Downloaded 10/25/17 to 80.82.77.83. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/ The application of a new static correction technique for seismic data in complex surface area: a case study of PD 3D area in Tarim Basin Guo Nianmin,Chen meng,Cui Yongfu,Fang Bing, Research Institute of Exploration and Development, Tarim Oilfield Company, PetroChina, Korla , China Summary Static correction technology is the core technology of seismic data imaging in complex surface area. Based on the traditional tomography static correction technology, we propose an improved workflow of tomography static correction technology, including picking up precise first break, progressive tomographic inversion from shallow to deep, picking up isovelocity surface as datum surface for static correction, etc. We apply the technology in the PD 3D area for practical application. The problem of static correction is effectively resolved and the seismic image quality is improved. Introduction Tarim Basin has always been a key area of oil and gas exploration in the Tarim Oilfield. For example, the surface conditions of the PD 3D area are complex. The PD 3D work area is high in south and low in north as a whole with elevation between 1700-3000m. The northern part is mainly Gobi, which local distributing a small amount of village farmland. The terrain is relatively flat; and the southern part is loess hill, with undulating topography. The characteristics of this area are shown as follows: (1) The surface structure of this area is complex and the relative height difference of the surface is up to 1000m from southwest to northeast; (2) The lateral variation of surface and near-surface lithology is large; (3) The weathering layer has a great change in the velocity and thickness. As shown in Fig. 1, the gathers in typical positions with static problems (thick loess area, high-speed surface area and Gobi area) are selected, where the loess is thick and the near-surface velocity near-offset is high and the first break is undulating. At the same time, the lateral variation of velocity near surface makes the static correction becoming more complex. Such surface conditions bring great difficulties to the problem of static correction. Fig. 1 The variation of surface conditions in this work area © 2017 SEG SEG International Exposition and 87th Annual Meeting Fig. 2 The source gathers under different surface conditions At the moment, there are three methods to solve the problems of static correction: the model method, the refraction static correction, the tomography static correction technology. The model method is a mathematical method of linear interpolation, which is more effective in areas with simple surface conditions. Refraction static correction is based on the hypothesis of horizontal layered model, and it can obtain the ideal effect with good stability and efficient in the area where the ground surface is smooth and the lateral velocity uniformity is good, and the obvious refraction surface has good stability. However, due to the simple model assumption, the method of refraction static correction is generally not ideal in the areas where the ground surface fluctuates strongly and the high-velocity layer is exposed. Because the difference of the elevation and the reverse of speed and the lateral variation of the velocity of weathered layer. In the complex mountainous area where the old stratum exposed, the surface fluctuates violently and the velocity of the surface layer changes greatly. The surface model does not accord with the layered assumption, and the refraction static correction is ineffective, while the tomographic static correction utilizes all the information of the first-arrival waves, including direct wave, refracted wave, inflected wave. It is not necessary to distinguish the type of the first wave and has no assumptions and limitations on the variation of the surface elevation and the velocity distribution of the subsurface. In theory, we can solve the problem of static correction in complex mountain areas by reliable inversion to build near surface depth velocity model. In order to solve the problem of static correction in this area, an improved tomography static correction technology workflow is proposed. Firstly, the initial near-surface LMI Page 5730 Downloaded 10/25/17 to 80.82.77.83. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/ Double click here to type your header layered model is obtained by first picking, and then the near-surface grids model is gradually obtained using reverse branch tomography inversion. Finally, the static corrections are calculated based on the final model. Technology and Method The first break picking up is the basement for tomography static correction and the key to get the right result. We adopt the self-developed neural network to pick up the first break for records before NMO, to avoid the large morphological changes in the phenomenon caused by the large variety of lateral velocity after NMO. The energy discrimination is made on the basis of the instantaneous amplitude. At the same time, the shape of the first arrival is identified by pattern recognition, and then polynomial fitting and multiple iterations are carried out, which can accurately and efficiently pick up the first arrival and greatly improve the efficiency, and ensure the accuracy of static correction and real reliability. The first break with offset gradually increasing roughly reflects the shallow velocity model from shallow to deep. Then start tomographic inversion from the small offset. And when the previous inversion of the final model has good convergence, gradually increase the larger offset for the tomographic inversion iterations. The last final model used as the initial model of the next tomographic inversion, can be relatively fixed the velocity of shallow layers, which equal to updating the shallow velocity model from shallow to deep gradually. In the case of complex near-surface conditions, the surface longitudinal and lateral velocities change greatly. Using the traditional isobathic surface as the static correction surface will bring large errors. In this paper, the isovelocity surface with replacement velocity of 2500 m/s was used as the static correction surface. The calculated static corrections are more accurate than those calculated from the isosceles surface. Example The tomography imaging technique is a reverse branch inversion method using first break, which treats the surface model as an arbitrary medium. It does not limit the difference of surface height, velocity and refraction interface and it can adapt to the vertical and lateral variety of velocity of weathering layer. It is more suitable for regions with complex surface structure, and can obtain accurate correction amount. First of all, an initial model is created by splitting the subsurface into grid cells, where the rays from the source to the receiver pass through the subsurface cells, where the velocity values for each element are constant. Then calculate the simulated first arrival time by ray tracing, and then modify the model to minimize the difference of the first arrival time between observations and calculations. The optimization process of the inversion is achieved by linearizing a large-scale nonlinear least squares problem and iteratively calculating the near-surface model. The iterative reverse branch tomography inversion with offset increasing can be used to constrain the shallow model. Aiming at the difficulty of first break picking up in complex area (Southern Loess Plateau, Middle Foothill Regions, Northern Gobi Desert), the method of automatic picking plus two manual modification is adopted to complete all offset first break picking up as shown in Fig. 3. ① Automatic pick up the first break from the beginning to 4500 offset and manual modification; ② Calculate the initial tomographic static correction using the first break in 4500 offsets; ③ Apply the initial tomographic static correction to the original single shot and manual pick up the first break of all offsets; ④Minus the initial tomographic static correction and obtain the true first break of all offset; ⑤Using the final first break to compute the tomography statics. Fig. 3 First break pick up After two manual picking up and two tomographic inversion iterations, it is a good solution to the correction problem in this work area due to the complex terrain, the © 2017 SEG SEG International Exposition and 87th Annual Meeting large undulating surface, the unevenness of the low velocity band and the large lateral velocity variation. The advantage of reverse branch tomography imaging technique is that the Page 5731 Downloaded 10/25/17 to 80.82.77.83. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/ Double click here to type your header tomographic inversion is not affected by topographic fluctuation, lateral velocity variation and underground interface inclination. It can better fit the original first arrival time and more fully excavate the first arrival time. Also it can accurately reflect the variety of the lateral velocity and make the stratification in the vertical direction more precise. It can solve the primary static correction problem of complex data and accurately retrieve the shallow velocity model. As a result, the accuracy of the seismic data is high, which can maintain the authenticity of small-scale structures and improve the imaging quality of seismic data. The thickness of the loess layer in this project area is not uniform and the variety of lateral velocity of the loess layer is large. The isovelocity surface with replacement velocity of 2500 m/s is more precise than that of the traditional isobathic surface. As shown in Fig. 4, the traditional 4500m offset singletomographic inversion is compared with the final model after the 15th-iterative tomographic inversion from 500m to 4500m offset in this paper. The velocity description of shallow low-velocity loess layer is more clearly described by the 15th-iterative tomographic inversion than the result of traditional single-tomographic inversion. Fig.5 shows the comparison of ray density with conventional tomographic inversion and 15th iterative tomographic inversion. The result shows that the ray density distribution using iterative tomographic inversion is more reasonable in southern Loess Plateau. (a) conventional method (b) modified method of this paper Fig. 5 The comparison analysis of ray density Figure 6 shows the comparison between the stacking section after the conventional tomographic inversion and the application of static correction using the 15 times iterated tomographic inversion. It can be seen that the stacking section after final static correction obtained by the improved tomographic inversion obtained better results in the southern shallow loess area (box area) and the PD tectonic zone (oval area). Figure 7 shows the single-shot comparison before and after tomography static correction. It can be seen that the first arrival is well improved and the static correction problem is effectively solved. (a) conventional method (a) applied the statics of conventional tomographic inversion (b) modified method of this paper Fig. 4 The final velocity model comparative analysis © 2017 SEG SEG International Exposition and 87th Annual Meeting Page 5732 Downloaded 10/25/17 to 80.82.77.83. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/ Double click here to type your header the traditional isobathic surface. And the imaging quality will be better. The practical application of the data from the three-dimensional work area in PD proves that the improved workflow of the static correction technique is very effective and this technique can be applied to the static correction of area with complex surface. （b）applied the statics of modified method of this paper Fig. 6 The comparison between the stacking section after the conventional tomographic inversion (a) and the application of static correction used in this paper (b) (a) before (b) after Fig. 7 The seismic shot data comparison before (a) and after (b) tomography static correction used method of this paper Conclusions Accurate first break picking up is the basis of tomographic inversion in mountain area. The method of automatic picking based on network plus two manual modification used in this paper solves this problem. The proposed partial offset multi-iterations tomographic inversion imaging technique can not only reflect the lateral velocity change finely, but also make the stratification in the vertical direction more precise compared with the conventional tomographic inversion static correction technique. For the complex mountain area data, picking up stable isovelocity surface as the static correction datum is more precise than © 2017 SEG SEG International Exposition and 87th Annual Meeting Page 5733 Downloaded 10/25/17 to 80.82.77.83. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/ EDITED REFERENCES Note: This reference list is a copyedited version of the reference list submitted by the author. Reference lists for the 2017 SEG Technical Program Expanded Abstracts have been copyedited so that references provided with the online metadata for each paper will achieve a high degree of linking to cited sources that appear on the Web. REFERENCES Marsden D., 1993, Static corrections — A review, Part 1, Part 2, Part 3: The Leading Edge, 12: 43–49, https://doi.org/10.1190/1.1436912; 115–20, https://doi.org/10.1190/1.1436936; 210–216, https://doi.org/10.1190/1.1436944. Jing X.-L., Tomography modeling for complex media and tests on models: Oil Geophysical Prospecting, 45: 66–71. Jing Y. H., 2009, Seismic first arrival wave travel-time tomography and near surface velocity model: Chang’an University. Lin B. X., 2002, Tomography for LVZ velocity inversion and statics: Geophysical Prospecting for Petroleum, 41: 136–140. Wang X, and Z. H. He, 2010, Technology of static correction for multi-information constrained tomographic inversion and its application: Natural Gas Geoscience, 21: 316–320. Chen S. J., 2006, The application of seismic first arrival ray tracing tomography static in the complicated near surface area: Geophysical Prospecting for Petroleum, 45: 34–39. © 2017 SEG SEG International Exposition and 87th Annual Meeting Page 5734

1/--страниц