Geoscientist Online

Seismic technology is advancing at a breathtaking pace. Here Rob Butler and Tony Pedley* pause to take a look at how new value can be added to old data – and how the insights gained can be communicated to a broader Earth science audience.

Geoscientist 19.6 June 2009

The Virtual Seismic Atlas (VSA) is generating an independent, community-based internet resource that captures and shares the geological interpretation of seismic data globally. But it can also show how improvements in seismic technology impacts on our understanding of sedimentary basin structure. In this article we look at a topical issue – the imaging of Mesozoic structures on the NW continental margin of Europe. Understanding these structures is important for investigating how the opening of the Atlantic was pre-conditioned by earlier tectonics, a general question for studies of continental break-up. But, as so often, this apparently academic issue has important commercial implications in the exploration for further hydrocarbon resources.

The catch is, the Mesozoic structures lie beneath layers of basalt that absorb and reflect much of the seismic signal. One approach is to use special seismic acquisition techniques such as low-frequency airgun sources and to set sources widely separated from the receivers (so-called “long-offsets”). But a lot of existing data were acquired in the past without these special methods. New surveys are expensive. Can more be made of the existing, heritage data using new processing approaches?

Stacking is a fundamental part of seismic reflection studies. It involves combining the records from large numbers of receivers and from multiple shots to increase the signal-to-noise ratio. But of course reflections are received from all over the place. The trick is to combine data to enhance the amplitude of reflections, rather like focusing a camera lens. Unfocused seismic signals are essentially filtered out, leaving, we hope, a crisp image.

But what happens if we’re getting less signal back from our target horizon than from spurious features? For analogy, imagine a two-way mirror – images from behind the mirror are normally swamped by the reflections from the mirrored surface, unless a light is turned on behind it. The deeper parts of some sedimentary basins are like this, lying beyond the mirror. How can we shed light on these?

Faroes-Shetland

Through the 1980s and 1990s a variety of companies, together with the British Geological Survey, acquired seismic profiles out on the NW UK continental shelf. Between 1995 and 1999 Fugro acquired a total of over 9500km of 2D non-exclusive seismic data across UK Quads 216-219 and 213-210 to the north and north-west of the Shetland Isles. This covers the region between the Faroes-Shetland Escarpment to the west, across the North Shetland Basin, to the North Shetland Platform in the east.

In general, the seismic data from the region processed in the 1990s reveal a Tertiary section that is effectively imaged down to the Palaeocene volcanic units that blanket much of the region. This volcanic sequence consists of basaltic flows which are highly variable in their nature and may be interbedded with volcaniclastic and sedimentary units. Intrusive volcanic units may also occur within the (potentially prospective) host sediments beneath the extrusive basalt flows.

This is where the problems start: we have encountered a two-way mirror. The highly heterogeneous nature and high reflectivity of volcanic sequences permits only weak penetration of acoustic energy. The low frequencies are absorbed, the high frequencies are scattered, and much of energy is refracted. Despite these problems, the original processing of the seismic data revealed tantalising reflections beneath the basalt. There was something down there but the poor imaging meant that only simplistic, localised interpretations could be made. A number of successful wells in the area have helped to increase industry interest in the potential of the region’s sub-basalt hydrocarbon plays and this provided the impetus to attempt to improve the imaging in the underlying section.

Reprocessing

In 1996 the Northern White Zone seismic survey (NWZ96) was acquired. The whole dataset straddles the UK / Faroese median line and consists of 2304km of data acquired using a relatively short 4.8km streamer towed at only 8m depth with a 2850 cubic in source delivering 100 bar m of output. Although the effectiveness of marine seismic sources and streamer lengths have increased greatly in recent years, the acquisition of the NWZ96 data was typical of the time. From the NWZ96 survey a test sub-set was chosen to investigate the potential of modern processing techniques improving the sub-basalt image.

The test reprocessing was carried out by Geokinetics UK Ltd. and was specifically designed to retain the low frequency content of the signal emanating from beneath the basalts, and recorded by the receivers that were arranged further out along the cable from the airgun sources (at so-called “long offsets”). This is not straightforward and requires rather elaborate processing. But as with so much in data processing, the rapid increase in the capabilities of computers has made complex processing strategies commonplace today. (Highlights of the reprocessing technique are described in the box; but the basic philosophy is to place the recorded seismic data in its correct place at depth (a process called “migration”) while maximising the quality of the signal.)

The test reprocessing achieved a major improvement in imaging beneath the basalts that had previously obscured the deeper geology. Following this successful trial on the NWZ96 data the entire Fugro 2D non-exclusive database in the region was re-processed. The surveys had been recorded with broadly similar acquisition parameters, and the processing was carried out using a generic scheme with similar parameters to ensure compatibility between datasets. This re-processed regional seismic dataset now allows a much more detailed interpretation of the sub-basalt section. But to take this further, additional data are needed to model the basin structure.

As part of their multi-client geophysical data library, Fugro Gravity & Magnetic Services (FGMS) have available a merged high-resolution aeromagnetic dataset extending across and beyond the area covered by the 2D seismic. This consists of 278,000 line kilometres of data with a predominant 500m line spacing. These data are particularly useful as a screening tool to identify the extent of the basalts, igneous centres and sedimentary basins, enabling the geometry, depth and magnetic properties of sub-surface rock structures to be mapped. The major features and trends of the area can be observed at a glance on such a magnetic anomaly map.

The magnetic data can however be used to help in interpretation of the reprocessed seismic data. The top of the magnetic basement can be picked on the reprocessed seismic lines when the seismic and magnetic data are integrated. To construct this model, FGMS proprietary Potential Fields Software was used. A magnetic susceptibility value is assigned to each stratigraphic layer, which has previously been interpreted from the seismic data. In the study area the sediments can be regarded as magnetically transparent, their susceptibility being several orders of magnitude lower than either the volcanics or crystalline basement.

Anomaly

The magnetic anomaly generated by the combined seismic/magnetic structural model is compared to the observed magnetic anomaly. The structural model is then “fine tuned” to achieve a best fit with the observed and calculated magnetic anomalies. The magnetic data therefore constrain the seismic travel-time basement pick, which would otherwise be difficult to interpret using the seismic data alone, given the limited well information available in the region. The success of this 2D modelling on selected seismic lines gave the impetus to undertake full 3D inversion of the magnetic data, permitting the generation of a depth to magnetic basement model across the entire region.

The re-processing of the seismic data and the modelling of the magnetic data has produced a dataset which greatly assists evaluation of the sub-basalt section and provides a significantly improved dataset. The improvement in data quality has helped to interpret the thickness of the volcanic units, to image potential hydrocarbon plays in the sub-basalt sediments, and to generate a depth-to-magnetic-basement model across the region. This powerful exploration dataset has been created using modern processing techniques but using “heritage” data, showing the potential improvements possible in older datasets that may often lie fallow. This process has also delineated basin structure, enabling tectonic geologists to model more accurately the structure and subsidence history of the NW European continental margin. The commercial advantage of this is clear. We are now able to sample this precious data archive and retrieve information in ways that are analogous to those sound engineers use to re-master acoustic recordings on shellac discs by long-dead musical geniuses of the past.

Postscript

Apart from the immediate scientific impact of this study, both for advancing hydrocarbon exploration in the frontiers of the European continental margin and for developing better models for continental break-up, this story illustrates the importance of multidisciplinary approaches for developing geological knowledge.

The catch is that as the various technologies advance it is increasingly difficult for individuals to stay abreast of the capabilities, limitations and implications of each component in an investigation, or even to have intelligible conversations with other members of the multidisciplinary team. The solution is better continual education in all sectors of the geosciences. The Virtual Seismic Atlas is designed to provide a small part of this solution, by sharing how seismic technology and the geological interpretation of seismic data improves knowledge of the solid Earth.

John Makin explains the technicalities...

The processing of these data followed general guidelines established for regions where an acoustic blanket such as a basalt flow overlies the target depths. The scheme was developed with special attention to the retention and enhancement of low frequencies, the inclusion of high angle reflections, the attenuation of long-period multiple reflections and random noise, and the accurate analysis of acoustic velocities
As very little high frequency energy penetrates to depth, care was taken to maintain low frequencies. This was achieved by applying band-pass filters that restricted high frequencies and avoiding any processes that could damage the low frequency content

The cable recorded primary reflections out to high offsets where contamination by direct arrival energy became a problem. Two-dimensional filtering of these slow arrivals based on velocity discrimination was required. Travel-time correction of this high offset data was based on a velocity field developed using sophisticated image focussing analysis.

The high reflectivity of the water bottom and the top of the basalt lead to multiples being the dominant form of recorded energy at depth. Modern removal processes such as Surface Related Multiple Elimination and High Resolution Radon demultiple were cascaded in the scheme. Any remaining random or localised remnant multiple noise was attenuated by applying coherency-based techniques.

John Makin (Geokinetics UK Ltd.)

Acknowledgment

All the images reproduced in this article, and more besides, are available for scrutiny on the Virtual Seismic Atlas, courtesy of Fugro. The VSA is freely accessible on-line at: www.seismicatlas.org. The content is available for use (subject to adhering to the published terms and conditions), especially for training and research. As a community resource, it has an open call on new content and, especially, interpretations of existing content. All necessary information is available via the website.

Further information on seismic processing can be gained from a series of specialised texts. One of the most accessible, and readable for non-geophysicists – is 3-D Seismic Interpretation by Mike Bacon, Rob Simm and Terry Redshaw (published in paperback by Cambridge, 2007). Much of the coverage is relevant to 2D seismic too.

* Rob Butler is at the University of Aberdeen and is Director of the Virtual Seismic Atlas; Tony Pedley is a Project Geologist for Fugro in the Europe, Africa and Middle East Region.