【作者】 戴小平；

【导师】 梅廉夫；

【作者基本信息】 中国地质大学 ， 油气田开发工程， 2014， 博士

【摘要】 四维地震是对在产油气田进行监测及推动深入开发的重要工具。四维地震使平面上大范围地监测油藏参数的变化,并将观测到的结果与从开发井中获得的实际数据进行对比成为可能。未衰竭／未水淹的含油区、压力隔夹层、注水超出范围等这些开发中发生的现象都可以通过对四维地震数据体的解释揭示出来。从四维地震上获得的这些信息对于进行油藏模型的更新、部署新的井位,最终提高油田的采收率都具有极其重要的价值。本文重点对四维地震在油藏描述及剩余油分布预测的实际应用方法进行研究。为了清楚地解释所做的研究,本文试图回答与此相关的5个问题：(1)什么是四维地震技术,它的发展历史及发展趋势是什么样的?(2)哪些类型的油藏适合采用四维地震技术?(3)四维地震在油藏描述及剩余油分布预测中应用的岩石地球物理的理论基础有哪些?(4)四维地震在油藏描述及剩余油分布预测上有哪些理论及经验方法?(5)在实际应用过程中,如何使用这些方法?本文分四个方面对第一个问题进行解答。首先,本文给出了四维地震技术的定义,并总结了该技术以下四个方面的价值：●提高油藏地质模型的合理性·监测油藏开发过程中的动态变化·对开发计划进行调整●HSSE监测本文对四维地震的发展历史进行的简要总结。其发展历程可以归纳为以下三个阶段：·阶段一,1983～1990年,这一阶段为四维地震技术的萌芽和起步阶段·阶段二,1990～2000年,这一阶段为四维地震技术的快速成长与发展阶段·阶段三,2000年～现今,这一阶段为四维地震技术的广泛应用与发展成熟阶段本文对主要的四维地震采集、处理技术进行了介绍。介绍的采集技术包括：·海上电缆采集·海底永久节点采集·海底电缆采集·陆地采集本文从全球各大地球物理公司中选择了一些最具代表性的采集技术进行了介绍。同时对各种类型的采集技术进行了一个比较。对于四维地震处理技术,本文主要介绍了四维地震资料处理响一股原则m主要的处埋流程,及四维地震资料处理区别与常规三维地震资料处理的独特处理环节。本文展示了BGP与CGG两家公司的四维地震处理流程。其中CGG的Fast-track与Parallel两种处理流程也是本文研究的AAA油田四维地震资料处理所使用的处理流程。本文也特别介绍了四维地震处理以下四个独特环节：·静校正及去条带噪音·规则化●4D面元划分●一致性处理本文对四维地震技术的发展趋势进行了预测,并总结归纳为以下五点：·维地震技术将会应用在更多类型的油田·海上应用前景好于陆上应用前景·四维地震采集将融合更多技术,从而获得更丰富的地下信息●海底、井下的永久、固定性的检波器将是下一代四维地震发展的趋势●使用四维地震对于油藏变化的预测将会更加准确和定量化第二个问题的实质是关于四维地震可行性的问题。可行性研究是四维地震技术的主要技术之一,也是油田开发生命周期过程中的地震采集部署设计的关键环：肖。四维地震可行性研究可以分为两个层次,第一个层次是以关键技术参数风险评估表为依据的评分筛选。这种评比打分方法的实质是利用地震响应最敏感的油藏参数进行定性评价的方法。本文列举了一些前人研究后总结的一些评价表格。另一种可行性研究的方法是使用地震正演模拟油藏参数变化进行定量评价的方法。使用正演的方法,我们可以得出四维地震的响应强度、可解释性,还可以帮助我们优化四维地震采集时间节点。正演模拟可以使用单井测井资料进行模拟,也可以使用油藏模型进行模拟。通过正演的可行性研究工作可以提高四维地震采集的效果,将技术原因造成的四维地震采集,失利风险降到最低,使我们清楚四维地震在目标油藏中是否有效,如何进行采集设计。本文以AAA油田为实例,使用模拟的方法进行了可行性的研究。通过AAA油田的可行性研究的实践,本文得出了以下儿点结论：●本地区四维地震在近道产生的变化比远道产生的变化更加明显●在理想情况下,本地区四维地震可以检测到的波阻抗变化最小值约为什1.5%·不理想的情况下(如噪声干扰、处理较差),波阻抗变化大约要增加到2.5%左右才能被四维地震有效地检测到·远道资料在理想情况下可以检测到2.5%的波阻抗变化(非理想条件下约为4%)有关第三个问题,本文对开发过程中四维地震检测到的油藏参数变化导致的岩石杠弹性变化的岩性物理学基础进行了阐述。四维地震响应的岩石物理学基础同样也是进行四维地震正演的理论基础。通常我们认为在油藏开发过程中,产生变化的主要有四个方面的参数,分别是流体饱和度、压力、压实性、温度。本文岩石物理学的角度出发解释了这儿项油藏参数变化所引起的四维地震响应。第四个问题包含了两个方面。第一个方面是四维地震在定性的油藏描述中的应用方法。本文介绍了一些利用四维地震属性解释油藏变化的方法,这些地震属性包括振幅、时差、阻抗、AVO、扩展弹性波阻抗。在四维地震定性解释方法的探索方面,结合AAA油田的实践,本研究提出了“色标指数”解释模版这样一种基于岩石物理及地区经验相结合的方法,可以非常有效地识别生产过程中油藏产生的不同的现象,这些现象包括：·压实·水驱油·孔隙压力降低●水驱油气·油驱气●孔隙压力增加●注气·溶解气分离●气驱油本文在四维地震应用探索的另一个方面,是研究其在定量描述油藏动态参数变化中的应用。已经有一些研究者介绍过一些利用四维地震属性的方法估算孔隙压力及饱和度的方法,不同方法中使用了不同的地震属性。Tura和Lumley (1999)提出了利用纵、横波阻抗定量计算孔隙压力及流体饱和度的方法Landro(2001)提出了利用AV0属性变化求取压力及饱和度变化的方法；Rojas (2008)提出了在砂岩气藏中使用纵横、波速度比指示岩性、流体饱和度、压力变化的方法等。本文提出了利用扩展弹性波阻抗(EEI)来预测油藏中流体饱和度及压力变化的这样一种方法。扩展弹性波阻抗实际是由Whitcombe等人在2002年才提出的一种概念,最初扩展弹性波阻抗(EEI)主要用于替代弹性波阻抗(EI)进行储层预测及流体检测。扩展弹性波阻抗(EEI)除包含了大量与岩石、流体等相关的叠前信息外,还可以进行多角度分析,从而拟合出与研究的目标油藏参数相关性最好的理论入射角；此外,由于扩展弹性波阻抗(EEI)提高了弹性波阻抗(EI)的稳定性,从而有利于使用该属性在四维地震中的变化更准确地求取流体饱和度及压力的变化。从理论层面来讲,扩展弹性波阻抗(EEI)具有成为四维地震定量解释属性的天然优势。虽然扩展弹性波阻抗(EEI)已经被提出了10余年,但是主要还是应用在勘探阶段更多。前人利用时移扩展弹性波阻抗(EEI)进行油藏参数变化的定量解释的例子几乎还没有见到。本文通过在AAA油田的实际应用,利用井中的流体饱和度、压力资料与相关最好的EEI x曲线,在此基础上进行EEI反演,求取流体饱和度、压力在平面上的变化,取得了很好的效果。本文也总结出使用时移扩展弹性波阻抗(EEI)求取流体饱和度、压力变化的方法流程：步骤1：使用已有的测井资料求取EEI x曲线；步骤2：交汇分析,求取与Sw,Sg相关性最好的EEI理论入射角x；步骤3：对△Sw,△Sg,△P和△EEI x进行回归分析,求取回归方程；步骤4：进行反演,求取△EEI x数据体；步骤5：利用△Sw,△Sg,△P和△EEI x之间的回归方程,从反演△EEI x数据体计算△Sw,△Sg,△P值。关于第五个问题,本文以AAA油田实例,阐述了使用四维地震资料进行油藏描述及剩余油分布预测的具体流程。提出了迭代逼近法这样一种以油田生产历史及四维地震属性作为约束条件,不断修改完善油藏模型的,以达到最低历史拟合要求的方法。在此流程中,地质统计建模、网格粗化、数值模拟、网格细化等都融入到历史拟合的过程中。数值模拟的生产史与四维属性都与井上获得的实际数据进行进行对比,并求得最终版本的油藏模型。本文从实际应用上也归纳了四维地震在油藏建模如下方面的应用：·构造模型优化(主要是确定断层位置,井点误差)●流动分隔单元划分(通过断层的传导性,储层孔隙度、渗透率趋势划分)·油气界面、气油界面及边界带的含油区·地震结构单元及相组合(控制砂体的展布趋势,通过数值模拟的输出结果与四维地震属性异常体匹配迭代,以达到最好相似性)·隔夹层的识别(通过井上及四维地震属性上反应的异常识别)●砂体连通性(四维地震上的异常识别不同砂体的连通性)更多还原

【Abstract】 Time-lapse seismic can be an important tool for monitoring and planning further development of a producing hydrocarbon field. Using of4D seismic make it possible to detect production changes over a relatively large area and match these changes with production data acquired from wells. Undepleted/unflooded patches, pressure barriers and injectors out of range can be revealed when interpreting a4D data set. This can be valuable information when updating reservoir models and planning new wells, and eventually will lead to an increased recovery from the field.This thesis focuses on the methodologies of4D application in reservoir characterization and residual oil distribution prediction. To get explicit explanation of the methodologies this study work on, this thesis try to answer five questions as follows:(1)What is4D technology, its history and its future holds?(2)Which types of reservoirs are suitable for4D implementation?(3)What is the rock physical basis of4D seismic application in reservoir characterization and residual oil distribution prediction?(4)What are the theoretical and empirical methodologies of4D application in reservoir characterization and residual oil distribution prediction?(5)How to use and what is the workflow of these methodologies in practical case?The answer for the first question is comprised of four parts. The first part gives a definition for4D seismic technology and its value which includes:●Improve geological model of reservoir●Monitor dynamic changes of producing reservoir●Adjustment to the development plan●HSSE monitoringThe second part is a summarization of the brief history of4D technology which includes three stages: ●Stage1is between1983-1990, which is the emerging of4D seismic technology.●Stage2is between1990-2000, which is the fast growing and developing phase.●Stage3is between2000-present, which is the widely implementation and maturing of4D seismic technology. The third part introduces some4D acquisition and processing techniques. Those acquisition techniques include:●Marine Streamer Acquisition●Ocean Bottom Nodes●Ocean Bottom Cable●Onshore acquisitionMost representative4D acquisition tools from different Geophysical companies are introduced. This thesis also gives a comparison between different4D acquisition techniques. The introduction of processing technique is comprised of general principals of processing, workflow of4D seismic processing and the unique elements of4D seismic processing different from conventional3D seismic processing. The processing workflows of BGP and CGG are illustrated in the thesis. The fast-track and parallel processing workflows of CCG used in AAA oil field are also introduced in this thesis. Four unique4D seismic processing elements are introduced in this thesis which include:●Statics&De-striping●Regularization●4D Binning●MatchingThe fourth part predicts the future of4D technology. This thesis predicts five possible directions related to the future development of4D technology, which are:●More widely used in various oil fields.●Offshore application seeing a brighter future than onshore.●Integrated with other technology to get more information from subsurface.●Permanent, fixed, downhole and seabed survey will be next generation’s tendency.●Precisely and quantitatively measurement of reservoir changes.The second question is about the feasibility of4D seismic. Feasibility studies are key elements of4D seismic and field life-cycle planning. Feasibility study of4D seismic can be further subdivided into two levels, the first level is screening with4D-technical-risk spreadsheet. The essence of the4D-technical-risk spreadsheet is qualitative assessment of the reservoir with key elements which4D seismic has the most sensitive response to them. This thesis illustrates several most recognized4D-technical-risk spreadsheets in4D screening. Alternatively, sophisticated simulation-to-seismic modeling can help us to assess the magnitude and interpretability of the4D seismic response and can help us to plan the optimal timing of repeat surveys. The modeling can be done with well logs or a3D reservoir model. For the most part, industry has used these tools effectively to high-grade the application of4D technology and, in so doing, to minimize the number of technical failures. We generally know whether4D will work in our reservoirs and when to acquire repeat surveys. This thesis also uses AAA oil field as a modeling example of feasibility study. Through the practice of AAA oil field, this thesis gets four conclusions:●4D changes will probably be more obvious on the near offsets than on the far offsets.●We should be able to detect4D signal as small as1.5%AI change.●In a worst case (i.e. noisy acquisition/poor processing) this might increase to2.5%AI change.●On the far-offsets, we should be able to detect a2.5%change (4%in a worst case).To answer the third question, this thesis briefly introduces the fundamental of rock physics and geomechanics which are used to translate production induced changes in the subsurface elastic properties to the seismic responses. We also explain how this is implemented in our modeling. In general, there are four main parameters that may change in a producing hydrocarbon reservoir, they are fluid saturation, pressure, compaction, temperature. This thesis explicitly narrates the rock physical basis of those four fundamental reservoir changes’ relationships with4D seismic responses.About the fourth question, there are two aspects included in it. The first part is4D seismic application in qualitative reservoir characterization. This thesis summarizes some theoretical approaches of using4D seismic attributes in interpretation of reservoir changes. Those attributes include amplitude, time-shift, impedance, AVO and EEI. For qualitative interpretation approach, this thesis also proposes a Colour Index (CI) template combined with geophysics theory and regional experiences to indentify different phenomena came out in the course of production. Those phenomena include:●Compaction●Replacing oil with water●Reducing pore pressure●Replacing HC with water●Replacing gas with oil●Increasing pore pressure●Gas injection●Gas come out of solution●Replacing oil with gas●Gas come out of solutionAnother part of the fourth question is about4D seismic application in prediction of the changes of dynamic reservoir parameters. Many researchers have studied different ways to estimate fluid saturation and pressure changes during production with different4D seismic attributes. Tura and Lumley (1999) present a method to map and quantify those changes utilizing P-and S-wave impedances. Rojas (2008) proposes to use the P-and S-wave velocities ratio as an indicator of lithologies, fluid saturation, and pressure changes in gas sandstones reservoirs. Landra (2001,1999) introduces an elegant, straightforward inversion scheme that solves for pressure and saturation changes from seismic amplitude-variation-with-offset, etc. This thesis proposes a new methodology which uses time-lapse extended elastic impedance (EEI) for estimation of fluid saturation and pressure changes. Extended elastic impedance is introduced by Whitcombe et al. in2002for the first time. It is used as a replacement of El (elastic impedance) in reservoir and hydrocarbon prediction at the exploration phase. Since it improves the stability of elastic impedance, it is good for a more accuracy quantitative estimation of fluid saturation and pressure changes. So from theoretical side, EEI has the natural advantage to become the right attribute adapt to quantitative4D interpretation. EEI has been used for almost ten years, but rare previous applications in4D monitoring cases were found. This thesis uses the regression equations and extended elastic impedance inversion to estimate the lateral changes of fluid saturation and reservoir pressure get a very good effect. This thesis also summarizes the workflow of using time-lapse EEI in estimation of fluid saturation and pressure changes as following steps:Step1:Compute EEIx logs with exist wireline logs.Step2:Cross-plot Sw, Sg with EEI at different theoretical angers and define the best correlation anger. After this work, we can get the best correlation theoretical anger X-Step3:Regressional analysis and get the quadratic equations between△Sw,△Sg,△P and△EEIX.Step4:Implement the EEI inversion and get AEEIx cube.Step5:Use the regression equation between△Sw, ASg,△P and△EEIχ t0compute the△Sw,△Sg,△P values with the inversion AEEIχ cube.About the fifth question, this thesis uses AAA oil field as a practical example, illustrates the workflow of using4D seismic data in reservoir characterization and residual oil prediction. This thesis proposes an iterative methodology based on advanced history matching solutions to constrain3D stochastic reservoir models to both production history and4D seismic attributes. In this approach, geostatistical modeling, upscaling, fluid flow simulation, downscaling and petro-elastic modeling are integrated into the same history matching workflow. Simulated production history and4D seismic attributes are compared to realizations, and finally decide the last version of geological model. This thesis not only explains the theoretical basis but also use AAA oil field as an example illustrates the4D seismic application in as follows:●Structural model modification (localised fault, well positioning errors highlighted by4D).●Compartmentalization (includes faults transmissibility and permeability porosity trend).●OWC, GOC or addition portion of hydrocarbon to the edge of the grid.●The seismic architectural elements and facies association (control the trend of sand distribution, specifically using the simulated output data to match with the4D anomalies with iterative process to get the best correlation).●The addition of baffles or barriers (particularly those suggested by pressure transient analysis or the4D seismic).●The addition of connectivity of different sandbody (particularly those suggested by the4D seismic).更多还原