New Innovations Continue Geophysics' Advance

Talk to geophysicists about innovation and common themes emerge: Increased computing power. Artificial intelligence for data analytics. Full waveform inversion. Enhanced acoustic sensing, with multiple sensors. And much, much more.

“One of the new technologies that’s playing a big role is fiber optics, so we’re capturing a lot more fiber-optic data for reservoir characterization,” said Ali Tura, professor of geophysics and director of the Reservoir Characterization Project at the Colorado School of Mines in Golden, Colo.

“A lot of companies are putting fiber optics into their onshore wells, and you can use this for capturing fracture properties stage by stage, and details of producing zones.” Tura noted. “We’re also using fiber optics offshore for reservoir monitoring and identifying where things are happening.”

Innovations also can be seen in established geophysical tools. And some geophysical technologies and techniques that have been developing for five years or more are finally reaching day-to-day application in the field.

One example of that is compressive sensing, according to Doug Foster, senior research scientist at the University of Texas at Austin’s Institute for Geophysics.

Compressive sensing basically developed as a way to fill out the seismic picture when only sparse signals are available. Today’s applications are based on “fairly well-established theory that was not available in the past,” Foster said.

“It’s been ramping up for five to 10 years and it’s coming on now. It’s a way to capture data with less effort and with better results,” he said. “We’ve talked about randomized sampling for a long time but nobody had figured out the theory.”

Custom-made Innovation

Overall, Tura observed, innovations in geophysics tend to come from addressing specific challenges faced by the industry.

“It really varies based on where you are. Each geographical region has different challenges,” Tura said.

“Offshore you need to remove the multiples in the water layer. If you have complex geology, like salt, you need to improve your velocity models and imaging,” he noted. “If you talk about, for instance, the Middle East, you have completely different challenges.”

Tura said the industry is developing waveform inversion for offshore seismic while pushing the limits of wide-azimuth seismic acquisition and other techniques.

In the Middle East, extensive sand dune topographical irregularities and near-surface complexities make seismic data acquisition, processing and interpretation difficult.

“To address these problems, very high fold and costly seismic acquisition is required, but with that we can do quantitative reservoir characterization,” Tura said.

Image Caption

St. Elias with Malaspina glacier in the foreground. This is one of the locations where researchers at the University of Texas’ Institute for Geophysics conduct field work. Photo by Ken Ridgway.

Please log in to read the full article

Talk to geophysicists about innovation and common themes emerge: Increased computing power. Artificial intelligence for data analytics. Full waveform inversion. Enhanced acoustic sensing, with multiple sensors. And much, much more.

“One of the new technologies that’s playing a big role is fiber optics, so we’re capturing a lot more fiber-optic data for reservoir characterization,” said Ali Tura, professor of geophysics and director of the Reservoir Characterization Project at the Colorado School of Mines in Golden, Colo.

“A lot of companies are putting fiber optics into their onshore wells, and you can use this for capturing fracture properties stage by stage, and details of producing zones.” Tura noted. “We’re also using fiber optics offshore for reservoir monitoring and identifying where things are happening.”

Innovations also can be seen in established geophysical tools. And some geophysical technologies and techniques that have been developing for five years or more are finally reaching day-to-day application in the field.

One example of that is compressive sensing, according to Doug Foster, senior research scientist at the University of Texas at Austin’s Institute for Geophysics.

Compressive sensing basically developed as a way to fill out the seismic picture when only sparse signals are available. Today’s applications are based on “fairly well-established theory that was not available in the past,” Foster said.

“It’s been ramping up for five to 10 years and it’s coming on now. It’s a way to capture data with less effort and with better results,” he said. “We’ve talked about randomized sampling for a long time but nobody had figured out the theory.”

Custom-made Innovation

Overall, Tura observed, innovations in geophysics tend to come from addressing specific challenges faced by the industry.

“It really varies based on where you are. Each geographical region has different challenges,” Tura said.

“Offshore you need to remove the multiples in the water layer. If you have complex geology, like salt, you need to improve your velocity models and imaging,” he noted. “If you talk about, for instance, the Middle East, you have completely different challenges.”

Tura said the industry is developing waveform inversion for offshore seismic while pushing the limits of wide-azimuth seismic acquisition and other techniques.

In the Middle East, extensive sand dune topographical irregularities and near-surface complexities make seismic data acquisition, processing and interpretation difficult.

“To address these problems, very high fold and costly seismic acquisition is required, but with that we can do quantitative reservoir characterization,” Tura said.

“Further, you may be looking at targets below carbonates. And high-velocity carbonates are excellent in generating inter-bed multiples that further complicate reservoir characterization,” he added.

A more general innovation in geophysics has resulted from the advent of artificial intelligence techniques for analyzing seismic data, including machine learning and deep learning.

In machine learning, statistical models and algorithms identify patterns in data and computer systems perform defined tasks, all without specific instructions on how to proceed. Deep learning typically utilizes artificial neural networks to enhance machine learning.

AI is “used now for solving all kinds of geophysics problems. One example is using machine learning for reservoir characterization, or mapping salt bodies, or faults in seismic data,” Tura said.

Offshore Geophysical Techniques

John Goff, also a UTIG senior research scientist in Austin, investigates the morphology of the seafloor, its pattern of acoustic reflectivity and shallow stratigraphy. He said his current work includes studying shallow waters offshore Texas for the U.S. Bureau of Ocean Energy Management’s National Offshore Sand Inventory project.

“Lately we’ve been studying the Trinity River paleovalley off Galveston,” Goff said.

“We also have a field class every year where we teach students how to use these (offshore geophysical) techniques,” the intensive Marine Geology and Geophysics Field Course, he said.

• Multibeam Sonar

Goff said his UTIG project uses multibeam sonar to measure the topography of the seafloor.

“It’s very sophisticated now – now we have arrays of transducers that listen to the sound coming back. We get what we call a swath of bathymetry,” he explained.

“Multibeam keeps really improving. I keep seeing great strides in the quality of the instruments,” Goff said.

• Chirp Sub-bottom Seismic

Chirp seismic, another established but still improving geophysical technique, is “one of our main tools. It’s essentially a very high frequency seismic system. That’s what the oil industry uses to find oil and for other purposes, but we use a much higher frequency,” Goff said.

• Parametric Seismic

In parametric seismic systems, the level-set determines the outline of target-body geometry and evolves during inversion in terms of identified, underlying parameters.

“I’m seeing them used in Europe, and I think they’re coming to the U.S.,” Goff said.

• Mid-frequency Systems

Characterized by their seismic signal sources, both sparker systems and boomer systems have made recent advancements, according to Goff.

“The images I’ve seen lately in the published literature have really improved,” he said.

• Multi-frequency Systems

Goff noted the trade-off between resolution and penetration when using different frequencies of seismic signals. An effective, one-pass, multi-frequency seismic tool would be a major innovation for the industry, he noted.

“That would be my ideal, if you had one tool that used multiple frequencies. That’s my pie-in-the-sky ideal,” he noted.

The chances of that kind of tool being developed?

“Who knows? There’s definitely a market for it,” Goff said.

Techniques for Polar Regions

The Earth’s polar regions are both a new focus of industry interest and a frontier challenge for geophysics. Ginny Catania is a glaciologist and professor in the Jackson School of Geosciences at the University of Texas at Austin.

“A lot of people in different geophysical disciplines are focusing their interest on polar areas now,” she said. “Having people come at this from different points of view is really helpful.”

Geoscientists studying the Arctic and Antarctic regions have to contend with both a scarcity of existing data and obstacles in acquiring new data.

“There’s a lot less data available because it’s more challenging to get data in these places,” Catania noted.

“Right now we have pretty good imaging below the ice in Greenland. We’re still trying to get it for the Antarctic region,” she said.

• GPR: Ground-penetrating Radar

Catania said her projects draw on remote sensing and space imaging as well as multibeam and borehole observations. Ground-penetrating radar, a tool more commonly associated with hydrogeology, is an important part of her work.

“Radar is a really useful tool. You can use radar for all sorts of things. Ice is very transparent to radar energy,” she said.

Other Developments

• DAS: Distributed Acoustic Sensors

Sensors are showing up everywhere, and Foster said distributed sensors now help assess the results and the effectiveness of hydraulic fracturing in unconventional plays.

“It’s really coming on in terms of monitoring the fracturing. They aren’t quite there yet, but it’s really coming on. They claim the wave field can be sampled down to a meter, sampling along the path of the well bore,” he said.

• Borrowed Algorithms

Foster told a story about a newly hired recent graduate who developed a field algorithm for an oil and gas company. Asked how the algorithm was created, the new employee explained it was based on an algorithm used by Netflix.

“There are a lot of things falling into geophysics now,” Foster observed.

• Estimated Rock Volumes

“The unconventional plays have been challenging for geophysicists. Traditional tools aren’t telling the engineers what they want to know, so they’re sort of carpet-bombing it,” Foster noted.

He said “the Holy Grail for unconventional reservoirs” is estimating the stimulated rock volume after fracturing, which can lead to better well placements, improved spacing and refrac’ing opportunities.

“How do you space these wells? They’re all commercial, according to the engineers, but some are better than others,” he said.

• Time-lapse Seismic

Tura noted the increasing application of time-lapse seismic, and identified enhanced oil recovery monitoring as a future area of growth.

“Further down the road we may be looking at using time-lapse seismic for monitoring EOR, which is being used more and more in unconventionals,” he said.

Foster said he had a long career in industry, including 13 years at ConocoPhillips, before moving into research at UTIG.

As advances in machine learning and other AI implementations occur, he sees a familiar pattern of early successes followed by some inevitable disappointments.

“We’re on an upward curve of, ‘This is going to solve all our problems.’ The question is, “How far can you take this thing?’ You can’t take people out of it entirely,” Foster said.

xfdsfqtvyausbbfeadvvzvwzudsczube

You may also be interested in ...