Federal Study Says U.S. Needs 4,000 Miles of Power-Line Replacements Each Year
The transmission finding places a hard infrastructure requirement beside AI data-center growth—and sharpens the question of who should finance the upgrades.
Listen to this story
The audio brief
Story brief
3 key pointsThe National Transmission Needs Study puts the U.S. grid’s maintenance burden at about 4,000 miles of line replacements annually for decades, alongside major capacity expansion through 2050. Electricity use is projected to rise 16%–25% by 2034, with AI data centers a significant driver. For AI infrastructure developers, the key constraint is not simply securing power: utilities and policymakers must decide how much...
- 01
Most U.S. transmission infrastructure dates to the 1960s and 1970s and is approaching the end of its useful life.
- 02
The study frames 4,000 miles as a recurring annual replacement pace, not a one-time construction target.
- 03
AI data centers are adding demand to a grid already requiring large-scale renewal, making interconnection costs more consequential.
The U.S. will need to replace roughly 4,000 miles of power lines every year for the next few decades, according to the National Transmission Needs Study. The finding ties the physical limits of an aging grid to rising electricity demand, with AI data centers a major contributor to that growth.
The central number is not a one-time construction target. It describes a replacement pace that would have to be sustained over decades, while the country also prepares for higher electricity use. The study says significant transmission investment is needed to meet future demand through 2050.
Much of the challenge is inherited. Most U.S. transmission infrastructure was built in the 1960s and 1970s and is nearing the end of its useful life, the study says. That means new demand is arriving as the network carrying electricity across long distances needs a broad renewal of its own.
The National Transmission Needs Study says approximately 4,000 miles of U.S. power lines will need replacement each year for the next few decades.
Federal projections cited in the study estimate that U.S. electricity consumption will rise 16% to 25% by 2034.
AI demand turns replacement into a cost-allocation question
The study’s cited projections put U.S. electricity-consumption growth at 16% to 25% by 2034. Rapid development of data centers that support AI technologies is a major contributor to that increase, placing AI expansion inside a larger investment problem rather than treating transmission as a separate utility concern.
The difficult question is not whether lines need replacing. The study’s age and demand findings point to a large need either way. The live policy dispute is how much of the incremental cost associated with large new electricity users should be borne by those users, rather than spread more broadly across electricity customers.
A voluntary principle beside a national rebuild
Katie Jereza, an assistant secretary in the Department of Energy’s Office of Electricity, pointed to President Donald Trump’s voluntary ratepayer protection pledge. She said technology companies, utilities and governors have signed it, with the principle that data centers should fund generation-related infrastructure upgrades instead of families.
That pledge offers a financing principle, while the study supplies the scale of the infrastructure challenge. Together, they make transmission a direct part of the AI buildout: data centers add to the demand the grid must carry, while the grid itself requires long-term replacement investment.
For data-center developers, utilities and policymakers, the study leaves a practical test. A national replacement need measured in thousands of miles per year will require investment regardless of AI demand. But as AI facilities become a major source of additional consumption, the terms used to assign upgrade costs will shape whether their expansion is seen as adding capacity—or adding bills.
Sources
- dailyjournalonline.comPower Grid Faces Costly Upgrades Due To AI Data Centers
Reader comments
Newest comments first. Replies stay oldest first.