大民屯凹陷变质岩潜山裂缝储层预测研究

发布时间：2018-06-28 14:20

本文选题：裂缝预测 + 储层预测　；参考：《长江大学》2015年硕士论文

【摘要】：裂缝型油气藏的油气产量占全世界油气总产量的一半以上,国内近些年勘探发现的裂缝油气藏也越来越多。随着勘探程度的不断提高,裂缝型等非构造油气藏的勘探已逐渐成为我国各大油田的主攻方向。大民屯凹陷地区位于辽河盆地的北部,三维资料基本覆盖了整个大民屯凹陷,从钻探情况来看,研究区具有较高勘探和研究价值。本论文主要是从地震属性基础理论知识出发,将裂缝预测的技术应用到大民屯凹陷潜山区块中,并以测井和地质两方面的资料为基础进行岩石物性分析,了解其储层物性参数的关系,为地震正演模型的建立提供有力的理论基础,通过地震模型正演和地球物理响应特征分析,研究地震方法对裂缝发育区的可识别性；并对地震属性进行分析,并筛选出敏感属性,将分析结果与地质、钻井资料进行对比分析,最后应用叠后相干体裂缝检测技术、蚂蚁追踪技术、多尺度边缘检测技术以及多属性神经网络等综合分析技术,对裂缝的有利发育区进行预测。在利用地震属性进行裂缝预测前,开展了裂缝段成像测井特征分析,利用成像测井刻度常规测井,提高常规测井识别裂缝的精度。同时,通过特征曲线统计结合测井原理分析,总结不同角度裂缝测井(速度、密度、电阻率、伽马、中子)响应特征,筛选对于裂缝敏感的测井资料,建立裂缝与测井响应之间关系。同时针对不同密度、不同发育级别的裂缝理想模型以及研究区实际地质模型进行正演模拟,以此为基础分析不同岩性、不同速度、不同裂缝级别等情况下叠后地震响应特征。论文在对地震属性进行简要分析后,有针对性的对目前较为常用的地震属性方法做了一些研究分析,进一步利用叠后相干体、蚂蚁体追踪技术、边缘检测技术对裂缝进行了预测研究。(1)相干体裂缝预测是对一定时窗内地震波形纵向和横向相似性的判别,揭示相对宏观的断裂信息；(2)用蚂蚁追踪进行了裂缝预测,在蚂蚁体属性上可见明显的裂缝分布,且裂缝倾角显示清晰,表现为高角度、近平行分布的特征；(3)利用小波多尺度边缘检测技术预测了裂缝发育区,针对实际工区的资料分析得出其小波边缘检测阈值分别为F1=0.012376及F2=0.137816,大于F2的为裂缝发育区；大于F1小于F2的为裂缝次发育区,小于F1为裂缝不发育区。(4)利用相关性和属性特征强度综合分析方法,优选出7种属性进行分析,并利用神经网络技术预测了裂缝的主要发育部位。最后论文通过一系列的研究在本区变质岩裂缝预测方法研究方面主要获得了以下几点认识：(1)总结出了该区对裂缝有敏感响应的测井曲线,在裂缝发育区,双侧向电阻率曲线存在大的正差异,声波时差变化大；而在裂缝不发育处,双侧向曲线差异小,声波时差变化小。(2)对裂缝段进行正演模拟,建立了5类23个不同裂缝参数的裂缝储层地震地质模型,并对这些模型进行了波动方程正演模拟研究和地震属性分析。结果表明：裂缝发育密度和裂缝等效速度是影响地震波场特征的首要因素,裂缝发育区的地震反射杂乱。裂缝密度越大,速度与围岩的差别越大,裂缝响应越明显。裂缝厚度和宽度也对其地震响应产生一定的影响,厚度和宽度越大则反射越强。(3)叠后属性分析中,沿潜山顶界面对地震数据体提取了振幅类、复地震道统计类、层序统计类、频谱类等39种属性,利用相关性和属性特征强度综合分析方法,优选出7种属性进行分析,并利用神经网络技术预测了裂缝的主要发育部位,并根据裂缝密度分区平面图,将工区划分为三类区块。作为验证井的哈20、哈30-28-28、沈303、沈269及哈31-16-18,其相对误差在2%～19%之间,平均误差8%,预测结果较好。(4)总体来看,属性预测方法较多,预测结果差异不大,裂缝主要发育在平安堡潜山、胜西低潜山、东胜堡潜山构造带。对于叠后多属性预测,预测误差相对较小,精度相对较高,是相对较好的裂缝预测方法。
[Abstract]:The oil and gas production in the fractured reservoir accounts for more than half of the total oil and gas production in the world. There are more and more fractured oil and gas reservoirs found in China in recent years. With the continuous improvement of the exploration degree, the exploration of fracture type and other non tectonic oil and gas reservoirs has gradually become the main direction of the major oil fields in our country. The damun depression area is located in the Liaohe basin. In the north, the three dimensional data basically covered the whole damtun sag. From the point of view of drilling, the research area has high exploration and research value. This paper, based on the basic theoretical knowledge of seismic attributes, applies the technology of fracture prediction to the Qianshan block in damun depression, and is based on the data of two aspects of logging and geology. The relationship between physical properties of rock and physical properties of the reservoir is analyzed, which provides a strong theoretical basis for the establishment of the seismic forward modeling. Through the analysis of seismic model forward and geophysical response characteristics, the identification of the seismic method to the fracture development zone is studied, and the seismic attributes are analyzed and the sensitive attributes are screened out and the analysis knot will be analyzed. The results are compared with the geological and drilling data. Finally, the fracture detection technique, the ant tracking technique, the multi-scale edge detection technology and the multi attribute neural network are used to predict the favorable development area of the fracture. The fractured section imaging logging is carried out before the fracture prediction is carried out by the use of the seismological property. Feature analysis, using conventional logging logging to improve the accuracy of identifying cracks in conventional logging. At the same time, through the analysis of characteristics curve statistics and logging principle analysis, the response characteristics of different angle fractures (velocity, density, resistivity, gamma, neutron) are summarized, and the sensitive logging data are screened for cracks and the response of fractures and logs is established. At the same time, in the light of different density, different development level of crack ideal model and the actual geological model in the study area, a forward simulation is carried out to analyze the characteristics of post stack seismic response under different lithology, different velocity and different crack levels. The more commonly used seismic attribute methods have done some research and analysis, further using post stack coherent body, ant body tracking technique and edge detection technology to predict the crack. (1) the prediction of the coherent body crack is the discrimination of the longitudinal and transverse similarity of the seismic waves in a certain time window, and reveals the relative macroscopic fracture information; (2) the ants are used to use ants. The fracture prediction is carried out, and the obvious crack distribution is visible on the ant body properties, and the fracture angle is clearly displayed, which is characterized by high angle and near parallel distribution. (3) the fracture development zone is predicted by the wavelet multiscale edge detection technology, and the threshold of the wavelet edge detection is F1=0.0 for the data analysis of the actual work area. 12376 and F2=0.137816, greater than F2 for fracture development zone; more than F1 less than F2 for the secondary zone of fracture development, less than F1 as the fracture undeveloped zone. (4) the use of correlation and attribute characteristics strength comprehensive analysis method, the optimization of the 7 properties of the analysis, and the use of neural network technology to pre test the main parts of the fracture. A series of studies have mainly acquired the following points in the study of fracture prediction methods of metamorphic rocks in this area: (1) the well logging curves that have sensitive response to cracks are summed up in this area. In the zone of fracture development, there is a large positive difference between the two lateral resistivity curves, and the variation of sonic time difference is large, while in the undeveloped cracks, the difference of the bilateral curve is small. The acoustic wave time difference is small. (2) the fracture section is forward simulated, and 5 types of fractured reservoir seismic geological models with 23 different fracture parameters are established, and the wave equation forward modeling and seismic attribute analysis are carried out. The results show that the density of fracture and the equivalent velocity of cracks are the primary factors affecting the characteristics of seismic wave fields. The greater the density of the cracks, the greater the density of the cracks, the greater the difference between the velocity and the surrounding rock, the more obvious the response of the cracks. The thickness and width of the cracks also have a certain effect on the seismic response, and the greater the thickness and width, the stronger the reflection. (3) in the post stack attribute analysis, the amplitude class is extracted along the interface of the submersible peak to the seismic data body. 39 attributes, such as seismic trace statistics, sequence statistics and spectrum, are used to analyze the 7 attributes by using the comprehensive analysis method of correlation and attribute characteristics strength. The main parts of the fracture are predicted by the neural network technology, and the area is divided into three types of blocks according to the partition plane map of the density of the fracture density. 20, ha 30-28-28, Shen 303, Shen 269 and ha 31-16-18, their relative errors are between 2% and 19%, the average error is 8%, and the prediction results are good. (4) in general, the property prediction methods are more, the prediction results are different, the cracks are mainly developed in Ping'an Qianshan, Shenxi low Qianshan, Dongsheng Fort Qianshan structural belt. Relatively small, relatively high accuracy, is a relatively good method of crack prediction.
【学位授予单位】：长江大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P618.13

【参考文献】