AI has moved out of the early adoption phase as smaller oil and gas companies follow the path set by larger firms, Ron Beck of Aspen Technology said March 19 in a panel discussion at Hart Energy’s DUG Gas Conference & Expo in Shreveport, Louisiana.

"The fastest adoption, earliest adoption has been by the largest companies,” said Beck, AspenTech’s senior director of solutions marketing. “The very largest investment in the world has been driven out of the Middle East,” specifically by Abu Dhabi and Saudi Arabia.

The beginning of phase two is here. Smaller companies are beginning to follow in bigger companies’ footsteps, and they are realizing results, Beck said. AspenTech, recently acquired by Emerson, provides asset optimization software across the energy industry. 

There’s more to come, said Troy Ruths, founder and CEO of PetroAI. The company, founded in 2013, uses AI to model shale reservoirs below the surface.

“We've been working on coming up with really good AI models that actually provide business value,” he said. "Now you're starting to see players and [executive leadership team] members come down, and say, ‘Hey guys, we need to operationalize AI, we need to start seeing business impact here.’ And where we're seeing business impact is really, really impressive results” that are boosting free cash flow to businesses.

For Peter Harding, founder and CEO of Kelvin AI, artificial intelligence has the potential to change business structures. Kelvin makes industrial automation and control software, focusing on the upstream.

“If you're starting an E&P today you can do it fundamentally differently with a different team makeup and a different approach than before, and frankly probably a lot more profitably,” Harding said. “We see some folks who are on that side of things, and then we see a number of companies that are dabbling but haven't really committed to using these tools to their full potential, whereas there's a handful that are really using it well and they're getting incredible returns.”

The panelists said the tech is already being used to generate more insights into underexplored formations like the Haynesville, to do more work per person with less administrative expense and to run well operations.

AspenTech applied deep learning powered by AI to older seismic data to “create a much more accurate picture” of the shale, finding the best places to drill new wells. A project that previously took three to six months is now done in a couple of weeks, Beck said.

And for KelvinAI, AI helped design and create dynamic balancing for a gathering system with multiple wells delivering fuel to three separators, Harding said. When one of the separators fails, the system adjusts output from the other wells to keep running without a full restart.

It also helped Legacy Reserves, since renamed Revenir, to design and drill a four-well pad in a tight space, Ruths said.

Trust the process

There are obstacles to speedier adoption of the technology. Legal departments may raise data governance issues, AI's answers and solutions can be flawed and top executives may not be fully on board.

“We're working through a lot of those data governance issues,” Harding said. “Now it’s a question of who within your organization wants to step up and take a chance. What we find is that groups that are willing to try this stuff, give it to their teams and utilize it are having a lot more success.”

Ruths said it’s important to fit the solution to the size of the problem and to have support for AI from the top of the organization.

“You need an executive team member or board member saying this is something strategic because it’s not cheap,” he said. “It's a rich cultural change to bring AI in, and then you need to have the business pressures at the high level to get value out of it.”

Comments