Efficiency gains through improved well designs, drilling and completions (D&C) and tech have led to higher production in the Permian Basin. And some changes have had a greater impact than others, according to panelists who spoke at Hart Energy’s SUPER DUG 2025 Conference & Expo in Fort Worth, Texas.

But even some of the most fundamental improvements in D&C have limitations.

Extended reach laterals are some of the lowest hanging fruit for efficiency gains, said Sean Hervo, co-founder and CEO of PrePad, which provides cloud-based software to optimize drilling and completions.

“If you have the contiguous acreage to actually push your well lengths longer, you are no doubt going to keep reducing your cost per foot as you go long and longer,” Hervo said. “But there is a point of diminishing returns where that is going to curve over. And depending on what kind of value drivers you’re chasing, you may not want to push it as far as you possibly can technically.”

Simultaneous fracking— the stimulation of multiple wells at once—is another lever being pulled to get a step-change in cost per foot, he added.

RELATED

Lower 48 Well Completions Doubled Since Late 2014, EIA Says

Service companies, working with operators, have been optimizing well designs to boost well productivity and ultimately, lower costs, as shale production growth slows.

David Reid, CTO for NOV, pointed to innovation in the challenging area of torsional vibration and motor improvements.

“Motor improvements have been a big deal, not just additive, but actually a long string where we’re having to run a couple of vibration devices and that’s really made a difference,” Reid said. “So, the distance that we’re able to go as well as to overcome all of the friction that builds up.”

Shankar Annamalai, senior vice president of Permian Geomarket for ChampionX, highlighted the intersection of lift, chemistry and digital for production optimization as well as eliminating waste. Using digital technology and software enables faster reaction, and “time to react is value for our customers,” he said.

“On the digital side, closed loop systems, algorithms that take all this great data that we’re getting from production and making interventions faster than humans would’ve done before is ways that we’re seeing production efficiency improve,” Annamalai said. “And then you tie that with chemistry, you tie that with other downhole mechanical systems, that’s where we’re seeing a tremendous amount of production uplift capability post-well construction.”

Making sense at $60

There doesn’t appear to be an emerging trend, however, when it comes to which completions technique—simul-fracs, trimul-fracs, zipper fracks—makes sense at $60 oil. Panelists agreed strategies depend on the wallet and the geology of the rock, among other considerations.

“Different operators have different balance sheets. They have different inventory. They’re fracking different rocks,” Hervo said. “So, some operators at $60 a barrel might just be steady as she goes because that’s fine. Whereas some operators, their breakeven may be fully burdened, or approaching that, and they may need to try different things.”

If the latter is the case, operators are seeking ways to turn over more rocks quicker and with confidence, according to Hervo.

Can AI be of further assistance? AI has transitioned from being a buzzword to more commonplace.

Predictive learning, which is considered a key aspect of AI, has been around for decades in the oilfield to forecast future outcomes or find trends based on historical data.

“In our software space, we’ve been doing basic AI for 20 years because we do predictive work, so we’re very familiar with it in a technical space,” Reid said. “But of course, in every part of the business we’re looking at [how] can we apply it? … If you look at drilling efficiency, we have a lot of AI products that are now making the decisions in the process and for us to optimize.”

He added that NOV is applying digital footprints into everything, which can prove beneficial considering the company is active on the acquisitions front.

“We buy a lot of companies. We end up with a lot of complex systems, and our challenge is to get data to talk to each other in the same language,” Reid said. “That’s the challenge across the whole industry. But … we’ve been able to teach models to read from different sources, which is quite important in the journey.”

Software business collapse?

Among the most interesting of Reid’s AI-related observations is less code is being written because AI is doing the task.

“What that means is the software business is going to collapse itself because the value of making software and the ability of AI to copy each other is actually going to destroy the value of software,” Reid said. “So, I’m quite excited for what’s going to happen in a business that generally has been taken money out of our pocket, and we’re going to be able to see it crushed down and take software costs to the very, very low prices.”

Hervo said he hasn’t seen any game-changers in terms of AI in the digital twin space for completions.

“We’re not going to be able to get rid of my developers anytime soon [by saying] ‘AI, design a better PrePad,’” Hervo said. “But what we can do is help our clients more easily use the power of our simulation engine, essentially. That’s a place where we’re continuing to lean into quite rapidly. … And there’s absolutely some machine learning aspects of some of the things we do, and you’ve seen clients leverage their data exceptionally. I just wouldn’t call it necessarily AI in the truest form of AI.”

The abundance of data and bespoke software is sometimes it’s too much for operators, Annamalai said.

“But there is a day coming, where you’re just going to ask it, Hey, what are my top five wells that should be producing better than they should? And what are the recommended interventions that we should do? Or which of these wells are the most problematic? And give me potential solutions,” Annamalai said.

With information such as the formation, geography, form of lift, bottom hole pressure and other data already known, efforts are improving interaction with tremendous amounts of data.

“Today it is overwhelming, and we’re already stretched thin in terms of our resources,” Annamalai said. “We’re really focused on how do we make it easier, less clicks for production engineers to get to something that delivers value.”

Comments