The equipment queue: transformer and switchgear makers are building to keep the wait, not to end it
Hitachi Energy is building or expanding more than 40 factories to make the heavy electrical gear that AI data centres need. Yet it has told its investors that, as they open, it still expects two and a half to three years of orders to be waiting.
Siemens Energy and Eaton tell the same story. Each is building new plants, and each has a longer queue of unfilled orders than when it began announcing them.
That is not a failure to keep up. It is the plan. The makers are building fast enough to keep the queue, not to clear it, and on their own figures the wait for this gear will still be measured in years in 2030.
The gear is unglamorous. A transformer at the grid connection steps the voltage down. Switchgear inside the site’s fence routes the power and cuts off faults. A data centre cannot switch on without both, and neither comes off a shelf. A large transformer is designed for its site, wound around a core of specialist steel and tested before it ships.
The wait cannot simply be bought away, for three reasons. A new factory takes two to three years from announcement to its first product. A plant runs for decades, and the industry remembers years when its halls stood half empty, so makers add floor space only once signed orders justify it. And every piece of gear must still be installed and tested on site by electricians, who are scarce too.
The clearest measure of the wait is in the makers’ own order books. Take the orders a maker has signed but not yet delivered, and divide by a year of its sales. That gives the years its factories would need to clear the pile if no new order arrived. At Hitachi Energy it is about three. At Siemens Energy’s grid business it is about four, and at Eaton, which makes the gear inside the fence, about one. All three have risen while the new factories were being announced.
Explain it like I don't work in finance
A popular hairdresser is booked weeks ahead. She adds a second chair, but only once the bookings justify it, and the diary still runs weeks ahead.
Counting her chairs tells you she is growing. Counting the weeks in her diary tells you whether you will get a cut sooner. For grid equipment, the diary runs in years.
The makers even say how long it lasts. Siemens Energy estimates that in Europe and North America, demand for large transformers ran about 40% ahead of every maker’s capacity in 2025. Its own chart still has demand ahead in 2030, by about 10%.
So the new factories, when they open from 2027, will meet a queue that has kept growing. The wait may get a little shorter in years, but it will not end this decade. For anyone planning a data centre, the transformer has to be booked before the site is chosen.
For investors
Every queue is longer than when the building began
The three makers that publish a backlog for this business all show the same thing: more years of work waiting, not fewer.
The wait is getting longer, and the new factories are still to come
Signed orders not yet delivered, in years of each business's sales
Hitachi Energy, by its own estimate the maker of one in six of the world’s installed transformers, has $66.9bn of orders waiting, about 3.1 years of work. Siemens Energy’s grid business has €51bn, about four years. Eaton’s Electrical Americas, which sells the switchgear and power rooms inside the fence, passed a full year of sales at the end of 2025.
The levels are not comparable, and the direction is the point. The two grid makers carry HVDC links, high voltage direct current lines that take years to build, so their ratios sit higher by design. What matters is that all three rose together, in the same two years the factory programmes were announced.
The new factories land from 2027
The building is real, and most of it is still ahead. A plant takes two to three years from announcement to first product, so the floor space announced since 2024 mostly arrives from 2027 on.
Most of the new floor space lands from 2027 on
Dated transformer and switchgear plants, from announcement to each maker's own start date
Until those plants run, extra orders can only lengthen the queue. When they do, they add output to a pile that has grown in the meantime.
The makers will not overbuild
The makers are not racing to empty the queue, and they say so. At its investor day on 10 June 2026, Hitachi showed its energy backlog growing towards about $100bn by fiscal 2030 and expected it to “remain stable at 2.5-3x” revenue. Hitachi Energy is “not building over-capacity”, in Hitachi’s own summary of the session.
The money is large and paced to orders. Hitachi Energy plans over $6bn of factory spending from 2024 to 2027, double the previous four years. Siemens Energy added about 20% capacity for €600m and is putting over €2bn more into transformers and switchgear by fiscal 2028. Both are building to the order book they can see, not to the one they hope for.
The industry’s memory explains why. A transformer plant runs for decades, and the makers lived through long years when it stood half empty. Capacity sized to today’s orders is safe; capacity sized to clear the queue would be idle the day demand cools.
By a maker’s own count, the gap narrows but never closes
The strongest evidence that the wait lasts comes from a maker measuring the whole industry. At its capital market day on 20 November 2025, Siemens Energy set demand for large and medium power transformers in Europe and North America against the capacity of every maker in the region.
Siemens Energy's own forecast: the shortage narrows, but never closes
Demand for large and medium power transformers above the capacity of all makers, Europe and North America
Read as yearly flows, each year’s shortfall joins the pile. That reading is mine; if the chart counts only factories in the region, imports fill part of the gap. Either way, Siemens Energy’s own grid backlog grew from €33bn in September 2024 to €51bn in June 2026, by more than half in under two years.
The wait already costs the makers
A queue sounds like pure profit for the seller, but orders priced months ago are delivered at today’s costs. In the second quarter of 2026, Eaton’s Electrical Americas margin fell two points year on year, to 27.5%, as higher metal costs met orders priced months earlier. The makers carry the risk of a long diary as well as its comfort, which is one more reason they will not build far ahead of it.
What could shorten the wait
The case against is real. A narrowing gap can lower years of work while the pile still grows: if orders run 10% above output and output grows 6% a year, Siemens Energy’s ratio drifts from 3.96 to about 3.7 within two years. Eaton’s ratio rose by only 0.01 in its latest quarter. And price rises, early bookings and a stronger euro can hold a ratio up with no longer real wait, since no maker splits its backlog into price and volume.
Some buyers also route around the queue. Schneider Electric sells more power rooms built and tested in its factories, which cut the work electricians must do on site; my claim on that is close to its line.
None of this ends the wait. A queue of three and a half years in 2028 binds a data centre developer as surely as four years today, and the makers’ own plans are built to keep it. The second chair has arrived, and the diary still runs years ahead.
Four readings will show whether the queue holds
My test is plain: at 30 June 2028, Siemens Energy’s grid business should still hold at least 3.6 years of work, and Eaton’s Electrical Americas at least one. Four readings between now and then will show which way it is going.
The ratios, every quarter. Siemens Energy’s grid backlog over a year of its sales stood at 3.96 in June 2026; Eaton’s stood at 1.05. Both are reported each quarter. A ratio that drifts down while orders still outrun output is the queue being kept; a ratio that falls because orders dry up is the queue clearing.
The plants against their dates. Eaton’s Jonesville and Bellevue plants are due in 2027, Siemens Energy’s new Nuremberg areas by 2028 and Hitachi’s Gallman plant in 2029. The strongest confirmation would be a plant that opens on time while its maker’s ratio still rises. The first real crack would be Eaton’s ratio falling two quarters running before its new plants start, or its book-to-bill over twelve months dropping below 1.0.
Price or volume. A backlog can grow because gear costs more, not because more is ordered. Watch for any maker splitting its backlog into price and volume, and for Eaton’s margin once price and cost are back in balance. A queue that earns is a queue the makers will defend.
The makers’ own chart. Siemens Energy’s estimate is the backbone of this piece. A revision that shows capacity meeting demand in Europe and North America before 2030 would weaken the case at its source.
Developers pay for the wait, and the whole grid shares it
If the queue is kept, the makers choose which orders to take and when, and the buyers who booked early are the ones who switch on first. The developer whose plan assumes gear can be ordered once a site is chosen waits behind everyone who booked before it. The makers carry a cost too: Eaton’s margin fell as orders priced months earlier met higher metal costs.
The four makers here are not alone in the queue. GE Vernova builds the same grid gear; ABB, Siemens and Vertiv make the equipment inside the fence; Prysmian, Nexans and NKT make the cable between them. Factory-built power rooms are the route around it, and Schneider’s Systems business, which sells them, grew faster than its loose parts in the third quarter of 2026: 22% against 17%.
Everything upstream depends on this gear. A utility can promise the power and clear a site’s place in the grid queue, and the site still waits for its transformer. That is why the equipment queue sits behind the grid stories in America rations the plug and the plug, not the chip.
The wait should outlast the first new plants
On the makers’ own plans, the queue lasts through 2030. Hitachi plans its backlog at 2.5 to 3 times revenue to fiscal 2030, and Siemens Energy’s chart still has demand about 10% ahead in 2030. My test reads it at mid 2028, early enough to be wrong in public.
The shortage itself is not news; the makers have shown it to investors since at least November 2025. What is new is the timing. The plants announced in 2024 and 2025 start running from 2027, so the next eighteen months show whether new floor space shortens the wait or only keeps it. Whether any of this is already in the makers’ share prices is not a question this journal answers; that work sits in the research notes below.
What would prove me wrong
The test fails if, at 30 June 2028, Siemens Energy’s grid backlog is below 3.6 years of its sales or Eaton’s Electrical Americas backlog is below one year. That is a genuine risk. The drift worked out above already takes Siemens Energy close to the line, and Eaton’s new plants should be running well before the date.
A pass would not prove a longer wait on its own, since price, early bookings, currency and HVDC contracts can hold a ratio up. Hitachi Energy is left out because its own claim already tests its ratio, at 2.5, to 2031. Schneider publishes no backlog, so it cannot be tested this way.
Research notes
Each maker has an initiation note: Hitachi, Siemens Energy, Eaton and Schneider Electric.
The claims on record, word for word
The test this piece stands behind, quoted word for word from the claims page:
The equipment queue is kept, not cleared (30 October 2026)
Claim: “The first wave of new transformer and switchgear plants will not shorten the equipment queue in years of work: at 30 June 2028, Siemens Energy Grid Technologies’ order backlog will still be at least 3.6 times its revenue over the previous four quarters (3.96 at 30 June 2026), and Eaton Electrical Americas’ backlog at least 1.0 times its sales over the previous four quarters (1.05), even though Siemens Energy’s own chart allows its ratio to fall while the pile grows.”
Breaks if: “At 30 June 2028, as each company reports it, Siemens Energy Grid Technologies’ order backlog in euros divided by its revenue over the previous four quarters is below 3.6, or Eaton Electrical Americas’ backlog divided by its sales over the previous four quarters is below 1.0.”
Four earlier claims on this queue:
Schneider Electric sells AI data centres the finished system, not the parts (7 October 2026)
Claim: “Schneider Electric’s AI data centre growth is arriving as finished systems, built and tested in its factories, rather than as loose parts, because skilled hours on site are now scarcer than equipment, so Systems will keep taking a larger share of its revenue.”
Breaks if: “Systems falls below 34 per cent of Schneider’s group revenue, its 2025 level, for the full year 2026 or 2027 while Schneider describes Data Centre and Networks sales as growing double digit; or Products outgrows Systems on an organic basis for two quarters running while data centres remain Schneider’s main growth driver.”
Where later evidence bears on it: not broken, but close. In the third quarter of 2026 Products grew 17% against 22% for Systems, and Schneider says Products growth was “majority volume-driven”: more loose parts, not just higher prices. The nine-month Systems share is about 34% from rounded shares (33.9% to 34.9%), on the line. The share test checks the outcome, not the cause: buyers may want prefabrication for speed as much as for scarce electricians.
Hitachi Energy is spending more than $9 billion on factories, and still plans a three-year queue (22 October 2026)
Claim: “Hitachi Energy is adding factories only as fast as signed orders justify, and its order backlog will stay at or above about 2.5 times annual revenue, roughly three years of work, through fiscal 2030, which ends on 31 March 2031, so the queue for its transformers and HVDC links will not clear.”
Breaks if: “Hitachi Energy’s order backlog divided by its revenue over the previous four quarters, both in US dollars as Hitachi reports them (about 3.06 at 30 June 2026), falls below 2.5 at two consecutive quarter ends before 31 March 2031, the end of fiscal 2030; or its orders for any full fiscal year up to fiscal 2030 fall below its revenue for that year.”
Where later evidence bears on it: the ratio rose to about 3.1 at 30 September 2026 (Hitachi, 28 October 2026). The claim holds. Its South Boston date came from trade press, not Hitachi.
Eaton is shipping more electrical gear than ever, and its order book is growing faster still (26 October 2026)
Claim: “Eaton is shipping more electrical gear than ever, yet its Electrical Americas backlog keeps growing faster than its sales, so the queue for the equipment inside a data centre’s fence, measured in months of sales on order, will keep lengthening until its new plants arrive in 2027.”
Breaks if: “Electrical Americas backlog divided by the segment’s sales over the previous four quarters (1.05 at June 2026, up from 1.04 in March) falls for two consecutive quarters before mid 2027, when the new plants start; or falls below 0.93, its June 2025 level, before the end of 2027; or the segment’s twelve-month book-to-bill falls below 1.0 for two quarters running; or Electrical Americas margin is still falling year on year after Eaton reports price and cost back to neutral.”
Where later evidence bears on it: no new figures by 30 October 2026. The rise slowed to 0.01, and some of it may be price.
Siemens Energy has measured the transformer shortage, and its own chart runs to 2030 (28 October 2026)
Claim: “By Siemens Energy’s own estimate, demand for large and medium power transformers in Europe and North America exceeds the combined capacity of all market players in the region in every year through its fiscal 2030, which ends on 30 September 2030, so unfilled orders keep accumulating even as the gap narrows: Grid Technologies’ orders stay above its revenue each fiscal year and its backlog in euros keeps growing.”
Breaks if: “Before 30 September 2030: Grid Technologies’ orders fall below its revenue for any full fiscal year; or its order backlog in euros, as Siemens Energy reports it (€51bn at 30 June 2026), falls for two consecutive quarters; or Siemens Energy publishes a revised estimate showing transformer capacity meeting demand in Europe and North America before fiscal 2030.”
Where later evidence bears on it: no new figures by 30 October 2026. Its Charlotte slip rested on a press report; Siemens Energy’s last stated date is early 2026, and its release of 3 February 2026 gives none.
The figures behind the charts
Three of the four makers publish a backlog for the business that sells this gear; Schneider reports only a “record backlog”. Dividing backlog by the same business’s revenue over the previous four quarters gives years of work, my arithmetic from each latest filing.
| Maker, business | Backlog | Revenue, previous four quarters | Years of work | Earlier point | Orders against revenue | Latest filing |
|---|---|---|---|---|---|---|
| Hitachi Energy | $66.9bn, 30 September 2026 | about $21.6bn | about 3.1 | about 2.8 (31 March 2025) | 1.66 times, fiscal 2025 | Hitachi results, 28 October 2026 |
| Siemens Energy, Grid Technologies | €51bn, 30 June 2026 | €12.9bn | about 4.0 | 3.56 (30 September 2024) | 1.88 times, nine months to June 2026 | Earnings release, 5 August 2026 |
| Eaton, Electrical Americas | $15.2bn, 30 June 2026 | $14.5bn | 1.05, about 12.6 months | 0.93 (30 June 2025) | book-to-bill 1.3, twelve months | Form 10-Q, 31 July 2026 |
The ratios are not like for like. The two grid makers carry HVDC links, high voltage direct current lines that take years to build. Eaton’s gear mostly ships within a year. Direction matters, not comparison.
The ratio rises when prices go up before deliveries catch up, when buyers book earlier for later delivery, and, for Hitachi’s dollar figures, when the euro strengthens. None of the three splits backlog into price and volume. It can also fall while the pile grows, as the test below shows.
The factories, with each maker’s own dates:
| Maker | Plant | What it makes | Announced | Production due |
|---|---|---|---|---|
| Siemens Energy | Charlotte, North Carolina, $150m | Large power transformers | 14 February 2024 | “Early 2026” as announced; no later date in its releases I found |
| Siemens Energy | Charlotte expansion and a new Greater Richland, Mississippi plant, in a $1bn US programme | Large transformers; high voltage switchgear | 3 February 2026 | Not stated |
| Siemens Energy | Nuremberg, Germany, €220m | Transformers, capacity up about 50% | 5 September 2025 | New areas by 2028 |
| Siemens Energy | Factory network, over €2bn | Transformers and switchgear | 14 November 2025 | By fiscal 2028 |
| Hitachi Energy | More than 40 factories: over $3bn in 2020 to 2023, over $6bn in 2024 to 2027 | Transformers, HVDC, switchgear | 10 June 2026 | Through 2027 |
| Hitachi Energy | South Boston, Virginia, $457m | Large power transformers | Ground broken 29 June 2026 | No date in its release |
| Hitachi Energy | Gallman, Mississippi, $528m | Transformers, more than double its current plant’s capacity | 15 September 2026 | 2029 |
| Eaton | Jonesville, South Carolina, $340m | Three-phase transformers | 12 February 2025 | 2027 |
| Eaton | Bellevue, Nebraska, over $30m | Switchgear | 8 April 2026 | First half of 2027 |
| Eaton | North Little Rock, Arkansas, $242m | Prefabricated power enclosures, doubling US capacity | 2 September 2026 | Not confirmed |
| Schneider Electric | US programme, over $700m | Medium voltage gear, switchgear, breakers | 25 March 2025 | Through 2027 |
Years of work is backlog divided by the same business’s revenue over the previous four quarters, my arithmetic from each latest filing. Corrections to the October pieces: Hitachi Energy’s South Boston plant “should operate by 2028” came from a trade press quote, and Hitachi’s own release gives no date; Siemens Energy’s Charlotte plant slipping a year rested on a press report, and its last stated date is early 2026.
Who is exposed if the equipment queue is kept, not cleared
This maps who operates in each layer. It is not a recommendation, and naming a company is not a view on its shares. Figures in the piece are disclosures used as evidence for the mechanism.
- The grid side
- Hitachi Energy, Siemens Energy and GE Vernova make large power transformers, high voltage switchgear and HVDC systems.
- Inside the fence
- Eaton, Schneider Electric, ABB, Siemens and Vertiv make switchgear, UPS systems and prefabricated power rooms.
- The cable between them
- Prysmian, Nexans and NKT make high voltage land and subsea cable.
- On the other side
- Developers whose schedules assume electrical gear can be ordered once a site is chosen, and the makers themselves if orders cool as their new plants open.
Sources: Hitachi results of 29 January, 27 April, 29 July and 28 October 2026, its Investor Day energy presentation and Q&A summary (10 June 2026), and Hitachi Energy releases of 29 June and 15 September 2026; Siemens Energy Capital Market Day presentations (20 November 2025), shareholder letters of 13 November 2024 and 14 November 2025, earnings releases of 11 February, 12 May and 5 August 2026, and releases of 14 February 2024, 5 September 2025 and 3 February 2026; Eaton's 2025 Form 10-K, Forms 10-Q for September 2024, September 2025, March and June 2026, second quarter release (31 July 2026) and plant releases of 12 February 2025, 8 April and 2 September 2026; Schneider Electric results of 20 February 2025, 26 February, 30 April, 30 July and 16 October 2026, Capital Markets Day transcript (11 December 2025), and its release of 25 March 2025. Years of work are my arithmetic. Personal research, not investment advice.