A Research Of 3D Seismic Data Processing For Shale Gas Reservoir In Jiaoshiba Area

In 2012,the ministry of land and resources disclosed that the potential of the shale gas reserve for land in our country is 134×1012m3, and the workable reserve is 25 x 1012m3. One of the largest shale gas reserve area in our country is Sichuan, Chongqing and Guizhou region. The Fuling shale gas field which is in Fuling District of Chongqing, has already into the shale gas exploration stage in 2009. Since now, it has become the first shale gas filed which has been successfully developed in commerce outside the North America.Double complex medium condition exists in research area:firstly, the geography of the area is mountainous, not only the elevation changes tremendously, but the lithology of near-surface is also various, both two conditions cause the complex near-surface condition; secondly, for the structure, the research area is located in the joint of several structural unit, structures such as concave、fold and fault developed in this area, in the mean time, the depth of the target horizon is deep, and the thickness of the target horizon is thin, these factors cause the complex condition of target horizon. Based on this condition, it is quite important for exploration and development that how to use appropriate methods to get high precision and high fidelity seismic section and gather without hurting amplitude and reducing SNR.In this article, using the Jiaoshiba area as our research area, building upon the research of predecessors, aiming at the trait that hard to image, using several special seismic data processing methods, and finally got the high-fidelity, amplitude-preserving and high imaging-precision seismic sections and gathers, laying a solid foundation for structural interpretation and reservoir prediction, providing some references to the works that aiming at double complex condition in future.First of all, based on reading and collecting enough articles, summarize the current research condition of double complex condition domestic and overseas, then analyse the seismic-geological condition of research area, find out the key and difficult points of seismic data processing, aiming at these points, select several methods, and after that, analyse the methods that will be used into the processing, after using these methods, evaluate the quality of the result, and make the summary of this research.Based on these researches, we get some understanding as follows:1) Tomography statics build the surface velocity model through grid tomographic inversion, in order to eliminate the influence of complex inspiring surface condition, achieve a good result in complex surface condition, eliminate the influence that low deceleration zone and high-speed layer made to de imaging of shallow and middle layer, make sure of the truth of imaging of seismic.2) Cross domain conical filter use the change between time domain and frequency domain in cross domain, make the surface wave that act nolinear in time domain change into linear noise in frequency domain. The method that suppress abnormal noise use different time gates to statistic amplitude and frequency. These two methods both use some way to make the irregularity noise into regulate, and suppress the noise by analyz their characteristics, in this way, the hurt of amplitude of effective wave can be reduced into least, get a good effect in amplitude protecting.3) In energy compensation, spherical spreading compensation mainly be used to solve the problem of energy imbalance which caused by the absorptive action of horizons, surface consistent amplitude compensation mainly be used to solve the problem of different amplitude level between sources and receivers which caused by complex inspire condition, residual amplitude compensation mainly be used to repatch the problem that spherical spreading compensation and surface consistent amplitude compensation cannot solve, it is based on statistics. These three methods supplement each other, and their result is nearly amplitude-preserved because of their arithmetic is reversible.4) The iteration of velocity analysis, NMO mute and residual static correction can focus energy group of velocity spectrum effectively, and can improve the accuracy of velocity analysis. High density velocity pick replace the traditional method by a brandy new mentality, it can make every point vertical match its velocity, it has higher accuracy, and can solve imaging problem caused by vertical anisotropy, the anisotropy field can also be used in prstack time migration.5) Sub-azimuth velocity analysis can get velocity ellipse of every point through high density velocity pick, by judging the direction of major and minor aixs, can get the trend of crack. Though this method, the correct NMO velocity for whole-azimuth can be got, and can also solve imaging problem caused by azimuth anisotropy.6) Anisotropic pre-stack time migration use the anisotropy field in Kirchhoff pre-stack time migration, and can solve imaging problem caused by vertical anisotropy further. And provide original velocity model.7) Original time-velocity model can be built by merging the near-surface velocity model which is built in tomography statics and the middle-deep velocity model which is built in pre-stack time migration. Through time-depth conversion, the original depth-velocity model can be built. During Kirchhoff pre-stack depth migration, the accuracy of the model can be improved by picking residual curvature and iteration of gird tomography.8) Reverse time migration is based on continuation of source wave-fields and backward continuation of receiver wave-fields. It image by searching the point of intersection of those two wave-fields in reflection horizon. Generally, there are three types of imaging conditions, in these conditions, co-correlation imaging condition is more suitable in complex structure area. Compared to other kind of PSDM methods, RTM can achieve much higher imaging accuracy for high-steep structural. The imaging accuracy can be improved tremendously by using the high-accuracy velocity model which formed in Kirchhoff pre-stack depth migration and make the truly high-accuracy imaging become true.