ÐÓ°ÉÔ­´´

REFLECTION SEISMOLOGY provides geophysicists with a powerful tool for imaging the inside of the Earth. The technique involves determining how artificial shock waves created at the Earth’s surface are reflected from rock layers deep within the crust. At the surface, geophysicists collect data on the reflected shock waves, using special microphones called ‘geophones’ that convert the reflections into electrical signals, and which they store digitally on magnetic tape. Powerful computers then process these signals with highly sophisticated software.

Some 15 years ago, geophysicists within the oil industry developed a seismic technique to produce three-dimensional images of the Earth’s interior. They combined dense seismic surveys of a region with an ability to position accurately details of the reflections they collected, something they could not do with conventional seismic methods. In recent years, the industry has refined the technique by developing seismic workstations. These systems are, in fact, sophisticated computer terminals with two or more colour video screens, like everyday television screens but with at least twice the resolution. We can now call up seismic profiles and merge them with other information, such as data on an oil well, geological maps of oilfields, and knowledge of the regional geology. The workstation also provides a quick way of converting seismic information into a geological map, and a quick and accurate way of seeking out the kind of geological structures that trap hydrocarbons.

The technique of seismic surveying that most oil prospectors use is much the same as that used by geologists studying the deep structures of the Earth, such as the British Institutions Reflection Profile Syndicate, or ‘BIRPS’ (see ‘Sound waves reflect Britain’s deep geology’, 4 February 1988). The difference is in the scale of the work. The oil companies collect massive amounts of data. Even with the price of petroleum and gas relatively low, they collect more than a million kilometres of seismic data a year worldwide. Their targets are small, though, compared with projects such as BIRPS, which aims to delve many tens of kilometres into the Earth.

In practice, industrial geologists are interested in only the first few kilometres of the Earth’s crust. Even quite large accumulations of oil are limited in depth and may be only 1 or 2 kilometres across. The reservoirs occur in porous rocks (usually sandstone or limestone), with the oil held in pores in the rock in much the same way that a sponge holds water. Because the oil is lighter than the water held in the porous rocks, it would rise to the surface if it had not become trapped. One way in which this happens is when the oil comes up against a layer of rock that is not porous, such as often occurs at a fault, or when the rock sequence is capped by an impervious layer of clay . The vertical displacement of a fault may be only a few metres, and if we are to map the structure it must show up on a seismic record. Furthermore, for a discovery to be worth exploiting, we need to know various things: how hot and how porous the rocks are; where the oil was created in the first place; whether it flowed into the trap; and whether it will flow out of the trap again if we drill a well. Such considerations limit exploration to the first few kilometres of the Earth’s crust.

When geologists look for hydrocarbons in a sedimentary basin, such as the North Sea, the first fields that they find are generally the largest ones, the easiest to spot. Geologists started to explore the North Sea with a new vigour in the 1960s following the discovery of the gas fields in the southern region, and subsequently the large oil fields to the north, such as Forties and Brent. As time went by, though, they were forced to seek out smaller structures, which required more detailed seismic coverage of a given area.

In the conventional method of ‘shooting’ seismic data, a series of profiles each provides a two-dimensional cross section, or slice, of the Earth’s crust. Geophysicists display such profiles by plotting a ‘trace’ for every measured interval along the line. These intervals would typically be 25 metres, with each trace appearing as a wiggle, where the size of a particular loop at a certain echo (reflection) time shows the amount of energy reflected within the crust after that length of time. By combining the traces in a continuous display, geophysicists can build up a picture of the rock layers. They can then create a map of the subsurface geology either by combining a number of these profiles crisscrossing a certain area, or by recording them in a grid.

Until a few years ago, this mapping process was a time-consuming, manual procedure. Geophysicists had to interpret the features, or ‘events’ as they call them, transfer the detail to a map, and then draw in the geological faults in the region by hand. To give a picture of the structure, they contoured the information on their map. They then added other geological data to see what chance there might be that a particular trap held oil or gas. Nowadays, geophysicists keep much of the data on computer, and view the profiles directly on colour video screens. From slices to structural pictures

For many purposes, a grid of two-dimensional profiles (‘slices’) does not provide geologists with enough detailed information of subsurface geology. They are likely not only to miss any features smaller than the grid spacing, but also to find that the imaging techniques used in conventional seismic processing are not good enough. The seismic reflections from boundaries in sequences of sediments, which a geologist may want to map are, in the raw form, most difficult to interpret. Each point on a reflector surface, or feature, scatters a spherical wavefront back to the geophones at the surface, and so the section appears to be out of focus. By means of a technique of digital processing called ‘migration’, which is analogous to optical focusing by lenses, geophysicists can now focus and position the reflectors. These techniques are much more effective when the geophysicist has sampled the emerging waves in both horizontal directions, preferably by an array of receivers rather than single lines. Computer systems are ideal for storing the resulting data on the three-dimensional geology, and the geophysicist can display them in a variety of cuts and slices to give a much better insight into the structural picture.

Geophysicists record three-dimensional surveys at sea in a series of closely spaced lines. They tow their detectors (hydrophones in this case) in a cable perhaps 3 kilometres long behind the vessel. The source of the seismic shock waves – usually an array of airguns – is towed close behind the ship. This arrangement constrains the recording of seismic reflections to slices of the Earth below the line of travel of the vessel. Covering a large area (over and around an oilfield, for example, which might be an area of 100 square kilometres) is time consuming. Nowadays, ships tow several cables and source arrays simultaneously, or the geophysicists use more than one vessel. This enables them to record several lines at once. The Nederlands Aardolie Maatschappij (NAM) recorded 12 lines simultaneously in this way during a survey which it undertook in 1988.

On land, geophysicists lay their seismic sources and geophones over a wide area, so that they can record data in three dimensions. But the surveys on land are considerably more difficult than at sea because of the problems created by access and unwanted noise caused, for example, from traffic. Certainly, seismic surveying is more expensive on land, where it is both more time consuming and labour intensive. Sometimes the problems are most delicate. Recently, geophysicists from NAM recorded a survey in Holland, where much of the area they were interested in was covered by large greenhouses. The survey crew recorded their data among the flower beds, using very small explosives charges. They managed not to break a single pane of glass.

The end product of a three-dimensional seismic survey is a vast amount of data. A survey of an area of 10 10 kilometres, for example, with 25 metres between traces, is the equivalent of 400 individual profiles, or a total of 160 000 traces, which when plotted is 400 pieces of paper each about 1 0.5 metres in size. Converting this to a geological image manually could take one person several months. In the early days of three-dimensional seismic surveys, geophysicists realised that if the technique were to become widely accepted, they would have to change the way in which they displayed their data.

Geophysicists have developed a variety of workstations for their use in interpreting seismic data. The workstations are based on small, powerful computers, such as the Vax 3000 series, IBM RT or Sun microcomputers, and include an image processor such as those manufactured by Ramtek or Vicom. Usually, the system includes two high-resolution video screens (1024 by 1280 pixels – a total of 1 310 720; the average home TV set has only 120 000) to display seismic trace data, and on which the geologist can undertake various interpretative work. Often together in consortia, sometimes singly, contractors and oil companies have developed the software for the interpretation and mapping work. In the past two years, such activity has produced one of the fastest growing sectors of the computer industry. The geophysicists store their trace data on magnetic and optical discs, and they can call them up on the video screen to work on.

So now, instead of plotting the data on paper, we hold it on a computer. We can call up seismic profiles on the screens and use them to generate other information, such as maps. We can highlight interesting features, such as folds, faults, or abrupt changes in the strata. But we are not limited to looking at conventional profiles. Because the three-dimensional survey can be considered as a cube of data, we can view it from any direction. Thus, for example, we can display a profile that passes through several oil wells; this does not have to be in a straight line. We often view the data in a horizontal plane, as a series of slices called ‘time slices’. By displaying a sequence of profiles or slices in quick succession, the interpreter can obtain an overall impression of the structure, or the way it developed in the geological past.

The most time-saving aspect of a workstation is the way it reduces the labour of actually ‘picking’ the seismic events and converting them to a geological map. We can use fast computer algorithms to recognise the ‘character’ of individual events, and follow them across the whole survey area, even jumping across faults. If the program gets into difficulties the geophysicists can intervene and undertake a traditional ‘interpretation’. The stored ‘events’ retain the characteristics of the original seismic data, such as the amplitude, indicating the strength of the reflection, so we can map these attributes. Sometimes we can reveal hydrocarbons directly in this way. A reflection from a sandstone layer filled with gas, for example, will show up much more strongly than one filled with water.

Such manipulations can produce startling results. In one survey by Shell Espana in the Mediterranean, Emile Gevers and Gerard Stampfli of the Shell Research Laboratory in Rijswijk, Holland, identified an ‘unconformity’ – a time gap in the deposition of sediments – as dating from the time when some geologists believe the Straits of Gibraltar were closed and the Mediterranean was dry. The land there was apparently quite mountainous, and it is difficult to follow the unconformity on the individual profiles. When water flooded the area, it deposited huge amounts of sediment on the Mediterranean seabed. The coarse sediments give much stronger seismic reflections than the sediments which settled later. The computer mapped the soundwaves reflected from the unconformity, and instead of the confused picture of the profiles, we can see a pattern of meandering river valleys. It looks rather like an Ordnance Survey map, except that it is of nearly 2000 metres below the seabed. It would have been impossible for geologists to unravel the complicated geophysical history of this area and locate potential oil reservoirs without the aid of a seismic workstation.

* * *

SOME OF THE ROCK STRUCTURES THAT CAN TRAP HYDROCARBONS

FOR OIL and gas to be trapped in commercially exploitable quantities, a number of things have to happen. First, the temperature and pressure in the subsurface have to be correct, or the oil will break down chemically. The oil must have moved in sufficient quantities into a location where the rocks are of the right type to provide a ‘reservoir’, and where it is trapped against rocks of other types that will seal the reservoir, so the oil does not move further. The diagrams show some of the possible types of trap.

The larger oilfields, discovered first, tended to be of the anticlinal type, where the structure is more easily spotted, although one of the largest in the North Sea, Brent, is in fact a trap of the unconformity type. Actually, most fields tend to be combinations of various types. As oil explorers try to locate smaller and less well defined fields, such as in the stratigraphic type of traps, they need more detailed and accurate ways of examining the structure beneath the surface. One of these is the use of three-dimensional seismic surveys, which in turn require computerised seismic workstations to interpret the data.

David Davies and Paul Wood are geophysicists working for Shell International in The Hague. Shell International Espana, and Nederlands Aardolie Maatschappij have provided the seismic data in this article.