Marine Four-Component Acquisition Boosts Role Of Shear-Wave Data In Seismic Work

Karl A. Berteussen, R. James Brown
PGS Reservoir
Oslo and Houston

Marine four-component (4C) acquisition has been gaining significant worldwide interest, especially in the North Sea where several 4C surveys have been acquired over the last 2 years. Still in its early days, 4C-or, more generally, multicomponent-marine seismic has already been identified as an important tool for both exploitation and exploration.

PGS Reservoir is one of the pioneers in marine 4C acquisition. In a large number of test surveys and several commercial surveys, its Dragged Array technique has reduced acquisition costs significantly without compromising quality of the data.

Marine S-wave surveys

Geophysicists for a long time have been aware of the potential benefit of acquiring not only one wave mode, compressional waves (P-waves), but several. In particular, shear waves (S-waves) have very promising applications.

The theory of S-waves was worked out in the last century, but practical problems have hindered their application. The shear-wave technique has therefore tended to remain very specialized, used in small land surveys and well seismic.

As a tool in the marine environment it was long considered impractical because of the needs for a marine S-wave source and a method for recording the waves on the seabottom (S-waves do not propagate through water). Some 10 years ago both of these requirements seemed, if not impossible, at least too costly to be seriously considered. Now we have moved from "impossible," via "much too costly," to "in operation" as an important tool in oil and gas exploration and exploitation.

There are several reasons for this. Instead of using a specially designed S-wave source, we have learned that we, to a large extent, can rely on converted S-waves-or waves generated by a P-wave source that convert to S on reflection. As for the receiver side, although modern marine hydrophone streamers still constitute the fastest seismic acquisition technique, geophone acquisition on the seabottom has immensely improved. Because the seabottom usually is relatively flat and free of cultural obstacles, the price per kilometer has been reduced drastically. Furthermore, any measurements that can reduce risk offshore, especially in deep waters, are of premium value.

It is in this context that we have seen multicomponent techniques emerging as key tools for both the exploration manager and the reservoir-evaluation manager.

4C data acquisition

Convinced that 4C was an important new technique, PGS Reservoir developed its own acquisition system. In achieving acceptable data quality, the two key areas in which practical improvements have been possible are in the coupling to the seabed and the speed of repositioning of equipment on the seabed.

The basic data acquisition technique is illustrated in Fig. 1 [76,332 bytes]. An ordinary marine airgun source is used. A big advantage of the converted-wave method for acquiring S-wave data is that it uses cheap conventional P-wave sources, typically airguns in the marine case, explosives or vertical vibrators on land.

The receivers consist of 4C systems: hydrophones and three-component (3C) geophones. The 3C geophones typically are oriented to measure the particle velocity of the seabed in the vertical, inline-horizontal, and crossline-horizontal directions.

Full use of the potential of this type of data requires that the responses of these instruments be properly matched and that the coupling to the seabed be sufficiently high. Fig. 2 [81,273 bytes] illustrates one of the tests used to check instrument response, orientation, and coupling: hodograms, or particle-motion diagrams.

PGS has not adopted the seabed cable approach favored by some geophysical contractors. Its Dragged Array system instead drags the array from one recording leg to the next quickly and efficiently, without any need for assistance from a remotely operated vehicle (ROV) during acquisition.

Coupling to the seabed is achieved with heavy pads housing the sensors. The weight of the pads ensures sufficient contact and coupling with the seabed for faithful recording of both S- and P-waves. The system is designed for single- or dual-vessel operation and has proved practical even at water depths well over 1,000 m.

The 150 km of multiclient 2D-4C surveys conducted by PGS this summer in three areas of the North Sea adopted a two-vessel approach. One of the ships was the company's customized Bergen Surveyor, a dynamically positioned vessel equipped for recording and for dragging the arrays. The Northern Seeker, a conventional vessel, provided the shooting source for the surveys.

In practice, offsets of 4-8 km are routinely sought, depending on the depth of the target, although greater offsets are easily attained in dual-vessel operation. Wide-angle reflections or refractions can be useful in a variety of special situations. These are not always available from towed-streamer arrays.

Why S-waves?

Generally, producing an S-wave section in addition to a P-wave section gives us a second independent image of that section or, in effect, an additional rock parameter.

The parameter can be expressed, for example, as Poisson's ratio, or simply the velocity ratio Vp/Vs (P-wave velocity to S-wave velocity). This usually affords better discrimination of rock properties than P-wave data alone do and gets us closer to images involving things like porosity and pore-saturant type.

For example, it is not possible to accurately identify lithology based on P-wave velocity data alone because many different rock types with different saturants have similar P-wave responses. But if we can combine the Vp data with Vs data, by plotting for example Vp versus Vs (or versus the ratio Vp/Vs) it is often possible to discriminate among rock types (e.g. sandstone, limestone, and dolomite), or between similar lithologies (e.g. sandstones) with differences in various parameters like porosity, pore saturant, and permeability.

Other parameter combinations can also be diagnostic. For example, laboratory data (Fig. 3 [47,750 bytes]) illustrate how Vp/Vs measurements can discriminate between high-porosity sandstones and low-porosity shales and shaly sands.1

Another way of getting at rock and pore properties from combining P and S is through amplitude variation with offset (AVO) responses, which can differ vastly on P- and S-wave sections. Fig. 4a [110,224 bytes] shows a P-wave (hydrophone) section with a reflection at around 2,500 ms zero-offset time. Fig. 4b [110,224 bytes]shows the corresponding P-S (inline) section and the same horizon around 4,500 ms zero-offset time. Whereas the AVO effect on the P-P section is rather featureless, it is quite variable on the P-S section with, among other things, clear phase reversals.

We can benefit in a more general sense from S-wave data combined with P-wave data. Once a gross correlation has been made between two such sections, other features that were not interpretable on one or the other section may be interpretable jointly on the two. This might be specific horizons (Fig. 5), faults, or direct hydrocarbon indicators (DHIs).²

In Fig. 5, [73,238 bytes] the inline (P-S) section has been compressed so as to effect a rough correlation with the vertical (P-P) section, indicated at two horizons by the arrows. But we can see better resolution on the P-S section and, in particular, a third clear reflector, about ¹ sec above the lower reflector on the inline geophone, that would be difficult to pick on the P-P section alone.

Reservoir changes

Changes in reservoir quality can often be detected by combined use of P-wave and S-wave recording. S-wave arrivals may also give useful information about fracture distribution and orientation because S-waves are much more diagnostic than P-waves of anisotropy in rocks. The recognition of seismic anisotropy and its close relationship to aligned cracks or fracture porosity as well as other lithologic causes (e.g. thin layering, grain alignment) has expanded enormously in recent years.

Another application is in imaging reservoirs with a gas-charged overburden, where the varying gas saturation plays havoc with the P-wave image but has much less effect on the S-wave image. Fig. 6 [113,632 bytes] shows the gas-affected vertical-geophone data plus the inline P-S data from a 2D line acquired over the Valhall field in the North Sea (operated by Amoco Norway). The main part of the Valhall reservoir is overlain by a gas chimney, which distorts the conventional image.

Another useful application of 4C seismic is noise suppression. The horizontal components can help us to produce a better P-wave section by, for example, enabling us to discriminate more effectively against undesirable events (e.g. sideswipe, surface waves) and to enhance signals of interest.

The dual-sensor method makes use of a combination of the vertical and hydrophone components to attenuate water-column multiples. And, because P-P sections and P-S sections have relatively different time-to-depth scaling, noise like water-column multiples will obscure the two sections at different geologic depths. This gives us a chance to see a zone of interest on one or the other section.

Complementary images

Real seismic sources produce a variety of wave types within the earth. Land sources generate both P- and S-waves and various surface waves. Marine sources can generate P waves and surface waves but not S-waves, at least not directly.

However, a marine P-wave source (like any other) will give rise to several wave types, including S, through conversions, chiefly upon reflection and refraction at the boundaries between layers-whether fluid or solid. Multicomponent acquisition makes it possible to record the entire vector wavefield, including the S-waves, which register mainly on the horizontal-inline geophone component.

Variations in parameters like pore fluid, pore shape, and grain size and shape can affect P-wave and S-wave propagation differently. Shear-wave propagation through porous rock involves change of shape but not of volume. This is controlled by rigidity, essentially a property of the rock matrix, not the pore fluids.

Compressional-wave propagation involves both change of shape and of volume and is controlled by both rigidity and incompressibility. Although pore fluids have essentially zero rigidity, their incompressibilities may vary widely (for example, gas versus water). The nature of the pore fluids thus plays a more important role in P-wave propagation.

This is essentially why S-wave velocities are usually much less sensitive to variations in the nature of pore saturants than are P-wave velocities, even though the actual interplay of rock properties and seismic responses is much more complex than can be deduced from such simple considerations.

Special processing

Processing converted-wave data has presented an array of challenges.

While ordinary P-waves in a layered section follow symmetric raypaths, P-S raypaths are asymmetric (Figs. 1 and 7), and the P-S reflection or conversion point does not lie halfway between source and receiver. In fact, the position of the conversion point depends both on the depth of the reflector and on the Vp/Vs ratio as it varies down through the section (Fig. 7 [86,683 bytes]).

For a particular layered geology with uniform Vp/Vs ratio, the conversion point approaches a fixed horizontal position for very deep reflectors, the so-called asymptotic conversion point (ACP). But this position still varies from case to case, or with varying Vp/Vs ratio. This velocity-dependent asymmetry has a number of consequences: in normal moveout (NMO) and velocity analysis, in gathering or binning traces for stacking,³ and in dip moveout (DMO) and migration.⁴ This invariably leads to more complex mathematical expressions and thus modified processing routines that are more computationally intensive.

Marine P-wave statics are often of minimal concern and even ignored. But, as a result of the very high Vp/Vs ratios often observed just below the seabottom, S-wave statics can easily be an order of magnitude greater and thus often require a near-surface or static-correction module. For the most part, the correction procedures parallel standard statics routines, with their sometimes good, sometimes not so good results.

Although it is often sufficient in 2D-4C to process just the horizontal-inline component for the P-S section, in general we need vector data involving both horizontal components. This allows us to perform a rotation of the horizontal axes to the desired orientation.

This could be true in a 3D-4C survey, where geophone axes cannot coincide with all source-receiver lines. Or in the event of significant azimuthal anisotropy, we can search the data to determine the lithologic axes (determined, for example, by the orientation of vertical crack sets). This allows us to separate the fast and slow shear waves, an essential step in interpreting the cause of the anisotropy.

The future

Less than 2 years ago, PGS carried out one of the first 2D-4C marine surveys in the North Sea at the Valhall field. The 8 km survey provided Amoco with the first good images of the field through the gas chimney.

When developments warrant it, PGS will move to 3D-4C surveys, although even now there are no insurmountable technical obstacles to this move. Still, the cost and complexity of 3D-4C technology mean that its use will be confined to specially targeted areas, where specific imaging problems require 3D solutions. Full development of the technology will take several years, with issues such as seabed coupling and shooting patterns still to be fully resolved.

Nevertheless, marine 4C seismic, both 2D and 3D, seems destined to become a significant addition to the toolboxes of seismic contractors and a valuable aid in identifying untapped oil and gas reserves worldwide.

Acknowledgment

We wish to thank Amoco Norway and their partners Elf Petroleum Norge A/S, Enterprise Oil Norge Ltd., and Amerada Hess Norge A/S, for permission to reproduce data from Valhall. We are also grateful to Helge Sognnes, Jan Langhammer, Sverre Strandenes, and Walter Sognnes of PGS for their assistance with this article.

References

1. Widmaier, M., Strandenes, S., Drivenes, G., "The feasibility of using P-waves and P-S converted waves for reservoir quality determination", 59th Conference, EAGE, 1997, Extended Abstracts, No. B012.
2. Granholm, P-G., Larsen, D.O., T rudbakken, B., Willersrud, K., "Data-focused exploration in a frontier area, including ocean bottom seismic - A case study from the Nyk High in the V ring Basin, Norway," 58th Conference, EAGE, 1996, Extended Abstracts, No. P162.
3. Eaton, D.W.S., Slotboom, R.T., Stewart, R.R., Lawton, D.C., "Depth-variant converted-wave stacking," 60th Annual International Meeting, SEG, 1990, Expanded Abstracts, pp. 1107-1110.
4. Harrison, M.P., "Processing of P-SV surface-seismic data: Anisotropy analysis, dip moveout, and migration," PhD thesis, 1992, University of Calgary, 229 pp.