Closing The Energy Gap To Power Clean Data Centers

In an increasingly AI-driven world, data centers are a necessity to power innovation. And it takes a lot of power.

The scale is quite remarkable. Today, data centers use more electricity than most countries. Only 16 nations, including the U.S. and China, consume more. Projections are that demand will increase 160% by 2030.

These figures sometimes lead to a myth that there’s not enough power on the grid to support this demand. That is actually not the case. In fact, there are times when certain grid operators have so much available power they will pay customers to use it, or give it away for pennies on the dollar.

The issue is that the supply of power on the grid is not well-matched to demand. And that the requests from industry for power or clean energy projects are skyrocketing faster than regional utilities can deploy ways to make, move or store that power for use at a different place or time.

The mismatch of supply and demand is especially evident with clean energy, which has unreliable production and gets even more complicated with powering something like a data center that requires significant power around the clock. There are very few utilities in the U.S. that can sign up to service that type of demand, regardless of whether it’s clean energy or fossil fuel-derived. And more demand than ever is being thrust upon utilities than ever before.

This vast demand, combined with a push toward switching to cleaner sources of energy, creates a huge opportunity and the promise of building data centers that are powered solely by clean energy. It also creates some significant challenges for traditional utility companies and clean energy developers alike.

Cracking the clean data center problem is not simply a nice-to-have. It’s an inevitable, necessary step that we as a society need to make a reality.

After decades of relatively predictable and stable power consumption patterns, the data center phenomenon is causing historic requests for power supply. From 2022 to 2023, requests for power or energy storage projects grew 30% from 2,000 gigawatts to 2,600 gigawatts. For context, a gigawatt of power is enough to run a medium-sized city, so it’s roughly like powering 600 more Albuquerques. Given regulatory policies and how long it takes to gain grid approval, this is leading to a growing backlog of projects waiting to join the grid.

Since utilities have an obligation to serve, they are falling back on fossil fuel-derived power, in some cases delaying the closure of coal power plants. Complicating matters further for utilities are regulatory restrictions that prevent them from greenlighting the production of fossil fuel-based power plants. Even if they were able to, permits take years. It’s the same with large-scale clean energy projects, which can take years to gain interconnection with the grid.

To understand the power conundrum we face today, it’s important to understand how the clean energy industry developed. For the past 25-plus years, clean energy has operated by asking questions like “Where can I get this project financed?” rather than “Where should we build this project to balance the grid?” This is because where there is the resource, say, abundant wind, there often isn’t the demand. This isn’t a problem if there’s a way to transport that clean energy to a market with demand, but transmission lines are often decade-plus projects.

The end goal for a clean energy future sounds simple enough: Fill every second of every day with clean electricity. Today, we reach that goal largely through a mix of clean and fossil fuel-based energy sources, and continuing to supplement with coal and gas plants remains a solution. But if a bolder move toward clean energy is the future we want, we need to be creative about how we achieve that. Because right now, regional utilities are up against what feels like an impossible situation. Double your supply. Make it clean. And get it done very soon.

Relying on the existing grid as the only source of power won’t work. To no fault of its own, it is too fragmented and encumbered by regulation and resources to meet the needs of every enhancement looking to tap into the grid. We need a new grid architecture that closes the gap between the existing clean energy and physical power industries. We need an architecture that can avoid getting placed at the back of a years-long queue to plug that clean energy into the grid. We need a supplemental power supply that can work to support the grid with little technological issue and operate much faster.

In short, we need power networks that can be stitched together with the existing grid so clean physical energy is available locally to meet the round-the-clock needs of data centers, but the grid is there to provide physical power at certain times of day or year.

If nothing else, one thing is true: The demand for power is changing dramatically, so we have to, too.