Oil and gas exploration is becoming increasingly challenging as oil reserves decline and both major oil companies and smaller regional and specialist exploration companies are compelled to develop and test new, cost-effective techniques and approaches.Seismic data processing is an essential step for improving the quality of raw data acquired in the field.
By considering knowledge of the area's geology, appropriate new and specialist techniques are being selected to improve imaging available for interpretation by petroleum geoscientists.
Some techniques such as reverse time migration (RTM) which have been proposed in the past are only now becoming practicable as computing hardware becomes more powerful, although hardware issues still remain.
This industry has, and will, continue to push hardware developments required to process seismic data in a timely manner. It is anticipated that RTM will become a method used in standard production use when hardware limitations are resolved in the future.
New techniques can evolve from either re-examining established methods and techniques and/or applying them in different scenarios.
The importance of compensating for source-receiver azimuth in processing 3D seismic data is also recognised in processes such as multiple attenuation and migration. Of course, this assumes that azimuths are adequately and properly sampled in the first place. The emergence of streamers with dual sensors offer the possibility of decoupling the up and downgoing wavefields when processing data recorded at the surface.
Travel time inversion will be replaced by full waveform inversion, for example, and 4D seismic will be performed in the depth domain.
Spectrum has also found that the reprocessing of data from existing surveys using new techniques, technology, and perhaps more up to date knowledge of a region can provide a powerful and cost effective dataset for reservoir characterisation.
Whether processing onshore, on board and or reprocessing older data to give it a new lease of life, with the right knowledge and understanding there are now techniques available to improve results. Some of these are outlined below.
Sub-basalt imaging
The first licensing round in the Norway-Iceland Jan Mayen agreement area opened on 22January and included bid rounds in the Northern Dreki area of Iceland.
Although little prospect information was known about the area, a well-documented correlation of source and reservoir rocks in adjacent Norwegian and Greenland margins gave promising indications of hydrocarbons.
A similar geological structure in the Icelandic prospects could be reasonably assumed, stimulating interest in the area.
The company reprocessed over 5000km of 2D seismic data that had been shot in Icelandic and Norwegian waters.
By using pre-stack time migration (PSTM) and surface related multiple elimination (SRME), for example, the images could be sharpened up. Geoscientists overseeing the reprocessing decided that these techniques, with the addition of radon combined with bespoke techniques developed in house would achieve better multiple attenuation and improved the imaging of sub-basalt reflectors.
As a result, the team working in the Woking office near London was able to offer an offshore Iceland survey with much more clearly defined Mesozoic geology.
Complex velocity models
In May, a recent 28 000km Gulf of Mexico seismic survey was processed using wave equation migration (WEM) techniques to reveal sub-salt geology.
WEM takes into account the multiple arrivals of seismic waves in the subsurface, overcoming the limitations of standard Kirchhoff PSDM which considers only the first arrival of seismic waves. Compared with normal Kirchhoff migrations, the WEM method increases the focusing of seismic images in areas with complex velocity models and excels in imaging below geology that is traditionally difficult to penetrate with seismic signals, such as salt formations. This dataset clearly illustrates the nature of seismic data processing in practice. There are no 'silver bullets' so it's not so much a case of 'the best method' but 'the right method'. In this case, both the Kirchhoff and WEM processed datasets held value to the petroleum geoscientist, because they helped to reveal different geological features.
However, the workflow was also quite different so it's not a 'push-button' exercise. For WEM, the input consisted of shot gathers.
Using both techniques, the seismic data processing team has been able to produce a dataset that helped the parties involved to learn important information about the sub-salt structure of Eastern Gulf of Mexico.
Impressed with the results from this survey, the team in the Houston office also tested the WEM technique on the company's Big Wave Phase 1 multi client project, also in the Eastern Gulf of Mexico.
The extra work paid off and the team concluded that it showed clear potential to improve the imaging of the Florida escarpment and the base of salt definition.
The main deep (>400m water depth) offshore exploration efforts offshore India have concentrated on the east coast, where a number of large discoveries have recently been made in the Krishna Godvari and other basins.
Attention turned to the deep offshore area of the West Coast of India as the Directorate General of Hydrocarbons (DGH) NELP rounds started. Here, outside the petroliferous (shallow water) province of the Mumbai High, exploration has been frontier in nature with very few wells drilled to date.
Imaging created using seismic techniques such as radon demultiple, PSTM, and PSDM applied to a 12000km regional 2D seismic survey collected for the DGH in 2002 upgraded the petroleum potential of the area - showing plays both in the Tertiary and in the deeper Mesozoic section where a potential petroleum province can now be recognised below the Deccan basalts.
Technology challenges
As a pure play seismic services provider, the company is now conducting research into new techniques and approaches that it feels can make a very valuable contribution the improvement of seismic datasets. Its top priorities include data regularisation, a 'hybrid' velocity model building technique for PSDM, integration of seismic with gravity and magnetic data for velocity model constraints and full waveform inversion. Many of the above techniques hold great potential for improving the imaging of the increasingly deep geological structures that oil exploration companies believe could hydrocarbon potential.
Improving data quality
Data regularisation is very important in seismic processing algorithms, as the application of these assume that the seismic data is regularly sampled. However, seismic data recorded in the field is usually sparsely sampled in source-receiver offset and azimuth. This can introduce serious artefacts to the seismic image in the form of processing-operator induced aliasing effects.
The Woking office is now working on full wavefield techniques and methods to overcome operator-induced aliasing and other artefacts to resolve this problem.
Noise attenuation is another practical consideration for newly acquired and older datasets. Again, it is not a case of 'one size fits all' and research is being done into the use of wavelet transforms, especially for ground roll attenuation and has plans to apply it to other types of noise as well. Going forward, more work is planned on characterising waveforms that are reflected back from the subsurface.
Building velocity models
Today, linear seismic processing methods are common, but a clear need has been indentified to continue development of non-linear inversion methods such as genetic algorithms in order to build velocity models for depth imaging.
This means increasing 'mix and match' techniques (linear and non-linear inversion) and domains (time and depth) to generate methods to solve the problem of depth imaging. This means continuing research into applying genetic algorithms for velocity model building.
A project with KACST in Saudi Arabia involves applying velocity modelling techniques in the tau domain, and combining this with the current approach of applying non-linear inversion methods for complicated areas that can reduce the velocity-depth ambiguity and improving depth imaging.
Of course, exploration is not limited to seismic surveys: the use of gravity and magnetic data to constrain velocity models, another example of hybrid methods for velocity model building, is also being explored.
Geologists in demand
Finally, there remains one last challenge on every oil exploration company and seismic service provider's list - the international skills shortage.
While the current economic climate has tempered concerns, there remains a long term problem. Where is the industry going to find the people to not only write all these algorithms and equations but also effectively process future data?
The oil industry is losing a valuable pool of expertise due to the cyclical nature of its business which needs to be addressed.
These are certainly interesting times in the field of geophysics as new technologies help oil companies explore increasingly challenging frontiers.
- Dr Ron Silva, Group Technical Manager, Spectrum spearheads the company's seismic technology research from Spectrum's headquarters in Woking, UK.
www.spectrumasa.com
The Spectrum group provides an integrated service to the international seismic industry. The group offers onboard and onshore data processing on data collected from all over the world and in a wide variety of geological conditions. It also offers an extensive multi-client library of strategic projects offering global opportunities to acquire data from all corners of the world.
Marine Acquisition is provided by our reliable and cost-effective vessel, the GGS-Atlantic. This recently refurbished vessel has undergone extensive technological upgrades and is equipped with state-of-the-art instrumentation and computing hardware. Offshore acquisition work is undertaken on a proprietary or client basis.
