油田压裂微地震地面监测速度模型校正及定位研究

发布时间：2018-01-11 08:11

本文关键词：油田压裂微地震地面监测速度模型校正及定位研究　出处：《吉林大学》2016年博士论文　论文类型：学位论文

【摘要】：低渗透油气藏微地震监测研究就是利用地下岩层错动或断裂而产生的声发射现象对微地震震源位置进行定位的一种技术方法。近年来,随着微地震监测理论不断发展成熟,基于微震监测的裂缝评价技术已成为低渗透油气藏开采过程中最直观而又可靠的技术。微地震监测方式主要有两种:井中监测与地面监测。与地面监测相比,井中监测具有微震事件能量比较强,数据信噪比较高,并且能够避免地面噪声干扰等优势,在国内外获得了广泛的认可。但井中监测对压裂现场条件要求较高,并且施工操作复杂,在没有监测井的情况下无法开展工作。而地面监测不受压裂现场井网条件及施工难度制约,将是今后主要发展方向之一,本课题针对地面监测,围绕微地震监测方法、理论、数值模拟以及野外实验等方面,对微震监测中的正演、速度模型构建、微地震定位以及三维微震监测可视化软件等关键技术进行了研究。以地震射线追踪为主的正演环节直接影响震源定位质量以及计算效率。因此,需要寻找一种计算灵活,高效而又准确的射线追踪算法。而地震勘探领域中广泛应用的最短路径法在地面检波器数量较少,地下震源较多的情况下,则存在处理射线不够灵活,提高精度的同时严重占用计算机内存且耗费大量计算时间等问题。针对三维层状结构,提出采用二分法对传统逐步迭代法进行改进,提高了射线追踪正演计算效率。针对三维低密度网格结构,将逐段迭代法与中垂线点搜索方式相结合,提出了一种新的两点间三维射线追踪方法,与最短路径法相比,该方法在保证精度的同时,能够大幅提高计算效率,为微地震监测引入较复杂的速度模型提供了可能。震源定位是微地震监测中的核心问题,传统定位算法是提取大量检波器所获微震信号的初至时间,并优化求解超定方程组以确定震源位置。但其主要缺陷在于该方法仅能采用均匀速度模型进行定位。本文采用基于网格搜索类的微地震定位技术对地下微地震事件进行定位,可引入更为精细的速度模型,以提高微地震定位可信度。但该方法存在计算效率与定位精度无法平衡的不足。因此,文中采用网格逐次剖分思想对常规网格搜索类微地震定位方法进行改进。该方法对目标区域进行首次剖分后,在最大能量聚焦的网格中心点附近进行进一步的细分搜索,寻找全局最大能量聚焦值,避免了为了提高定位精度而进行的全局化细分,以较小的计算代价获得了更高的定位精度。文中还对速度模型对该算法的影响进行了分析与讨论,最后,通过模型试算及野外实际数据处理验证了改进后的微地震定位算法在计算效率以及计算精度上的优势。速度模型是影响微地震定位精度主要因素。本文首先对基于初至时差地面微地震速度模型构建进行了研究,压裂射孔信号是速度模型校正的重要先验信息,而实际监测过程中射孔起震时刻无法准确获取。为解决此问题,提出基于初至时差的双差分均方根误差来描述理论速度模型与实际速度模型之间的差异,采用极快速模拟退火法尽可能的降低双差分均方根误差值,在双差分均方根误差最小值一定阈值范围内选取对应的速度模型,并依次对射孔点进行重定位,进而筛选出速度模型最优解。当储层较深时,射孔地面记录通常具有低信噪比的特征,很难准确拾取纵横波初至信息。本文基于偏移振幅叠加法,并结合极快速模拟退火方案,提出一种新的速度模型构建方案,该方案不需要拾取震相初至信息,通过监测射孔点处能量聚焦情况,并以射孔点重定位精度,判定速度模型是否可用于后续微地震定位。由于地震数据反演非唯一性,在利用该方法时将速度模型校正作为一种误差补偿方案以提高射孔重定位精度。合成数据试算结果表明:采用振幅叠加方法校正后的速度模型,对两种主要误差都有较好的补偿效果。在射孔记录信噪比S/N=0.1时,100次定位误差均在2m以内,而基于初至时差法则完全失效。鉴于目前国内适用于微地震监测三维可视化软件的匮乏,本课题采用Open GL绘图技术,将Qt Creator植入Visual Studio 2010作为Open GL开发环境,利用C++语言开发了微地震监测三维可视化模块,利用坐标变换将处理好的微地震数据导入到计算模块,通过分析处理,在输出设备上直观的显示微地震事件,为微地震监测工作提供了较为便利的人机交互工具。本文将震源识别,射线追踪,微地震定位,速度模型构建,及三维等模块嵌入,设计开发了三维微地震地面监测软件系统,并将该系统应用于山西省娄烦县宁武盆地野外压裂实验,实际应用表明该系统在地质条件简单的地区用于水力压裂裂缝监测是可行的,未来通过对速度模型校正方法的改进,可推广至地质条件较复杂的地区。创新之处在于文中在目前较为常用的射线追踪算法的基础上进行改进,在不影响计算精度的前提下,计算效率更高,更适用于微地震监测领域;在微地震地面监测速度模型构建方面,针对射孔记录信噪比较高的情况,提出了基于初至时差的微地震速度模型校正方法,避免了射孔信号起震时刻未知,测井曲线初值不准等问题,解决了速度模型反演非唯一性的问题。针对射孔信噪比较低的情况,本文提出采用振幅偏移叠加微地震速度模型校正方法,避免了在低信噪比情况下,地震波初至拾取不准确的问题;在微地震定位方面,提出了网格逐次剖分法对常规的基于网格搜索类微地震定位算法进行改进,平衡与兼顾了计算效率与计算精度,为微地震实时监测提供了可能;最后设计研制了微地震三维可视化模块,为微地震监测工作提供了便利的可视化工具。
[Abstract]:Low permeability oil and gas reservoir study on microseismic monitoring is the use of underground rock dislocation or fracture caused by acoustic emission phenomena of a technique of micro earthquake locations are located. In recent years, with the continuous development of the theory of micro seismic monitoring, crack evaluation of microseismic monitoring technology has become a process of low permeability oil and gas reservoirs in the most intuitive and reliable based on technology. There are two main ways: micro seismic monitoring wells monitoring and ground monitoring. Compared with the ground monitoring, monitoring wells with microseismic event energy is relatively strong, data with high SNR, and can avoid the ground noise and other advantages, has been widely recognized at home and abroad. Monitoring wells fracturing requires high field conditions, and the construction of complex operation, unable to work in the absence of monitoring wells. The ground monitoring is not affected by the field well network condition and construction difficult The degree of control, will be the main development direction of the topic for the ground monitoring, around the micro seismic monitoring method, theory, numerical simulation and field experiment, the forward speed in microseismic monitoring, model construction, micro seismic location and three-dimensional microseismic monitoring of key technologies of visualization software is studied. The play links affecting the quality and efficiency of source location calculation directly to seismic ray tracing. Therefore, there is a need for calculation of flexible, efficient and accurate ray tracing algorithm. They are widely used in the field of seismic exploration of the shortest path method in ground detector number of underground source under many circumstances, there is not enough X-ray treatment flexible, improve the accuracy and take up serious problems of computer memory and consume a large amount of computation time. The three-dimensional layered structure, based on the traditional dichotomy gradually The iterative method is improved and the improved ray tracing forward calculation efficiency. According to the grid structure of 3D low density, the iterative method and the combination of perpendicular search, proposed a new 3D ray tracing method between two points, compared with the shortest path method, the method of ensuring the accuracy and greatly to improve the computational efficiency, provides the possibility for micro seismic monitoring into the complicated velocity model. The source location is the key problem of micro seismic monitoring in the traditional location algorithm is extracted from a large number of detector received the first arrival time of microseismic signals, and the optimization solution of overdetermined equations to determine the source location. But its main drawback is that the method only by using the uniform velocity model for positioning. This paper adopts micro seismic location technology of grid search class to locate underground microseismic events based on the introduction of more precise velocity model, to Improve the micro seismic location reliability. But this method has less calculation to balance efficiency and positioning accuracy. Therefore, the ideas of conventional grid search micro seismic location method is improved by using the grid in this paper. The method of successive section for the first time to partition the target area after further subdivision in the vicinity of the grid center's search high energy focusing, finding the global maximum energy focus value, avoid the Global Subdivision and in order to improve the positioning accuracy, with low computational cost to obtain a higher positioning accuracy. The velocity model are analyzed and discussed, the effect of the algorithm. Finally, through the model test and field data processing results of micro seismic location algorithm improved the calculation efficiency and calculation accuracy. The advantage of speed model is the influence of micro seismic location. Firstly, the main factors Based on the first arrival time difference of ground micro seismic velocity model is studied, fracturing perforation signal speed is important prior information of model calibration, and the actual monitoring process of perforating shock moment cannot be obtained accurately. To solve this problem, put forward to the early time of double differential RMS error to describe the differences between the theoretical model and speed the actual speed based on the model, using very fast simulated annealing method as much as possible to reduce the double differential RMS error value, the double differential root mean square error of the minimum velocity model corresponding to the selected threshold range, and turn on the perforation point positioning, and then screened the optimal velocity model solution. When the reservoir is deep, perforating records often has the characteristics of low SNR, it is difficult to pick up the wave first break information. In this paper, offset amplitude superposition method based on, combined with the very fast simulated annealing scheme is proposed. A new velocity model is proposed, the scheme does not require picking phase initial information, by monitoring the perforating point energy focusing, and point to the perforation position precision and speed to determine whether the model can be used for subsequent micro seismic location. Because seismic data inversion of non uniqueness, in the use of this method will speed model as a kind of correction error compensation scheme to improve perforation accurate positioning. The test results show that: using synthetic data amplitude superposition method for velocity model after correction, has good compensation effect of the two main error. In perforating record SNR S/N=0.1, 100 times the positioning error is within 2m, and based on the first arrival time law of failure. In view of the lack of suitable for the micro seismic monitoring and 3D visualization software, this paper adopts the Open GL graphics, Qt Creator Visual Studio Open GL 2010 as the implantation of open Development environment, C++ language is used to develop the micro seismic monitoring and 3D visualization module, micro seismic data into the calculation module by using coordinate transformation will be dealt with, by analyzing and processing, visual display of micro seismic events on output devices, micro earthquake monitoring provides more convenient tools for human-computer interaction. The source identification, ray tracing, micro earthquake location, velocity model, and 3D module design and development of embedded 3D micro seismic ground monitoring software system, and this system is applied to the Shanxi Ningwu basin in Loufan County, field fracturing experiment, the practical application shows that the system used in the simple geological condition area monitoring hydraulic fractures is feasible the future, through the improvement of the speed model correction method, can be extended to more complicated geological conditions in the region. Innovations in this paper are often used after Ray Based on the improved tracking algorithm, the computational accuracy is not under the influence of more efficient, more suitable for the field of micro seismic monitoring; micro seismic ground monitoring speed model, aiming at the perforating record high SNR, the first arrival time difference of micro seismic velocity correction method based on the model. To avoid the vibration signal of perforation is unknown, the initial value of logging curves and other issues are not allowed, solves the non uniqueness of the model inversion problem. For perforating the SNR is low, the amplitude of migration stack micro seismic velocity model correction method is avoided in case of low SNR, the problem is not to pick up accurate seismic wave; in the micro earthquake location, this paper proposes a grid partition method of successive improvement of conventional grid search based on micro earthquake location algorithm, and taking into account the balance calculation efficiency and precision. It provides the possibility for real-time monitoring of microseismic events. Finally, a 3D visualization module for microseismic design is developed, which provides a convenient visualization tool for microseismic monitoring.

【学位授予单位】：吉林大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：P631.4;TE357

【相似文献】