The Global Impact of AI in Energy


Moderated by Houston Chronicle business columnist Chris Tomlinson, the Time Machine Interactive panel on AI’s impact in global energy addressed a wide-ranging set of topics, from carbon emissions management and additive manufacturing to the sharing of operational information among competing companies and the importance of trust.

Tomlinson was joined on the stage by Francesco Menapace, GM Exploration Technology, Shell; Wafik Beydoun, the International Association of Oil & Gas Producers (IOGP) Director of Americas; and Scott Parent, CTO Digital Solutions, Baker Hughes.

Tomlinson opened with a couple of broad questions: “How do we reduce emissions, improve efficiency, and simultaneously facilitate the transition to a cleaner energy world? And how does AI fit into this world of increasingly variable resources and generation asset types?”

Parent responded by observing that over the past 10 years, we’ve seen that the only fast way for the industry to turn on more supply is to become more efficient.

“How hard is it,” he asked, “to convince customers that a little more spent on AI is worth it? We have an opportunity to help customers upgrade their capabilities. A 1% combustion efficiency improvement gives the world a 2-3% improvement in carbon footprint. As an OEM we have that opportunity right now.”

Beydoun countered with the assertion that given the constant pressure to develop more affordable solutions, we have to demonstrate more innovation than we have to this point.
“We can’t do it on the platforms we have now,” he said. “It’s very hard to deploy at scale; it just doesn’t happen. Yet we want to be well-placed to capture innovation over the next decade.”

Beydoun referenced his role as a trade association leader to say their organizational role is to recommend and provide best practices for the industry around safety and emissions, saying: “This work now has a great deal to do with energy transition. But the technology needs to be mature enough to test in different contexts, environments, and regions. Certain technologies and processes may not work as well, for example, in the Gulf of Mexico as they do in the Middle East. We need robustness to write good practices. In fact, we are now more involved than in the past in such work.”

Menapace added that despite the availability of many promising technologies, the ultimate issue is one of trust, echoing a common theme for other conference presenters.

“Trust,” responded Parent, “is a really important word in this context.”

He then cited the example of Aramco, which has shared well data with three major service companies for more than 18,000 wells over 14 years in the Saudi area. “If I want to scale up more quickly,” he said, “I need to have a level of trust that translates into actionable data. If it were possible to bring everyone up by just half a percent, that would translate to tens of gigatons of emissions removed from the world.”

Tomlinson then shifted the direction of the discussion more directly toward the topic of AI.

“How do we get past the hype cycle? I get a dozen press releases every day talking about AI,” he said. “Yet in the past five years, the world hasn’t changed all that much.”

Menapace expressed optimism, noting the industry has moved beyond much of the original hype and that real value is now being demonstrated, with industry leaders much more inclined to trust the technology to amend their workflows.

Parent agreed with Menapace’s assessment, adding, “because we can now collect larger data sets and do more with them, we can take a mechanical system, know the inputs and outputs, and be quite successful developing accurate models and doing simulations. In this space, there is still a lot to learn, but we’re coming down the hype curve pretty quickly.”

Tomlinson probed further, querying as to the greatest opportunities and the greatest promise.

Menapace took up the challenge: “There are definitely some home runs in maintenance and production. It’s just good business now; you can see it happening. Most exciting for me is subsurface evaluation. The data is multi-dimensional, and as soon as we can analyze it and turn it into actionable insights, that’s where I see the most value. There’s a hyperdimensionality of things—how can I get inside our seismic well data and look at it through an AI lens?”

Parent went on to express excitement around additive manufacturing. “By combining it with design, we can print three, four, or more alloys simultaneously to create a component that’s soft in the middle and hard on the outside. When you add in changing topography and structure, with super light parts, we see a 60% reduction in materials. And that’s not just in oil and gas; it also affects aerospace.

“I’d challenge anyone on the planet to build a diving board using nonhomogenous material so that, on the fly, I can change the material from the beginning to the end of the board to create variable physical geometry. That’s not something that can be done by humans alone. It’s going to be a gamechanger, with as much as a 30% compound annual growth rate possible. Given the complexity of parts in oil and gas and the amount of material required, this will be disruptive.”

Beydoun indicated that he tends to look at the AI opportunity in a somewhat more incremental manner, particularly as it relates to emissions, carbon capture, and similar challenges.

Tomlinson countered by observing that it is in companies’ natures to hoard their competitive advantages and resist sharing with competitors.

“How important,” he asked, “is cooperation in successfully innovating?”

Menapace answered that policies and platforms need to come together in a more coherent manner. “What is it we need to give people?” he asked. “Data and a common platform. We don’t have that today, and we need to fix this. Put it in a platform so people can access it. That should accelerate innovation. When you realize that every other supermajor is doing the same thing, you have 7,000 platforms talking to each other, and 7,000 custom implementations. That’s why we started to use OSDU (Open Subsurface Data Universe) to come together so that we all can use and deploy it. It hasn’t gone quite as fast as I would like, but it has definitely generated momentum. There are hundreds of companies on the platform now.”

Agreeing, Beydoun added, “The industry is currently in a state of “coopetition” (citing a term first coined as the title of a 1996 book by Brandenburger and Nalebuff). If a particular operator can be competitive, he won’t naturally be inclined to share with others, unless sharing will open up new opportunities. The energy transition to sustainability is a very new competitive area.”

The lively conversation ended with Tomlinson acknowledging that regardless of sustainability goals, the global demand for energy will continue to grow due to the enormous number of people on earth who still don’t have access to transportation, light, and heat.

“But at the same time,” he said, “we’re trying to fight climate change. The energy industry is forced to wrestle with both of these challenges. We need a strategy that manages to optimize for profit and for the planet.

Latest blogs

SparkCognition is committed to compliance with applicable privacy laws, including GDPR, and we provide related assurances in our contractual commitments. Click here to review our Cookie & Privacy Policy.