Speculative Data Center Requests Still Clog Utility Queues
A May 2025 warning that only a fraction of proposed data centers would ever be built still stands. Community opposition, labor and equipment delays are why.
More than a year after Utility Dive reported in May 2025 that only a fraction of proposed U.S. data centers would be built, OilPrice reports it remains unclear how much has changed, with community opposition now a major barrier alongside construction labor shortages, long lead times for electrical equipment and limited power availability.
The pitch that reached American electric utilities over the past two years was close to unprecedented in the industry's modern history: requests for grid connections measured not in megawatts but in gigawatts, arriving from developers whose projects were, in many cases, little more than a land option and a slide deck. In May 2025, Utility Dive put the problem plainly in a headline — "A fraction of proposed data centers will get built. Utilities are wising up." More than a year on, according to OilPrice, it is not clear how much has actually changed.
What has changed is the character of the obstacles. Community opposition has emerged as a major barrier to data center development — a constraint that sits outside the control of both the developer and the utility, and one that no amount of capital can reliably clear. Alongside it sit construction labor shortages, long lead times for critical electrical equipment, limited power availability in key markets, and a general uncertainty that makes every published pipeline number softer than it looks.
Why the interconnection queue overstates what gets built
The core distortion is structural. A developer who is unsure which of five candidate sites will work has every incentive to file interconnection requests at all five. Queue positions are scarce, the wait is long, and the cost of holding a place is usually far smaller than the cost of losing one. The result is that the same computing project can appear as several separate load requests across several utility territories, and the sum of those requests bears no reliable relationship to the megawatts that will eventually be energized.
Utilities know this. The phrase "wising up" in the Utility Dive headline referred to the countermeasures that began spreading through the sector: larger and non-refundable deposits, proof of site control, phased energization schedules tied to demonstrated construction progress, and tariff structures that make a customer pay for the capacity it reserves whether or not it uses it. The logic is that a speculative request becomes expensive enough that only serious projects file.
That is good regulatory hygiene, but it does not resolve the underlying question facing every integrated resource plan in the country: how much generation and transmission to build against demand forecasts that may be double-counted. Build too little and a utility fails its industrial customers and its reliability obligations. Build too much and its ratepayers — households, not hyperscalers — are left paying for stranded capacity across a multi-decade rate base.
The constraints money cannot shorten
The four non-political bottlenecks named in the reporting are worth separating, because they resolve on very different timescales.
- Electrical equipment lead times. Large power transformers, high-voltage switchgear and gas turbines are made by a small number of manufacturers with order books already extended. Adding factory capacity is itself a multi-year construction project, which means the supply response lags the demand signal by years.
- Construction labor. Electricians, pipefitters and high-voltage linemen are trained through apprenticeship programs measured in years, not quarters. A region can bid wages up, but that mostly reallocates a fixed pool of tradespeople between competing sites rather than enlarging it.
- Power availability in key markets. Data centers cluster where fiber, land and tax treatment align. Those same clusters are where headroom on the existing grid has already been consumed, pushing new entrants toward either new generation or markets with thinner infrastructure.
- Community opposition. This is the newest of the four and the hardest to model. It plays out in county zoning hearings and local ballot fights over water use, noise, land conversion and what residents suspect they will pay on their own utility bill.
Only the first two have a clear, if slow, path to resolution through capital spending. The last two are decided by geography and by voters.
The bill ends up on somebody's rate case
Only the first two have a clear, if slow, path to resolution through capital spending.
The financial stakes here are not confined to technology balance sheets. When a utility commits to new generation, transmission upgrades or long-lead equipment orders to serve an announced data center campus, that spending enters a regulated rate base and is recovered from customers over decades. If the campus is built and operating at scale, a large new load spreads fixed costs across more kilowatt-hours and can, in principle, hold rates down. If the campus never materializes, the same investment becomes a cost with no offsetting revenue.
This is why the phrasing in the original Utility Dive piece mattered so much to regulators. "A fraction" is an assertion about the gap between announcement and construction, and every state commission now weighing a large-load tariff or a certificate of need is effectively pricing that gap. The instruments being adopted — minimum take provisions, collateral requirements, staged capacity releases — are all attempts to shift the risk of non-completion back onto the party that made the announcement.
What to track from here
The useful indicators are unglamorous and mostly public. Interconnection queue withdrawal rates show how many filed requests quietly disappear. Utility capital expenditure guidance, and the share of it explicitly attributed to large-load customers, shows how much of the announcement flow has actually converted into committed spending. State commission dockets on large-load tariffs show where the risk-allocation fight is being settled. County zoning calendars, tedious as they are, are where community opposition either stops a project or does not.
Equity markets have largely traded the AI buildout as an assumption rather than an audit. In the most recent session before this writing, the benchmarks closed higher across the board: the S&P 500 tracker (SPY) finished at $771.35, up 0.54% on the day from a prior close of $767.18; the Nasdaq 100 tracker (QQQ) closed at $744.50, up 0.46%; and the Dow tracker (DIA) closed at $517.49, up 0.94%, all as of the last trade on Friday, 25 September 2026. Those levels reflect confidence in the demand story. They do not, by themselves, say anything about how many of the proposed facilities clear a zoning board.
The gap between proposed and built has been the central uncertainty in this sector for well over a year. Nothing in the current reporting suggests it has closed. What it suggests instead is that the binding constraint has migrated — from financing, where capital is plentiful, to transformers, tradespeople, substation headroom and local consent, none of which respond quickly to a larger check.
Key facts
- Original warning: Utility Dive, May 2025: 'A fraction of proposed data centers will get built. Utilities are wising up.'
- Newest barrier: Community opposition to data center development
- Other constraints: Construction labor shortages, long lead times for critical electrical equipment, limited power availability in key markets
- Market backdrop: SPY closed $771.35 (+0.54%), QQQ $744.50 (+0.46%), DIA $517.49 (+0.94%) as of 25 Sep 2026, 20:00 GMT
Frequently asked questions
Why do utilities receive more data center requests than projects that get built?
Developers frequently file interconnection requests at several candidate sites for a single project, because queue positions are scarce and the wait is long. The cost of holding a place is usually lower than the cost of losing one. As a result, the total megawatts requested across utility territories can substantially exceed the capacity that will ever be energized.
What did the May 2025 Utility Dive article say?
It was headlined 'A fraction of proposed data centers will get built. Utilities are wising up,' and it examined the problem that speculative data center requests create for U.S. electric utilities. More than a year later, according to OilPrice, it remains unclear how much has actually changed in the way those requests convert into completed projects.
What is community opposition doing to data center projects?
Community opposition has emerged as a major barrier to data center development. It typically plays out at the local level, in county zoning hearings and land-use decisions, over concerns including water use, noise, land conversion and the effect on residential utility bills. It is a constraint neither the developer nor the utility can resolve with capital alone.
Why do electrical equipment lead times matter so much?
Large power transformers, high-voltage switchgear and turbines are produced by a limited number of manufacturers whose order books are already extended. Expanding factory capacity is itself a multi-year construction project, so the supply of this equipment responds to a demand surge only after a long lag. Long lead times are among the constraints named in the reporting.
Who pays if a utility builds capacity for a data center that never opens?
Utility investment in generation, transmission and long-lead equipment generally enters a regulated rate base and is recovered from customers over decades. If the anticipated large load arrives, fixed costs spread across more kilowatt-hours. If it does not, ratepayers can end up covering infrastructure with no offsetting revenue, which is why state commissions are scrutinizing large-load tariffs.
What indicators show whether the data center pipeline is real?
Interconnection queue withdrawal rates reveal how many filed requests are abandoned. Utility capital expenditure guidance, especially the portion attributed to large-load customers, shows what has converted into committed spending. State commission dockets on large-load tariffs and local zoning calendars show where projects are approved, delayed or stopped outright.
