Permian Basin AI use grows in oil and gas operations

Artificial intelligence continues to be adopted throughout the oil patch, and companies across the energy spectrum are finding a variety of uses for the technology.

Artificial intelligence was the focus of a panel discussion at the recent Permian Basin Environmental Regulatory Seminar presented by Midland College’s Petroleum Professional Development Center and the Permian Basin Petroleum Association.

The discussion began with moderator Robert Henkhaus, vice president of product at Enverus, asking where AI is showing up in the industry.

Yogashri Pradhan, founder of Iron Lady Energy Advisors, said she has seen it being used for monitoring geologic features and subsurface fault mapping.

“On the technical side, it’s being used proactively,” she said. AI is used to examine historical data and extrapolate sites that could be seismically sensitive.

Akash Sharma, vice president of product management at Enverus, called AI a means to an end for companies seeking specific outcomes.

“How do we leverage these tools to meet outcomes?” he asked. AI could be used not just to solve problems but to connect AI-generated data with other data.

University Lands uses AI to maximize outcomes and the efficiency of data, said Wil Vark, business solutions and application development manager. He said the data and documents AI can generate enhance his team’s ability to analyze the information.

AI can extract information from the documents, permits and plats uploaded by operators active on University Lands, allowing analysts to analyze the information rather than enter the data, he said.

Effectiveness starts with understanding the geographic area and connecting data, Sharma said.

“If I can understand where a fault can be triggered, I can perform the decision-making I need to be making,” he said.

Still, the panel agreed AI is no replacement for human judgment, especially in the approval cycle for authorizations for expenditures, or AFEs. But Sharma said it can be used to provide a more complete picture of expected production and potential opportunities for well participants.

They also praised AI’s ability to extract information from contracts and present summaries for landmen or analysts.

“One thing we’re having to do is educate the staff about the nature of AI and what you really get in a response,” said Vark. “The mission of AI is to give a response. It never says ‘I don’t know’ even if it doesn’t understand the question. Be careful and check against the response.”

Henkhaus asked what may be preventing companies from adopting AI.

Pradhan cited concerns that AI would replace jobs. There is also the lack of understanding about just what AI does and what causes it to generate the answers it provides.

Vark said the need to protect sensitive information needs to be part of AI training.

Henkhaus listed the need for transparency and citations in using AI.

Panelists said the biggest misconception about AI is that it will replace jobs.

“How do you transfer 30 years of experience to AI?” asked Pradhan. “It’s not AI, it’s you making decisions. AI is helping you make those decisions.”

AI will not solve all problems, Sharma said. “I wish that was the case. It requires understanding. Not everything can be solved by AI.”

AI is changing so fast, Vark said. “The best strategy is to be aware of AI and what’s happening, stay aware of the trends. I think AI will become a layer that’s invisible but a part of how we use technology.”