How long does a new large power transformer take to arrive today?
Up to four years. An IEA survey published in February 2025 found that large power transformers take up to four years to procure, twice as long as in 2021, and Wood Mackenzie measured US averages of 128 weeks for power transformers and 143 weeks for generator step-up units in the second quarter of 2025. With replacement this slow, the useful question moves from which unit to replace to which units can keep running safely until a new one arrives.
For most of the last thirty years, a failed power transformer was an expensive problem with a known end date. The replacement arrived within a year, sometimes sooner, and the spare strategy was built around that rhythm. That rhythm is gone. The order you place this year may be delivered after the next budget cycle, the one after it, or later.
This changes what an asset manager has to decide. Replacement is still the answer for some units, but it is now a multi-year project with its own risk. For the rest of the fleet, the work is to know which transformers can be trusted to run until their turn comes, and which need attention before then. The data on lead times and prices explains why; the oil reports you already receive are where most of the answers are.
On this page (7 sections)
What do the lead-time figures actually say?
Three sources measure the wait in different ways, and they agree on the direction. The IEA asked industry participants across markets; Wood Mackenzie tracks quoted lead times in the United States quarter by quarter. Read the figures with their scope: a US benchmark is not a European quotation, but the IEA, looking at transmission grids worldwide, describes the same direction.
| Source | Scope | What it reports |
|---|---|---|
| IEA, Building the Future Transmission Grid (2025) | Industry survey, several markets, conducted in 2024 | Up to four years to procure a large power transformer, about twice the 2021 lead time |
| Wood Mackenzie, April 2024 analysis | United States | From around 50 weeks in 2021 to about 120 weeks on average in 2024; 80 to 210 weeks for large substation and generator step-up units |
| Wood Mackenzie, executive summary, September 2025 | United States, Q2 2025 | Power transformers 128.0 weeks (138.4 in Q1); generator step-up units 143.4 weeks (147.7 in Q1) |
The quarterly figures show a small easing in 2025, ten weeks for power transformers. They still describe a market where two to three years is the normal wait, and where a unit ordered after a failure is a unit missing for most of the next planning period.
Why are lead times and prices so high?
Demand grew faster than factories. In the United States, Wood Mackenzie estimates that demand for generator step-up transformers grew 274% between 2019 and 2025 and demand for power transformers 116%, mostly driven by new generation, grid reinforcement, data centres and electrification. The same study puts the 2025 market deficit at 100% for step-up units and 39% for power transformers, counting the order backlog that long lead times create.
Materials added to the pressure. A transformer core is made of grain-oriented electrical steel; its price roughly doubled after January 2020, while copper rose about 50% over the same period, according to Wood Mackenzie. Skilled labour for winding and testing is the other constraint that manufacturers cite when they explain why new capacity takes years to come online.
| Indicator | Change | Source |
|---|---|---|
| Power transformer prices, real terms | About +75% since 2019 | IEA, 2025 |
| Transformer prices, average | +60% to +80% since January 2020 | Wood Mackenzie, 2024 |
| Power transformer prices, United States | +15% from January 2023 to 2025 | Wood Mackenzie, 2025 |
| Grain-oriented electrical steel | About doubled since January 2020 | Wood Mackenzie, 2024 |
| Copper | About +50% since January 2020 | Wood Mackenzie, 2024 |
New factories are being built. Wood Mackenzie counts close to $1.8 billion of transformer manufacturing investment announced for North America since 2023, and expects the deficit to improve over the coming years. Trade policy and material prices pull in the other direction, so planning on a quick return to pre-2020 lead times is a bet, not a baseline.
From which to replace to which to keep running
When replacement takes a year, the fleet plan can be simple: rank units by age and condition, replace from the top, keep a spare for the worst surprise. When replacement takes three or four years, three things change.
First, an unplanned failure costs more than the unit. It costs the years of operation without it, the temporary arrangements and the load that cannot be served. Second, the order of replacements has to be fixed earlier, because a production slot is booked long before the old unit is removed. Third, the units that will not be replaced in this cycle have to be kept fit for longer, and the money spent on them, from oil treatment to tap changer work and refurbishment, competes with new orders for the same budget.
The decision is no longer about age. Two units of the same vintage can be years apart in condition, and the difference is usually written in the laboratory results you already receive. Our guide on repairing, refurbishing or replacing a transformer covers the economic side; the evidence side starts with the oil.
Which evidence tells you a unit can keep running?
No single number does. Each question has its own evidence and its own reference document, and the answer is only as good as the sample behind it: date, laboratory, fluid type and sampling point must be known before any result is read.
| Question for the next years | Evidence | Reference |
|---|---|---|
| Is there an active fault? | Dissolved gases, their level and their rate of change between samples | IEEE C57.104-2019 for mineral oil; IEC 60599:2022 for fault interpretation |
| Is the oil still fit for service? | Breakdown voltage, water content, acidity, interfacial tension and dissipation factor, read property by property | IEC 60422:2024 |
| How far has the paper aged? | Furans, mainly 2-FAL, read as a range for the degree of polymerisation, never as a date | IEEE C57.140-2017 |
| Can the unit carry more load in an emergency? | Moisture in the solid insulation and the thermal history of the unit | IEC 60076-7 loading guide |
| Is the tap changer behaving? | Gases in the diverter compartment, read with methods meant for tap changers, not for the main tank | IEEE C57.139-2015 |
| Where does the unit stand in the fleet? | A health index that combines these results with documented weights and shows what is missing | CIGRE Technical Brochure 761 |
Two habits make this evidence usable. Read trends, not single samples: one result describes a moment, while the change between dated samples shows whether a condition is developing. And keep the reason with every conclusion: a unit flagged for attention should carry the gas, the clause or the missing value that put it there, so the next engineer can check it. Our guide to dissolved gas analysis explains how the main methods work and where each one stops.
How should the fleet be ordered for the next budget?
Condition alone does not set the order. A unit in fair condition with no spare and a critical load can deserve attention before a unit in worse condition that a mobile transformer can replace for a week. A workable order combines three columns: what the evidence says about condition, what a failure would cost in service and safety, and how long a replacement would really take for that rating today.
From there, each unit gets one of a small number of outcomes: keep running on the normal sampling plan, confirm with a new sample, act in the next planned outage, or start the replacement project now. The point is to decide in a planned window rather than after a protection trip, and to move production slots, refurbishment work and spare allocation to the units where the evidence says they matter.
What can be done this quarter?
- List the units without a spare or an alternative supply, and the ratings for which a new order would take more than two years.
- Collect the latest oil reports for those units and check that each result has a sampling date, a laboratory, a fluid type and a unit of measure.
- Look at the change since the previous sample, not only at the latest value, and order confirmation samples where a gas is rising.
- Rank the fleet on condition and consequence together, and record the reason for each position. The free Fleet DGA Triage does the first DGA pass on a laboratory CSV.
- Book production slots early for the units that will need replacing, and plan refurbishment for the units that will have to wait.
Four years is long enough to see most problems coming. It is too long to discover them after the fact.
Sources
- IEA, Building the Future Transmission Grid, executive summary, 2025.
- IEA, Rising component prices and supply chain pressures are hindering the development of transmission grid infrastructure, 25 February 2025.
- Wood Mackenzie and American Clean Power, Making the connection: meeting the electric T&D supply chain challenge, executive summary, September 2025.
- Wood Mackenzie, Supply shortages and an inflexible market give rise to high power transformer lead times, 2 April 2024.
