Why Are Data Centers Waiting Years for a Grid Connection?

We asked our network who's most responsible for closing the gap between data center demand and grid capacity. The results were clear. 37% pointed to the government and regulators, nearly double the next-highest option. Utilities & grid operators and hyperscalers & developers tied at 22% each, with the training and education pipeline seen as the least responsible at 19%, based on 27 votes. 

That instinct, that this is fundamentally a policy and regulatory problem, is only half right. The average wait time for a grid connection in primary US markets now exceeds four years. In Northern Virginia, the world's largest data center market, that figure rises to seven years, according to the 2026 JLL Global Data Center Outlook. Meanwhile, a modern data center shell can be built, from groundbreaking to commissioning, in 18 to 24 months in most US markets.

The mismatch is stark. Buildings are going up faster than the electrons needed to power them can be secured. Power availability, not location, not cost, has become the primary site selection criterion for data center developers. But regulation and permitting are only part of the story. Underneath the connection delay sits a second, less visible crisis: a shortage of the engineers needed to build the grid in the first place. Regulatory reform can speed up approvals. It can't manufacture engineers that don't yet exist.

How big is the interconnection backlog?

The scale of the problem is difficult to overstate. As of the end of 2025, roughly 8,200 projects representing 1,312 GW of generation and approximately 749 GW of storage capacity were actively seeking grid interconnection across the US, a figure covering ISOs, RTOs and non-ISO utilities representing nearly all of the country's installed generating capacity. Separate analysis puts the broader queue, including load interconnection requests, at closer to 2,600 GW.

Time in that queue has grown substantially. The average time from initial interconnection request to commercial operation rose from under two years in 2008 to nearly five years by 2024, and only 19% of projects that requested interconnection between 2000 and 2019 had reached commercial operation by the end of 2024. In PJM specifically, the regional transmission organisation covering 13 US states and Washington, D.C., the average time from application to commercial operation has climbed from under two years in 2008 to more than eight years in 2025.

The bottleneck has also shifted. New PJM data shows delays are no longer concentrated in the interconnection queue itself. Projects now spend an average of more than three years reaching an interconnection service agreement, followed by another four years waiting to come online after approval, as transmission buildouts, substation capacity and strained supply chains become the primary obstacles. This is a structural issue that regulatory reform alone doesn't resolve.

Why can't utilities just move faster?

Part of the answer is physical, not administrative. Large power transformers, the equipment that steps voltage up for transmission and back down for distribution, are in acutely short supply. Wood Mackenzie data shows power and distribution transformers facing supply shortfalls of 30% and 10% respectively, with lead times for large power transformers averaging 128 weeks and generator step-up transformers averaging 144 weeks as of mid-2025. More recent figures show those lead times still climbing. Substation transformer lead times averaged roughly 140 weeks in 2023, rose to around 150 weeks in 2025, and now exceed 160 weeks in 2026.

Regional pressure points illustrate the scale of the shift. In Texas, CenterPoint Energy reported a 700% increase in large load interconnection requests, growing from 1 GW to 8 GW between late 2023 and late 2024. Arizona Public Service has roughly 4.5 GW of committed large-load demand in its queue, with active discussions underway for a further 19 GW. ERCOT voted in June 2026 to introduce a batch-based review system for queued interconnection requests, though the process is still being finalised. Maine became the first US state to ban new data center construction outright, in April 2026, largely as a direct response to the mismatch between grid capacity and demand.

The second crisis: who builds the grid?

This is where the story most people miss the second half of the problem, and where the 37% who blamed regulators in our poll only had part of the picture. Closing the connection gap doesn't just require more transformers and faster permitting. It requires more engineers, and utilities and data center developers are drawing from the same limited talent pool.

The skillset in shortest supply sits at the intersection of two disciplines: liquid-to-chip cooling and HVDC power distribution. The industry has started calling this "Grid-to-Chip" expertise, now considered the gold standard for 2026 infrastructure delivery. Data center developers, backed by hyperscale capital, are able to offer compensation packages and delivery schedules that compete aggressively with traditional utility employers. The result is a steady pull of HV/EHV engineers, protection specialists and commissioning talent away from grid-side programmes and into data center delivery teams.

The arithmetic is unforgiving. Every experienced HV engineer hired onto a data center project is functionally one fewer engineer available to build the transmission and substation capacity that same data center is waiting years to connect to. Utilities are not just competing against other utilities for this talent anymore. They're competing against the very customers whose demand is driving the queue in the first place. No amount of regulatory reform changes that dynamic.

What does this mean for the industry?

Our poll suggests most of the industry sees this as a policy problem first, with over a third pointing squarely at regulators. The data suggests it's at least as much a workforce problem. Interconnection reform, faster permitting, and transformer supply chain investment all matter, but none of them solves the underlying constraint if there aren't enough qualified engineers to execute the work once approvals come through. Solving the queue without solving the talent pipeline underneath it doesn't remove the bottleneck. It just relocates it.

For utilities, developers and the recruitment partners working across this space, the practical implication is straightforward. Engineering talent strategy now needs to be treated as seriously as transmission planning and equipment procurement. The organisations that get ahead of this, building talent pipelines, partnering on training, and competing on more than compensation alone, will be the ones that actually close the four-year gap, not just talk about it.

---

Spencer Ogden partners with utilities, grid operators and data center developers across the energy and infrastructure sector to build the engineering teams closing this gap. Get in touch to talk through your grid and data center talent strategy.