【作者】 崔杰；

【导师】 韩立国；

【作者基本信息】 吉林大学 ， 地球探测与信息技术， 2011， 博士

【摘要】 应用振幅随偏移距变化(AVO)等叠前地震信息预测含油气储层已得到了广泛的实践,在国内外油气地球物理勘探与开发领域中取得了大量成功实例,但由于影响地震波振幅变化的因素非常复杂,常规的AVO分析也存在着多解性问题,解决这一多解性的有效途径是利用多波多分量地震勘探信息。纵波和P-SV转换波的联合AVO分析及应用可以大大减少多解性,这点已得到了地球物理工作者的普遍认同；同时,综合二者的AVO属性还可以获得更多更准确的储层岩石物理参数,因此,转换波AVO技术的研究与完善是储层地震勘探亟待解决的课题之一。科学合理地构建地质-地球物理模型也是储层预测和流体识别的关键问题。从南海深水区等油气藏实际数据分析来看,基于双相介质和各向异性介质的储层建模比常规的各向同性介质储层模型更加科学合理。为了更好地利用叠前各种地震信息来精确预测储层和识别流体,人们必须首先研究各向异性等复杂介质中的纵波及转换波的AVO的理论及分析技术,并据此来获取储层的岩性及空间展布信息等。本文从实效性就弱各向异性介质中转换波的反射系数公式进行了系统研究,为了提高计算效率推导了简化公式,建立了各种属性参数计算方法,同时推导了弱各向异性介质中P-SV波弹性阻抗,构建了各向异性介质中广义流体识别因子,得出了一系列有意义的研究成果。本文首先给出了各向同性介质中转换波的反射系数线性近似公式,系统分析了基于该公式的AVO分析算子,认为AVO分析的不同近似式对模型都有不同程度的依赖性。根据四种典型的含气含水层实例分析,P-SV波的AVO响应特征较PP波要复杂,不能同时区分四类含气含水层,尤其是对第一类和第二类基本没有识别效果,对第三类的识别效果较好,说明了转换波的AVO对储层的流体成分不是很敏感。而通过归一化后的反射系数AVA响应曲线呈单调增加或减小同时还放大了异常,对四类AVO类型的含气含水层识别效果都有所提高,对前三类,归一化后的反射系数随着入射角的增加呈单调递增现象,对第四类,反射系数是呈单调递减的,但是反射系数均为正值。最为流行的各向异性介质中转换波反射系数近似公式是Vavrycuk(1999)给出的,但该公式在AVO分析中的适用性不强,因此Petr Jilek(2002)给出了PS波反射系数近似公式的另一种表达形式,适用于任意半空间各向异性介质,由于假设了界面两侧纵横波速度和密度的弱差异性和任意半空间的弱各向异性,所以称此公式为弱差异弱各向异性近似公式,它是关于Thomsen各向异性参数、入射角和方位角的函数,并适用于以下介质组合：各向同性、VTI、HTI和正交各向异性,垂直对称平面的任意方位角适用于HTI介质和正交介质。但是Petr Jilek的表达式比较复杂,因此我们对正交各向异性介质中P-SV波反射系数公式进行了简化,对各向异性参数进行了转换,将其退化到VTI介质和HTI介质两种常用的典型各向异性介质中,结合各向同性中转换波反射系数的近似方式,尝试将反射系数表达成sinθ奇函数的形式,在保证精度的基础上分别保留到sinθ的一次幂、三次幂、五次幂和七次幂,本文将其称之为一阶、三阶、五阶、七阶新近似公式,均为各向同性反射系数与各向异性反射系数扰动项加和的形式,通过不同的理论模型对比了几种表达形式的精度,并做了相应的误差分析,同时针对不同模型做了反射系数的影响因素分析,得出了相应的规律。在得出新近似公式之后,本文首次尝试将归一化反射系数思想用于VTI和HTI介质的P-SV波反射系数三阶新近似公式,利用各向同性情况下四类经典AVO含气含水砂岩模型的具体参数值,在此基础上增加了Thomsen各向异性参数,计算并绘制含气含水砂岩AVA的响应特征,获得了与各向同性介质情况下类似的结果。仿照各向同性条件下的P-SV波AVO截距梯度属性分析方法,以VTI介质为例讨论了P-SV波AVO截距梯度属性,计算结果表明,各向异性介质转换波截距梯度的正负规律跟各向同性介质情况下的规律是类似的,由各向异性参数引起的区别在于截距和梯度取值的大小。不同储层模型的交会图的特征是不同的,但是它们的背景趋势可以拟合成一条直线,由直线的斜率可以求取平均横纵波速度比。在进行AVO属性交会图分析之后可以根据需要得到横波速度剖面,密度剖面和剪切模量剖面,为AVO定性分析提供了有利的途径,在横纵波速度比k一定的情况下对各向异性参数可以进行定量分析,利用该公式也可以进行参数反演。各向同性储层模型弹性阻抗公式也具有多种形式,并且大部分已经应用于叠前实际数据处理中。而对转换波弹性阻抗以及各向异性介质中弹性阻抗的研究及成果都比较少,在研究过程中需借鉴各向同性和各向异性介质中纵波弹性阻抗的方法和理论。因此,利用本文给出的VTI和HTI介质转换波反射系数近似公式推导出了最新的转换波弹性阻抗公式,通过不同模型的正演模拟,对公式精度进行了误差分析,并结合whitcombe的归一化思想,形成了弱各向异性介质中扩展的转换波弹性阻抗公式,最后对影响弹性阻抗公式的主要因素及规律做了总结。通过弹性阻抗的交会分析,也表明了在储层识别方面弹性阻抗比声波阻抗更灵敏。为了从实际地震资料中提取更多地反映储层流体的信息,人们提出了许多流体因子,地震纵波、横波和转换波的综合利用为此也提供了一条有效的途径。这些方法都是使用线性Zoeppritz方程来提取物性参数,例如P波和S波阻抗、速度、泊松比等,同时也可以获得一些利于解释的参数,如弹性模量、拉梅常数以及一些流体因子。通常这些流体因子都可以较好地识别烃类区域。从流体因子建立的岩石物理基础Biot-Gassmann方程出发,分析了填充不同类型的流体时岩石对速度和波阻抗的影响,在此基础上列举了常见的阻抗类流体识别因子,使用四类典型的砂岩模型做数值试验,总结了各个流体因子的对含气含水砂岩的识别能力及敏感性。为了将各向异性参数、入射角和方位角引入到流体识别因子公式中,本文尝试把弱各向异性介质中PP波和P-SV波弹性波阻抗公式代入到已有的流体因子公式中,构成新的广义流体因子公式,这里已有的流体识别因子选择的是高灵敏度流体识别因子Fdanj和灵敏度较弱的μρani识别公式,便于进行灵敏度比较,分析了流体因子Fdani和μρani对三种岩性组合模型的识别效果,同时也分析了Thomsen各向异性参数的影响,得出了相应规律。转换波信息是纵波的补充,同时AVO、弹性阻抗和流体识别都是储层预测的关键技术,在加入各向异性因素之后,可以提高它们对储层预测和流体识别的精度。更多还原

【Abstract】 Application of AVO and other pre-stack seismic data to predict oil and gas reservoirs has been widely used, which has made a lot of success stories in domestic and international oil and gas geophysical exploration and development. But due to the complex factors of AVO changes, conventional AVO analysis also exists multiple solutions, and multi-wave and multi-component seismic exploration information is an effective way to solve multiple solutions. The combination using of converted wave would greatly reduce multiple solutions, which is generally accepted by geophysicists. At the same time, comprehensive AVO attributes can get more accurate physical parameters of reservoir, therefore, the research and improvement of converted wave AVO technology is an urgent subject of reservoir seismic exploration.Building the geological-geophysical model scientifically and reasonably is the key issue for reservoir prediction and fluid identification. From the actual oil and gas reservior data analysis of the South China Sea deep water, the reservoir modeling based on two-phase media and anisotropic media is more scientific and reasonable than the conventional isotropic media. In order to make better use of pre-stack seismic information to accurately predict reservoir and identify fluid, we must first study PP wave and P-SV wave AVO in anisotropy media, according to which we can get lithology information and spatial distribution of reservoir. In this paper, the weak anisotropic media converted waves linear approximation formula is simplified, and AVO attribute analysis is done, while the weak anisotropic media P-SV wave elastic impedance is derived, the generalized anisotropic medium fluid identification factors are constructed, which are very meaningful.The paper first gives the advantage of converted wave AVO analysis, and summarizes converted wave reflection coefficient linear approximate formula in isotropic medium, recognizing that even with the same formula, accuracies of different models are not the same, which means that approximate formulas have some dependence on the model. For the four typical gas aquifer, P-SV wave AVA response characteristics are more complex than PP wave, and cannot distinguish the four types of gas-bearing aquifer, particularly did not identify first and second base, has better identification of the third category, indicating that the converted wave AVO is not very sensitive to the fluid composition of the reservoir. Reflection coefficient AVA response curve after normalized is monotonic increase or decrease, but also enlarge the exception, recognition effect for four types of AVO gas aquifer has increased, for former three categories, normalized reflection coefficient decreases with the increasing angle, and for the fourth category, the reflection coefficient is monotonically decreasing, but the reflection coefficient is positive.The most common converted wave reflection coefficient approximate formula in anisotropic media is given in 1999 by Vavrycuk, but the applicability of formula in the actual of AVO analysis is not strong. Therefore. Petr Jilek (2002) gives another PS wave reflection coefficient approximate formula expression for arbitrary anisotropic dielectric half-space, the derivation of the formula is based on the following assumptions:weak contrasts of elastic medium parameters across the interface and weak anisotropy in half-space, so this formula is called weak contrasts and weak anisotropy approximations, this formula is the function of the Thomsen anisotropic parameters, incident angle and azimuth. This resulting expressions can be applied for any combination of isotropic. VTI. HTI and orthorhombic halfspaces.. But it is complex, so we simplified orthotropic medium P-SV wave reflection coefficient formula, converted the anisotropy parameters, and degradated it to the VTI and HTI media, which are two typical anisotropic media.we commonly used it in combination with converted wave reflection coefficient in the isotropic approximation method, try to express the reflection coefficient as odd functions of sinθ.on the base of accuracy, we reserved it to one power.three power, five power and seven power, the paper called it a first order approximation, the third order, fifth-order, seven-order new approximate formula. Both are the combined-items of isotropic reflection coefficient and anisotropic disturbed item, through different models we compared the accuracy of several forms of expression, and made the corresponding error analysis, at the same time for different models the impact of factors of the reflection coefficient obtained corresponding laws.After obtaining the new approximate formula, we firstly attempt to normalize the P-SV wave new third-order approximation reflection coefficient in VTI and HTI media. Based on using of four classic case of isotropic AVO gas sandstone model parameters, adding to the anisotropy factors-Thomsen anisotropic parameters of the assignment, we draw gas/water sandstone response characteristics of AVA to get the effect of identification, obtained the similar results under the isotropic medium. Similar to the isotropic P-SV wave AVO gradient and intercept method, an example of P-SV AVO gradient attribute intercept is discussed in the VTI medium wave, and we obtain anisotropic converted wave positive and negative gradient case law with the law of isotropic media is similar, while the difference caused by the anisotropic parameter is the size of the intercept and gradient values. Forms of cross plots of different models are different, but their background trend can fit a straight line. The slope of the line can strike the average k value. After the research on AVO attributes analysis by new equation,we can get S wave velocity profile, density profile and shear modulus profile, which can provide a favorable way for qualitative analysis. When attribution parameters are constant, quantitative analysis and inversion of anisotropic parameters can be carried out.Many scholars have proposed different types of elastic impedance formula, the P-wave elastic impedance inversion of pre-stack data has become more sophisticated, while converted wave elastic impedance and elastic impedance in anisotropic media research is less than P-wave, so it needs to draw on the research based on the isotropic elastic impedance method and theory. Based on this, P-SV wave elastic impedance formulations in VTI and HTI were proposed by approximation formula.Then we present shale/gas sand of models to test the accuracy of P-SV wave elastic impedance.We normalized it with whitcombe’s ideas, and formed converted wave elastic impedance formula in weak anisotropic media, and elastic impedance of the final formula of the main factors and rules are also summarized. The intersection of elastic impedance analysis also shows that recognition of reservoir types in terms of elastic impedance is more sensitive than the acoustic impedance.In order to predict reservoir by using AVO and multi-components seismic data, some fluid factors and fluid identification methods have been studied, seismic wave, shear wave and converted wave of comprehensive utilization of this also provides an effective way. These methods are used to extract linear Zoeppritz equation parameters, such as P wave and S wave impedance, velocity, Poisson ratio, but also can get some help to explain the parameters, such as elastic modulus. Lame constant, and some fluid factor. Generally, these fluids can better identify hydrocarbon region. Starting from the physical basis of Biot-Gassmann equations, we analyze the impact of filling the rock with different types of fluids on velocity and impedance, and in this basis list the common impedance fluid identification factors, which are used to identify four typical sandstones. Here recognition ability and sensitivity of the various factors on gas/water sandstone are analyzed. In order to introduce the anisotropic parameters and azimuthal angle into the equation of fluid factor, we insert the PP and P-SV wave elastic impedance into the fluid factor. The high sensitive fluid identification Fdam and low sensitive fluid identificationμpam are compared, which is easy to compare the sensitivity analysis of the fluid factor Fdam andμρam. Though three lithological model identification, we also analyze the Thomsen anisotropic parameters’effect and obtain the corresponding laws.Converted wave is the supplement for P-wave information, and at the same time AVO. elastic impedance and fluid identification are the key technology,joining the anisotropy factors for reservoir prediction, which can improve the accuracy of their reservoir interpretation, so our study is very meaningflul.更多还原