Okay. You ever stand near something so hot it changes how you think about heat?
No.
Me neither.
That's the honest answer, and it's a strange thing to admit on a show, because most of what we talk about we've never actually touched.
Right. Okay. Daniel went down a rabbit hole. YouTube, steel mills, as he does. Two videos, and they couldn't have been more different. One's an Indian mill, and the shot is a man standing maybe three feet from molten steel coming out of a furnace, prodding it, keeping it moving, and the room is orange. The other one is a British facility, a guided tour, and the person filming is just repeating the word "enormous" in different tones.
I know the genre.
And what got him was the comments under the first one. Men who'd worked a summer in a mill saying you can't imagine the heat. Hosing themselves down with water to keep standing. Fire-resistant clothing that weighs you down. Molten metal that would put you in a burn unit if a splash got under your collar.
Right.
So he wants the bird's-eye view. Where the industry actually is, what the health and safety divide looks like between those two worlds, and whether there's a route to producing this stuff sustainably. And there's a personal thread too. He grew up in Ireland, in Cork, and there's a row of huge factories along the marina there that used to be the town's employment before it turned into a knowledge economy. He thinks that's the real policy knot.
It is the real policy knot. Everything else is a detail hanging off it.
Then let's start with the map. Where does this metal actually come from?
China. That's the answer, and everything else is a rounding error by comparison. Roughly half of all crude steel on Earth comes out of China. Somewhere in the low fifties percent, depending on the year you check. India is number two, and India's the growth story, the fastest-growing major producer, because they're building infrastructure and cities and they need steel for all of it. Japan, the States, Russia, South Korea, Germany, Turkey, Brazil, Iran, that's roughly the top ten. But the headline is China.
Half of it. One country.
And it's not spread evenly inside China either. Hebei, Jiangsu, Shandong, Shanxi, Inner Mongolia. A handful of provinces carry a huge share of the output and, as we'll get to, a disproportionate share of the damage.
And the UK, which is the second video?
The UK is the counter-example and it's a striking one. A country that essentially invented the modern steel industry, now down to a couple of sites. Port Talbot in Wales, Scunthorpe. Port Talbot's blast furnaces shut down in 2024. That's the last of the traditional ironmaking in Wales, gone. The site is pivoting to an electric arc furnace, which is a recycling operation, it melts scrap. Which sounds like the happy ending, and it is not that simple.
Because an electric arc furnace doesn't employ the same number of people.
It doesn't. A blast furnace complex is a city. An electric arc furnace is a large building. And we'll come back to what that does to Port Talbot, because that's the Cork question.
Before we get to the towns, I want the two processes clear, because everything downstream depends on them.
Two routes. The dominant one is blast furnace to basic oxygen furnace. BF-BOF, if you're reading the industry literature. You take iron ore, you take coal, you essentially cook the oxygen out of the ore using carbon, and the carbon leaves as CO2. That's the traditional route, it built the twentieth century, and it's still the majority of global production. The other route is the electric arc furnace. You take scrap steel, you melt it with an electric arc, and you pour it. That's the EAF route.
And the two have almost nothing in common when you look at them as environmental objects.
Almost nothing. Which is the thing most coverage gets wrong. Steel isn't one industry. It's two industries wearing the same word. And once you see the split, the health and environmental numbers stop looking random.
Okay. Now take us into the first video. The Indian mill, the man three feet from the molten steel.
The reason that video exists, and the reason it looks like that, is that a lot of the world's new capacity is being built where the industry is still in that mode. And I want to be careful here, because the danger isn't one dramatic thing. It's the accumulation. The research on Chinese steelworkers is the best-documented version of this, and it reads like a list of slow injuries.
Go through it.
Hypertension. A nested case-control study found an incidence of about thirty-four percent among steelworkers. Thirty-three point nine five, if you want the decimal.
I do not, particularly.
Fine. Thirty-four percent. And when you break it down, high-temperature exposure raised the risk about one point nine times on its own. Dust raised it one point three seven. Noise one point three three. Carbon monoxide one point three one. Shift work one point four three. Every one of those is a feature of the room in that video.
So the heat isn't just uncomfortable. It's a cardiovascular risk factor.
And it doesn't stop at blood pressure. Hyperuricemia, that's elevated uric acid, incidence of seventeen point three percent, with heat exposure and shift work as independent risk factors. Abnormal bone mineral density in twenty-seven point six percent of steelworkers, and the drivers there were night-shift work and combined exposure to heat, dust and noise.
That one surprised me. Bone density. I'd have guessed the bones were the one thing that survived.
The bones are not fine. And then there's the metabolic side. In a Tangshan cohort, pre-diabetes prevalence was sixty-three point four percent. Nearly two-thirds. And it rose with years of service and with heat, noise and carbon monoxide exposure.
So the longer you do the job, the worse every number gets.
Dose-dependent. Every single one of these is dose-dependent.
The cataracts.
The cataracts are the detail I'd put in front of anyone who thinks of this as "a hard job." It's a classic steelworker injury and it's been documented for decades. Intense infrared radiation from the furnace, and from molten metal, cooks the lens of the eye over years of service.
Cooks it.
That's the mechanism. The infrared lands on the lens and the proteins in it denature, the way a lens protein denatures in any heat cataract. You don't notice it happening. You notice it when you can't drive at night.
And that's the thing Daniel's commenters were circling. They kept saying "you can't fathom the heat," and the honest version of that is, no, you can't, because you've never had years of it.
Right. And the cumulative picture is what those comments were reaching for. It's not one burn. It's thirty years of a body being held at a temperature it was never built to run at, with dust in the air and noise in the ears and carbon monoxide in the blood.
Now, the same hazards exist in the British plant. So what's actually different?
The management of them. The technology is largely the same. A furnace in Wales is a furnace. What differs is enforcement, capital investment and regulatory pressure. In the UK and the EU you have exposure limits that are written down and inspected. You have heat-stress protocols, so there's a rule about how long a man can be in a given heat load before he's rotated out. You have ventilation capture systems pulling dust and fume at the source rather than letting it drift through the shed. PPE that's actually specified and actually replaced.
And that costs money.
All of it costs money. Which is the whole story.
I want to push on the comfortable version of this, though. Because the tidy story is "wealthy countries protect their workers, poor countries don't," and that lets everyone off. Is it actually that clean?
It is not clean. And there's a good study for this. A Romanian case on electric arc furnace plants, operating under Best Available Techniques, the EU's own standard for what a well-run plant should be doing.
Best Available Techniques. That's the regulatory ceiling, not the floor.
That's the ceiling in EU law. Dust capture, dioxin and furan capture, PCBs. And the finding was that even under that regime, workers in the field still show a considerable rate of morbidity.
Considerable.
Considerable is their word.
So the developed-world model manages the risk. It doesn't remove it.
It manages it. You move from "this job will break you at fifty" to "this job will leave marks, and we are arguing about how many marks is acceptable." Which is a real improvement and it is not the same thing as safe.
And that's worth sitting with, because Daniel's second video, the awe-inspiring tour, the feeling in it is "look how far we've come." And the honest version is "look how much is still in the air."
Yes. The tour guide is looking at scale. Scale is impressive. It is also the thing that makes the hazard hard to control, because you're managing a heat load the size of a cathedral.
Hold on. I want to go back to something you said, because I don't think I followed it, and I don't want to nod along.
Go on.
You said sintering is the top air-pollution source and blast furnaces produce most of the CO2. So the pollution and the carbon aren't coming out of the same part of the plant?
And it's the thing that makes the policy interesting. Sintering is the step where you take fine iron ore and fuse it into lumps you can feed a furnace. It's the front end. And it's where the air pollution comes from. Seventy-one percent of the sulphur dioxide, seventy-three percent of the nitrogen oxides, fifty-four percent of the fine particulates, all out of sintering. Meanwhile the blast furnace is where eighty-one percent of the carbon dioxide comes out.
Different chimney. Different problem.
Different chimney, different problem, which means a plant can clean up its air quality and still be a carbon disaster, or cut its carbon and still be poisoning the neighbours. Those are two separate fights and they need two separate sets of rules.
Okay. So that's the human cost on the ground. Zoom out and there's a second ledger. The environmental one.
Steel is responsible for somewhere in the range of seven to eight percent of global carbon dioxide emissions. Put that next to the fact that aviation gets all the attention and sits at around two and a half percent, and you see the distortion. We argue about flying. We don't argue about steel.
Because we don't see it.
And in China specifically, the iron and steel industry alone emits about one point five six petagrams of CO2 a year, plus zero point eight five teragrams of fine particulates.
Those units are impressive and meaningless.
Then here's the one that isn't. That particulate load has been linked to about one hundred and forty-seven thousand premature deaths a year.
A hundred and forty-seven thousand. From one country's steel industry.
Per year. And here's the part I actually cannot get past. The health burden varies by a factor of seventy-seven thousand five hundred between individual plants.
Between plants?
Between individual plants. Same country, same industry, same product. One facility is thousands of times cleaner than another. Which means this is not some unavoidable consequence of making steel. It's a consequence of how a particular plant is built and run, and who is checking.
I want that number unpacked, because "seventy-seven thousand" sounds like a typo you'd let stand.
It's not a typo. The point is that the worst plants are doing something the best ones aren't, and the gap between them is what you'd get if you compared a hospital to a latrine. Think of it this way. Imagine two bakeries on the same street, same ovens, same flour, same bread. One of them vents its chimney into a scrubber and the other vents it straight into the playground next door. That's the scale of the difference we're talking about, except it's not bread, it's fine particulate, and the playground is a city of four million people.
And you grew up in Connecticut, you said. That's not a steel town.
No, Storrs is a university town. But I spent a lot of my doctoring years in Jerusalem, and I want to be careful not to moralise, so I'll just put the fact down. There's a second inventory of eight hundred and eleven Chinese iron and steel enterprises, and it found the sector added three point six micrograms per cubic metre to national population-weighted particulate levels, and that caused roughly fifty-nine thousand premature deaths in two thousand twenty.
Fifty-nine thousand.
In a single year. And that's the estimate for the particulate alone, without even counting the carbon.
So Daniel's question about "is there a path to sustainable production" isn't a climate question with a health asterisk. The health is the bigger number.
The health is the immediate number. The climate is the long number. And here's where it gets hopeful, which I don't get to say often.
Go on.
The mortality intensity of the two routes. Blast furnace to basic oxygen, versus electric arc furnace. Per unit of steel, BF-BOF carries one hundred and seventy-two deaths per thousand gigagrams. EAF carries forty-five. That's a factor of three point eight.
Almost four times the death rate for the same tonne of steel.
For the same tonne of steel. Same product. Different route.
And the policy consequence of that?
Relocating three point six percent of China's eastern blast furnace capacity to scrap-based electric arc furnaces, sited on the low-carbon grids in the southwest, would prevent about twelve thousand three hundred deaths a year. And generate twenty-four and a half billion dollars in combined climate and health benefits.
Three point six percent of one region's capacity.
Twelve thousand three hundred deaths a year. That's the argument I'd make to anyone who thinks decarbonisation is a luxury for rich countries. You don't have to believe a word about climate change to want that trade. It's public health with a carbon side effect.
So what actually gets you there? Because "electric arc furnace" only works if you have scrap.
That's the constraint, and it's a real one. The near-term pathway is scrap-based EAF plus energy efficiency, and it works, but it's limited by how much scrap steel exists in the world at any moment, and by the fact that scrap is a globally traded commodity with its own market.
Which means the countries with the most scrap are the countries that already built everything.
Which is the developed world. So the near-term fix is a fix for countries that already have the steel in their buildings and cars and bridges. For India, building out from scratch, the scrap isn't there yet. The medium and long-term answers are hydrogen metallurgy, hydrogen direct-reduced iron, and carbon capture and storage.
Hydrogen DRI. What's the actual mechanism?
You reduce iron ore with hydrogen instead of carbon. The hydrogen takes the oxygen off the iron and you get water instead of CO2. The physics works. The problem is that green hydrogen at scale is expensive and the plants that make it are being built slowly. And there's a middle option. Natural gas direct-reduced iron feeding an electric arc furnace, which doesn't need scrap and carries about half the mortality of the blast furnace route.
Half.
Half. So even the compromise route, the one you'd get if you couldn't afford the good one, is a large improvement on what's running now.
And the demand side, which I notice you skipped.
I skipped it and you caught me. The demand side is the one everybody skips. Material efficiency and scrap recycling. Using less steel per building, designing so the steel can come back out again, not over-specifying. The literature is explicit that this is equally critical, and it's the cheapest lever, and it's the one with no new technology in it at all.
Because it's not a technical problem, it's a procurement problem.
It's a procurement problem, and procurement problems don't get ribbon-cuttings.
Give me the example, though. Because "use less steel" sounds like a slogan.
Fair. Here's the concrete one. A building's structural frame is usually specified to a standard load, and then engineers add a safety margin, and then the fabricator adds a margin on top of that, and then the supplier rounds up to the nearest available section size. Each step is rational on its own. Stacked together, you can end up with a frame that's thirty or forty percent heavier than the physics actually requires. Multiply that across every office block in a city and you've built a steel mill's worth of demand that nobody needed.
And the fix is just someone checking the stack.
Someone checking the stack, and a client willing to pay for the checking, and a code that rewards it rather than treating the extra tonnage as free. No new furnace. No hydrogen. Just arithmetic.
So now bring it back to the towns. Because all of this is a spreadsheet until you put it in Cork.
Port Talbot is the live case. The blast furnaces are down, the site is moving to an electric arc furnace, and the honest arithmetic is that an arc furnace employs far fewer people than a blast furnace complex. So you can decarbonise a steel town and shrink its workforce at the same time, and both of those are true.
Which is the thing Daniel's actually circling. He's not asking about steel. He's asking what a town does when the thing it was for goes away.
And the answer from every steel town on Earth is: badly, at first. Pittsburgh, Gary in Indiana, Scunthorpe, Tangshan, Port Talbot. The identity is built into the place. The whole town is arranged around one gate.
And the knowledge economy that replaces it doesn't want the same people.
It doesn't want the same number of people, and it's often not in the same place. Which is why the transition literature keeps saying "region-specific pathways," which is policy language for "there is no general answer and someone is going to have to do the hard job town by town."
And the carbon trading and green finance levers you mentioned.
Those are the tools. A carbon price makes the dirty route cost more. Green finance makes the clean route cheaper to build. But neither of them writes a new identity for Port Talbot, and neither of them answers a man who's fifty-four and has done nothing but tap a furnace.
Which is why the neat version of this, where the technology solves it and the towns sort themselves out, is wrong.
The technology is real and the towns don't sort themselves out. And Daniel got to that in two sentences about Cork, which is why I wanted to do this one.
Right.
You keep saying "sintering" like it's a noun that means something. It's a mat. It's a slab of hot ore and it comes off the line and it's still glowing.
My uncle tapped at a mill, and he said the colour told you everything. Cherry red, that's a cold pour, you're going to fight it. Salmon pink, that's a good heat, that's where you want it. White hot, it's running, and if you're smart you step back and let it run. He could read a pour off the colour from twenty yards.
A furnace tapper.
Thirty-one years. And he'd have laughed at the idea that the heat was the worst of it.
That's the bit I want.
He said it was the dust. Fine grey dust, gets into the weave of your shirt, into your hair, into your teeth. He coughed up little grey pearls every morning. That's what he called them. Little grey pearls. Said he could tell how heavy the week had been by how many he got.
That's silica.
That's what the doctor said, eventually.
And the mill would have had it in the fume off the charging floor and the tapping floor, and without capture it goes everywhere, because it's fine enough to behave like a gas.
It went everywhere. My aunt used to find it in the cups in the cupboard.
In the cupboard.
Sealed cupboard.
How does it get in a sealed cupboard?
Nobody ever explained that to me and I've stopped asking.
Dust of that fineness carries on clothing and hair and it goes home with you. That's a known thing. It's how take-home exposure works. The mill doesn't stop at the gate. There are documented cases of workers' families showing elevated heavy metals, because the dust rides home on the coveralls and gets into the laundry and the carpets. The plant follows the man.
That's a hell of a thing to bring home with you.
It's the part of the story that gets left out, because the exposure doesn't end when the shift does.
He had a yellow crane. Number four. If four was down, they didn't pour. That was the rule, and nobody wrote it down, and it was the rule.
Why?
Four was lucky.
The crane was lucky.
The crane was lucky and you didn't argue with it. There were eleven cranes on that floor and four was the one you wanted on the pour, and if four was down you found something else to do that shift.
What happens when you need four and four is down and the heat's ready?
You don't pour. You wait, or you lose the heat. I saw them lose a heat rather than run it with nine.
That's a lot of steel to throw away over a paint colour.
It wasn't the paint. It was four.
He sang, you said?
Part-time. He was in the pub choir and he did the Anvil Chorus. Il Trovatore. He said the mill was good practice, because you learn to project over an orchestra and a mill floor is louder than an orchestra.
The Anvil Chorus is literally about striking an anvil.
He was aware of that. He thought it was funny. The foreman would put it over the loudspeaker sometimes when a heat was going well. Just the one recording, he had a tape of it.
The foreman would play an opera recording over the plant loudspeaker.
On a good pour, yes.
And nobody stopped him.
Nobody stopped him.
So the two things worth keeping from that are the pearls and the crane.
The pearls are the thing. He died at seventy-six, which everyone said was old for that floor, and he'd say it was the pearls that got him and not the heat. He was probably wrong about that.
Probably.
He'd have told you he was right.
Hilbert, the one number I want to keep is that thirty-one. Thirty-one years is longer than most of this conversation has been alive.
That's right.
So all of that, and then we get to the closing. Which I want to keep short, because I think the honest thing here is to leave it unresolved.
Then here's my one thing. The technology to clean this up largely exists. The scrap route, the gas route, the hydrogen route, the capture systems. The question isn't whether we can. It's whether the economics and the politics will let it be deployed before the men in that first video have finished their careers.
And mine is the thing Hilbert's uncle would have recognised. The forty-five deaths per thousand gigagrams against the hundred and seventy-two. That's not a climate argument, it's a worker argument, and it's the strongest case anyone has made for the transition. The towns are the difficult part. The furnaces are not.
There's a version of this where green steel preserves the community the way the old industry did, and I want that version to be true.
You want it to be true.
I don't know that it is.
That's the honest place to leave it. Steel built the modern world with physics that hasn't changed since the nineteenth century, and we have the route to make it humane and we're not sure we have the route to make it fair.
And we'd like to thank our producer, Hilbert Flumingtop. This has been My Weird Prompts.
If you've got a minute, a review wherever you listen helps other people find the show.
And stay warm, or whatever the correct version of that is.
Go inside.
We'll be back soon.
