History matching time-lapse seismic data using the ensemble Kalman filter with multiple data assimilations

The ensemble Kalman filter (EnKF) has become a popular method for history matching production and seismic data in petroleum reservoir models. However, it is known that EnKF may fail to give acceptable data matches especially for highly nonlinear problems. In this paper, we introduce a procedure to improve EnKF data matches based on assimilating the same data multiple times with the covariance matrix of the measurement errors multiplied by the number of data assimilations. We prove the equivalence between single and multiple data assimilations for the linear-Gaussian case and present computational evidence that multiple data assimilations can improve EnKF estimates for the nonlinear case. The proposed procedure was tested by assimilating time-lapse seismic data in two synthetic reservoir problems, and the results show significant improvements compared to the standard EnKF. In addition, we review the inversion schemes used in the EnKF analysis and present a rescaling procedure to avoid loss of information during the truncation of small singular values.     

CO2地质封存时移垂直地震监测技术

神华CO2捕获和封存（carbon capture and storage,CCS）示范项目是我国首个全流程煤基CO2捕获和在低孔低渗深部咸水层进行多层注入的CO2封存示范项目,意在通过时移垂直地震（VSP）监测CO2地质封存状况。介绍了CO2地质封存时移VSP技术和监测效果。通过注气前后3次VSP观测,得到了高品质时移VSP资料,研究了一致性处理技术、矢量波场分离和多波成像技术,获得了高分辨率的多波时移VSP剖面,通过深度域标定和多种属性解释对比,分析了注气前后VSP地震响应特征,预测了CO2地下运移范围,达到CCS项目预期监测目标。结果证明,时移VSP技术是监测CO2地质封存与运移的有效技术手段。     

时移电阻率法归一化数据反演分辨电阻率结构微小变化

时移电阻率法能够应用于监测地下水污染物运移、衡量坡面稳定性等工程和环境问题.本文利用归一化时移电阻率法数据反演结果识别地下微小电阻率结构变化.首先,将初次采集数据作为背景数据,利用其归一化其他时刻时移电阻率法数据;再次,实现背景数据、时移数据以及归一化后时移数据的非线性共轭梯度反演.在合成数据算例中,保持观测系统不变,用相同反演参数和均匀半空间参考模型参与反演.反演结果表明,相对背景电阻率而言,归一化数据反演结果能够直接地、有效地分辨较为微小的电阻率结构动态变化,而传统电阻率反演不能分辨.     

Use of the geometric average for effective permeability estimation

The geometric average is often used to estimate the effective (large-scale) permeability from smaller-scale samples. In doing so, one assumes that the geometric average is a good estimator of the geometric mean. Problems with this estimator arise, however, when one or more of the samples has a very low value. The estimate obtained becomes very sensitive to the small values in the sample set, while the true effective permeability may be only weakly dependent on these small values. Several alternative methods of estimating the geometric mean are suggested. In particular, a more robust estimator of the geometric mean, the j^th Winsorized mean, is proposed and several of its properties are compared with those of the geometric average.     

Using time-lapse 4D seismic to monitor saturation changes in carbonate reservoirs

In the past ten years, time-lapse (4D) seismic has evolved as a standard way of monitoring reservoir performance. The method is now being used as good reservoir management practice to provide evidence of saturation changes within the reservoir at field scale. 4D provides a new piece of data describing the dynamic behavior of the reservoir fluids between the wells, often limited to small scale monitoring at the borehole scale. Thus, it provides sophisticated techniques for reservoir monitoring and management relying on the integration of geological models, static and dynamic properties of the reservoir rock, and detailed production and pressure field data.While 4D seismic data has been very successful in monitoring hydrocarbon production from clastic reservoirs, this work has focused on implementing 4D time lapse to monitor saturation changes in carbonate reservoirs and it’s capability to be used as enhanced oil recovery (EOR) tool that can help in enhancing the recovery factor for the filed and help to locate new drilling to sweep more oil out of the reservoir and locate the by-pass oil.The principal goal of this research was to detect the maximum change in seismic attributes (amplitude, acoustic impedance, travel time) that could occur as a result of oil production, water and gas injection in carbonate reservoirs by using time-lapse 4D seismic.     

Bayesian estimation of seismic hazard for two sites in Switzerland

Seismic hazard analysis is based on data and models, which both are imprecise and uncertain. Especially the interpretation of historical information into earthquake parameters, e.g. earthquake size and location, yields ambiguous and imprecise data. Models based on probability distributions have been developed in order to quantify and represent these uncertainties. Nevertheless, the majority of the procedures applied in seismic hazard assessment do not take into account these uncertainties, nor do they show the variance of the results. Therefore, a procedure based on Bayesian statistics was developed to estimate return periods for different ground motion intensities (MSK scale).Bayesian techniques provide a mathematical model to estimate the distribution of random variables in presence of uncertainties. The developed method estimates the probability distribution of the number of occurrences in a Poisson process described by the parameter . The input data are the historical occurrences of intensities for a particular site, represented by a discrete probability distribution for each earthquake. The calculation of these historical occurrences requires a careful preparation of all input parameters, i.e. a modelling of their uncertainties. The obtained results show that the variance of the recurrence rate is smaller in regions with higher seismic activity than in less active regions. It can also be demonstrated that long return periods cannot be estimated with confidence, because the time period of observation is too short. This indicates that the long return periods obtained by seismic source methods only reflects the delineated seismic sources and the chosen earthquake size distribution law.     

地震反射数据Q值估计及其在油气勘探中的应用

油气地震勘探中,地震波在实际介质中传播时会产生明显的吸收衰减现象,导致地震信号中的主频向低频移动,频带宽度变窄,相位发生畸变,制约了地震勘探识别薄层的分辨能力。为了获得更高分辨率的地震成像,文章介绍了一种在地震反射数据成像域进行Q值估计建模的方法与流程,并利用估计出的Q值通过偏移成像对数据进行衰减补偿,实现高分辨率成像。该方法在时间域引入等效Q值的概念,首先在初步黏弹性时间偏移成像域的时窗内,通过数据在时窗内的补偿效果来确定时间域的等效Q值参数,接着通过在深度域层速度上计算成像射线获得时深转换关系,进而对转换到时间域层Q值的等效参数进行时深转换,完成最终深度域Q值建模,最后将该关键参数作为黏弹性叠前深度偏移的输入,进行复杂构造的黏弹性补偿成像。同时使用中国东部某实际地震数据来验证方法的有效性,验证结果表明发展的流程和方法可较好实现深度域负责构造Q值建模和成像域补偿,实现复杂构造高分辨成像。     

A piecewise exponential model for seismic well-logging data

Conventional methods of analyzing sonic log data do not always yield accurate information on each velocity segment of a well. It is shown here that the velocity-depth parameters and the sections of approximately constant velocity may be more precisely defined by using an exponential spline to model the data.     

带断层地震数据的Delaunay三角剖分算法

数据网格化是三维地质建模的常用方法。断层是普遍的地质现象,带断层的海量数据三角剖分在建模中较为复杂。提出了一种简单的可以解决带断层地震数据Delaunay三角剖分的有效算法。该算法首先对数据边界进行初始Delaunay三角剖分,然后将其变尺度加密,生成较均匀三角网,最后将断层边界强行嵌入三角网,实现了三角网格化。     

地震数据重建方法原理及运用

在数值分析领域中,许多方法都是借助于插值公式导出的.几乎所有的经典数值微分、数值求积和常微分方程的数值积分公式,都可以从插值公式中推导出来.地震勘探领域常见的插值重建方法按照数学原理大致可以分为三类:适用于均匀采样的预测误差滤波插值方法,基于算子的插值方法(如DMO和偏移算子)和在最小二乘意义下估算特定数学变换系数进行插值的方法.这里从数学原理上系统地阐述了这几种插值方法的基本原理,并利用数值试验和实际数据进行了检验.     

Bayesian techniques for seismic hazard assessment using imprecise data

Some Bayesian methods of dealing with inaccurate or vague data are introduced in the framework of seismic hazard assessment. Inaccurate data affected by heterogeneous errors are modeled by a probability distribution instead of the usual value plus a random error representation; these data are generically called imprecise. The earthquake size and the number of events in a certain time are modeled as imprecise data. Imprecise data allow us to introduce into the estimation procedures the uncertainty inherent in the inaccuracy and heterogeneity of the measuring systems from which the data were obtained. The problem of estimating the parameter of a Poisson process is shown to be feasible by the use of Bayesian techniques and imprecise data. This background technique can be applied to a general problem of seismic hazard estimation. Initially, data in a regional earthquake catalog are assumed imprecise both in size and location (i.e errors in the epicenter or spreading over a given source). By means of scattered attenuation laws, the regional catalog can be translated into a so-called site catalog of imprecise events. The site catalog is then used to estimate return periods or occurrence probabilities, taking into account all sources of uncertainty. Special attention is paid to priors in the Bayesian estimation. They can be used to introduce additional information as well as scattered frequency-size laws for local events. A simple example is presented to illustrate the capabilities of this methodology.     

Wavefield analysis of crosswell seismic data

Despite the numerous advantages of crosswell seismic data over surface seismic data, crosswell seismic geophysics is still underutilized and underdeveloped. The factors limiting the full utilization of crosswell data include the lack of standardized methods for processing and imaging the data. This is because crosswell data is not completely understood. To improve the understanding of crosswell data, we performed acoustic and elastic modeling of a west Texas carbonate oilfield data using finite difference methods and crosswell geometry. To account for the different wave modes in the field data, we decomposed the full data into its constituent wave modes. Results of the forward modeling show that elastic synthetic data is a better representation of crosswell field data than the popular acoustic synthetic data. Wavefield decomposition gave insight into the time-space kinematics behavior of the different wave modes that constitute the full data. Overall, the study improved our understanding of crosswell field data. The learning from this study has been utilized to perform data-driven reflection enhancement processing where the discerned characteristic of different seismic arrival is utilize to suppress unwanted and enhanced the desired wave modes. The processing reduced the complex data to only up-going P-P reflections that can be imaged to reveal the subtle geological structures of the oilfield.     

适于大规模数据的逆时偏移实现方案

针对海量地震数据逆时偏移所面临的问题和挑战,解决逆时偏移海量计算和海量存储需求在CPU/GPU异构集群平台上实现时的技术瓶颈,重点介绍了基于GPU?RDMA直接通信技术和GPU?Stream流技术的并行算法,形成适合于GPU集群的大规模分布式并行计算,即高密度采集地震数据的三维叠前逆时偏移计算技术;应用高密度实际地震数据测试,计算效率可提升15~40倍。     

矿井槽波地震数据处理系统GeoCoal软件开发与应用

定位于矿井槽波地震勘探,开发一套针对槽波地震数据的处理解释平台,以期推动矿井槽波地震勘探技术更好地为煤矿安全生产服务。构建了软件设计架构,介绍了软件主要子系统和每个子系统包含的功能模块和软件运行环境,最后进行了部分软件界面和成果的展示。     

大炮检距优化6次NMO校正

常规2次NMO动校是假定炮检距与深度比值小，并且射线为直线。计算的旅行时精度随着炮检距与深度比值的增加而大大降低，NMO校正能通过高次(4次、6次)项逼近得到改善。因此，在大炮检距情况下，若直接忽略高次项，则计算旅行时会产生很大误差。所以在大炮检距条件下进行2次NMO近似不利于速度分析、AVO研究和CMP道集动校叠加。笔者引入了一个更精确的旅行时近似计算，称为优化6次NMO动校方程式，因为截断误差较小，故其精确度高于泰勒6次展开式。通过模型和实际数据的试验，证明大炮检距优化6次NMO方程精度高，更具有实际应用意义。     

地震勘探高分辨率资料处理的挑战与对策

地震勘探发明以来,其分辨率的极限准则一直是波长的四分之一。在探索压缩地震子波方法的过程中,前期最突出的矛盾是信噪比和分辨率;后期的矛盾,则是高分辨率与高保真。与此同时,利用井资料信息,研发了跨越四分之一波长的很多反演方法。这些方法依赖于地震资料的保幅性、地震解释的正确性和井震资料的匹配性;后来又摆脱了井资料的约束,实现了无井约束反演,分辨率达到了反射系数的尺度,此时问题焦点又转化成阻抗信息的可信性。在梳理地震勘探过程中处理部分的相关信息基础上,借助于地震沉积学的思路和方法,提出了不用井资料约束的储层精细分离方法,既超越了四分之一波长的极限,又有了精度的保证。以实际资料为例,给出了四分之一波长范围叠置河道分离结果。这对石油和煤炭企业老资料挖潜,以及新资料开发具有极大的应用前景;特别是对超薄储层超小断距的确定,采空区及废弃巷道的探测等迫切的生产难题,提供了新的解决思路。      

Deterministic ensemble smoother with multiple data assimilation as an alternative for history-matching seismic data

This paper reports the results of an investigation on the use of a deterministic analysis scheme combined with the method ensemble smoother with multiple data assimilation (ES-MDA) for the problem of assimilating a large number of correlated data points. This is the typical case when history-matching time-lapse seismic data in petroleum reservoir models. The motivation for the use of the deterministic analysis is twofold. First, it tends to result in a smaller underestimation of the ensemble variance after data assimilation. This is particularly important for problems with a large number of measurements. Second, the deterministic analysis avoids the factorization of a large covariance matrix required in the standard implementation of ES-MDA with the perturbed observations scheme. The deterministic analysis is tested in a synthetic history-matching problem to assimilate production and seismic data.     

地震剖面匹配闭合差校正

当用不同施工时期采集的地震剖面对同一地区进行综合对比解释时,可能会遇到相交剖面产生闭合差情况.通常这些闭合差是由于激发因素、接收条件的不同,导致交叉点上振幅、时间及相位存在差异造成的,这将给地震资料的综合解释带来不便.因此,随着四维地震技术的推广,做好地震剖面的匹配校正是一件很有意义的工作.这里给出了2D地震测线上各个交点的振幅、时间和相位差匹配校正的原理公式,利用迭代最小平方算法,通过模型试算,详细阐述匹配流程并进行了效果分析.     

对后续震相的认识促进了地震层析成像的发展

与地震初动震相相比,后续震相包含有更为丰富的地球内部信息,因此逐渐地在地震层析成像研究中得到重视。文中以赵大鹏等最近的层析成像研究为例,介绍了通过增加后续震相到时信息得到的新的层析成像研究成果,包括通过使用反射波,仅用两个地震台站就可以进行地壳成像研究;深度震相到时的增加,拓宽了区域成像范围,成功实现了台网外部层析成像;随着国际地震中心(ISC)提供的更丰富的各种后续震相到时,对后续震相的认识和使用促进了全球体波走时成像的发展和对地球深部结构及动力学过程的认识。另一方面,地震学相关领域的进步,使得更多的后续震相能够被拾取并被应用到成像研究中,这将大大提高对地震波形信息的挖掘,提高对研究地区下三维结构的认识,极大地促进地震层析成像学的发展。     

地震资料的多准则速度分析方法

速度分析是地震数据处理中至关重要的一环,常规速度分析方法中,给定一系列相同间隔的速度进行扫描,以叠加能量或相似系数等作为速度分析的准则制作速度谱.速度谱的能量团通常在正确的动校正速度附近存在一定的发散,从而影响速度的拾取.通过对方差为准则的速度分析方法进行分析和研究,认为方差对同相轴的相干性比较敏感,因此,方差速度谱具有较高的精度.拾取方差速度谱中最小方差数值对应的扫描速度为动校正速度,然而实际资料处理时,由于精度太高,最小方差数值对应的区域太小,不利于速度的拾取.为此笔者提出了以方差倒数、叠加能量和相似系数三者的乘积为目标函数的多准则速度分析方法.噪声会对方差产生影响,然而噪声只是提高方差的整体数值,不会改变方差的相对大小关系,因此地震数据含有一定噪声时本文方法仍然适用.通过模型数据和实际资料的处理,验证了本方法的正确性和实用性.