准确求取各向异性初至波走时与射线路径对偏移成像及层析反演具有重要意义，目前基于图论的最短路径方法虽然能同时得到走时与射线路径且无条件稳定,但仅适用于弱各向异性介质。利用牛顿下山法求解群相关系方程,由群角得到较为精确的相角,进而求得较为精确的群速度和走时,使得最短路径射线追踪方法可以适用于复杂各向异性介质。将此射线追踪方法应用于均匀各向异性介质及层状介质模型,并将试算结果分别与理论走时和弹性波动方程有限差分结果进行比较,验证了方法的准确性;通过复杂各向异性介质模型试算,验证了方法对复杂介质的适用性。

The accurate calculation of anisotropic first arrival traveltime and raypath is critical for migration imaging and tomographic inversion.The shortest-path method based on graph theory can get the traveltime as well as the raypath with high stability.Thus,we proposed a new shortest-path ray tracing algorithm for anisotropic medium.To compute the traveltime in a grid,Newton downhill method is used to solve the equation relating group angle with phase angle.By this way,a more accurate phase angle can be achieved from a group angle,and then a more accurate group velocity and traveltime can be obtained.Correspondingly,the shortest-path method can be applied to strong anisotropic medium.Finally,this ray tracing algorithm was applied to homogeneous anisotropic medium,a layered model and a complex model,and its effectiveness was proved through comparing the computed traveltimes with those using the finite difference implementation of elastic wave equation.

针对我国南方山地探区复杂地表和复杂地下构造导致常规各向同性叠前深度偏移难以准确成像问题,研究了基于TTI各向异性的全速度建模技术,建立了一套起伏地表条件下各向异性全速度建模的技术思路和实现流程。首先采用井约束初至层析反演方法建立准确的起伏地表速度模型,将其与常规处理获得的中深层速度模型进行匹配拼接,建立起初始的各向同性起伏地表全速度模型;然后通过井震联合的方法获取各向异性参数,将其加入到各向同性起伏地表全速度模型中,结合倾角和方位角数据,实现TTI各向异性起伏地表全速度建模。四川盆地DXC工区地震资料的成像处理验证了各向异性全速度建模技术的有效性;基于TTI各向异性全速度模型的起伏地表Kirchhoff叠前深度偏移结果有效地消除了起伏地表的影响,同时降低了复杂构造与速度各向异性对地震成像的影响,使得地下构造成像更合理也更精确。

The complex surface and the complex underground structure lead to inaccurate image from the conventional isotropic pre-stack depth migration in the mountainous area of southern China.Aiming at the problem,a full velocity modeling technology based on TTI anisotropy was studied,a series of technical ideas and workflow of anisotropy full velocity modeling from rugged topography was established.Firstly an accurate irregular surface velocity model was built by the well-constrained first arrival tomographic inversion method.Secondly,an initial irregular surface isotropic full velocity model was built by matching and stitching the near surface model and the mid-depth velocity model was obtained by conventional data processing.Then the anisotropy parameters was obtained through borehole-seismic joint method.Finally,the TTI anisotropic irregular surface full velocity modeling was accomplished by integrating the dip,azimuth and anisotropy parameters into the existing initial irregular surface isotropic full velocity model.Apply the velocity modeling technology to the seismic image of DXC survey in the Sichuan Basin.The processing results show that anisotropic full velocity modeling can achieve available full velocity modeling in the mountainous area; TTI anisotropy velocity model based Kirchhoff pre-stack depth migration from rugged surface can eliminate the influenc of rugged topography and reduce the impact of complex structure and velocity anisotropy on the seismic imaging.

本文基于球谐展开群速度表达式计算走时关于各向异性参数的Fréchet核函数,利用共轭梯度法对两种参数化方法进行了VTI介质中多参数联合反演方法研究.经过理论分析和数值试验发现,与经典的Thomsen参数化方法相比,垂直慢度、水平慢度与动校正慢度的参数化方式更有利于VTI介质多参数联合走时层析反演.为了克服走时对ε参数的不敏感性,我们采用了两步法进行双参数反演,理论模型试验反演得到了与垂直速度精度相当的ε参数.可以将两步法扩展到三步法以同时反演各向异性介质中的三个参数,数值试验展示了该策略的应用潜力.

YKL气藏开发迫切需要高品质地震成像资料的支撑,尤其是要求有高精度的构造成像资料。由于该区块三维地震勘探资料存在缺失近炮检距、覆盖次数低等不利因素,因此,常规单一的速度建模方法难以解决该区井震误差大的问题。为此,针对性地提出并联合应用了井控各向异性网格层析和逐层速度建模技术。在选好区内标志层、建好初始速度模型的基础上,利用区内多口已知井信息求取各向异性参数,先消除浅部大套稳定地层速度对气藏深度的影响,再由浅到深逐层对标志层反演,更新迭代各向异性参数、速度及深度成像,直至各标志层的井震误差降低到规定范围内。3口验证井的应用结果显示,实际资料最终成像成果的气藏构造深度预测值与实钻井深度误差均在5 m内,气藏构造成像的精度得到了显著提高,断裂成像也得到明显改善,降低了开发钻井的风险。

<p>The urgent demand for high quality imaging data,especially high-precision geological structure,is a key problem in the development of YKL gas field.Because of missing near-offset data and low fold of seismic data,well-seismic misfit for structural depth prediction in this region can&rsquo;t meet the exploration requirement if we only use conventional velocity modeling.Therefore,well-controlled anisotropy grid-tomography technique was adopted to conduct velocity modeling layer-by-layer.Based on precise choosing the reasonable markers and building the initial velocity model in the selected area,we use known logging data to calculate anisotropic parameters within this region.Then we eliminate the influence of large set for stable formation velocity on gas reservoir depth,and from shallow to deep layer by layer inversion on marker layer,update iterative anisotropic parameters,velocity and depth migration imaging until the predicted depth errors from well log and seismic data for each marker layer to be reduced.In addition,it has been confirmed by 3 drilling wells that the misfit between the predicted structural depths respectively from the final seismic migration results and logging depth measurements in this gas reservoir are all within 5m.So we can draw a conclusion that our technique can significantly improve the imaging accuracy for structure and fracture of gas reservoir so as to reduce the risk of production drilling.</p>

地质导向是水平井钻井中保障沿着优质储集层钻进的重要支撑手段。传统的基于地层对比的地质导向技术所构建的导向模型通常难以提供精细且具有预测能力的地层构造模型、岩性分布模型及含油性分布模型,往往使导向工作出现偏差,引发优质储集层钻遇率低的问题。地震数据纵向分辨率虽低但具有层面横向控制能力,测井数据纵向层位虽精细但因井位少使其纵向地层延展性控制能力差,基于二者优点的井震联合构建三维地质导向模型技术,则可以提供更准确的三维地层模型和剖面导向模型,进一步结合随钻数据实时校正模型,即能够确保钻井轨迹位于薄砂层目的层中,指导钻井施工实现精准地质导向。三维建模技术、地层剖面建模与分析技术以及模型更新技术研究结果表明,在勘探开发中后期,钻井、测井、录井资料丰富的情况下,基于井震联合三维地质导向模型构建的地层剖面模型能够满足地质导向的精度要求,有效提高油气层钻遇率。

Geosteering is an important support to ensure drilling along high quality reservoirs in horizontal well drilling. The guidance model constructed by the traditional geosteering technology based on stratigraphic correlation is difficult to provide a fine and predictive stratal configuration model, lithologic distribution model and oil-bearing distribution model. It often leads to deviation in guiding work and causes low rate in drilling high-quality reservoirs. Vertical resolution of seismic data is low but has horizontal control ability. Log data are precise in the vertical horizon, but the control ability of their vertical formation ductility is poor due to the lack of well location. 3D geosteering model technology was constructed jointly based on the advantages of well logging and seismic. It can provide more accurate 3D stratigraphic model and profile guidance model and further combine with data while drilling to real-time correct model, that is, it can ensure that the drilling track is located in the target bed of thin sand layer, and guide drilling operation. The accurate geosteering is realized. The study results of 3D modeling technology, stratigraphic profile modeling and analysis technology and model updating technology show that stratigraphic profile model based on 3D geological model of well logging and seismic combination can meet the precision requirement of geosteering with abundant drilling, well logging and mud logging data in the middle and late stages of exploration and development, effectively improving the rate of drilling oil and gas reservoirs.

高精度速度场的建立对渤南洼陷隐蔽油气藏勘探十分重要。综合大量测井、测试和地震资料，分析和讨论了测井、VSP和地面地震速度间的差异;利用VSP测井优化技术，对叠前偏移速度做了标定和校正，在此基础上对离散速度数据进行空间插值，获得了全区地震层速度、均方根速度三维数据体。不同类型速度资料的对比分析表明，研究区浅层的井孔速度大于井旁地震层速度，中、深层则呈现正好相反的特点;浅层速度弥散、深层介质的各向异性是导致井、震速度误差的主要原因。利用VSP速度对地震层速度的局部误差进行闭合差校正，经空间插值后建立的全区速度模型可从任意方向切片观察速度空间展布。三维速度资料为研究渤南洼陷整体的地层密度、压力系数等岩石和流体参数的空间展布和结构特征提供了可能。

Seismic velocity is one of the most important parameters for subtle reservoir exploration in Bonan Sag.The difference among logging,VSP and ground seismic velocity are analyzed and discussed based on a large amount of logging,testing and seismic data.On the basis of analysis for discrete velocity in target area,a method with VSP logging optimum technique is proposed to calibrate and correct prestack migration velocity, and the 3D data volume of seismic interval velocity and RMS velocity is obtained.Comparison between different types of velocity data reveals that,velocity obtained from borehole is larger than seismic interval velocity in the upper stratum,of which the fact is opposite in the middle and deep stratum.Velocity dispersion of upper stratum and rock anisotropy of deep stratum could explain the cause of the logging and seismic velocity error.The local error of seismic interval velocity is conducted mis-tie correction by using VSP velocity.By spatial interpolation,we get an overall velocity 3D model and it offers a comprehensive understanding of velocity distribution at any direction slice.3D velocity data reveals an overall variation view of formation density,pressure parameters and some other spatial distribution as well as structural characteristics of rock and fluid.