AI data centers and the role of renewable energy

The demand for electricity is accelerating as data centers expand in the U.S. and across the world. U.S. data centers consumed 183 terawatt-hours (TWh) of electricity in 2024 alone according to IEA estimates, or 4% of the country’s total electricity consumption. In fact, the Department of Energy estimates that this will only increase, as data centers are projected to consume up to 12% of total U.S. electricity by 2028. As AI continues to grow at massive speed and scale, it is critical to ask what role renewable energy can play in this expanding industry.

While there is no federal requirement for data center registration, between 4,000 and 5,000 data centers are currently known in the U.S., the most of any country worldwide. One third of these data centers are in just three states: Virginia (643), Texas (395), and California (319).

AI-specialized data centers cluster servers close together to accelerate communication, which requires constant cooling. These data centers range in size from a single room in an office building to massive, “hyperscale” campuses. A typical size for a new data center is 50-150MW, but hyperscale campuses can draw as much as 1GW of electricity load.

What is a data center?

Data centers are facilities used to house servers, storage systems, networking equipment and associated components that are installed in racks and organized into rows. Data centers are composed of:

Servers: The computers that process and store data account for around 60% of electricity demand in modern data centers.
Storage systems: Devices used for centralized data storage and backup account for around 5% of electricity consumption.
Networking equipment: Switches to connect devices within the data center, routers to direct traffic, and load balancers to optimize performance account for up to 5% of electricity demand.
Cooling and environmental control: Equipment that regulates temperature and humidity to keep IT equipment operating at optimal conditions. Cooling systems vary from 7% to 30% in energy demand depending on efficiency.
Uninterruptible power supply (UPS) batteries and backup power generators: These keep the data center powered during outages. Both UPS and backup generators are rarely used, but necessary to ensure the extremely high levels of reliability that data centers must meet.

The environmental impact of AI-specialized data centers can’t be overstated. Because renewable energy like solar and wind is intermittent, and these data centers require 24/7 power, they are overwhelmingly powered by fossil fuels. Only 25% of AI-specialized data centers are currently powered by renewables, primarily solar and wind; 20% of these data centers are powered by nuclear energy, with additional capacity in development. Data centers also require fresh water for cooling––up to 5 million gallons of water per day. However, global investment in sustainable solutions to power these data centers is expanding, with several key roles for renewable energy to transform this industry in the future.

Power Couples

One potential opportunity for renewable energy to power data centers in the future is co-locating solar, wind, and storage with fossil fuel plants. These “Power Couples,” a term coined by policy experts at RMI, can help AI firms supply data centers with sustainable energy without risking grid reliability. In a Power Couple, new renewable energy resources, such as wind, solar, and battery storage, are built near the site of an existing generator, sending power not required by the co-located load to the grid using its point of interconnection.

Energy matching

Another sustainable opportunity for data centers is energy matching, the practice of ensuring that the electricity consumed by a facility is balanced with an equivalent amount of clean or renewable energy generated elsewhere on the grid. Technology companies that consume massive amounts of energy procure renewable energy capacity through power purchase agreements to offset their energy consumption.

Long duration energy storage

For hyperscale data centers demanding gigawatts of power, long duration energy storage (such as batteries that can store 10-100 hours) is emerging as a core component of a sustainable power system. Data centers require electricity every hour of every day, which the intermittent nature of wind and solar generation can’t account for. While data centers are not ready to rely on renewable energy entirely, long duration energy storage options can store excess solar or wind power to function as a source of energy for the grid.

Long duration energy storage technology, such as pumped hydro, compressed air, thermal storage, flow batteries, metal-air systems, and liquid CO₂, is engineered to operate for eight to one hundred hours or more, which makes it uniquely capable of replacing fossil-fuel peaker plants. The long duration batteries can store energy for longer periods, which removes the issue of intermittent power from renewables.

The AI industry is investing in renewable energy

America’s AI leaders are investing in solar and storage to satisfy data centers’ massive energy demand because solar and wind farms can be completed in half the time of nuclear and gas plants, making renewable energy an ideal addition to support grid flexibility.

“The industry is experiencing a transformative shift with the rise of AI, with unprecedented growth and increased energy demand nationwide. Blymyer partners with EPCs and developers on the forefront of innovating renewable solutions,” says Blymyer President Mike Rantz. “As a multidisciplinary firm, we are uniquely qualified to meet the engineering demands of this new energy landscape. Solar, wind, and storage will be the keystone to powering the future sustainably.”

Blymyer is committed to supporting critical enterprises and ensuring grid reliability across the country with over 25GW in engineered solar and 11,630MWh in installed energy storage capacity.