浅薄层油藏储层预测技术研究与应用

发布时间：2018-05-19 17:09

本文选题：薄层 + 灰质砂岩　；参考：《中国石油大学(华东)》2015年硕士论文

【摘要】：本文依托于胜利油田项目《春风油田浅薄层稠油油藏储层预测技术研究》,结合春风油田工区实际资料,对浅薄层油藏储层进行预测技术研究与应用。本文首先在大量查阅该工区文献的基础上,针对工区的地质情况、测井数据以及地震相特征进行分析,对全工区资料具有全面的认识,这是我们进行构造解释和岩性解释的关键,也是储层预测和油气储层综合评价的基础。目标处理是地震勘探中针对不同的目的进行的特殊处理过程,本工区由于灰(砾)质砂岩的影响,使得储层的反射受到了一定的干扰。本文利用基于地质统计学反演的方法对该工区进行灰质进行去除,取得了一定的处理效果。储层的定性描述是储层预测的基础,也是后续进行定量预测的关键因素,在地震勘探岩性解释中占据着不可替代的位置。其中属性的提取、波阻抗反演正日趋成熟,成为地震勘探岩性解释的不可缺少的元素。本文利用多种属性提取,结合调谐体分频技术、波形聚类等对工区的储层进行定性描述,并取得较好的应用效果,为后续的储层定量预测奠定了良好的基础。在完成了储层的定性描述以后,开发人员更加关心的则是储层的厚度和砂体的边界分布。本文首先利用了多种统计理论的回归分析方法,结合录井资料和前面定性预测的结果,进行储层的定量预测。其中BP神经网络和支持向量机的回归方法在砂体厚度预测中取得了较好的应用效果。最后结合储层定性描述和定量预测的结果,利用层次分析法对储层分布进行综合评价,得到有利砂体的空间展布,为开发人员钻井提供很好的依据。
[Abstract]:Based on Shengli oilfield project "study on reservoir prediction technology of shallow and thin layer heavy oil reservoir in Chunfeng Oilfield", combined with the actual data of Chunfeng Oilfield, this paper studies and applies the prediction technology of shallow and thin layer reservoir. On the basis of consulting a large number of documents in this work area, this paper analyzes the geological conditions, logging data and seismic facies characteristics of the working area, and has a comprehensive understanding of the data of the whole working area. This is the key to structural interpretation and lithologic interpretation, and is also the basis of reservoir prediction and comprehensive evaluation of oil and gas reservoirs. Target processing is a special processing process for different purposes in seismic exploration. Due to the influence of grey (gravel) sandstone in this area the reflection of reservoir is disturbed to a certain extent. In this paper, the grey matter is removed from the work area based on geostatistical inversion, and some results are obtained. The qualitative description of reservoir is the basis of reservoir prediction and the key factor of quantitative prediction. It occupies an irreplaceable position in the interpretation of seismic exploration lithology. Among them, the extraction of attributes and wave impedance inversion are becoming more and more mature, and become an indispensable element in lithologic interpretation of seismic exploration. In this paper, we use a variety of attributes extraction, combined with tuned volume frequency division technology, waveform clustering and other qualitative description of the reservoir in the working area, and achieved good application results, which laid a good foundation for the subsequent quantitative reservoir prediction. After completing the qualitative description of the reservoir, the developer is more concerned with the reservoir thickness and the boundary distribution of the sand body. In this paper, the quantitative reservoir prediction is carried out by using multiple regression analysis methods of statistical theory, combined with logging data and the results of qualitative prediction. The BP neural network and the regression method of support vector machine have been applied to sand body thickness prediction. Finally, combined with the results of qualitative description and quantitative prediction of reservoir, the comprehensive evaluation of reservoir distribution is carried out by using analytic hierarchy process (AHP), and the spatial distribution of favorable sand bodies is obtained, which provides a good basis for drilling by developers.
【学位授予单位】：中国石油大学(华东)
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P618.13

【参考文献】