Optimizing seismic operations on the Atlantic Margin

Pre-stack migration seems to improve AVO reliability

• Areas in the Main Porcupine Basin have been extensively surveyed. [19,936 bytes]
• Pre-stack migration on Atlantic Margin. [97,678 bytes]
• Pre-stack migration on Atlantic Margin. [98,323 bytes]
• An 800-meter wide seismic survey swath allows high productivity. [22,581 bytes]

With the summer seismic shooting window in the challenging Atlantic Margin area off western Europe being so short, it is necessary to establish good overall productivity. The area has a technical need for long streamers for correct imaging of deeper targets and to ensure sufficient offset is available for amplitude versus offset (AVO) analysis. This results in a clear industry need for long streamers (4,500 meters or longer).

Most contractors are able to offer this, but it means more complex operations and longer line changes. Equally, in order to provide good productivity, clients normally prefer at least six streamers per vessel, which impacts mobilization times for those not used to configuring such large spreads, often under adverse conditions. Combined with the issue of timesharing and other environmental downtime factors, the actual time available for shooting out of the short summer season is extremely limited.

How can productivity be improved? In an example from the deepwater Gulf of Mexico, one Geco-Prakla vessel has been operating with an 800-meter wide, 6,000-meter long streamer configuration, averaging over 1,000 square km per month throughout the winter season.

Looking at the statistics, total percentage external downtime is the same for the summer (May-September) in the Atlantic Margin as it is over winter in the US Gulf. Potential productivity should therefore be similar. Two factors combine to limit productivity:

• A higher proportion of mobil izations caused by smaller individual surveys
• Technical downtime as a result of older technical specifications derived from 2D.

Optimizing specifications

New approaches to technical specifications are being reviewed by Geco-Prakla, where the data quality, rather than the instrument status, is used as the main criteria. These developments, (called quantified quality assurance or QQA) demand some understanding of data from the area, and require some pre-survey work to determine the critical parameters and values.

In the meantime, a more areal approach to conventional instrument based specifications helps significantly in ensuring that data quality is not affected while avoiding missing successive weather windows in an attempt to achieve the perfect spread.

The issue of survey size and its impact on overall cost has been highlighted during the last couple of years. This comes about partly because the mobilization time is independent of survey size and therefore constitutes a larger proportion of total time for a small survey, and partly because wide and long spreads require more time to change line and thus the production to line change ratio becomes inefficient.

With such conditions, the reliability of the in-sea equipment becomes very important. Geco-Prakla's experience in the Atlantic Margin has been that as operational experience has improved, mobilizations have shortened and the need to pick up streams for repair. The firm has built experience in both 2D and multi-streamer 3D operations with streamer lengths of 4,600 meters or more, and all the 3D with either 6 or 8 streamers per vessel.

The weather is certainly a challenge, since the pattern along the Atlantic Margin is a set of short good weather windows interspersed by storms. Under such conditions, being able to keep the equipment deployed is much better than retrieving and re-deploying it.

However, as streamer counts rise, so does the total spread. The safety of the crew and the vessel itself becomes an issue. Therefore, a contractor's preference as to whether to recover, or keep deployed and sail for calmer waters should always be respected by the client.

AVO in Rockall

During the summer of 1996, Geco-Prakla acquired 1,543 km of non-exclusive seismic data in and around the Irish Rockall Trough. The majority of the lines formed a grid around the southeastern margin of the Rockall Trough, where old seismic lines shot in the 1970s and early 1980s showed evidence of Mesozoic rift basins separated from the Rockall Trough by a basement ridge.

The new survey confirmed the presence of three possible Mesozoic basins located between the Porcupine Bank and the Rockall Trough, and a smaller half-graben located near the SW corner of Quad 17.

The northernmost of the three basins, which has not previously been described, consists of a complex series of tilted fault blocks. The center basin is called the Bean Basin and contains a sedimentary section in excess of 3 seconds with clear evidence of very strong uplift and erosion. The southern basin is similar to the Bean Basin - its southern limit is not defined by the new survey.

The stratigraphy in the Bean Basin frontier region is unknown but the sediments sub-parallel to the seabed in the first 200 - 300 ms are believed to be Oligocene or Miocene and the underlying strata below the prominent unconformity are possibly of Jurassic and Permo-Triassic age.

There is an obvious faulted anticlinal structure with a possible flat spot below the crest of the anticline at 3.67 seconds. The same line shows other possible flat spots near the boundary fault at the east side of the basin. This section was therefore chosen for further study to determine whether the flat spots exhibited AVO effects.

Data-processing sequence

Geco Prakla's method was to re-process the data preserving amplitudes and maximize resolution in the zone of interest. The processing sequence included radon transform demultiple, amplitude preserving dip-moveout, and common offset pre-stack time migration.

Two benefits of pre-stack time migration are that the traces within a common mid-point (CMP) gather are in their migrated position before AVO analysis and that it can also resolve imaging problems associated with velocity conflicts caused by conflicting dips. AVO analysis was performed using angle stacks and difference plots followed by intercept and gradient analysis.

There is a clear improvement in both the vertical resolution and the fault resolution allowing more confident interpretation. The AVO analysis shows that the flat-lying events near the eastern boundary fault have no AVO anomaly and are thought to be spurious reflections, perhaps from out-of-plane of section.

The event below the crest of the anticline does show an AVO anomaly which, in some conditions, can be associated with hydrocarbon accumulations. The AVO analysis thus provided a useful qualitative tool in this frontier area despite lack of well data or knowledge of the stratigraphy.

Copyright 1997 Oil & Gas Journal. All Rights Reserved.