Marine geophones: noise suppression and seafloor coupling

Advances make four-component OBC seismic effective

Lawrence B. Sullivan
Arco International Exploration
Dan Ebrom
Texaco EPTD

Scanning electron micrograph of seabottom sediments at Teal South. Note dominance of platy minerals (clay-minerals, or muds) over sands. This is consistent with the soft water-bottom conditions at Teal South.

Sandbag wrapped around receiver cable. Texaco engineer Robert Cross is verifying the geophone signal polarity prior to deployment [13,614 bytes].
Receiver gathers: hydrophone, east-west horizontal geophone, vertical geophone, north-south horizontal geophone [19,711 bytes].
Seismic data acquisition in shallow water using sensor-equipped sea floor cables has been performed commercially for over five decades. Introduced to extend land-based exploration methods seaward, shallow water ocean bottom cable (OBC) operations initially used waterproof cabling and geophones deployed on the seafloor by hand. Hydrophones rapidly replaced geophones in these operations as surveys progressed into deeper water. This required the use of small boats for cable layout and eliminated the manual coupling of receivers to the seafloor.

Using hydrophones as seismic receivers, bottom cable operations continued to be extended seaward to depths where deepwater seismic vessels towing hydrophone streamers could maneuver safely. In water depths where streamer acquisition techniques were feasible, it was thought that OBC systems would not be competitive with streamers. Suppression of water-column reverberation noise and acquisition of mode-converted shear waves, however, has sparked greater interest in multicomponent-geophone OBC even in deepwater environments.

Conditions, motivation

In both marine and land operations, the direct contact (coupling) of seismic sources and receivers to the sound transmission path is the key to successful seismic exploration. Defective coupling of sources and receivers leads to distortion of the seismic wavelet with data amplitude loss, phase change and frequency reduction.

Conventional marine sources and hydrophones are perfectly coupled to the sound path. Land seismic sources can usually be buried (explosive charges) or pressed down on the surface with weight (vibrators). The receivers, geophones, responding to particle velocity, must be attached to the earth's surface with spikes or other means. Conditions like soft sand, rock, mud, and snow create obvious geophone coupling difficulties.

Receiver signal-to-noise (S/N) ratio is another factor that is crucial to obtaining seismic records. Noise is constantly encountered in seismic surveys whether at sea through wind, waves and vessel-generated sounds or on land where wind, rain, electrical disturbances and cultural activity can overwhelm the recorded data.

Electronic instrument self-noise S/N ratio is initially addressed by careful design and manufacture of receivers and instrumentation systems. Data noise can be improved in field operations by:

Careful receiver deployment
Use of receiver arrays
Acquiring redundant data.

Data processing techniques have been developed to further enhance data S/N ratio, but noise suppression prior to recording is always desirable.

Production decline in aging offshore oil fields and the increasing acceptance of 3D seismic imaging techniques have created a demand for geophysical surveys in waters congested with production rigs and wellheads. These obstructions are extremely hazardous to large vessels towing long streamers or arrays of multiple streamers. Interest has renewed in using stationary receiver spreads on the seafloor in deeper water. This limits the size of towed equipment to relatively compact air gun arrays.

The major drawback to OBC exploration is the loss of important segments in the frequency spectrum of hydrophone data. This is due to reverberation in the water column at depths typically greater than 10 m. Of particular concern is receiver ghosting, the initial reflection of signal from the sea surface. Further, use of multicomponent seismic recording, especially observation of shear waves, requires well-coupled multicomponent geophones rather than acoustic-only hydrophones.

Development, application

In 1989, Barr and Sanders demonstrated the complementary effect of reverberation on the frequency content of data from seafloor-located hydrophones and geophones. In essence, if data were collected independently from co-located OBC hydrophones and geophones, the data from these dual sensors could be corrected and summed in processing and the missing frequency components recovered regardless of depth.

So, ocean bottom cable exploration could be performed at any depth, limited only by recording methods, equipment design and the ability to handle the bottom gear. Subsequent field experience showed that more factors than receiver sensitivity were involved in reconstructing missing frequencies in OBC data, most notably the seafloor composition and the geophone coupling to it.

Three years later, Arco Exploration was asked if dual-sensor bottom cables could be used to collect a 3D seismic survey in the Cook Inlet of Alaska. This hostile body of cold and murky water presented many potential problems from both an operations and instrumentation viewpoint. The inlet has maximum tidal elevation changes of 10 meters and maximum currents in excess of six knots.

Initial tests at the proposed survey site indicated that suitably weighted bottom cables would remain in place on the flat rock seafloor at 35 meters nominal depth through several tidal cycles. The cables were deployed and retrieved with a special mechanical cable handler. However, geophone noise levels in the range of 4,000 microvolts RMS were observed during periods of peak current flow, a value over 130 times higher than the normal limit for allowable geophone noise.

After much deliberation, it was decided that a sandbag wrapped around each geophone station was an attractive method for controlling the noise in a high current condition. It provided maximum geophone stability, maximum seafloor coupling and permitted effective deployment and retrieval.

The design was intended to contain each geophone and its protruding cabling against the bottom cable and to press the assembly firmly against the seafloor. Subsequent field experiments involving a few sandbag designs and several commercially available marine geophones confirmed the sandbag concept for use in the Alaskan 3D program. The Cook Inlet OBC 3D survey was conducted in 1993 and produced high quality seismic images of the subsurface.

Advances

Wrap-around sandbags were used several times thereafter in dual-sensor bottom cable surveys on a sandy seafloor in Indonesian waters with uniformly good results. In all these instances, the data being recorded were produced by compressive seismic waves, P-waves, with geophones oriented vertical to the seabed. In 1994, Berg et al, published a paper describing the reception of seismic shear waves, S-waves, with horizontal seafloor geophones. These S-waves were produced by conversion at the seafloor of the seismic source's P-waves, which were subsequently reflected from subsurface layers in the earth. In 1995 and 1996 wrap-around sandbags were used by Arco on an experimental eight-station four-component (vertical, horizontal-inline, horizontal-crossline geophones and hydrophone) bottom cable in the Java Sea with good results.

Coupling for 3D multi-component seismics: The success of sandbag coupling made this technical approach appealing when Texaco and Input/Output were deciding just how to acquire a 3D multicomponent survey offshore Louisiana (Teal South, Eugene Island 354). This area of the Gulf of Mexico has a seafloor composed of fine muds. An electron micrograph from Eugene Island 354 showed that it was 95% clays, 3% sands, and 2% foraminifera.

This soft bottom raised some concerns. Although mode-converted arrivals (P to S conversions) had been successfully acquired and imaged in the hard seabottoms of the North Sea (Berg et al), it was not obvious that the Gulf of Mexico would be so generous.

The sandbags were wrapped around each of the multicomponent geophone/hydrophone (I/O BCS 4) receiver groups. The 3D survey at Teal South was, to our knowledge, the first 3D/4C commercial seismic survey collected offshore in the Gulf of Mexico. The survey was conducted during a 1-week period at the end of July-August 1997.

Seas were especially mild, with wave heights frequently below 0.3 m (1 foot). The source effort was a 3-km by 3-km square, with source spacing every 25 meters. The source was a small array of airguns towed at 3 meters depth. The survey was conducted by Western Geophysical with the source provided by SeaScan.

The receiver effort consisted of 24 4C receiver groups, deployed along four east-west cables. Each cable contained 6 receiver groups. The groups were spaced 200 meters apart (east-west) and the cables were spaced 400 meters apart (north-south). Each cable was connected to a remote recording buoy of unique design, both of which were provided by Input/Output.

Mode conversions at Teal South: The seismic data were recorded and processed as receiver gathers, as opposed to the common streamer-seismic notion of shot gathers. The receiver gather has four portions to it, corresponding to the outputs of the hydrophone, the east-west horizontal component, the vertical component, and the north-south horizontal component. Clearly evident on the north-south horizontal component are the mode-conversions that the survey hoped to capture. The mode conversions are also present on the east-west component, in part because the boat sail line was slightly to one side of the receiver group.

The pleasant surprise here is that the sandbag technology originally developed for reducing flow-noise (associated with strong currents) is also effective at enhancing the coupling of the horizontal component geophones to the muddy bottom.

Challenges for sandbags: The survey at Teal South was a success in acquiring P-wave and mode-converted S-wave data for imaging the subsurface (Ebrom et al, 1998). The availability of two independent seismic images allows the interpreter to determine additional rock characteristics, such as Vp/Vs and lateral lithology changes. Time-lapse seismic for reservoir monitoring may also benefit by interpreting mode-conversion amplitude changes in terms of pore pressure changes.

Vector fidelity

The goals for multi component data acquisition are greater, however. Geo physicists want not just to capture the P-wave and mode-converted S-wave events, but to exactly describe the motion of elastic waves at the seabottom. This concept is called vector fidelity. This means, in practice, that events, which should appear on just one component, will not leak into any of the other components.

An indication that the Teal South 4C data can be improved is that the mode converted waves are present on the vertical component. These events should be entirely confined to the horizontal component geophones. It is possible that near-surface anisotropy is partially responsible. An experiment to test that is planned for this year. It is also possible that some of the undesirable cross-coupling between the vertical and north-south horizontal components is due to cable torsion.

As it turns out, there may be a data processing solution to the cross-coupling problem. Jim Gaiser has shown that a receiver deconvolution technique can remove much of the mode-converted energy from the vertical component data in cases where there has been cross-coupling. Use of receiver deconvolution, and other acquisition and processing techniques, pursuing vector fidelity, will ultimately lead to more reliable seismic-based tools for reservoir characterization.

Conclusions

Ongoing field experiments, like Teal South, are the "proving grounds" that show not only what is desirable, but also what is possible now. As these technologies are proven in mature regions, we can expect to see increased use worldwide of OBC 4C systems even in very deep waters.

With oil prices at recent historic lows, technology is a key component in maintaining profitability. Multicomponent reflection seismic methods will provide value by optimizing production costs and increasing ultimate recovery of oil and gas.