|VOLVE - HOW CLOSE COMMUNICATION BETWEEN THE ASSET TEAM AND RESEARCH WAS AN ENABLER FOR FIELD DEVELOPMENT|
|Field Developments and Case Studies|
From the early days of SUMIC (SUbsea seisMIC), by Berg et al (1994), to the development of fibre optic ocean bottom seismic (OBS) solutions for permanent reservoir monitoring (Thompson et al, 2007), continued and consistent development of new technologies for ocean bottom seismic have long generated value for Statoil where these technologies have been applied to real business settings.
In 1997, Statoil acquired a 3D OBC dataset at Statfjord that led to significant improvements in fault definition and a better resolution of small scale structures (Rognø et al, 1999, Osmundsen et al, 2002). This was mainly attributed to the full azimuth, high fold, and rich offset distribution, which result in better illumination (Thompson et al, 2002). In addition, combining geophone and hydrophone measurements led to a better suppression of water-layer related multiples. Further studies into the role of azimuth using advanced depth imaging techniques (Arntsen and Thompson, 2003), whereby both conventional 3D marine seismic and 3D OBS data were compared, corroborated the importance of azimuth and fold after the survey was acquired. Through extensive practical experience with the Statfjord survey, Statoil soon fully realized and took advantage of the full azimuth acquisition solution offered by 3D OBS. Since 1997, OBS surveys have been carried out over a large number of Statoil’s geologically complex
One of the fields that have seen the value of continued and consistent development of technology for OBS is the Volve field.
The Volve field is located in the southern part of the Viking Graben in the gas/condensate rich Sleipner area approximately 8km east to the Sleipner B platform. It is a small field with a dome-shaped structure formed by the collapse of adjacent salt ridges during the Jurassic. The structure is bounded to the north, east, and south by faults that were mainly formed by salt tectonics. The western faults are more influenced by regional extension. The reservoir in the Middle Jurassic Hugin sandstone formation is a structural trap with four-way closure (see Fig. 1). Its thickness can vary substantially over short distances from circa 20m on the crest up to 100m on the flanks. Oil was found in a well in 1993 and appraisal wells have been drilled in 1997 and 1998.
Fig. 1: A representative geological cross-section through the Volve structure
In 1998 three marine streamer datasets, previously acquired around the Volve structure to map the reservoir, were merged and processed with prestack time migration. The imaging results below the base of the cretaceous unit (BCU) were in general poor. Neither the top nor the base of the reservoir was clearly visible due to the ambiguity in the seismic data.
Due to the geological complexity below the BCU, an agreed upon interpretation of the reservoir was not possible. This was leading to large uncertainties in the calculated reservoir volume and its distribution. As a main consequence, the planning for field development was postponed.
|OCEAN BOTTOM SEISMIC SURVEY|
With experience from other 3D OBC surveys a survey planning project was initiated to find out if this technology would address the imaging problems encountered. In 2002 an OBC survey was acquired using a parallel geometry with a receiver spread covering approximately 27 km2 (Fig. 2). Six swaths were shot using two receiver cables of 6km length with 400m separation. Dual source flip-flop shooting gave a 50-by-50m shot separation.
Fig. 2: The OBC cables are marked with blue lines. The background shows the top Hugin interpretation of the 3D OBC after time-processing in 2002.
|IMPROVEMENTS IN RESERVOIR IMAGING|
In 2002 the processing of ocean bottom seismic was still in its infancy. The only commercial imaging option, which was widely available for ocean bottom seismic, was pre-stack time migration (PSTM). This in itself was a fairly new addition to the imaging tool box, which had previously relied upon common conversion point (CCP) binning and converted wave dip move-out (PSDMO) for converted wave imaging.
Initially a pre-stack time imaging sequence was applied to the data, and first results were delivered at the end of 2002. As in the Statfjord case, the P-wave volume showed a significant uplift of the data quality. It confirmed existing interpretations of the eastern flank and led to a consensus about the base Hugin interpretation in the eastern part. The structural interpretation of the west flank was still a challenge, even though there were clear indications of rotated faults blocks in the western flank.
However, there was still discussion about the top Hugin interpretation despite the uplift in the seismic, which left the reservoir thickness unresolved.
|SIMULTANEOUS PP AND PS DEPTH IMAGING|
At the time Statoil R&D initiated an imaging project that aimed to use simultaneous P-wave and PS-wave pre-stack depth imaging (PSDM) technologies to improve imaging even further, building upon the success of previous imaging work (Sollid and Ettrich, 1999, Ettrich et al, 2000). A layer-stripping tomographic approach was used to estimate vertical P- and S-velocities and the anisotropy parameters e and d. Sonic log and check shot data from five wells were used to constrain the anisotropy (Szydlik et al, 2007).
The depth imaging generated uplift to the quality of the PP data. Most significantly, this led finally to a consensus about the top and base Hugin interpretation in the eastern part. This reduced the uncertainty in calculated volumes leading to the delivery of a plan for development and operation (PDO) in February 2005. The PS data was significantly better than the time-migrated PS data. The PP and PS depth imaged data correlated well above the BCU. However, the interpretability of the PS data below the BCU was still challenging.
After the depth imaging, several problems remained: Significant uncertainties in the west flank, imaging problems in the south part and thus uncertainty about the reservoir thickness and distribution in the south, and the quality of the PS data below the BCU.
A following-up R&D project, using the previous workflow, was initiated with the goal to further improve the joint depth model: One layer would be added at the very top to accommodate a high-velocity channel and thus to avoid a static solution and a thick layer in the middle was divided into two layers (see Fig. 3). Additionally, the anisotropy was allowed to vary laterally in the layers with the strongest anisotropy.
Fig. 3: Depth velocities after the 2008 processing with additional layers. a. The P-velocity at an arbitrary inline, with visible high velocity channel in the shallow part. b. Shear velocities at the same inline. The channel is not visible due to low contrast. c. The e values show clear lateral variation in several of the layers in the upper half.
The final imaging results were delivered early 2008 and proved to be an uplift compared to the previous results. In particular, the PP images were less noisy and showed more continuous events, which changed the fault interpretation (compare Fig. 4a and b), especially in the east flank.
In the east flank, normal faults have a down throw towards the east as a result of tilting of the east flank. Originally, the fault offsets were interpreted to be relatively small and thus having no adverse effect on the fluid flow between segments. The newly imaged data questioned this due to the emergence of a much larger offset, which would potentially restrict the fluid flow to the planned producer. In addition, the planned path penetrated only two fault blocks. In summary, the original well planning would not result in a proper drainage strategy. Thus, well path was updated and extended to a third segment. In this way, it would not only increase the chances to drain the reservoir better, but help to understand the communication between the two fault blocks after drilling.
The 2008 processing improved the PS images enough to allow some limited interpretation below the BCU. A detachment fault and some intra reservoir events are better visible, and the top salt is better delineated than before. In the sum, the PS images can now be used to support the interpretation of the PP images within the reservoir.
Fig. 4: The fault interpretation within the reservoir. Displayed is the top Hugin interpretation, a. based on the depth processing from 2004, and b. based on the depth processing in 2008. Note how the fault patterns in the west flank and the north east are changing.
The Volve field is a good example how a continued and consistent development of new imaging technologies for ocean bottom seismic has generated value. The quality of the original marine seismic surveys did not allow imaging of the geologically complex reservoir at Volve. Only after a full azimuth 3D OBC was acquired and processed using new PSDM technologies, was planning for the field development possible. Renewed efforts to improve the imaging allowed for improved well planning decisions. The factors that were essential for such a fast implementation of research results are close communication between the asset and R&D teams, and the close collaboration with the contractor to get access to and pilot new technologies.
The authors would like to thank Statoil and the partners ExxonMobil and Bayerngas for permission to publish the paper. The authors would like to thank Dave Underwood, who was responsible for the latest reprocessing, and many others, Theresa Szydlik, Einar Magerøy, and Lars Aamodt to name a few, that contributed to this work.
Arntsen, B. and M. Thompson, 2003, The importance of wide azimuth in imaging: 65th Annual International Meeting, EAGE, Expanded Abstracts, A40.
Berg, E., Svenning, B., and Martin, J., 1994, SUMIC: Multicomponent sea-bottom seismic surveying in the North Sea: Data interpretation and applications: Presented at the 64th Ann. Intl. Mtg., Soc. Expl. Geophys., Expanded Abstracts, 477-480.
Ettrich, N.,A. Kristensen, L. Amundsen, L. Claesson, F. Audebert, A. Herrenschmidt, and P.-Y. Granger, 2000, PS Pre-Stack Depth Migration case study of the 3-D 4-C OBC survey over the Statfjord field: 70th Annual International Meeting, SEG, Expanded Abstracts, 1154-1157.
Osmundsen, I.K., E. Magerøy, and R. Sørbel, 2002, Multi-Azimuth PZ data, a step forward in fault definition on Statfjord: Presented at the FORCE Seminar “4C/OBS Data Processing and Interpretation - Old Myths, New Insight”, www.force.org.
Rognø, H., A. Kristensen, and L. Amundsen, 1999, The Statfjord 3-D, 4-C OBC survey: The Leading Edge, 18,1301-1305.
Sollid, A. and N. Ettrich, 1999, Coherency Optimization of transversely isotropic velocity models via PP/PS prestack migration: 69th Annual International Meeting, SEG, Expanded Abstracts, 1707-1710.
Szydlik, T., P. Smith, S. Way, L. Aamodt, and C. Friedrich, 2007, 3D PP/PS prestack depth migration on the Volve field: First Break, 25, no. 4, 43-47.
Thompson, M., L. Amundsen, and K. Duffaut, 2002, Acquisition geometry versus 4C image quality. A study of Statfjord, and Gullfaks Sør: 64th Annual International Meeting, EAGE, Expanded Abstracts, F025.
Thompson M., L. Amundsen, H. Nakstad, J. Langhammer, and M. Eriksrud, 2007, Field Trial of a Fibre Optic Ocean Bottom Seismic System: