Use of satellite can speed seismic surveys

The development moves the up- stream oil and gas industry toward its ideal of "real-time" seismic exploration: the ability to process and interpret data while acquisition is under way.

A 20 min demonstration in Cleveland accomplished work that might normally have taken 1 year or more to complete. It moved a seismic dataset from one place to another via satellite in just a few seconds, then processed and shared it with simulated joint venture partners in just a few minutes.

The transfer mechanism was an experimental National Aeronautics & Space Administration (NASA) satellite known as the Advanced Communications Technology Satellite (ACTS), using networking technology known as asynchronous transfer mode (ATM).

ACTS, among other things, is testing a high data rate (HDR) antenna able to relay digital information at rates as high as 622 megabits/sec. Conventional communications satellites operate at about 1 megabits/sec.

Operated by NASA's Lewis Research Center in Cleveland, ACTS is the first communications satellite that can transfer data at rates equal to those of fiber-optic networks on land-so-called terrestrial systems.

ACTS thus can link sites on point-to-point fiber optic networks with remote locations not on terrestrial networks without the bottlenecking that occurs with current transfer methods. ATM technology, a leading developer of which is Amoco Corp., helps by providing for continuous flow of data through changing media, such as space, fiber-optic cable, and wire.

Demonstration overview

In the Cleveland demonstration on Sept. 13, seismic data moved across ACTS from Colorado to a Houston processing center linked via terrestrial systems to separate oil company offices.

Oil company interpreters reviewed 2D depth-migrated data on interactive workstations and discussed ways to improve the survey design based on results. The data exchange and decision-making easily could have occurred during the course of a typical, 60-90 day marine seismic survey.

In normal practice, survey partners often do not view migrated seismic data until acquisition is complete. It can take a year or more to move magnetic tapes from remote survey sites to processing and interpretation destinations, schedule and complete computer processing, distribute output tapes to joint venture partners, conduct individual processing, schedule and hold meetings, and reach decisions based on the interpreted data.

If results show problems in survey design or parameters, another year may be needed to reshoot and process the new survey.

Delays like that are expensive. Raymond E. Cline Jr., American Petroleum Institute director of information systems, cites industry estimates that the time from acquisition to drilling decision for a modern 3D survey costs $200,000/month.

Based on the number of 3D surveys now shot, Cline estimates that high speed communications and computing might save "tens of millions of dollars per yearif we can make the model work."

Demonstration details

The Cleveland demonstration transferred about 8 gigabits of stacked seismic data from the National Center for Atmospheric Research in Boulder, Colo., to Schlumberger/Geco-Prakla's parallel processing center in Houston.

Data were from the Department of Energy's synthetic seismic dataset, simulating a 3D survey over a salt sill in the Gulf of Mexico. Geco-Prakla completed processing to a 2D depth-migrated section in less than 5 min.

Seismic interpreters from Amoco and Shell Oil Co. in their respective offices in Houston and Chevron Corp. at the demonstration site in Cleveland were interactively linked to the dataset.

They simulated a collaborative interpretation that showed the seismic data to be fairly closely matched to the velocity/depth modelor set of assumptions about sound speed at various depths and lateral locations in the vertical plane below the seismic line. But imaging below the salt sill was poor.

The interpreters concluded that subsalt data would benefit from longer offsets - or distances between source and receiver points.

If a real survey had been under way, they could have ordered the vessel, which would still have been on location, to use longer streamers or different shooting geometry.

Without satellite HDR technology, interpreters probably would not have received processed data in time to reach the decision with the vessel still on location and under contract.

Randall A. Odom, data processing manager for Schlumberger/Geco-Prakla's North and South America region, said the demonstration thus highlighted two broad advantages promised by the new technology: the abilities to adjust survey designs and to confirm or adjust geologic models on which designs are based, both while surveys are in progress.

Snowballing program

The demonstration was the second of three planned in a snowballing development program.

Amoco held the first demonstration in Washington, D.C., last December to showcase ATM technology. It worked with a group of vendors and communications providers in conjunction with a partnership of research laboratories of NASA and DOE in an effort to apply ATM technology to its global corporate network.

In its December demonstration on Capitol Hill, the company used a 1.544 megabit/sec satellite link to transfer data from a fixed platform in the Gulf of Mexico to its nationwide interactive network.

The next month, API and its members agreed to develop a business plan to extend what by then was known as the ATM Research and Industrial Enterprise Study (Aries) into an industry-wide project.

The Cleveland demonstration evolved from progress on the business plan. Its data transfer rate100 times the rate in the December demonstrationis about 20 times the speed at which data can be read off magnetic tape.

David Beering, Amoco program manager and codirector of the Aries project, says important technical capabilities tested in Cleveland were technology for pointing the relay signal and very high data transfer rates.

ACTS achieves its high data transfer rates and another advantage - broad bandwidth - partly through "spot beam" technology in which transmitted energy is concentrated on specific ground receivers rather than whole sections of the earth. Because of this technology, antennas that communicate with ACTS don't have to be as large as those used with conventional satellites, but they must be aimed well.

The next step will be to establish the HDR data link with a moving station. The next ACTS demonstration, planned for next February, will relay data from a moving seismic vessel.

Key steps, says API's Cline, will be stabilizing an HDR antenna aboard ship, developing software to improve throughput on the satellite, processing data at whatever transfer rates prove feasible, and coordinating efforts of the variety of companies, government, and research groups involved in the project.

Beering of Amoco says the ship-satellite link in February will transfer data at a rate slower than the one demonstrated in Cleveland - probably about 3 megabits/sec. He says a data transfer rate of 8 megabits/sec, about tape speed, may prove best.

A goal of future tests will be finding the best balance among the amount of processing done aboard ship, the amount done at the processing center, and speed of transfer.