What’s trending in exploration geophysics these days, in eight words: Machine learning, artificial intelligence, data, data, data and data.
Those concepts fit together into one reality. As the oil and gas industry acquires larger and larger masses of geophysical data, it leans even harder on machine learning and AI to sort it all out.
“The recent growth of machine learning and AI-like tools is advancing quite rapidly in geophysics,” noted Paul Sava, a professor and department head in geophysics at the Colorado School of Mines in Golden, Colo.
Consequently, geophysics students today get a heavy dose of computing instruction.
“We have students who are quite advanced in this. We have some wizards in understanding trends in computer processing,” Sava said.
Also, “more and more (geophysics) students are minoring in data science,” he added.
New and Advancing Tools
In data acquisition and instrumentation, distributed acoustic sensing systems using fiber-optic cable are still a trendy topic. The technology provides for continuous monitoring and measurement along the entire length of the cable line.
“One of the unfortunate by-products of fiber optics is that we have an enormous amount of data,” Sava observed.
He said one of his CSM colleagues is now grappling with the problem of moving those enormous datasets out of the field for processing.
“It’s not that trivial to move humongous amounts of data into the processing center. Then that brings up the question whether processing should be moved toward the field, in what is called edge computing,” he noted.
“The other aspect of instrumentation is that sensors are becoming cheaper and cheaper,” he said.
Inexpensive sensors have allowed the industry to deploy huge numbers of recording nodes in the field, some of them no more complex than chips in cell phones. Those individual nodes replace connected networks of sensors.
“Until a handful of years ago, we were going into the field with several trucks full of hardware. Now we might go with one truck, full of nodes,” Sava observed.
Another emerging area in instrumentation involves robotic data acquisition, including the use of drones to gather geophysical information. Robotics applications are still in a fairly early stage in geophysics, Sava noted.
“I’d personally like to see that advance even faster,” he said.
As a result of the industry’s expanded acquisition abilities, geophysical data sets have grown even more immense. Machine learning and AI are now tools of choice for data analysis, although they can’t provide a full answer. Sava said understanding and evaluating the data is just as important as being able to manipulate it.
“Machine learning has this drawback. It’s all about data, or it seems to be all about data, (but) it’s also about where the data comes from,” he said.
Seismic application work has now expanded to included geophysics for carbon capture, utilization and storage. Some recent surveys have focused exclusively on identifying and analyzing reservoirs for CO2 injection.
“In terms of applications, there is a broadening of the application space” to include evaluation for CCUS, Sava said.
“You have to know where to inject. For sure, that is the flip side of production. There is also a safety aspect to it. You don’t want to inject (CO2) and then see it leak to the surface,” he said.
Major Projects Ongoing
It now appears that a world/Brent oil price above $80 per barrel will sustain global exploration, and prices were well above that level going into the end of 2023’s third quarter.
A recent uptick in seismic work reflected the market’s strength. Geophysical companies were at work in Asia, the Gulf of Mexico, Brazil, Argentina, Africa, the North Sea and other regions, with offshore acquisition dominating activity.
Energy-data provider TGS has received funding for several multi-client projects in Southeast Asia, set to start late this year. The program covers 2-D and 3-D acquisition and is expected to last for six months.
The company also reported a 2-D-cubed project offshore Brazil targeting the Brazilian equatorial margin area, as well as an expansion of its multiclient 3-D coverage in the Malvinas Basin offshore Argentina.
PGS, a marine geophysical company, announced a 3-D exploration acquisition contract offshore Africa. That program has a duration of four to five months, beginning in fourth-quarter 2023.
In July, offshore seismic specialist Shearwater launched a project with OMV for a towed-streamer survey over the Berling gas and condensate discovery in the Norwegian Sea.
Geophysics Job Market, Education
Sava said the job market is strong for geophysics students right now, although the upturn in industry exploration hasn’t impacted the university level yet.
“Our students are in high demand and recruiters want them. And that’s great. The exploration part is less obvious to me,” he said.
Rapid advances in geophysics-related computing and analysis present a challenge for instruction. Sava said that amounts to preparing students for an unknown, and unknowable, industry future.
“The professional life of a student is far longer than our predictive ability. Knowing what a student will be doing in 15 years, 20 years, 25 years is impossible,” he said.
His department’s response has been to emphasize core knowledge and key concepts in exploration geophysics while exposing students to the latest tools and skills.
“We are trying not to throw away things that work, but to build on our successes,” Sava said.
Artificial intelligence plays another part in university studies today. Sava noted that anyone can sign up for a ChatGPT account and begin using AI. He doesn’t mind, and it’s one way for students to familiarize themselves with AI applications.
“Students are going to do that. And they should. Why should they hide themselves from tools?” Sava said.
But it’s important for students to assess AI output and to develop their own judgments, he noted. ChatGPT has become famous for its occasionally puzzling outcomes.
“Sometimes the answers are spectacular. Sometimes the answers are ridiculous,” Sava observed, so students need to develop an understanding of the tool they’re using.
He has noticed a split in mindsets between undergraduate geophysics students and graduate students. Instruction at CSM has moved beyond the pandemic and “right now our behavior is business as usual,” but “the return to the office is not rapid,” he said. Graduate students, especially, work quite a lot from home.
In undergraduate life, “students have an interest in sustainable living, in some sense. They are concerned about climate. They are concerned about human effects” on the environment, Sava observed. Graduate students typically receive industry funding and are much more task oriented.
To Sava, the biggest change in geophysics over the past 15 years has been “the availability of large-scale computing in the hands of almost everyone. I think that is impacting every part of our profession, including exploration.”
“What I’ve seen in the past is that we were far more split by skill sets,” Sava said.
He hopes multidisciplinary approaches will lead to more creativity and the use of a broader selection of tools. Sava compared the change to having a bigger pencil box for drawing, with a lot more pencils to choose from.
Today, “we draw with more colors than in the past,” he said.
“More interesting results pop up.”