The global knowledge network for professionals in the energy and industry

Data Centers Could Save the Power Grid Up to $150 Billion

Demand response would reduce peak electricity consumption and avoid costly expansions of energy infrastructure.
Los centros de datos podrían ahorrar miles de millones en energía

The growth of AI-driven data centers is increasing electricity demand in the United States. In response, technology companies and utilities are exploring new ways to adjust the energy consumption of these facilities to alleviate pressure on supply grids.

The strategy, known as demand response, involves temporarily reducing or shifting electricity consumption during periods of high demand. According to a study by the Nicholas Institute at Duke University, its broader application could prevent investments of between $40 billion and $150 billion over the next decade.

Demand Response Could Reduce Electricity Costs

Data center operations require a continuous electricity supply to power servers, cooling systems, and IT equipment. However, part of their consumption can be adjusted when the power grid faces high-demand situations.

Through demand response, these facilities could temporarily reduce certain loads or shift computing tasks to off-peak hours, easing pressure on the energy system.

This would provide utilities with an additional tool to manage consumption peaks without relying exclusively on new generation plants or transmission grid expansions.

Lauren Shwisberg, director of the RMI energy think tank, told Reuters that this strategy is attracting growing interest among states, federal agencies, and U.S. electricity market operators.

Data Center Electricity Consumption Projected to Grow Until 2030

Projections from the Electric Power Research Institute (EPRI) illustrate the scale of the energy challenge associated with the expansion of digital infrastructure.

According to their estimates, U.S. data center electricity consumption, which ranged between 177 and 192 terawatt-hours (TWh) in 2024, could reach between 383 and 793 TWh by 2030.

This growth is linked to the expansion of digital services and the development of artificial intelligence systems that require high processing capacity.

Furthermore, data centers surveyed by EPRI identified a potential for peak electricity demand reduction of between 10% and 30%, depending on the characteristics of each facility.

Arin Kaye, EPRI’s technical lead, explained that some large data center operators, known as hyperscalers, could achieve higher percentages.

These figures demonstrate the potential of flexible energy management, although its application will depend on the operational and technological requirements of each facility.

OpenAI Considers Reducing Up to 1 GW of Electricity Demand in Georgia

One of the most recent examples of this strategy involves OpenAI, which agreed to temporarily limit the electricity it would draw from the grid during periods of electrical system stress.

The commitment includes a reduction of up to 1 gigawatt (GW) at a planned 3.2 GW facility in Georgia, United States.

Through this mechanism, the company could help reduce pressure on local energy infrastructure during peak demand times.

Likewise, the agreement shows how large technological projects are beginning to incorporate electricity flexibility mechanisms into their planning.

This approach could also facilitate faster connection of new facilities to the grid, provided operators can ensure the security and stability of the supply.

Regulators Study New Electricity Connection Conditions

Until now, the application of demand response programs in data centers has primarily focused on pilot projects and individual agreements.

However, interest in expanding these initiatives is growing among energy planning bodies.

In June 2026, U.S. federal regulators ordered grid operators to consider new rules for connecting large electricity consumers.

Among the alternatives analyzed are procedures that would accelerate the connection of facilities capable of temporarily reducing their demand during critical situations.

For this strategy to achieve a larger scale, it will be necessary to establish technical and operational conditions that allow coordinating the needs of data centers with the available capacity of the electricity grids.

The main opportunity lies in better utilizing existing infrastructure while the energy demand for artificial intelligence continues to grow. However, projected savings will depend on how many facilities can participate and how frequently they adjust their consumption.

Source: MarketScreener

Photo: Shutterstock

Written by
Verified Author

I am a civil engineer and I work in the writing area. I have a deep passion for civil works. I constantly seek to update myself on new engineering technologies to integrate them into my work, since my main interest is exploring how innovations in the industrial sector can improve and transform the world.