Subsalt Steep Dip Imaging Study with 3D Acoustic Modeling
- 1. Introduction Subsalt imaging in the deepwater Gulf of Mexico environment is challenging due to wave field distortion by complex salt geometries. Wide-azimuth (WAZ) towed-streamer acquisition is rapidly replacing narrow-azimuth (NAZ) acquisition as the industry standard (Michell et al., 2006; Corcoran et al., 2007, Howard 2007), because of superior subsalt image. The benefits to subsalt imaging were demonstrated in a recent WAZ processing project covering the Jack discovery in Walker Ridge OCS area. Figure 1 shows a comparison of final NAZ (Figure 1a) and final WAZ (Figure 1b) subsalt images. The four-way closure shown in the green circled area is much better imaged on the WAZ image than on the NAZ image. However, the WAZ image still has problems. First, as shown by the red circle in Figure 1b, the subsalt sediments against a salt keel are not imaged. A clear image of up- dip truncation is needed to define the potential three-way closure. Second, pointed to by the red arrow in Figure 1b, poorly imaged faults appear above the four-way closure. Their poor definition makes interpretation difficult. Motivated by the problems identified above for imaging subsalt steep dips, we performed a 3D modeling study in Jack area to understand the reason and come up with a solution. Method To perform a 3D modeling requires: 1) 3D velocity and density model, 2) acquisition geometry, 3) high-fidelity modeling and migration algorithms. The final salt model of a production WAZ project, which had gone thought sediment tomography updates and salt interpretation, was used in the modeling. A profile of the velocity model (Figure 2a) shows a typical salt sheet in the study area. Thick and undulating salt sheets are prone to cause steep dip imaging problems. To create the density model, we first derived a density background from the velocity model using the Gardener equation, then inserted density perturbation to create impedance contrast for sediment layers and target steep dips. The final density model is shown in Figure 2b, consisting of eighteen sediment layers; steeply dipping subsalt faults (500 , 600 and 700 ), a subsalt four- way closure, a three-way closure with up-dip truncated against salt and a sinuous channel structure. A more precise density could be constructed if density logs were available. The model dimension is 40x40x15km on a 25x25x10m grid. Two types of acquisition configurations: surface seismic and vertical seismic profile (VSP), were simulated. Figure 3a shows surface seismic configurations. A full-azimuth (FAZ) surface seismic was simulated with both inline and crossline offsets of +-8km. From the FAZ data, we extracted a NAZ data with inline offset of 8km and crossline offset of +-500m and a WAZ data with inline offset of +- 8km and crossline offset of 4km. WAZ data followed the acquisition geometry of the production WAZ project with the following: 45 degree NE-SW shooting direction; 150m source spacing and 500m sail line spacing. Inspired by the successful cases of VSP to image surface seismic “blind spots” (Hornby et al, 2007; Burch et al, 2009), we simulated a 3D spiral VSP acquisition to evaluate the VSP imaging capability in comparison to surface seismic. Figure 3b shows the source pattern of the 3D VSP acquisition. In our 3D VSP modeling a well is placed in an up-dip position from targeted subsalt steep dips. This is to ensure that reflection energy from the targets can reach the receivers in the borehole. The same reflected energy from subsalt steep dips can refract at the base of salt and may not reach surface receivers due to the critical angle condition (Liao et al, 2009). Eighty-one VSP receivers were placed at depths from 6km to 10km with a 50m interval. The 3D VSP has 10km maximum source offset from the wellbore and a 500m radial increment. Maximum frequency is 15Hz for both the surface seismic and VSP for fair comparison. The surface seismic and VSP acquisitions were simulated using a two-way acoustic finite-difference (FD) modeling algorithm. To complement the high-fidelity modeling algorithm, we used reverse time migration (RTM) to migrate the synthetic shots with the exact velocity model. The migrated images were utilized to analyze the modeling results. By using RTM instead of ray-tracing based or one-way wave equation migration, we eliminated the possibility of losing steep dips due to a dip-limiting 73rd EA GE Conference & Exh ibition incorporating SPE EUROPEC 2011 Vienna, Austria, 23-26 May 2011
- 2. imaging algorithm. In addition, RTM is able to handle the complex overburden such as complicated salt geometry. To limit our scope of the study to only illumination-related effects, surface multiples are excluded in the acoustic modeling by adopting an absorbing free-surface boundary condition. In field data, surface multiple contaminations can severely interfere with and mask subsalt steep dips. We conducted this model study in Jack area with ~40 OCS blocks, as shown in Figure 3c, where we overlaid a block map with top of salt horizon. The yellow star represents the surface location of the VSP well. 3D two-way acoustic wave-equation FD modeling is a computationally intensive process. Ample resources were allocated to ensure this modeling study was executed in a timely manner. Results First, we compare RTM inline sections at the VSP well location for real WR WAZ (Figure 4a) and synthetic WAZ (Figure 4b). As pointed by the red circle, the dim zone we observed in the real WAZ image has been re-produced in the synthetic WAZ. Modeling confirmed that WAZ could not image the steep dips of three-way closure due to poor illumination. We then compare migrated image for synthetic WAZ (Figure 4b) and synthetic FAZ (Figure 4c). Overall, FAZ improves the illumination for lower dips, but it is still unable to image the three-way closure, which indicates even FAZ surface seismic still suffers from poor illumination. Next, we compare RTM images for FAZ surface seismic (Figure 5a) and 3D VSP (Figure 5b). As shown by the blue oval in Figure 5a, FAZ surface seismic failed to illuminate the three-way closure and subsalt steep dipping faults, which indicates surface seismic data may not contain any reflected energy returning from subsalt steep dips regardless of surface acquisition geometry. If the returning reflected energy hits the base of salt at an angle higher than the critical angle controlled by the velocity contrast between salt and sediment, energy from subsalt steep dips is refracted downward and cannot reach the surface. For a velocity contrast of 2 to 1, the critical angle at the base of salt is only 30°. As a result, any subsalt dips greater than 30° will be missing from surface seismic data. If the shape of salt varies significantly spatially, certain azimuths of surface seismic can undershoot the impeding salt body. In our study area, no such window exists. Generally, subsalt structures that dip opposite to base of salt are more prone to have poor illumination. Unfortunately this is the case for our study area. In contrast, 3D VSP is able to image all the subsalt steep dips , as shown in Figure 5b. VSP downhole receivers record direct arrivals from steeply dipping reflections as well as refracted energy from the base of salt. Neither the direct arrival nor the refracted energy can be recorded by surface receivers. However, VSP illumination from a single borehole is limited to flat reflectors near the well location and steep dips in the adjacent area. Consequently, neither surface nor VSP image can be used by itself for a complete subsalt interpretation. Therefore, a combination of surface seismic which fully images four-way closure and 3D VSP which illuminates subsalt steep dips is desired, as shown in Figure 5c. Finally, migration amplitude was extracted from a subsalt horizon for surface seismic WAZ (Figure 6a), FAZ (Figure 6b) and 3D VSP (Figure 6c) with strong amplitude in red and weak amplitude in blue. FAZ image show more homogeneous illumination than WAZ image. Consistent with RTM images in Figure 5, the migration amplitude map confirms illumination of all surface seismic is poor at the three-way closure (red ovals in Figure 6a and 6b). 3D VSP image (red oval in Figure 6c) shows improved illumination of the three-way closure. Amplitude map again confirms that VSP illumination is localized to area near the borehole and steep dips in the adjacent area. Conclusions To summarize, we have presented a 3D acoustic wave equation modeling study for imaging subsalt steep dips (faults and three-way closure). Results show surface seismic cannot image subsalt steep dips due to poor illumination resulting from energy refracted at the base of salt. On the other hand, a 3D VSP manages to image subsalt steep dips by placing receivers below the allochthonous salt to record both the reflected and refracted energy. The success of 3D VSP in modeling subsalt steep dips demonstrates its effectiveness for specific “target-oriented” imaging objectives. A 3D VSP can be 73rd EA GE Conference & Exh ibition incorporating SPE EUROPEC 2011 Vienna, Austria, 23-26 May 2011
- 3. used as a good complement to surface seismic acquisition. However, due to its localized illumination, VSP modeling study is needed to understand the objectives and to design an effective VSP acquisition program for the intended subsalt targets. Acknowledgements We thank CGGVeritas for the permission to publish this work. References Burch T., Hornby B., Sugianto H., Nolte B., Subsalt 3D VSP imaging at Deimos Field in the deep water Gulf of Mexico: 79th Annual International Meeting, SEG, Expanded Abstracts, 4139-4142. Corcoran, C., Perkins, C., Lee, D., Cattermole, P., Cook, R., Moldoveanu, N., A wide-azimuth streamer acquisition pilot project in the Gulf of Mexico: The Leading Edge, 26, 469-468. Hornby, B., J. A. Sharp, J. Farrelly, S. Hall, and H. Sugianto, 2007, 3D VSP in the deep water Gulf of Mexico fills in sub-salt `shadow zone': First Break, 25, 83–88. Howard, M., Marine seismic surveys with enhanced azimuth coverage: Lessons in survey design and acquisition: The Leading Edge, 26, 480-493. Liao, Q., Ramos, D., Cai, M., Ortigosa, F., Subsalt illumination study through seismic modeling: Subsalt imaging workshop 2009, EAGE Michell, S., Shoshitaishvili, E., Chergotis, D., Sharp, J., Etgen, J., 2006, Wide azimuth streamer imaging of Mad Dog; Have we solved the Subsalt imaging problem?: 76th Annual International Meeting, SEG, Expanded Abstracts, 2905-2909. (a) (b) Figure 1: NAZ (a) and WAZ (b) image comparison from the Jack discovery. Reflections within the green ovals show the benefits of WAZ in delineating four-way closure. The lack of coherent reflections within the red oval in (b) indicates the image is still poor around potential three-way closure. (a) (b) Figure 2: Profile of velocity (a) and density model (b). Allochthonous salt sheet with varying thickness causes illumination problem for steeply dipping subsalt faults and three-way closure. 73rd EA GE Conference & Exh ibition incorporating SPE EUROPEC 2011 Vienna, Austria, 23-26 May 2011
- 4. (a) (b) (c) Figure 3: Acquisition configurations and study area. Inline and crossline offsets for surface seismic simulation is shown in (a). 3D VSP (b) has a maximum source radial offset of 10km. Our modeling study area (c) is ~40 blocks around Jack discovery in southern part of Walker Ridge OCS area. (a) (b) (c) Figure 4: RTM image comparison of real WR WAZ (a), synthetic WAZ (b) and synthetic FAZ (c). Dim zone on real WR WAZ image has been re-produced on synthetic WAZ image. FAZ still suffers from poor illumination at three-way closure. (a) (b) (c) Figure 5: Image comparision of FAZ surface seismic (a), 3D VSP (b) and combined image (c). 3D VSP illuminates the three-way closure. A combined image with fully imaged four-way closure and well illuminated subsalt steep dips is desired for a complete subsalt interpretation task. (a) (b) (c) Figure 6: Comparions of migratoin amplitude extracted from a subsalt horizon. Illumination of both WAZ (a) and FAZ (b) surface seismic are poor at three-way closure. 3D VSP (c) illuminates the three-way closure and the illumination is localized around wellbore and steep dips at adjacent area. 73rd EA GE Conference & Exh ibition incorporating SPE EUROPEC 2011 Vienna, Austria, 23-26 May 2011