Seismic is an enabling technology for hydrocarbon exploration, field development and production monitoring. Traditionally the technology development within this field has been evolutionary, where improvements have been introduced at a slow, but steady rate. Examples are the move from 2-D to 3-D acquisition, which took more than a decade, and the steady increase in streamer spread size on towed seismic vessels.
However, recent trends seem to indicate not only evolution, but also a few revolutions. Broadband acquisition has quickly gained widespread adoption and a move from hydrophone one-component (pressure) to four-component (pressure + u,v,w) data recording seems to just have started. These rapid changes were lead by ocean bottom node (OBN) acquisition, but are currently also being introduced into towed streamers in various forms.
Another potential revolution on the marine seismic side is the adoption of blended acquisition (BA). Unlike on land where blended acquisition already has gained widespread use, marine BA has for various reasons not taken off. To our knowledge, several contractors have performed small scale tests, but no one has yet deployed multi-vessel marine BA commercially.
In this paper we will present data examples from recent BA multi-vessel field trials and show that this technique has the potential to reduce costs and to improve both acquisition efficiency and quality.[Page Break]
What is blended acquisition?
Blended acquisition involves firing more than one source array at a time, accepting that energies from separate sources overlap in the recorded data. This incresed source effort can be translated into either efficiency or quality improvements through increased sampling, multiple azimuth or long offset illumination. The method is applicable for nearly all types of towed marine and OBN/OBC acquisition. Figs.1 and 2 show examples of BA setups that we have tested with towed streamers. In Fig.1, a normal seismic vessel is followed by a source vessel that is shifted 25m sideways in the cross-line direction. Both vessels employ conventional flip/flop source geometry. This setup will improve cross-line sampling from 25m to 12.5m while maintaining the inline fold.
Alternatively, one could of course drop the 25m sideways shift of the source vessel and use the acquisition setup to double the inline fold.
Fig. 2 illustrates a side-by-side setup that allows us to acquire two full fold CMP-swaths simultaneously, effectively doubling acquisition efficiency. This setup can easily be extended to a multi-vessel WAZ survey, where it could be used to significantly increase the fold and thereby improve imaging quality. Our field testing confirmed that vessel positioning and streamer steering was sufficiently accurate to allow this type of surveying. We would also like to mention the potential of using BA in connection with OBN surveys. By deploying multiple source vessels there is significant potential for increased efficiency or improved imaging. The former in particular could prove valuable in demonstrating the value proposition of the full azimuth node technology, which has traditionally been considered to be less cost efficient than other seismic technologies.[Page Break]
De-blending
A key enabling component in blended acquisition is our ability to separate the overlapping data (de-blending). Figs.3 and 4 show examples of a shot gather before and after de-blending. The de-blended images were produced using an iterative algorithm which gradually builds estimates of the individual source contributions.
The algorithm relies on variations in the source firing times around a regular shooting sequence. It then exploits the fact that when the data is sorted to the common-offset domain; each source response can be viewed in one time configuration as random noise and in a different time configuration as coherent signal.
We can use the known time delays to jump between these two configurations, enabling better separation than can be achieved by random noise removal or dip filtering alone.
As can be seen in Figs.3 and 4, the de-blending is almost perfect. This de-blending step is typically done at the very beginning of the processing flow, and once beyond this point, the data is indistinguishable from normal seismic data.[Page Break]
Economy
The economic benefits of BA setups can be illustrated with the following example.
A modern multi streamer seismic vessel will typically use around 30 days to acquire a 1000km2 survey. With a hypothetical day rate of US$ 300K this gives a baseline cost (excluding mobilization) of US$ 9M. A source vessel can approximately be chartered at US$ 100K/day. By carrying out BA, as illustrated in Fig. 2, the survey area could be covered in half the time, with a cost of (100+300)^*15 = 6M, potentially saving 30 per cent of the survey cost. Furthermore, it is easy to envision more than one extra source vessel. With two source vessels in addition to the master seismic vessel, the survey could potentially be acquired in 10 days at a cost of only US$ 5M.
With extra source vessels it is also possible to acquire more data over a longer period to improve imaging. With blended acquisition one can easily double or even triple the effective trace count in a given survey with only a moderate cost increase.[Page Break]
Summary
We have recently performed both acquisition and de-blending of blended data from OBN surveys, WAZ-surveys and various forms of long offset towed streamer surveys, all with excellent results. From our perspective as a seismic contractor, marine BA has now matured to the point where we no longer need small scale tests. Instead, BA is ready to be deployed large scale commercially.
It now only remains for a client to take advantage of this opportunity and (to the best of our knowledge) be the first user of full-scale commercial marine multi-vessel blended acquisition.
Enter √ at www.engineerlive.com/ihss
Thomas Elboth, Margherita Maraschini and Richard Dyer are with Fugro Seismic Services in Oslo, Norway and Swanley, UK. www.fugro.com. All the authors work on R&D related to acquisition and processing of seismic data.
