Data Centers as a New Challenge for Gas Utilities

Growing Energy Demand on Gas Utilities

The rapid expansion of data center development, driven largely by artificial intelligence (AI), cloud services, and digital infrastructure, is introducing a new class of large, fast-moving energy users for local natural gas distribution companies (LDCs). While much of the public discussion has focused on the strain on electricity systems, the impacts of data centers on gas utilities appear to receive far less attention. The demand on gas utilities tied to data center development, including on-site generation and supply for data-center-dedicated power plants, is increasing (IEA, 2026, pp. 49-51; Shehabi et al., 2024, pp. 5-7).

Over the past several months, the authors have been studying the impacts of these new loads and gathering perspectives on current conditions and the practical challenges gas utilities are facing. Based on firsthand discussions with gas utilities and data center industry players, this article examines how data centers are beginning to affect local gas utilities, what challenges are emerging, and where LDCs may find opportunities to adapt planning and decision-making approaches.

Data Center Growth and Industry Demand

Recent industry analysis highlights that data center demand differs from conventional commercial load growth. AI-driven facilities are designed with extremely high uptime expectations, rapid load ramps, and limited tolerance for curtailment. Load growth associated with data centers is often sharper and more time-sensitive than historical growth patterns used in gas system planning (IEA, 2026, pp. 10-14; Shehabi et al., 2024, pp. 5-7, 16-17).

A 2026 International Energy Agency (IEA) report indicates behind-the-meter natural gas generation is increasingly being considered not just as backup power but as a primary or bridge solution while the electric grid catches up (pp. 10-11). These systems can create large, sustained gas demand that may resemble industrial load profiles more than traditional commercial service.

At the same time, AI infrastructure itself is evolving. While hyperscale training facilities currently dominate development, industry projections suggest that inference workloads may drive growth in smaller, distributed data centers located closer to population centers already served by local gas distribution systems (IEA, 2026, pp. 28-29, 36-38). This evolution could diversify the ways gas utilities interact with data centers, from single, very large loads to multiple moderate-sized customers.

Data centers offer the potential for unprecedented growth for LDCs. Utilities representatives conveyed that they are accustomed to organic growth, typically described as 2-3% annually. This organic growth comes from a mix of residential, commercial, and industrial customers. Multiple LDCs indicated that data center requests regularly represent 20-50% growth in their systems.

How Data Center Requests Are Reaching Gas Utilities

Gas utilities reported during discussions that data center-related inquiries are arriving with increasing frequency, in some cases weekly or more. These requests are often characterized by limited technical detail, with developers often uncertain about load profiles, service duration, or long-term operational plans. Utilities note that only a small fraction of these inquiries ultimately proceed to construction, but the volume of requests alone is creating planning and screening burdens.

Importantly, these requests are not limited to large investor-owned utilities. Municipal utilities and smaller public gas systems report receiving similar inquiries and are actively evaluating them. In some cases, gas utilities are also approached indirectly. They are asked to supply fuel to (new) electric generation facilities that are effectively dedicated to serving a single data center campus.

Several utilities emphasized that developers are generally willing to pay for required infrastructure upgrades and are not particularly sensitive to interconnection or extension costs. However, utilities expressed concerns related to recovering costs for ongoing operations of these facilities and infrastructure.

Planning Challenges Uncertainty, Permanence, and Risk Allocation

Load Permanence and Customer Longevity

A recurring concern raised during discussions with gas utilities is uncertainty about the durability of data center demand. Utilities are evaluating how potential changes in technology, energy supply arrangements, or operating needs could alter future natural gas usage, which may affect the timing, scale, and cost recovery of infrastructure investments.

Unlike traditional industrial customers with long operating histories, data centers may face different market dynamics, including mergers, rapid technology turnover, and changing siting strategies. The utilities are beginning to account for potential changes in load and demand in their contract design to protect their ratepayers and support infrastructure decisions.

Capacity, Supply, and Price Implications

While the utilities report limited immediate concern about firm capacity, often noting priority access to transmission-served supply, they have broader questions about long-term system implications. These include managing transmission, storage, and upstream supply as data center-driven demand grows, and how expanded large-load commitments could affect procurement strategies and price stability for residential and small commercial customers.

The utilities also noted that large data center loads are being considered in operations and broader long-term planning frameworks. Critical operational characteristics include safety, reliability, and resilience. Important priorities include affordability, sustainability, and customer relations. As utilities evaluate data center opportunities, they are working to continue providing safe and reliable operations while balancing future planning and an evolving business context.

Operational and Contractual Complexity

Data centers that use on-site gas generation raise operational and contractual questions that gas utilities do not routinely encounter. Utilities are asking data centers about the possibility of interruptible or premium-priced service during extreme weather to protect system reliability. Some utilities have proposed contracts that include an early-termination penalty and an upfront early-termination deposit. Other open questions include how on-site generation interacts with the electric grid, such as whether data centers can export excess electricity without violating gas or electric service agreements. These issues could span gas, electric, and regulatory jurisdictions, increasing the complexity of coordination.

Opportunities for Gas Utilities: Planning, Structure, and Coordination

During discussions, gas utilities identified approaches that may help manage both near-term uncertainty and longer-term opportunity. Several utilities emphasized the value of early-stage engagement and standardized screening processes to evaluate data center requests. Developing internal tools to assess the likelihood of project completion, expected load duration, and infrastructure implications can help utilities prioritize engineering and commercial resources.

These utilities are also observing alternative structural responses in other sectors. For example, spin-off generation entities have been created by some utilities to isolate their ratepayers from concentrated data center-related risks: an approach that could inform future gas utility strategies depending on the regulatory context. Rather than assuming permanence, these utilities are considering phased or modular infrastructure upgrades aligned with contractual milestones. This approach may reduce stranded-asset risk while still enabling projects to proceed.

Finally, these utilities consistently emphasized that data center-related planning increasingly requires coordination across gas, electric, regulatory, and local planning entities. Aligning assumptions, timelines, and contingency plans can help reduce surprises and improve system-level outcomes.

Looking Ahead

Data centers are emerging as a planning consideration rather than an isolated edge case for local gas utilities. While many questions remain, early experience suggests that traditional approaches to load forecasting, contract design, and risk allocation may need to adapt.

For natural gas LDCs, the challenge lies in balancing responsiveness to economic development opportunities with fiduciary responsibility to existing customers. Utilities that invest early in structured assessment frameworks, cross-sector coordination, and flexible planning approaches may be better positioned to navigate data center-driven demand without compromising safety, reliability, affordability, or long-term system goals.

Want to learn more? GTI Energy will be leading a panel discussion with local distribution companies during UAI Community & Innovation Week about the impact of data centers on LDCs and their operations on Thursday, July 16, 2026, at 11:00 am – 12 pm ET. You can sign up to attend below.

Register for UAI Community & Innovation Week

Session Details

Energy demand from AI and data centers: The role of natural gas and local distribution companies

Speakers

Description

Data center development driven by AI and cloud computing is creating significant increases in energy demand. The impact of this increased demand frequently focuses on electric utilities and the grid, yet local distribution companies (LDCs) for natural gas are also facing significant potential impacts from data center growth. Increased demand is beginning to surface new planning questions for LDCs. Data center load requests often arrive quickly, with limited definition, and outside traditional load growth assumptions. In this panel-style webinar, GTI Energy will lead a discussion with representatives from two natural gas distribution companies to discuss what LDCs are seeing on the ground today. The conversation will explore how data center-related requests are reaching gas utilities, where uncertainty and risk are emerging, and how utilities are beginning to adapt planning approaches. The session is designed as a practical exchange of observations, lessons learned, and open questions relevant to utilities navigating data center-driven demand.

About this Summary

Generative AI was used to summarize the authors’ discussions with industry experts in natural gas utilities and data center operations.

References

International Energy Agency (IEA). (2026, April 16). Key questions on energy and AI. International Energy Agency. https://iea.blob.core.windows.net/assets/3179f7f8-01f6-4dd6-bffa-c9f7b73f1dc9/KeyQuestionsonEnergyandAI.pdf

Shehabi, A., Smith, S. J., Hubbard, A., Newkirk, A., Lei, N., Siddik, M. A. B., Holecek, B., Koomey, J., Masanet, E., & Sartor, D. (2024). 2024 United States data center energy usage report (LBNL-2001637). Lawrence Berkeley National Laboratory. https://eta-publications.lbl.gov/sites/default/files/2024-12/lbnl-2024-united-states-data-center-energy-usage-report_1.pdf

About the Authors

All authors are employed by GTI Energy, a nonprofit research and technology development organization located in Des Plaines, IL.

Ansh Nasta is a Principal Energy Systems Analyst leading GTI Energy’s investigation of the impacts of data centers on LDCs. He researches and analyzes low-carbon solutions to transition to the net-zero energy system of tomorrow. His areas of expertise include energy systems modeling, GIS, hydrogen, mobility, and data centers. Ansh is passionate about energy and climate issues and has been in this space for ten years. Ansh received a B.Eng. in Mechanical Engineering from the Hong Kong University of Science and Technology. He earned his M.S. in Energy Science, Technology and Policy from Carnegie Mellon University. You can contact him at anasta@gti.energy.

David Van Wagener is an institute energy systems engineer, with a focus on energy system decarbonization. As a chemical engineer with a background in process modeling, energy optimization, and data science, he is helping to expand GTI’s expertise in life cycle analysis. Many of his roles provide vision for the future integration of digital tools with energy supply and decision-making. David has extensive experience in carbon capture technology and greenhouse gas emissions measurement and analysis. He holds a PhD in chemical engineering from the University of Texas at Austin.

Iana Iacob is a principal scientist at GTI Energy. Her work focuses on energy system decarbonization, market and policy analysis, stakeholder engagement, community benefits planning, and life cycle assessment. She has experience evaluating hydrogen, carbon capture, methane emissions, and other emerging energy technologies, with particular expertise in the social, policy, and market factors influencing technology deployment. She holds a PhD in Engineering and Public Policy from Carnegie Mellon University.

Zach Weller is a senior data science manager who leads GTI’s Data Science and Applied Statistics Hub (DASH). He specializes in collaborating with subject matter experts to develop data-driven solutions and analyses using data science. He has extensive experience in methane emissions, sample planning, uncertainty quantification, and analysis of energy usage data from advanced metering infrastructure. He holds a PhD in statistics from Colorado State University.