Special Report: POINT OF VIEW: New SEG chief helped bring vector seismics into practice

Oct. 2, 2006
A set of techniques known as vector seismics has migrated from research to practice, changing how geophysicists study the subsurface and improving information about the physical characteristics of oil and gas reservoirs.

A set of techniques known as vector seismics has migrated from research to practice, changing how geophysicists study the subsurface and improving information about the physical characteristics of oil and gas reservoirs. Leon Thomsen, who becomes president of the Society of Exploration Geophysicists at the group’s annual meeting this week, has helped this once-esoteric idea become a practical tool of geophysics.

Twenty years ago, Thomsen wrote an article in SEG’s technical journal, Geophysics, that proved to be fundamental to vector seismics and that SEG identified last year as the most frequently cited work in the profession. Since then, the subject has developed to the point that Thomsen, principal geophysicist and senior advisor in the Exploration and Production Technology unit of BP America Inc., now predicts, “In the future, all geophysicists will be vector seismologists.”

With experience in both research and application of vector seismics, Thomsen speaks with authority when he says, “To get new ideas into practice is maybe the hardest part of exploration geophysics.” Yet he holds a view of the role of research in oil and gas companies that’s not wholly in line with pessimism often expressed on the subject.

An appreciation for Thomsen’s perspective requires an understanding of how vector seismics is influencing geophysical work and how the concepts have evolved.

Reshaping geophysics

Vector seismics advances a well-established trend in which seismic techniques provide information not only about locations of boundaries between subsurface rock layers but also about physical characteristics of the rocks themselves, such as lithology, fluids, fracture content, and pore pressure.

The general trend progresses as geophysicists learn to extract more and more information from seismic reflection data. Vector seismics treats a seismic wave as a propagating package of particle displacements, with both magnitude and direction, instead of as a simple pressure pulse.

Conventional seismic work focuses on the compressional mode of the seismic wavefield, or “P-waves,” in which material vibrates in a direction parallel to the direction of propagation. Until recently, exploration geophysicists have paid much less attention to shear waves, in which material vibration is perpendicular to the direction of propagation, and to converted waves, which are shear waves generated upon reflection of a downward-moving P-wave.

Conventional P-wave analysis is scalar, depicting reflection time, location, and signal strength without indicating the vector direction of particle motion within the wave. Scalar P-wave analyses therefore omit important information about the wavefield-and thus about subsurface features.

Thomsen’s 1986 article laid the foundation for work in this area by clarifying the concept of anisotropy. The traditional assumption, he explains, had been that subsurface media are isotropic for sonic propagation, meaning that the velocity at which sound energy moves through them is the same regardless of the direction of travel. Yet velocity fundamentally is a vector (with both magnitude and direction) rather than a scalar, like pressure (with magnitude only).

“Casual inspection of any sedimentary outcrop reveals a layered sequence, indicating that the velocity must be different in different directions: that is, that velocity is anisotropic,” Thomsen says. On closer inspection, “one usually finds joints and cracks, further indicating that the velocity must vary in all directions.”

The crucial point: “Seismic waves propagate as vectors. If we don’t take full advantage of that in our acquisition practice and our theory and our processing and our interpretation, then we’re just missing out on some of the information content that they’re bringing back.”

Vector concepts are both enriching the reservoir information available from interpretation of seismic data and improving established seismic tools. One such tool is amplitude variation with offset (AVO), in which interpreters infer rock type and other reservoir characteristics from observations of how reflection signal strength (amplitude) changes in relation to the distance between seismic energy sources and receivers (offset).

Until recently, most AVO has been based on isotropic assumptions. But the increasing use of wide-azimuth seismic surveys, in which source-receiver offsets cover a wide range of directions, has produced data that challenge the isotropic view.

“We normally find that there is an azimuthal variation with AVO, that the AVO response changes with a change in direction between sources and receivers,” Thomsen says. “As soon as you start acquiring wide-azimuth datasets you see it all the time.”

How it evolved

Thomsen’s work in subsurface anisotropy began 6 years before Geophysics published his article on the subject, shortly after Amoco Corp. hired him out of academia to work at its research center in his hometown, Tulsa. Thomsen’s father Erik had been an interpretive geophyicist with Amoco forerunner Pan American Petroleum Co.

During a visit to Amoco’s Denver office, Thomsen was presented with a question emerging in 2D paper sections from crossing seismic lines. Although reflections aligned as expected, their amplitudes differed. Why?

From previous work with crystals, Thomsen could see a possible answer. Crystals are all anisotropic. Maybe, he thought, anisotropy, contrary to normal assumptions, could be at work in subsurface media, too.

During the next couple of weeks, Thomsen worked out the concept eventually published in his famous 1986 article.

“What we were seeing in these 2D sections was an expression, in stacked 2D data, of what we now call AVO,” he explains. Because the data were stacked, with reflection values from recordings at different offsets combined to improve signal-to-noise ratios, “We didn’t see the angular variation with offset. But we saw how that summed up in the stack.”

The researchers also saw that the effect was different with different offset directions.

“That’s what we now call azimuthal variation of AVO-or AVOAz, a hot topic these days,” Thomsen says. But that was the 2D era; practical use of the finding with P-wave data had to await wide-azimuth 3D data, which is only recently becoming available.

Shortly after this initial work on anisotropy, Thomsen’s Amoco colleague Heloise Lynn provided him a shear-wave dataset in which velocities on crossing lines differed in value.

Thomsen says, “The interpretation was clear: cracks in the subsurface, fractures and joints and cracks of all scale with preferential alignment, which meant not only that P-waves were propagating in different directions but also that vertically propagating shear waves with different polarizations would propagate at different speeds.”

The shear-wave analog to what the Amoco researchers had seen in the P-wave data is now called shear-wave splitting: the dependence of shear-wave velocities on polarization of the wave as well as on the direction of propagation. Because interpretation didn’t depend, as it does with P-waves, on availability of wide-azimuth 3D data, early research in anisotropy focused on shear waves. But shear-waves attenuate more than P-waves do, so practical application of the shear-wave findings had to wait.

“It took maybe 20 years for the science and practice to mature to where we can do it with P-waves,” Thomsen says. “But now we’re there.”

The payoff? From wide-azimuth data recorded with multicomponent instruments, which detect shear as well as compressional energy, interpreters are increasingly able to describe the intensity and direction of subsurface fractures and thus to assess such key reservoir characteristics as permeability.

In addition, Thomsen points out, what the Amoco researchers had learned about shear waves helped them use converted waves to image through shallow gas accumulations, which obscure deeper reservoirs in many parts of the world. Amoco, which merged with BP in 1999, applied the lesson at Valhall oil and gas field in the Norwegian North Sea.

Science vs. technology

Experience moving vector seismics from concept to application has sharpened Thomsen’s views of putting ideas into practice.

“The hard part is not the science; the hard part is the technology,” he says. The difference? “Science is hard to keep secret, and technology is hard to propagate, even to yourself. The hard part is to get your own colleagues to understand it and to use it every day.”

Thomsen has come to believe strongly in the need for researchers to work in close physical proximity to people applying technology. The communication not only helps professionals in the field learn about new ideas but also boosts creativity of researchers, he says.

“Intelligently applying what you already know how to do is a big challenge,” he adds. “Every company struggles with it.”

But the reward can be great.

“A lot of research is useless. Maybe most of it is useless,” Thomsen says. “However, I believe that the small fraction which turns out to be useful can pay for the whole program and a lot more. But it’s a very inefficient process, whereas when you’re taking up the challenge of applying what you already know, that’s really an efficient process. You can’t lose with that.”

He acknowledges the lament of many observers that oil companies have reduced their commitment to research and concedes that most companies no longer have their own research laboratories. But he says companies to some degree have simply integrated research with operations.

Like many companies, for example, BP has set up disciplinary interest groups that foster learning and communication. And it has designated advisors, of which Thomsen is one, able to use some of their time to pursue research in areas of their choice. Inevitably, the research yields unpredictable benefits.

BP recently commissioned the design, construction (by Fairfield), and deployment of 900 autonomous, four-component recording nodes in the Gulf of Mexico to test the utility of wide-azimuth seismic recording in deep water. The nodes are designed to record continuously while on the seafloor for about a month. The first survey using the new equipment recorded densely spaced shots for subsalt imaging over Atlantis oil field. BP will report results at SEG’s New Orleans meeting.

During the test, the largest earthquake in the gulf’s recorded history occurred just 10 miles from the seabottom array of recording instruments, which proved to be sensitive to the low-frequency data used in earthquake seismology. By luck of timing, the active shooting didn’t interfere with the earthquake.

“It’s an extremely interesting data- set,” Thomsen says, and will be discussed at an SEG workshop at the annual meeting. “The importance of it has yet to be demonstrated. Understanding it might be really important for the safety of our operations in the Gulf of Mexico or maybe not. The experts haven’t yet put their heads together.”

Whatever the practical value of the earthquake data, Thomsen adds, “this dataset will be taught to university students for decades.”

Oil and complexity

Thomsen cites the earthquake workshop as an example of SEG’s role in advancing geophysical knowledge and technology. That know-how will be crucial to a “painful transition” to sustainable energy sources.

“It will be up to geophysicists,” he says, “to help find the oil to buy the time to pave the way for the transition.”

What can be said with confidence about that transition is that simple ideas of the past won’t be enough.

“I think we’ve found all the oil that’s going to be found with simple ideas,” Thomsen says. “It takes more-complex ideas and more-complex practice to find the more difficult oil.”

The challenge of the geophysical profession, then, is “to find the oil that wasn’t findable by the technology of today.”

Career highlights

Click here to enlarge image

Leon Thomsen is principal geophysicist and senior advisor in the Exploration and Production Technology unit of BP America Inc.

Employment

Thomsen joined Amoco Corp., later merged into BP PLC, in 1980, after serving as a member of the faculty of the State University of New York at Binghamton. During his 8 years at the university, he spent a sabbatical at the Australian National University in Canberra and was a visitor at the Goddard Space Flight Center in New York. He earlier worked at the Centre Nationale de la Recherche Scientifique in Paris, at the California Institute of Technology in Pasadena, and at IBM Corp. in San Jose, Calif.

Education

Thomsen holds a BS in geophysics from Cal Tech and a PhD in geophysics from Columbia University.

Affiliations

He has been a member of SEG’s Research Committee since 1987 and was committee chairman in 1998-2000. A 1994 recipient of SEG’s Reginald Fessenden Award, he served as editor of the group’s Geophysical Developments Series of 1994-98. He now serves as a trustee associate of the SEG Foundation and chairs the foundation’s Project Review Committee.

Thomsen served as an SEG distinguished lecturer in 1997 and as an SEG/European Association of Geophysicists and Engineers (EAGE) distinguished instructor in 2002.

He’s an honorary member of the Geophysical Society of Houston and of EAGE. He was appointed a foreign member of the Russian Academy of Natural Sciences and received the group’s Kapista Medal in 2004.