New SEG chief: 'Virtual reality' hones upstream-industry skill

How does an industry cutting staffs and budgets hone the geotechnical capabilities it needs to solve increasingly complex problems of the subsurface?

"Learn from virtual reality," suggests William L. (Bill) Abriel, incoming president of the Society of Exploration Geophysicists. His reference is to a series of subsurface models and synthetic data sets developed through cooperative, public research by the SEG Advanced Modeling Corp. (SEAM).

But how does virtual reality apply to real work?

"Imagine you could learn from a reservoir similar to your field-but one that has all the data you ever hoped for and has absolutely no uncertainty," explains Abriel, who worked 36 years as a geophysicist with Chevron Corp. until retiring in May 2015. He helped found SEAM 19 years ago and remains a director. "You could test work flows and procedures for accuracy and efficiency and then use those best practices on your property."

SEAM addresses what Abriel sees as a problem for the upstream oil and gas industry: "Our knowledge management for complex subsurface projects has to improve. A single, smart new computer program is not going to be a solution."

Using SEAM, geophysicists, geologists, and engineers can test theories and transfer what they learn to real subsurface environments. And professionals of all experience levels can use the program for training.

"In the newly restructured industry we need to have valuable crew members," says Abriel, now consulting as principal of Orinda Geophysical in Orinda, Calif. "But I think the question is not 'how many' but 'what type?'"

Modern upstream workers must possess "very deep technology roots" in one of the core disciplines as well as the ability to work in integrated teams.

"You've got to have depth, or you can't contribute to the team," he says. "You have to contribute to the team, or depth doesn't count."

To become a top integrated production geophysicist requires perhaps 20 years of work on a variety of subsurface problems with elusive solutions, Abriel says.

"So how do we get people to that position as rapidly and as cheaply as possible?" he asks. "More time in the classroom probably isn't the answer. Working project after project after project so you've collected a lifetime of experience is inefficient, expensive, and risky.

"How do we enable it? The answer is to employ an efficient virtual simulator," analogous to simulators in the aviation industry.

"Flight simulators are very effective tools for rapidly getting experience in complex situations, where the consequences of making a wrong choice are not damaging or expensive," Abriel says.

SEAM development

SEAM projects, in order of their development, cover subsalt imaging in Tertiary basins, core challenges in land seismic, pressure prediction, and "life of field" work flows integrating geology, engineering, and geophysics (see box). The pressure-prediction project focuses on predrill pressure and hazard prediction with surface seismic and electromagnetic data. An offshoot project examines the evolution of pore pressure during production and uses time-lapse (4D) seismic data and reservoir characterization.

Abriel summarizes potential application of SEAM's products by describing the pore-pressure extension project in scientific terms: "We have done the forward problem (generating synthetic data from a managed reservoir) so that people can now do the inverse problem (modeling reservoir dynamics from simulated field data). We can provide the blind tests for people to take well data and seismic data and try to manage a dynamic reservoir."

Thirty-five oil and service companies have participated in SEAM, which has receives about half its funding from the US government, including the Research Partnership to Secure Energy for America (RPSEA) and National Energy Technology Laboratory. Government involvement accelerated during SEAM's first project.

"We got a certain distance in the subsalt project, and RPSEA came in and said, 'That's great. What more can you do?'" Abriel explains.

SEG patterned SEAM after the Subsalt Multiple Attenuation and Reduction Team (SMAART), a joint venture founded by Abriel and formed in 1998 by BP, BHP Billiton, Chevron, Mobil, and Texaco that operated until 2002.

Abriel says the SMAART JV received credit from Delft University for advancing surface-related multiple elimination (SRME), a standard method of mitigating multiples, or reverberations, in seismic data. The SMAART JV also received the SEG Distinguished Achievement award in 2008 for its work on wide-azimuth and multiazimuth data acquisition, essential in deep water and for complex geology.

Unlike SEAM, SMAART was a private cooperative. But it demonstrated the power of cooperative research among competitors in the oil and gas industry.

"And a fundamental vehicle that enabled results was the power of large-scale, numerical models that represented key challenges," Abriel says.

Appreciation for models

An appreciation for models comes easily to the new SEG president, who used modeling extensively while part of a Chevron team active in early-day application of 3D seismic methods offshore. The team experimented with methods now routine, such as circular and two-boat surveys, ocean-bottom acquisition, and time-lapse (4D) recording.

Led by Robert Wright, Chevron's first development geophysics manager whom Abriel considers a mentor, the team recommended redevelopment rather than divestment of Bay Marchand oil field in the Gulf of Mexico-a 1949 discovery then well into decline. The team acquired, processed, and interpreted 3D seismic data over the field-a risky investment for a mature asset in the mid-1980s-and applied results to development decisions that reversed the production decline (OGJ, Nov. 4, 1991, p. 50).

"Production increased 300%," Abriel recalls, noting the field remains under development.

Abriel then returned from a research-lab assignment to the Gulf of Mexico-thought to have few remaining exploratory targets and disparaged as the "dead sea." The applied research focus, subsalt imaging, became crucial to exploration and development in deep water and opened huge volumes of the subsurface to exploration around the world. The essence of subsalt seismic imaging is model-based processing and analysis.

In that work, and later as geophysical lead on projects elsewhere in North America and in China, Australia, South America, Kazakhstan, Turkey, and Russia, Abriel continued to use models.

Why not have that?

To Abriel, the combined potential of virtual reality and cooperative research should be irresistible to an industry pressured by a slump in commodity prices.

His vision is clear.

"You can conceive of somebody just coming into the business and wanting to be a world-class interpreter. They want to work on carbonates, subbasalt, arctic permafrost, compressional environments, extensional environments, all of the geological environments that are difficult," he says.

"Now picture a library where you check in and check out the training data, say, for carbonates. It comes with instructions that says, 'Welcome to the training library of interpretation. Your job is to interpret this data set. Here are the rules. Here are the basic data parts. Here is the recommended work flow. You don't have to follow this work flow, and if you're creative and you win, congratulations. But if you can't do the recommended work flow then you may have issues becoming a professional interpreter.

"And your reaction would be, 'That's really interesting. I wonder what else I can learn about carbonate interpretation.' Well, here is a bibliography of all the other published examples that relate to this particular type of carbonate. Also, here are other people's answers to this data work flow you just completed.

"It's an open library."

And the "library" is accessible not only by geophysicists but also by geologists and engineers.

"If you have a multitrillion-dollar subsurface management business like oil and gas, why would you not have that?" Abriel asks. "With the proper cooperation, the industry can do this."

Geophysical advance

The new SEG president, who holds a BS in geosciences and MS in geophysics from Pennsylvania State University, won't predict the next major breakthrough in geophysics, saying he can't see beyond about the next 5 years.

Breakthroughs of the past, he says, include digital processing, 3D imaging, and integration of geology with engineering. But the next one is unpredictable.

"If I knew what the next big thing is in geophysics," Abriel says, "I'd make a lot of money."

But he does see an area needing improvement, and it relates-inevitably-to models.

"We don't actually go out every day and dig," he notes. "We sit in our offices and manipulate projects from afar, so we plan and execute with a model. Maps, spreadsheets, and conceptual thinking are definitely models."

Models progress from the conceptual to the illustrative to the functional, Abriel explains.

"Let's assume we've got a project. In the beginning, I've got a mental image of the subsurface, and you've got a mental image. I don't know if my model is the same as yours. I can't tell what's in your head without us illustrating those conceptual models.

"Shockingly, that communicative step is not well handled in our industry. It's not that easy to adequately draw your conception of the subsurface and describe the uncertainties-especially with limited data and a large number of unknowns."

Abriel also sees a problem with operational, numerical models of the subsurface and reservoir: "Once a team gets a complex model built, they are very reluctant to make major changes or rebuild. It's just too time-consuming."

Overall, improvements to integrated subsurface management are crucial to expansion of what Abriel describes as the "social contributions" of his profession.

"Geophysicists are oil finders. We will still do that well into the future," he says.

"But we will also have a greater role in integrated subsurface dynamic management-ground water, CO2 sequestration, gas storage, oil and gas extraction, hydraulic fracturing, and induced seismicity: We're going to be there."

SEG ADVANCED MODELING CORP. (SEAM) PROJECTS

Phase 1: Subsalt Imaging in Tertiary Basins

• Began in March 2007

• Developed numerical earth model representative of a 60-block area of the deepwater Gulf of Mexico

• Simulated 65,000 shot records containing as many as 450,000 traces each for an acoustic data set

• Developed and adopted a data-compression scheme

• Developed multiple classic data subsets for research

• Initiated complementary geophysical simulations (controlled-source electromagnetic, gravity, magnetic, etc.) over the earth model

• Simulated anisotropy and enabled commercial elastic modeling

• Established process for efficient storage and distribution of data and classic data sets

Phase 2: Land Seismic Challenges

• Simulated geophysical data sets for three land challenges:
Unconventional shale plays (including fractures)
Arid karst near surface geology
Foothills overthrust with near-surface topography

Pressure Prediction

• Began in late-2014

• Evaluating and advancing methods for predrill pressure and hazard prediction

• Focuses on deepwater Gulf of Mexico, but findings will be more broadly relevant

Pore Pressure Extension

• Began in fourth-quarter 2015

• Focuses on understanding evolution of pressure, fluids, and geomechanics during production

• Uses time-lapse seismic data, reservoir characterization, reservoir simulation, and geomechanical simulation

Life of Field

• Began late in 2015

• Integrates work flows involving geology, engineering, geomechanics, and geophysics

• Industrial-scale synthetic data set will benchmark full work process for interpretation accuracy and efficiency

• Data can be used to estimate the inverse problem of interpreting reservoir dynamics with only well and geophysical data, yet everything about the underlying reservoir is fully controlled with no uncertainties