How AI and natural gas can power a sustainable data center revolution
In an increasingly digitalized global economy, the world’s power systems and electricity consumption patterns are undergoing profound shifts that are difficult to ignore. Consumers and industries are becoming more reliant on digital tools like artificial intelligence (AI). Operating the growing number of power-hungry data centers that underpin this digital shift in a sustainable manner has become a defining challenge of our age.
Data center capabilities grow in line with our computational needs, as does their power consumption. We have come a long way from the earlier versions of data centers pioneered in the 1980s.
As internet usage, e-commerce and handheld digital devices became ubiquitous from the turn of the millennium to the last decade, we have seen even larger data centers call on a grid capacity of around 5 MW per center.
That is now rapidly changing with AI evolving as a digital tool with two distinct strands: generative AI and agentic AI. Generative AI refers to systems capable of using models that learn inputted data characteristics and structures to produce bespoke outputs, something most of us would be familiar with courtesy of X’s Grok and ChatGPT.
Agentic AI goes beyond merely reacting to instructions to being proactive and capable of complex problem-solving with limited human intervention. Both require 'hyperscale' or large data centers.
Scale of the task ahead. AI usage has typical power demands of 100 MW or more per center, with an annual electricity consumption equivalent to the demand from 350,000–400,000 electric cars, according to the International Energy Agency (IEA).1
Even a basic ChatGPT search requires 2.9 watts of electricity. That is nearly ten times as much as the 0.3 watts needed for a routine Google search.2 This is why demand forecasters are upping their projections based on assumptions of a significant scaling up of hyperscale data centers for consumer and industrial AI use.
Global power demand from data centers is currently projected to double to > 1,000 TWh by 2027.3 The projected surge would represent a rapid increase from the estimated 460 TWh consumed in 2022 and is equivalent to the annual electricity use of Japan.
At a time of rising global commitments to achieve net-zero carbon emissions, it also places immense pressure on existing modes of power generation, traditionally predicated on fossil fuels.
Nuclear and renewable energy can only service a portion of the additional demand in a carbon neutral fashion. The rest must come from elsewhere within the energy mix. In the eyes of many, that happens to be natural gas—an abundant low-carbon alternative.
A pragmatic and synergistic natural gas-AI nexus. An embrace of natural gas to power the AI revolution is driven by realism and necessity. At our organization, we recognize our dual role in shaping both the demand and supply sides of the energy landscape. Our commitment to achieving net-zero by 2050 is reflected not only in how we optimize energy consumption across our operations and customer engagements, but also in how we innovate and deliver sustainable energy solutions. This is reflected across our assembly lines, products, solutions, corporate actions and customer engagements.4
We are also acutely aware of the critical role of electricity and digitalization in uplifting living standards and economic progress. As a child who grew up in a developing country like South Africa, I can personally testify to this empowering feeling.
Ensuring the spread of round-the-clock electricity and connectivity in developing economies, where some may not enjoy the benefits of electrification many in the developed world take for granted, should remain mission critical.
We cannot have mismatched energy demand and infrastructure that may lead to localized power shortages with potential global ripple effects, all caused by wilfully ignoring viable solutions. That is where natural gas comes in.
Stakeholders in the energy ecosystem must manage the complex interrelationship between AI, natural gas consumption and sustainable development.
IThis is a challenge we readily embrace in our organization, based on our belief that AI and natural gas industries can drive each other’s growth.
Proponents of the industry note that as a low-carbon energy feedstock, natural gas is a stable source for powering hyperscale data centers. It can also serve as a backup for renewable sources as part of the wider energy mix with adequate planning.
Concurrently, based on our decades of industrial products and software domain expertise, we can confidently assert that the AI tools we offer for natural gas production and liquefied natural gas (LNG) terminals can improve efficiencies, bolster the robustness of supply chains and cut emissions.
These tools range from advanced analytics to real-time process optimization software, predictive maintenance platforms and digital twins. In short, we offer a comprehensive solutions suite for industrial process electrification that is designed to cut carbon emissions and lower operating expenditure. This is particularly true for LNG trains deemed integral to an industry tipped to see a 60% growth in demand by 2040.5 Ultimately, a collaboration between AI and natural gas industries would boost, and not sidestep, decarbonization.
Leading on decarbonization. In line with this thinking, we are helping to facilitate change for some of the biggest energy and chemical players in the world. We have helped global companies across the energy value chain to identify and curb carbon emissions and deliver measurable and meaningful impacts.
Often, the challenge for many of our customers is where and how to begin this daunting task. It is why we have pulled our industry knowledge and automation expertise together into our decarbonization e-guide. The guide offers actionable insights and solutions for companies across the upstream, midstream and downstream value chain. It also illustrates how streamlined operations can enhance efficiency and minimize the costly impacts of leaks, and provides a checklist of quick wins that can help operators reduce carbon dioxide (CO₂)‑equivalent emissions by up to 30%.
That is in addition to actions that require mid- to longer-term investment. Lastly, the guide offers real-world examples of how our organization is successfully reducing carbon emissions for key energy players across the world.
The publication also continues our tradition of providing thought leadership in the energy and heavy industries segments to foster wider collaboration on decarbonization.
These insights are all part of developing effective partnership models and innovative industry-wide frameworks. They hold the key to unlocking the tremendous potential of AI and natural gas for accelerating dual-sector growth across industrial electrification and clean fuels.
Of course, there is still much work to be done. Finding the perfect balance for dual-sector progress will always be a moving target with every gain we make.
However, with decarbonization as the ultimate objective in a world conscious of climate change, aligning the strengths of the AI and natural gas industries to deliver sustainable growth—while supporting the energy transition—is an opportunity that we cannot afford to miss.
LITERATURE CITED
1 Spencer, T., “What the data center and AI could mean for the energy sector,” IEA, October 18, 2024, online: https://www.iea.org/commentaries/what-the-data-centre-and-ai-boom-could-mean-for-the-energy-sector
2 Golman Sachs, “AI is poised to drive 160% increase in data center power demand,” May 14, 2024, online: https://www.goldmansachs.com/insights/articles/AI-poised-to-drive-160-increase-in-power-demand
3 IEA, “Electricity 2024,” May 2024, online: https://www.iea.org/reports/electricity-2024/executive-summary
4 Schneider Electric, “Sustainability Development Report 2024: Technology for impact,” 2025, online: https://www.se.com/ww/en/assets/564/document/513141/2024-sustainability-report.pdf
5 Shell, Shell LNG Outlook 2025, online: https://www.shell.com/what-we-do/oil-and-natural-gas/liquefied-natural-gas-lng/lng-outlook-2025/_jcr_content/root/main/section_125126292_co/promo_copy_copy_copy/links/item0.stream/1740577530786/97f8aacf1e8b27cd8dd32f6070d6463f60cff411/lng-outlook-2025-full-report.pdf
ABOUT THE AUTHOR
Devan Pillay is the Segment President of Heavy Industries within Schneider Electric’s Industrial Automation business. In this role, Pillay leads a global team, driving Schneider Electric’s strategic impact across hard-to-abate sectors, including energies, chemicals, mining, minerals and metals. His mission is to help these carbon-intensive industries navigate the complexities of decarbonization while maintaining operational resilience and efficiency.
With > 30 yrs of experience spanning business strategy, engineering and transformational leadership, Pillay brings a proven track record in spearheading growth initiatives and building new business streams in emerging markets. Prior to his current tenure at Schneider Electric, Pillay held several senior leadership roles at Signify Africa, 3M Switzerland, Anglophone Africa, General Electric and Eskom. He earned an MBA from the Gordon Institute of Business Science, and a Bch degree in electrical engineering from the Durban University of Technology, South Africa.
Comments