It’s a political hot potato for state and federal bureaucrats. Invest taxpayer dollars to cut the red tape, replace aging grid infrastructure, and speed up the Ai revolution that will replace millions of jobs with Ai.
It's hard to protect the jobs by throttling the grid, because the grid bottleneck only determines where the disruption originates, not whether it happens. The workers' best immediate structural ally was the bottleneck. Grid diversification among regions, or grid deregulation is no friend of labor in this case. States will compete.
Fascinating article! Would the Grid benefit from a Federal initiative that could cut red-tape and direct resources at the bottlenecks that create these huge queues? How does America's Grid compare with other developed economies?
Thanks, and great question — it's actually two questions, and they cut in different directions.
On federal intervention, the bottleneck isn't primarily red tape.
The interconnection queue backlog — 2,600 GW nationally, with wait times averaging 5 years — is a physics and engineering problem as much as a regulatory one.
You can streamline permitting (and FERC's Order 2023 is trying to do exactly that with cluster studies and financial commitment requirements to flush out speculative applications), but the binding constraint is physical: transformers are on 2-3 year lead times, 345kV breakers are scarce, and there aren't enough lineworkers to build what's already been approved.
Federal coordination could help — especially on interregional transmission, where no single ISO has the authority or incentive to plan across boundaries. But 'cutting red tape' without solving the equipment and labor shortage just moves the bottleneck downstream. Everyone is chasing the same crews needed to do the work, and the workforce is aging. People who retired 5 years ago are being brought back on a consulting basis.
On international comparison: the U.S. grid is actually several grids, which is part of the problem.
ERCOT operates independently in Texas. The Eastern and Western Interconnections are loosely coordinated across dozens of utilities and ISOs.
Compare that to the UK's National Grid ESO, which plans and operates a single integrated system, or Germany's four TSOs coordinating under a unified federal energy transition framework.
The EU is building cross-border interconnection capacity as a strategic priority — the European grid is increasingly designed to move power across national boundaries (which has its own issues such as Germany sucking power out of well-supplied geographies such as Norway, and socializing Germany’s nuclear phase-out mistake into higher prices for those who don’t have their problems) .
The U.S. has roughly 1.4 GW of interregional transfer capacity per million people; Europe has roughly 3x that ratio.
And then there's China, which is the uncomfortable comparison.
State Grid Corporation built out over 40,000 km of ultra-high-voltage transmission in about 15 years — connecting remote wind and solar in the west to load centers on the coast at a pace and scale that no market-based system has matched.
They can site, permit, and build a 1,000 km transmission line in 2-3 years. In the U.S., that same project takes a decade or more.
The tradeoff is obvious — centralized authority vs. property rights, local permitting, and multi-jurisdictional coordination — but the output gap is real.
When people ask why the U.S. can't build fast enough for AI load growth, part of the answer is that the American grid was designed for distributed decision-making, not centralized mobilization.
Where the U.S. does lead is in market design — ERCOT and PJM run sophisticated wholesale markets that attract private capital for generation.
But the transmission planning to connect that generation to load is where the U.S. falls behind, because it's fragmented across state and regional jurisdictions with no single entity empowered to plan nationally.
Short version: the queue problem is real, federal coordination would help on transmission, but the deepest bottlenecks are physical — equipment, labor, and the fact that America doesn't have one grid, it has several. That fragmentation is both the grid's resilience and its constraint.
It’s a political hot potato for state and federal bureaucrats. Invest taxpayer dollars to cut the red tape, replace aging grid infrastructure, and speed up the Ai revolution that will replace millions of jobs with Ai.
It's hard to protect the jobs by throttling the grid, because the grid bottleneck only determines where the disruption originates, not whether it happens. The workers' best immediate structural ally was the bottleneck. Grid diversification among regions, or grid deregulation is no friend of labor in this case. States will compete.
This is such an eye opening article. The part that feels underappreciated:
Demand is moving on a 2–3 year timeline.
Infrastructure is moving on a 7–10 year timeline.
That gap doesn’t resolve smoothly.
It forces prioritization.
And that’s where pricing, location, and access start to diverge.
It will be interesting to see how this gap progresses in the next 12-24 months as demand for AI explodes.
Fascinating article! Would the Grid benefit from a Federal initiative that could cut red-tape and direct resources at the bottlenecks that create these huge queues? How does America's Grid compare with other developed economies?
Thanks, and great question — it's actually two questions, and they cut in different directions.
On federal intervention, the bottleneck isn't primarily red tape.
The interconnection queue backlog — 2,600 GW nationally, with wait times averaging 5 years — is a physics and engineering problem as much as a regulatory one.
You can streamline permitting (and FERC's Order 2023 is trying to do exactly that with cluster studies and financial commitment requirements to flush out speculative applications), but the binding constraint is physical: transformers are on 2-3 year lead times, 345kV breakers are scarce, and there aren't enough lineworkers to build what's already been approved.
Federal coordination could help — especially on interregional transmission, where no single ISO has the authority or incentive to plan across boundaries. But 'cutting red tape' without solving the equipment and labor shortage just moves the bottleneck downstream. Everyone is chasing the same crews needed to do the work, and the workforce is aging. People who retired 5 years ago are being brought back on a consulting basis.
On international comparison: the U.S. grid is actually several grids, which is part of the problem.
ERCOT operates independently in Texas. The Eastern and Western Interconnections are loosely coordinated across dozens of utilities and ISOs.
Compare that to the UK's National Grid ESO, which plans and operates a single integrated system, or Germany's four TSOs coordinating under a unified federal energy transition framework.
The EU is building cross-border interconnection capacity as a strategic priority — the European grid is increasingly designed to move power across national boundaries (which has its own issues such as Germany sucking power out of well-supplied geographies such as Norway, and socializing Germany’s nuclear phase-out mistake into higher prices for those who don’t have their problems) .
The U.S. has roughly 1.4 GW of interregional transfer capacity per million people; Europe has roughly 3x that ratio.
And then there's China, which is the uncomfortable comparison.
State Grid Corporation built out over 40,000 km of ultra-high-voltage transmission in about 15 years — connecting remote wind and solar in the west to load centers on the coast at a pace and scale that no market-based system has matched.
They can site, permit, and build a 1,000 km transmission line in 2-3 years. In the U.S., that same project takes a decade or more.
The tradeoff is obvious — centralized authority vs. property rights, local permitting, and multi-jurisdictional coordination — but the output gap is real.
When people ask why the U.S. can't build fast enough for AI load growth, part of the answer is that the American grid was designed for distributed decision-making, not centralized mobilization.
Where the U.S. does lead is in market design — ERCOT and PJM run sophisticated wholesale markets that attract private capital for generation.
But the transmission planning to connect that generation to load is where the U.S. falls behind, because it's fragmented across state and regional jurisdictions with no single entity empowered to plan nationally.
Short version: the queue problem is real, federal coordination would help on transmission, but the deepest bottlenecks are physical — equipment, labor, and the fact that America doesn't have one grid, it has several. That fragmentation is both the grid's resilience and its constraint.