Regardless of which side you fall on in the debate about recent tariffs, there is the potential for accelerated reshoring of manufacturing operations back to the United States. While this shift holds promise for domestic job growth and increased supply chain resilience, it brings with it a question about our infrastructure that is too often overlooked: Can the power grid handle the load?
Beyond the expectation of reshoring, the rise of artificial intelligence (AI) has suddenly placed unprecedented pressure on our energy infrastructure. AI is driving a rapid expansion of data centers and it’s transforming how goods are produced by enhancing manufacturing speed, precision and scalability. Even before the discussions of tariffs, utility companies that are forecasting rising demand due to AI began taking a proactive approach.
For example, Duke Energy recently announced a 13 percent increase in their five-year CAPEX plans, citing the need to support large load growth, including data centers and manufacturing sites. Likewise, Southern Company is now expecting electricity sales to grow six each year through 2028, which is up from a projected one to two percent, due to data center demand. This same scenario is taking place among utilities across the country.
For utilities and manufacturers alike, these implications are massive. It’s a dilemma that demands a new approach to two fundamental questions: Where will the electricity come from, and how can we keep it cost-effective and clean?
The Hidden Consequence of Reshoring
Reshoring brings with it a host of advantages: job creation, enhanced supply chain security and economic stimulus. However, most conversations overlook the energy implications. Even before tariffs entered the discussion, U.S. manufacturing was already expanding, with the sector seeing it’s fastest growth since June 2022 in February of this year. A broader return to domestic production will only accelerate this trend while also requiring gigawatts of additional power. And it will do so quickly.
The problem is that much of the aging U.S. grid is not ready to handle the anticipated growth in demand for industrial energy. Adding capacity isn’t as easy, and permitting, infrastructure upgrades and transmission line buildouts can take years to materialize.
AI: Powering the Cloud and Factory Floor
While its impact on energy demand is widely recognized when it comes to data centers, AI’s influence goes far beyond cloud infrastructure. Manufacturing facilities are becoming increasingly AI-driven. From predictive maintenance to smart robotics and supply chain optimization, AI tools are revolutionizing production capabilities. It brings a welcome increase for efficiency but also raises the energy intensity of operations.
Now, consider the challenge: We must power both AI-enabled factories and the data centers that support them. Data centers alone are expected to triple their electricity consumption by the end of the decade. Add to this the growing electrification of transportation and industry, and we’re looking at an energy landscape under intense strain. Those interconnection queues will only grow longer, and our grid’s capacity remains finite.
You can’t scale what you can’t power. Manufacturers need fast, reliable, and clean electricity, but today’s grid constraints don’t align with that need.
Some large power users are being told to wait five years or longer for a firm grid connection. Renewables like solar and wind, while clean, are intermittent. Battery storage is costly and limited in duration. And gas turbines? Manufacturing backlogs for gas turbines, power transformers and switchgear are reportedly extending out until 2028, which means they may not arrive any faster than the grid connection itself.
The market needs a realistic, scalable solution that attempts to balance speed, reliability and environmental stewardship.
A Smarter Path
Enter flexible onsite power generation. These modular, scalable generating systems are powered by natural gas, offering faster deployment, dispatchability and the ability to reduce pressure on the grid. Rather than waiting years, large load customers can deploy flexible microgrid systems within 12–18 months, securing reliable power exactly when and where it’s needed.
These systems enable utilities to maintain grid stability by helping them manage peak loads more efficiently, and can defer costly infrastructure upgrades. Flexible generation fills the gap between current grid limitations and future clean energy aspirations while providing improved resilience and accelerating speed to power. Onsite generation can also provide backup power during emergencies or unplanned outages as well as emission reduction strategies.
Companies that proactively address their own energy strategy will be best positioned to thrive in this new industrial era. With reshoring incentives and AI productivity gains on the table, those that invest in flexible power infrastructure now can avoid the energy bottlenecks that are certain to occur in the future.
This strategic energy planning is no longer the domain of utilities alone. It’s becoming a core pillar of operational excellence for business operations big and small. From uptime and cost control to sustainability and scalability, energy strategy is now a critical business decision that directly impacts both safety and a company’s balance sheet. Resiliency solutions like flexible power generation offer the agility manufacturers need to keep pace with demand while managing their costs and emissions profiles.
The combined impact of AI and reshoring has the potential to reinvigorate U.S. manufacturing. But without expedited access to reliable power, that vision may be delayed or even lost. It’s time to rethink the role of energy in industrial planning. Our aging grid alone can’t shoulder the load. Manufacturers must lead in adopting smarter, flexible power solutions that enable growth, control costs and protect the environment.
In this next era of industrial innovation, the biggest competitive edge may come not just from what companies make—but how they power their operations.
Thomas McAndrew founded Enchanted Rock in 2006. He brings more than 36 years of experience in energy asset management ranging from large, complex nuclear plants to small distributed generation.