I spend most of my day on the phone with EPCs, general contractors, and data center developers. By the time they call, the site is usually controlled, the capital is committed, and the tenant demand is there. Then we get to the electrical equipment schedule, and the room gets quiet.
What comes next depends on who I'm talking to, but it's always some version of the same problem.
An electrical contractor tells me the transformer used to sit in a laydown yard waiting on the building. Now the building waits on the transformer. A GC walks me through a requote where the price held, but the delivery date moved out a full quarter, with no recourse, because in this market the factory queue doesn't negotiate. A developer explains, almost apologetically, that they released transformer orders before the lease was signed, because the alternative was letting a supply chain they don't control set their energization date.
Then there's the call I got this week: a team a few months out from energization, asking whether anyone, anywhere, has a unit on the ground. At that stage, the honest answer is that they're shopping for someone else's canceled order.
The number I hear over and over is 40-plus weeks for the three-phase padmount transformers data center projects actually need. That number doesn't show up in most of the headline market data, which is exactly why so many pro formas miss it.
Bloomberg reported that of the roughly 12 gigawatts of US data center capacity slated to come online in 2026, only about a third was under active construction, and pointed to a shortage of transformers, switchgear, and batteries as a major reason. Let’s dig deeper into the lead times on data center infrastructure and how it impacts your timelines.
Why average transformer lead times are misleading
A recent Wood Mackenzie survey put average distribution transformer lead times at roughly 30 weeks, down from a peak above 100 weeks in 2023. If you read that quickly, it might sound like the problem is solving itself, but that’s not the case. Let me explain why.
That average doesn’t just include the three-phase padmount units you need for a hyperscale campus. It includes everything from those large padmount units to small, single-phase pole-top units designed for residential feeders.
Those categories are not experiencing the same market shifts.
Read more: Transformer lead times: What to expect and how to get faster delivery
Since 2019, demand for padmount three-phase transformers has grown 77%. Pole-top units grew only 30%. So, while pole-top units are running on halfway-decent lead times, padmount distribution transformer lead times are running double or triple their pre-pandemic norms.
The units data centers need sit in the exact segment where demand growth is steepest. The blended average tells you almost nothing about the quote your EPC is going to receive.
If you move up the voltage classes, the problem only gets worse. Power transformers averaged 128 weeks in the same Wood Mackenzie survey. Generator step-up units averaged 144 weeks, with some specialized orders stretching to four years. If your project needs substation-class equipment, your wait is measured in years, not weeks.
The National Infrastructure Advisory Council has called the transformer shortage a clear strategic risk to grid reliability, and NERC's Summer Reliability Assessment flags elevated transformer lead times as a reliability concern for the bulk power system.
When the body responsible for grid reliability starts writing about your procurement problem, it has stopped being just a procurement problem.
Where average lead times stand today
.png)
Switchboards and switchgear are a second long-lead problem
Your transformer lead times are one of your chief timeline concerns, but you also need to consider the switchboards that sit next to them.
A recent survey put switchgear lead times at an average of 44 weeks. Contractor-side reporting in early 2026 runs harsher for the classes of data centers actually specified: standard switchboards near 52 weeks, power circuit breaker switchboards past 84 weeks, and medium voltage gear approaching two to three years for data center specifications.
To be fair to the supply side, switchgear is not as constrained as transformers, and some manufacturers argue their capacity expansions have moved it off the critical path, but the quotes crossing my desk say otherwise. It has not left the critical path; it has just stopped being the longest bar on it.
A switchboard at 52 weeks behind a transformer at 40 does not save your energization date. It gives you a second long-lead package to manage in parallel.
If you step back and observe the pattern, you see that the transformer shortage is not an isolated event. It is the most visible symptom of one buyer class becoming the dominant customer of the entire electrical supply chain. Every package downstream of the meter is repricing around that fact.
Three factors driving transformer lead times higher
There is a temptation to read this as a post-pandemic supply chain hangover that will sort itself out, but the evidence is pointing the other direction. Let’s take a look at the three demand waves impacting these timelines.
The first is data centers. The Lawrence Berkeley National Laboratory projects US data center consumption growing from 176 TWh in 2023 to between 325 and 580 TWh by 2028, or up to 12 percent of national electricity use. US data center power demand is projected to more than triple to more than 80 GW by 2030. Every one of those megawatts arrives through a transformer.
The second is replacement. More than half of the distribution transformers in service in the United States, roughly 40 million units, are already beyond their expected service life. That fleet was largely installed in the postwar buildout, and it is aging out on its own schedule regardless of what AI does.
Distribution transformer capacity may need to grow 160 to 260 percent by 2050 just to meet demand. Utilities buying replacement units are standing in the same order queue as data center developers buying new ones.
The third is everything else that plugs into a wire. Industrial electrification, renewable interconnection, and grid modernization, each consuming transformers and switchgear from the same manufacturing base. These categories do not share a peak cycle. They are all active at once, which is why one study modeled a 30% supply deficit for power transformers and 10% for distribution units.
Supply also can’t flex quickly. Transformer plants take years to build, staff, and qualify. Core materials and copper windings carry their own bottlenecks.
Imports have come in to fill the gap in the meantime. An estimated 80% of US power transformer supply and 50% of distribution transformer supply came from abroad in 2025. That keeps the market moving, but it also means US project schedules are exposed to trade policy, shipping, and every other buyer on the planet competing for the same factory slots.
Read more: Data center electrical distribution system: Key elements
Transformer prices and the impact of tariffs
If the lead time data were not enough, the price data confirms the patterns we’re seeing.
Power transformer prices are up 77% since 2019, and distribution transformers are up 78 to 95% in the same time frame.
Trade policy is adding pressure rather than relieving it. New tariffs, including 50% on copper and elevated rates on imports from several transformer-producing countries, are expected to raise costs and tighten supply even more.
Put rising prices next to lengthening queues, and you get the real takeaway: the market is rationing by time as well as money.
Paying more no longer reliably buys speed. What buys speed is position in the queue, and queue position gets allocated months or years before steel gets cut.
Why equipment procurement now drives your schedule
Most data center development schedules are built around land, permitting, interconnection, and construction labor. Equipment shows up as a line item, not a timeline constraint. That worked when a padmount transformer was a 12-week commodity purchase, but nowadays, the procurement and construction clocks don’t overlap as neatly.
A developer who waits for financial close and permits before releasing equipment orders has added most of a year to energization without changing a single construction assumption. On substation-class equipment, the same sequencing mistake adds two years or more.
And a slipped energization date can impact everything from the tenant's ramp to financing assumptions and, in some cases, exit timing. In a market where hyperscalers and neoclouds are competing on time to compute, the developer who delivers power six months late is often delivering a different, less valuable project.
Which is why the industry conversation has shifted from land-led site selection to power-led. Theoretical power availability is no longer enough. The real screening questions are whether the utility can deliver service inside the required window, whether the transformer capacity is already allocated or orderable, and where your project sits relative to competing load requests in the same territory.
Positioning your build using lead times
The market is responding with real investment in new manufacturing capacity, and that investment is genuine, but none of it reaches a project energizing in the next two years.
A transformer plant takes years to build, staff, and qualify. Capacity announced today won’t become shipped product for a long time. In short, expansion can narrow the deficit we’re seeing, but it still doesn’t hand you a 2019 lead time.
Before Giga, I worked in energy trading. One habit transfers directly: an exposure you haven't hedged is still a position you hold. A developer who hasn't secured transformer delivery is short equipment in a market where demand is outrunning supply. The market doesn't care whether that exposure was intentional.
The sophisticated buyers already know this. Some are even buying manufacturing slots before they've finalized a site. Equipment procurement has moved to the front of the development sequence, ahead of decisions it used to trail.
A playbook for managing transformer lead times
To survive and thrive in the modern energy infrastructure market, you need to follow this playbook:
- Break energization into real milestones. One blended energization date hides the problem. A credible schedule shows separate milestones for the utility application, load and facility studies, service agreement, equipment order release, factory acceptance testing, delivery, installation, and commissioning. Build the schedule this way and the equipment order date moves itself forward.
- Screen sites on equipment reality, not theoretical power. Substation configuration, spare bay availability, transformer age on the serving circuit, the utility's planning cycle, and the volume of competing load requests in the queue all belong in site diligence. Two sites with identical nameplate availability can be a year apart in deliverable power.
- Release equipment orders ahead of full project certainty. This is the uncomfortable one, because it means putting capital at risk earlier. But the alternative is letting a supply chain you do not control set your energization date. The buyers securing slots pre-site have decided that equipment optionality is worth paying for, the same way land optionality always has been.
- Know whose factory floor you are standing on. In a slot-constrained market, the manufacturer relationship is more important than ever. Visibility into the production schedule and a direct line to the people building your units are worth real money when the market average is measured in quarters. Instead of just asking suppliers about prices and specs, you need to be prepared to talk about vertical integration, communication timelines, and what their build process looks like.
None of this makes the shortage go away, but following these steps converts an invisible risk into a managed one. That is the entire difference between the projects that energize on schedule and the ones that explain themselves to their lenders.
Frequently asked questions
How long are transformer lead times for data center projects this year?
The three-phase padmount transformers data center projects need are quoting at 40-plus weeks, based on what EPCs and developers share with me directly. Published averages sit near 30 weeks, but those blend small residential units with data center-class equipment. Power transformers average 128 weeks and generator step-up units average 144 weeks, with some specialized orders extending to four years.
Are switchgear lead times as long as transformer lead times?
Not as long, but no longer a commodity purchase either. Switchgear averaged 44 weeks in a recent survey, with data center class switchboards quoting 52 weeks and beyond per contractor reporting. For a project schedule, switchgear and transformers are now two parallel long-lead packages, and both need order release dates that sit well ahead of financial close.
Why are transformer lead times so long?
Three demand waves are hitting the same manufacturing base at once: AI data center construction, replacement of an aging fleet of roughly 40 million units past expected service life, and industrial electrification. Supply cannot flex quickly because plants take years to build and qualify, and roughly 80% of US power transformers are imported.
When will the transformer shortage end?
Not before most projects currently in development need their equipment. New US manufacturing capacity is being added, but plants take years to build and qualify, and that capacity is arriving into a market that is already structurally short. Expansion can only narrow the deficit rather than restoring pre-2020 lead times.
How much have transformer prices increased?
Power transformer prices have risen 77% since 2019, and distribution transformer prices have climbed 78 to 95% over the same period. Analysts have flagged prices for essential components up roughly 80% over five years, with new tariffs on copper and imported units adding further pressure.
How can data center developers reduce transformer lead time risk?
Move equipment procurement ahead of financial close, break energization into discrete milestones so the order release date surfaces early, screen sites on equipment reality rather than theoretical power availability, and work with manufacturers who control their own production queue. Sophisticated buyers are already purchasing manufacturing slots before finalizing sites.
.png?w=1280)
