和预测检验,结果显示对储层开展分类识别的正确率达90 %,产油量预测的符合率为80 %,由此验证了该方法模型的有效性与实用性。该方法突出了致密砂岩储层的砂体结构、物性和有效厚度等因素对产能的综合影响,有效提高了测井曲线的产能预测能力,对开展致密砂岩储层的分类及产能研究具有较好的参考价值。

关键词: 致密砂岩, 常规测井, 属性融合, 储层分类, 产能预测, 三叠系, 姬塬地区, 鄂尔多斯盆地