A transformer supply crisis bottlenecks energy projects | Hacker News Reader

A transformer supply crisis bottlenecks energy projects

spectrum.ieee.org

115 pointsTaurenHunter108 comments

"Ibrahima Ndiaye, a senior principal engineer at GEVAR who led the project, says the breakthrough was figuring out how to give a conventional transformer the capability to change its impedance (that is, its resistance to electricity flow) without changing any other feature in the transformer, including its voltage ratio.

"Impedance and voltage ratio are both critical features of a transformer that ordinarily must be tailored to each use case. If you can tweak both factors independently, then you can modify the transformer for various uses. But altering the impedance without also changing the transformer’s voltage ratio initially seemed impossible, Ndiaye says.

"The solution turned out to be surprisingly straightforward. The engineer added the same amount of windings to both sides of the transformer’s core, but in opposite directions, cancelling out the voltage increase and thereby allowing him to tweak one factor without automatically changing the other. “There is no [other] transformer in the world that has a capability of that today,” Ndiaye says."

From what I can gather, this[1] is the research paper behind the described transformer, however it's not open access. However there is an available technical report here[2] with details.

The design is based on an autotransformer[3], where the primary and secondary windings are connected in series, so effectively one coil. Section 2 in the technical report has details on how the flexible impedance windings.

It's not as trivial as just adding some more windings, though perhaps in practice that's what it looks like.

[1]: https://ieeexplore.ieee.org/document/9959916

[2]: https://www.osti.gov/biblio/1527031

[3]: https://en.wikipedia.org/wiki/Autotransformer

It is hard to believe that no other transformer has used opposite directions in the past. It is such a big oversight.

i am not much into transformers, but isn't so-called zero-flux thing something similar?

https://electronics.stackexchange.com/questions/511880/measu...

> An LPT can weigh as much as two blue whales

I'm sure glad they used such a commonly understood unit of weight. It really helps me understand the size of the item better than pounds or tons.

Unfortunately blue whales are a terrible, archaic measure. There’s too much variability in mass, and then you’ve got the buoyancy problem and different salinity, plus pressure at depth.

Any modern scientific discussion should be done exclusively using standardized measures, like 747’s or 10-story office buildings.

Don’t forget the standardized unit for volume: Olympic swimming pools.

An iconic (and widely-known) skyscraper with a mass in a usable range (because it is not SO huge) might be the Flatiron building.

As a metric European, blue whales give me a lot more intuition to work on than pounds.

Tons would be nice though.

I'm sorry, only long tons or short tons. Why have one unit when you can have two? In metric it's a Tonne.

In many European languages the spelling like "ton" is used for 1000kg.

It's primarily in English where there's still a need to distinguish between metric tons and older units.

In UCUM it’s disambiguated as MetricTon

Ton v Tonne?

"metric ton" vs "short ton" is a little long but precise. Lately I've been thinking about the economics of plastic materials which are made from monomers which cost about 50 cents per pound or, as it turns out, $1000 per short ton. It is all approximate (prices go up and down) and might as well be per metric ton.

2 pounds = 1 kg

Just like 3 feet = 1 m and 4" = 10 cm.

Those are not "good enough for XKCD" approximations, but they are good enough to understand what people are saying.

Now, if I had something similar for Fahrenheit! One could use the hash of the temperature number and it would mean the same to me.

"Now, if I had something similar for Fahrenheit!"

Subtract 30, divide by two.

100 Farenheit, a very hot day, is 35 on this scale, but 37C on the real scale. 32F, the freezing point, is 1 on this scale but 0C for real. -10F, freaking cold, is -20 on this scale and -23C for real. For human temperatures the diff is negligible versus the real conversion. Obviously that freezing point is a big problem for real science, but who does real science with Fahrenheit anyhow, even in the US?

Still an ugly conversion, certainly, but you don't need the "2" in the 32 or the 9/5s ratio to get a good-enough-for-the-weather result.

(Likewise in reverse; multiply by two, add 30. Order matters, of course.)

How do you go through life without once deciding to learn the rough conversion between kgs and pounds (just double it, 1kg~=2lbs) or miles and kms (~1.5x it, 1mi~=1.5km) and so on?

Do you consume american media and just shrug every time you see a number and go "oops there's no way I can understand this now" unless you use Google?

Also european here.

Miles to km, yes. As another European, I know that because it's relevant to understand when they talk about speed limits, top speeds of cars, etc.

Pounds to kg... If it's written media, I just paste it to Google and get the conversion. If it's movies or shows, honestly I don't think I have found an instance while it actually matters at all. If a movie character says "Oh, no, I've gained 10 pounds in a month, I need to go on a diet" you know that they're sad because that's a non-negligible amount of weight, and knowing how much it exactly is doesn't affect understanding of the movie at all.

It's a similar situation to old novels talking about money... When you read a novel from 200 years ago talking about dollars, pounds or whatever, do you look at the historic inflation rate to check today's equivalence? You might, if you're a perfectionist, but the truth is that you can perfectly understand such novels without knowing that just by the context in which they mention money.

Historic inflation numbers are constantly changing depending on the year, pounds to kg is just one number.

2.2 is good enough for most practical purposes, even 1 kg = 1 lbs isn’t that bad and it’s easier to calculate.

It’s like not knowing how many months are in a year…. I mean you don’t need to remember, but people will look at you funny if you’re constantly looking it up.

Pounds is just one mostly-American unit. Add to that miles, feet, inches, ounces, cups, pints, quarts, gallons, barrels, acres, bushels and Fahrenheit, each of which might be relevant at most once a year, even for someone who watches a lot of American media.

It's about as useful as knowing how many sickles in a galleon, or knuts in a sickle — 1 point gained in the trivia quiz!

No-one in Europe will "look at you funny" if you don't know the conversion, except maybe if you live in Britain and it's miles, feet, inches or pints.

Mostly-American or historical European. The Count of Monte Cristo isn't that old yet it was written long before France went metric. Critically measurements are frequently meaningful in works of fiction. Back to the Future's 88 miles per hour would mean something very different if it was the equivalent of 40 kph or 400 kph, but nobody is looking it up while watching.

Sickles in a galleon shows up in one franchise, inches shows up in millions.

Months in a year is very frequently relevant, though. What I'm trying to say is that pounds in a kilo is not something most Europeans know (or remember) because it's not relevant for us. The example with historic dollars was just to illustrate that you typically don't really need to understand measurements in a movie or show to understand the narrative.

Right now I know pounds in a kilo because we're talking about it, but next time I encounter pounds in a context where the value is actually relevant might be years from now, and I'll probably have forgotten by then. Mentions in American media (which was what my parent poster was referring to) aren't enough to prime the memory and keep it fresh because they aren't relevant. If I hear in a show "I've gained 10 pounds", my mind just goes "Oh, they're unhappy because they gained some weight", I don't even stop to think how much a pound was.

It just seems strange to me that you realize you will regularly encounter something and then just not know it is strange to me.

Knowing common constants like the square root of 2 is 1.414, square root of 3 is 1.732, e is 2.718, etc isn't about the difficulty in looking them up it's about not needing to derail your thoughts by looking them up.

There is more to know in life than I will have time to learn. I memorize things I need often, or things I don't need often but in an emergency I will need to know them instantly - everything else I look up on need. It gives me more time to work on the things that are important to me. Unless your goal is to win a trivia contest you shouldn't memorize everything. (nothing wrong with winning a trivia contest if that is your goal, but only a few people can do it across all contests). Every one of your examples are things that I'd look up if I need to - in every context where I'd need them I'd already be using a calculator that has those built in.

Everybody should learn CPR, hopefully taking a day every year for the class turns out to be a waste of time at the end of your life but I still encourage everybody to do it. Some people need to know what e is - but it turns out I haven't needed that since I got out of school (learning to use e was in general good for learning rigorous thinking, but no need to memorize e as that value was someplace in my notes when I needed it). There are obscure things I use all the time I have memorized - but they are specific to my job or hobbies.

Looking stuff up isn’t free, which means it both costs time and frequently doesn’t happen.

Dismissing stuff as trivia doesn’t make that fundamental tradeoff go away. Do you need to know what 7 x 8 is, or the leader of China? No, but the cost of learning commonly used facts is far less than the non existent benefits of ignorance.

The key is commonly used. I don't think I have ever needed the square root of 3 in a context where I wasn't using a calculator anyway, so memorizing it would have been far more effort than looking it up (since any calculator can look this up instantly). Does it matter who the leader of China is - for some people it does, for other it is trivia not worth the bother of looking up. Back when I was getting my degree 7x8 was important to know (even if a calculator was allowed doing the math mentally was a trick I learned - most college calculus problems are selected to have easy math so if the math is hard I probably need to back up a step and fix my mistake - this trick didn't apply to other classes though), now it is mostly useful in context of proving to my 4th grader that I can do her level of math (7x8 might come up in real life, but only in contexts of a much harder problem so I'd be using a calculator anyway)

For me, the identity of the leader of China is genuinely as commonly used a fact as the lb-kg conversion rate, or that the circumference of the earth is almost exactly 4/30ths of a light-second.

Pounds is definitely not something I "regularly encounter" -- at least in the sense that I need to know more than that it is more than 100g and less than 900g. It just doesn't come up.

Perhaps different European countries are different here..

There's a big gap between needing something and benefiting from knowing something.

If a book/movie/whatever describes someone carrying a 60lb backpack is that a big deal? There's a big gap between a 6kg vs a 54kg one.

The cost of ignorance here isn't missing a question on a test, it's not getting the joke or similar in the moment thing.

I think you are overestimating the relevance of minor bits of American (or historic) culture.

We can easily work out from the context what 60lb means. The precise figure is less relevant than the rest of the scene anyway — is a soldier carrying this backpack? An injured child?

All kinds of knowledge are relevant when reading a book. Is the backpack to be carried from Warsaw to Prague? Maybe I have a greater understanding of what that entails than you, even though I'm less sure about the 60lb.

> overestimating the relevance of minor bits of American (or historic) culture.

That’s a judgement you can’t make because by your own admission you’re unaware of the subtleties involved.

That’s the core issue. It’s not that you can’t know what’s happening it’s that you’re assuming you do know what’s happening even when you’re missing significant elements and then in a fit of circular reasoning assuming because you don’t know it must not be important.

Clues embedded in a murder mystery, signs that someone was cheating, implications of fraud, etc etc all depend on concrete details you simply gloss over.

To us the 60lb is the joke. Same for 960 ounces. It stays funny pounds for pounds.

Which then ruins more serious scenes…

Ignorance simply is a net loss no matter the subject. Token for example uses both furlong and league as a means of setting the settings tone, but also to convey information.

I did actually do this the other day, to get some idea of how much Scrooge is offering to pay the boy in the street to go and buy the turkey for him at the end of A Christmas Carol.

So how much was it? It would take time some time to look that up (find the story, find that part of the story...) when you know it anyway and can thus allow me to be lazy.

I think about $25, twice that in the Muppets Christmas Carol because they doubled the amount he offers for some reason.

I feel like you're being contrarian on purpose. Surely remembering that 1kg ≈ 2lb is not that hard...

Well after reading this thread I will remember 2:1 forever...

Really is is just a measurement of how often you encounter facts. What you encounter often you cache and remember. Trivia you never use you don't.

As a 41 year old Norwegian, I have not encountered pounds often enough in my life to remember this fact.

It's not like I have ever used Google to convert either. It just isn't relevant. It is like..what is the currency of Malaysia? I really have no clue. Pounds is just something I almost never encounter.

Closest real use for pounds I can remember having I think was in "pound per square inch" pressure vs Newtons; but in that conversion there is so much else.

(Yes, before this thread I did have a vague sense. I would have said a pound was more than 200g and less than 800g I think.)

Although I agree with you, the phrasing of your reply is needlessly snarky.

I agree, I tend to come out like this. Working on it. I do think it's a huge time saver for your whole life after to remember these 2 numbers. I use this information dozens of times per day.

It took me until early 30s to get bored enough to learn lb-kg conversion factor, despite also being the kind of nerd that memorised π to 11 decimal places (and who now has a greek keyboard on the phone); I managed to go this long just fine as nothing important is ever conveyed in this unit, only people telling me how much they weigh, it's never mattered.

And that's despite being British, so it's conventionally used by normal people, and (when I was a kid) a common unit of measure for produce and recipes.

Well I do I know that pounds are same order of magnitude as kg, miles same order of magnitude as km, and so on.

That is really enough for most reading. As another commenter said, you can read novels from the 70s involving dollars just fine without doing inflation adjustment to present value.

wrote this 100 years ago. Violentmonkey still runs it.

https://userscripts-mirror.org/scripts/show/130277

> Metric intervention Script Summary: Converting that old French system the Brits are still using to the metric standard of science. ~ foot, inch (00',00",00'00,00'00"), yard, mile, stone, Pound-mass/Lbs., Gallon ~ It will totally convert something heretical like: 1'23 1/4" x 2'12 5/8" into something ellegant and civilized like 89.535 cm x 93.0275 cm

Pounds is so imperial. I prefer measuring using Loaves of Bread (lb) to communicate mass.

2 * bw ≈ 630,000 lb

Unleavened bread?

Gave me a much better frame of reference than some huge scalar, personally.

Does 2.8 billion honeybees help?

i miss a banana for scale

The article was interesting until thar line, now reads like a joke..

They should have used Olympic swimming pools.

How many olympic swimming pools do you need to accommodate a blue whale?

A blue whale can be 30m long, 5m high, and 10m wide, so to fit in Olympic-size swimming pools (50m/25m/2+m) you'd need 3 pools stacked on top of each other.

(Technically there's no limit to the depth of the Olympic pool, so you might just need 1 really deep one.)

A blue whale floats in water, does the LPT?

The business is tiny in terms of number of units built, and they're basically hand-built. Some of the interior components are laminated wood. The number of people with the skills to make these things is small, and, since they're expected to last 30 to 50 years, nobody wants to buy from a startup. You want to be able to get replacement parts in 2050.

Plus, General Electric broke up, and GE Vernova now has the transformer business.[1]

[1] https://www.gevernova.com/grid-solutions/hvmv_equipment/cata...

> Plus, General Electric broke up, and GE Vernova now has the transformer business.

Slight clarification, GE Vernova has the (most of) the medium and high voltage transformer business. ABB bought GE Industrial Solutions which includes all of GE’s low-voltage (under 1000 volts) transformers (plus all other sub-1000 volt GE switchgear, breakers, panel boards, etc) plus some 5-15kV medium voltage lines.

Right, but it's the big semi-custom stuff that's hard to get. Distribution transformers (the things on poles) are a commodity.

Part of the problem with generator step-up transformers is that the input side generator specs are determined by the generator, and the output side transmission grid specs are determined by the grid operator. The transformer is the component stuck with the job of making those compatible. That's the kind of problem which leads to N x M variants being needed.

    They’re now reworking transformer designs to use different or less sought-after materials, to last longer, to include power electronics that allow the easy conversion between AC and DC, and to be more standardized and less customized than the transformers of today.

Or maybe they need to work out a new design that is not patent encumbered yet ?

They mention a bunch of new designs: hollow cores, high-temperature insulation, adjustable impedance.

It's sort of hard to replace a grid-frequency transformer (mammoth-sized because of the low frequency) with something like a switched power supply that uses much higher frequencies and thus much smaller coils. The semiconductors required for that barely handle a couple of kV, while the grid uses tens and hundreds of kV.

But a crisis like this is a good thing for the progress of technology. Demand by far outstrips supply, there's obviously good money to be made. Many new approaches are going to be researched and tried.

We know well how to, and we have ways that allow for making very efficient DC/DC converters for the voltage levels that the grid utilizes.

They're quite expensive and big. Building sized. Sure these transformers are technically also building sizes, but it's a garden shed, not a warehouse.

Or maybe the power grid is the problem.

Can you imagine what would happen if we had another carrington effect (which we will eventually) and were suddenly faced with having to build thousands of transformers....

Or on a smaller scale, hybrid and open warfare, domestic terrorism [1] or just rednecks shooting at power lines [2] and transformers [3] already cause a lot of damage to power infrastructure.

An AR-15 or Kalashnikov is enough to cause serious damage, a hit squad or two can take out unprotected infrastructure on a wide scale in a very short time - and most substations consist of barely more than a fence and barbed wire to deter wildlife from entering, not a concrete wall that would thwart attackers.

[1] https://www.schneier.com/blog/archives/2014/01/military_styl...

[2] https://newtoncountytimes.com/stories/shooting-power-and-fib...

[3] https://www.sasktoday.ca/southeast/local-news/carlyle-rcmp-s...

One would expect the electrocution risk would be enough to deter the human attackers.

A sniper team with an anti-material rifle could wreak havoc very effectively and never even be noticed while doing so.

That's the thing with high powered guns, you don't need to get near enough to be electrocuted. Snipers are even better but take more skilled attackers and are harder to explain away in a random traffic spot.

> carrington effect

Event*

https://en.wikipedia.org/wiki/Carrington_Event

Russia specifically targets transformers in Ukraine. 3 years of bombing gets to cause a supply shortage in such a slow capex heavy industry.

Semiconductors have advanced a lot in the last few decades so they also enable different transformers and transmission lines. Very low level materials science. China already has megavolt level UHVDC deployed: https://en.wikipedia.org/wiki/Ultra-high-voltage_electricity...

It's unfortunate that the article doesn't point at the most obvious cause of the shortage: Russia deliberately and systematically destroying Ukrainian power grid for three years, and Western unwillingness to put a stop to that.

The scale of necessary repairs is mind boggling. Here's what just Europe alone shipped to Ukraine to repair the damage [1]:

- 8 827 power generators… with an estimated 1 258 590 kVA in power generators

- 3 597 transformers + 5 autotransformers

- 3 364 generators and 14 transformers from the rescEU stockpile

Then there is US support:

- 18 autotransformers "with more on the way" [2]

- 59 transformers just for Kharkiv [3]

…and that is just scratching the surface and is barely enough to keep the lights on, with Ukrainians installing private batteries and generation up and down the country.

Just last night Russia launched another massive attack at the energy infrastructure [4]. More expensive, rare, long lead time transformers getting blown up, and the raid doesn't even get any attention in the West.

[1]: https://erccportal.jrc.ec.europa.eu/ECHO-Products/Maps#/maps...

[2]: https://x.com/USAmbKyiv/status/1819094722754797600

[3]: https://x.com/USAmbKyiv/status/1792791129513218539

[4]: https://www.reuters.com/world/europe/russia-launches-large-s...

So Transformers caused a shortage of transformers.

Only partly. Also sun, cars, and wars.

Not to mention AI and coin-mining of course.

Transformers is the AI

Everybody wants them.

They are semi bespoke.

Ramping up supply chain takes time.

Something novel we want to sell you in ieee space.

The “Ramping up the supply chain takes time” has an interesting back story:

https://electronics.stackexchange.com/questions/563786/why-d...

> They are semi bespoke.

Thats interesting, surely there are millions of them deployed out there. Was there something that prevented standardization onto few types/sizes?

It's said in the article, I think it's because the exact boundaries of operation depend so strongly on local conditions the specification space has too many parameters in variable space: easier to make them for the situation to spec than build to a standard?

Down at the domestic supply level I am sure they buy thousands of mostly identical units.

Sure, but how much of that is strictly required? It makes sense that transformers have traditionally been completely custom designs, but does something like cable orientation really matter all that much when you've got dozens of projects being delayed by many months? Why not modify the substation design to fit the transformers which are available?

Over my lifetime they seem to have both shrunk and become much more reliable so someone is iterating and improving the designs, just not at a high speed we are accustomed to in the IT world.

My dream for the power grid would be for every home connected to it to have at minimum a battery that could handle two hours of peak load. The operation and maintenance of the grid would be so much easier if parts of it could be disconnected for an hour or two when necessary. The end users wouldn't need to know a disruption even happened. Even better would be capacity for two days, but even just two hours should get us several sigma of improvement.

> Over my lifetime they seem to have both shrunk and become much more reliable so someone is iterating and improving the designs, just not at a high speed we are accustomed to in the IT world.

The reliability is one thing, that got improved mostly by better monitoring (you can spot early signs of degradation by analyzing the oil for residues indicating arc discharge), but the size is in the end fundamentally dominated by physics - you need a minimum gauge of copper wire to transfer a given current.

The only way to improve the size of transformers will be the widespread availability of room-temperature superconducting material, which is as of now not even theoretically predicted, much less proven in practice.

> The only way to improve the size of transformers will be the widespread availability of room-temperature superconducting material

I'm not sure this is enough. You certainly do not need it to be room temperature, but the current density of your superconductor is critically relevant, and it tends to not be great.

There are a few standardised sizes.

I don’t do transformer installations (anymore), but generally only energy producers use bespoke transformers.

That's what the ieee say is in short supply

I work for a high voltage switchgear manufacturer.

Everything is in short supply, transformers, cables, switchgear…

This is the kind of thing that I sort of have the knowledge to jump into (I do power electronics by day) but have no will to actually pursue since hardware manufacturing in the US is very expensive to start, very hard to develop, very punishing to maintain, and very low margins even when you accomplish all that.

Trump talks a ton about "Bringing back manufacturing", but the dude is clearly using it purely for political points. Anyone in the industry knows the supply chains here are weak as hell, and anyone young with a brain is going to SWE or Finance. The industry today is held up by grey beards past retirement sticking with it because no else can do it.

Don't be so pessimistic. There is a lot of manufacturing and supply chains in the US. However US manufacturing needs to be highly automated to compete and so it only make sense for products that are easy to manufacturing in a highly automated way. This implies high volumes, or something else that isn't always easy.

There are a lot of grey beards in industry, but that means there is room for a young person to learn what they know (they do know a lot of great things!) and then take over as their retire/die.

EM said in one of the podcasts that chip shortage -> transformer shortage -> energy shortage. That was about a year ago or so. Asked Chat to make me a portfolio to target the transformer market. Have to say, it works.

Even before the crisis you had lead times of over a year.

This would be a cool yc startup.

And an energy shortage bottlenecking Transformer projects

This seems like exactly the sort of problem Elon is good at solving (or more accurately incentivizing engineers to solve).

xAIs cluster buildout (powered by generators while the grid was being ramped up) would’ve had a lot of lessons.

It seems like the kind of thing Musk would say he’d do but never deliver.

https://elonmusk.today/

That website needs an update. He did deliver a pick-up truck (though that was neither a better truck than the F150, nor a better sports car than a standard 911, like he promised).

Time blindness is a well known common symptom of autism, which he has openly disclosed having. Anyone dealing with time estimates from someone with autism and time blindness should simply adjust for this known attribute of their being instead of relying on or attacking them for their disability.

Naw.

He doesn’t get a pass when it comes to time-related criticism of him if he doesn’t also afford the same pass to his employees and the regulators of his businesses.

> It remains unclear if autism is characterized by a fundamental time perception impairment.

https://pmc.ncbi.nlm.nih.gov/articles/PMC6852160/

Some other reviews state there is evidence, but only on studies for very small time scales. Not saying this definitively states anything, but it is interesting.

From what I can tell, Elon's biggest impact is pushing aside traditional and conservative management and being willing to take larger business risks, allowing the already incentivized engineers who already know the problems they need to solve at paypal, tesla, spacex, solarwinds etc... tackle the problems they needed to to make a change.

In this case, the incentive is there, engineers are working on it, but businesses are hesitant to scale up manufacturing and supply if they can't predict the future demand. Musk might actually be able to help here, if only because his willingness to lose money to get things done.

I thought the same, especially Space X.

I think with space, there's a near infinite unrealized market. If you build a lot of cheap rockets, people will come and pay you to put things on them.

If you build cheaper transformers, will anyone buy them? They will buy some, but is there enough market demand to buy 2 or 3 times the number of transformers over a long period?

Yeah right, but who needs transformers when there are solar roofs, and hyperloop is all solar-powered anyway?

There are massive global multinational companies that have been working on engineering a better transformer for over a hundred years. ABB/GE, Siemens, Schneider Electric, Eaton.. you really think Elon has a fresh take on something that thousands of engineers have been working on for hundreds of thousands of man years?

I realize that SpaceX solved problems that massive companies had been working on for a long time, but electrical transformers are dead ass simple. It’s a bunch of iron/steel and copper/aluminum submersed in oil. The principles of electromagnetism and induction are extremely well understood and have been for a couple hundred years.

Found the cyber truck owner

Reply on news.ycombinator.com