MSE_F2017_06

Materials Selection and Economics Fall 2017 Session · 8 sections 18 cases · Watch on YouTube ↗ all files
Layer 3 — readable edition

§1. Course logistics and module transition [00:00]

§1.p1

This is my sixth lecture plus the introductory lecture, and we're going to change things now because of the students' schedules and other obligations — you do six lecture modules. We're starting now to make those six lecture modules; we're going to take the old modules and break them up from twelve to six. There's no class after today until next class. Next live class is Tuesday October 11th, and that should be Dr. Bellman. And then on the 12th I may just do a recitation, so you can bring questions and ask whatever.

§1.p2

To finish up your twelve lecture modules, I did lecture this material back in 2016 in the fall. If you do those four lectures — 6, 7, 10, and 12 — I looked at them briefly this morning, there's not going to be too much overlap. There will be new material in there. And if you look at spring 2016, I did something called Structural Materials Selection where I didn't spend as much time on externalities and economics. If you look at 11 and 12, that would be your next six lectures to complete this module for this semester. It's a little weird because you're on twelve lecture modules and we're in this transition where we're going to six lecture modules. It doesn't really matter — you're not going to be quizzed on it anyway.

§1.p3

Some of my travel got changed, but Dr. Bellman will finish up some things on the 11th and maybe I'll do something on the 12th. We'll let you know. And then we're done with the live lectures. I told you we'd be done before Halloween and before it starts getting too dark out early in the morning. Even though we're not doing anything live for the next week and a half, I would encourage you to watch some of the modules just as if it were being done live. I understand a couple of times the video was dark. I have to clean up some of these old recordings, and I did look at some of them this morning. I didn't watch the whole thing, but I think those six lectures — there's not a whole lot of redundancy. I tried to cut it out.

§1.p4

Any questions on that? Hopefully you got your half-page write-up on what your topic is going to be. Brian's supposed to be checking on this. I responded to people who sent them directly to me; Brian's going to check through them and we'll get that straightened out this week, and you can start writing your papers, which are due I think October 31st. The first draft has to be turned in, and then you have your little summaries to write up. A summary is due sometime around mid-October. If you watch these six modules you'll be done with this module, and you may even be done with Dr. Bellman's module. So you may be done with a couple of modules. Any questions on the assignments?


§2. From Bessemer to the basic oxygen furnace [03:45]

§2.p1

We were talking about melting of metals, and specifically one of the problems — a serious problem — was how do you melt steel, which melts at 1600 degrees centigrade. You can't get but about a thousand degrees centigrade out of a normal fire burning in air. If you take hydrocarbons and burn them in air, you're going to get 2,000 degrees Fahrenheit, a thousand degrees C. So with a little bit of help, you can melt cast iron, you certainly can melt copper alloys, but you can't melt steel. Before Henry Bessemer, the way they made steel, they would take cast iron — which is four or five percent carbon — and they would blow air through it. The carbon would burn off as carbon monoxide, and it would leave a lower-carbon version of iron, which is wrought iron. Wrought iron is malleable, can be forged, as opposed to cast iron, which is not malleable. You try to forge it, it'll crack. It's just brittle.

§2.p2

Henry Bessemer came along with his converter, which was a better way to blow oxygen through the iron. They called it puddling iron — if you want to learn about puddling you can go look it up on Wikipedia, but guys would just stand there blowing compressed air through a bath of steel, and it was a very slow process. Andrew Carnegie came along, the railroads were being built around the world, and he started making steel in a big way. If you want to learn about how he got into this business, there were a couple of shady deals in Pittsburgh, and he ended up owning a steel company which is now U.S. Steel as it's come down through the last 150 years. He became the richest man in modern history, even richer than Bill Gates or Warren Buffett or Carlos Slim or any of the guys today that are worth a lot of money. In constant dollars he was the richest.

§2.p3

Carnegie actually built what they called integrated steel mills. You start out with coal and you have a coke oven where you put the coal into a rarefied atmosphere — just like we made charcoal. You didn't have a lot of oxygen, you burn off all the organics out of the coal, and you end up with something that has some structural integrity, basically a graphite, which is called coke. It's not the kind you drink, it's not the kind you shoot, it's basically this black lump of stuff. The coke ovens in Pittsburgh were one of the major reasons for the Air Quality Act of 1972. The coke ovens burning off all the organics in the coal just created a terrible atmosphere from a human health perspective. Just going off in the air. And that was acceptable back then.

§2.p4

When I worked for Bethlehem Steel, they had changed their iron ore charge from something that was basically just a rock that came out of the Mesabi Range in Minnesota. It was virtually pure iron ore — wonderful material — but we used it all during World War Two and just after. So they went to sintered iron oxide, which they got from a bunch of rock in Minnesota but they had to do some processing. When they started putting the sintered iron oxide in with the coke and the limestone in the blast furnace in Bethlehem Pennsylvania, they found they were producing five thousand pounds a day of cyanide going up the stack. Didn't bother to tell anybody because that would upset some people, but it didn't matter legally — they had no duty to clean up. They actually did fix it; they realized it was something that should be fixed, and they figured out what the chemistry was. Cyanide is just carbon-nitrogen, right? If you start burning carbon in air with a lot of oxygen, you can get a chemistry that gives you CN. But people living in Bethlehem just had to breathe the air. That was the different rules they had back in the old days.

§2.p5

Because of those types of things, eventually Congress came up with the Air Quality Act of 1972, which is one of the beginning laws for the Environmental Protection Agency. It's just like the labor unions. Whether you're pro-labor or not, if you go back and look at what the management was doing to the workers before unions, back a hundred years ago — they basically hired Pinkerton guards when they had a problem. These guys could fire at will, and they would murder them. There are a number of stories at Alcoa and U.S. Steel of Pinkertons being hired to come in and shoot down some of the rowdy labor force. They were all immigrants anyway — you know what the president thinks of immigrants. Sorry. Again, it was a different standard.

§2.p6

Integrated steel mills got to the point where they would consume tens of square miles and would cost about five billion dollars — today would be about twenty billion. The last integrated steel mill to be built in the world by a private company was Bethlehem Steel Burns Harbor in Indiana, started in 1964, finished about eight or ten years later. It could produce five million tons per year, which was five percent of the U.S. steelmaking capacity at the time. It had two blast furnaces, so if one went down you didn't shut down the whole mill, because it can take three months to reline a blast furnace. When I lived in Bethlehem Pennsylvania, they relined the blast furnace, and that's when everybody got new grills for their patio. Everything went — it was a huge project, cost a couple hundred million dollars, and they were building all kinds of things in the shops, many of which actually went to the steel mill but most of which went to the people's homes. [laughter] Extra stuff — they call it a government job, if you've ever heard that.

§2.p7

There was a lot of nepotism — not necessarily family nepotism, but people scratching each other's backs. When you got into management in the steel companies, they were pretty corrupt. In the 1930s, Bethlehem Steel had six of the top ten paid executives in the United States working for Bethlehem Steel, and Eugene Grace, who ran Bethlehem Steel through the '30s and '40s, got paid a bonus on every ton produced. Even though during the Depression Bethlehem was losing money, he got paid a bonus for every ton produced. Whether they were making money or losing money, he made a fortune. People just kind of accepted this as the way it was with management. A lot of these companies like Bethlehem went bankrupt a few years later in the '80s. When General Motors went bankrupt and someone asked, well, what do you think about General Motors' bankruptcy, my comment was, well, they earned it. The steel companies earned the bankruptcies they went through. There's a lot of arrogance.


§3. Blast furnaces, open hearths, and the BOF [12:10]

§3.p1

So the blast furnaces — we talked about this before. You've got this big tall furnace which might be ten stories tall nowadays. You pour coal in the top — the fuel you put in could be oil or natural gas, just to speed up the process — you have all your coke, limestone, and iron ore, and those three components go in the top and they work their way down. Every now and then, about once or twice a day, you tap the heat, which means you drain off some of the liquid cast iron at the bottom of the blast furnace. You have to preheat the air, so you have these two big brick ovens, one of which is being preheated by the exhaust gas and the other one is preheating the air that's going into the furnace. About twice a day you switch over your ovens, so the one that has cooled off now gets heated up and the other one that's now hot from the exhaust gases will preheat the air. You just switch it over.

§3.p2

We did the same thing with open hearths, which we don't have anymore in the world, but that's how Andrew Carnegie made his steel, after the Bessemer converter. We used open hearths for about a hundred years. You have this big brickwork — same type of thing in glass front melting furnaces. Instead of a bed of steel you've got a bed of glass, and in order to get the temperatures you have to preheat the air coming in. You can't get more than 2,000 degrees with normal hydrocarbon fuels, whether it's natural gas or oil or coal. You're only going to get about 2,000 degrees. Take my welding course if you want to learn why.

§3.p3

In the late '50s, early '60s, some Austrians decided, well, I can go back to the puddling process — instead of blowing in air, I could blow in pure oxygen. In Germany they had developed the Linde process for liquefying air and distilling it, so they had lots of excess oxygen. They had lots of nitrogen, and 1% argon — you basically distill the air into those three components. You can blow liquid oxygen down through a water-cooled copper lance, and in this vessel that might be four or five stories tall, you base it with just five or ten feet of liquid cast iron in the bottom. You blow this in, and in twenty minutes you can burn off all that carbon. When it's going, it's really exciting. It's the same thing — you've got a flux of limestone and sand and an oxygen lance. But the volume inside is about twenty or thirty times the volume of the steel, and it foams up as a big froth. Sometimes people commit suicide by diving into it. It's happened — when I was at Bethlehem Steel — you can tell because the body floats. Steel is heavier than water, so you float right on top. You see this depression in the shape of a person.

§3.p4

Why doesn't oxygen — because there's carbon, and the carbon burns to carbon monoxide, which is a reducing gas. If you keep blowing forever, after you've burned off all the carbon, you'll start burning off the iron and form iron oxide. How do you tell? They know from experience how long to blow — how many tons of liquid oxygen do you put in to how many tons of cast iron, you know how many tons of carbon you had. From experience they don't blow below about 0.05 percent carbon in the steel, because at that point you get to burning off iron oxide and you're going back to where you started from. That's not very efficient.

§3.p5

The person who worked all this out was a guy named John Chipman. If you go right across the hallway through the courtyard here, you hit the Chipman Room. John Chipman was a physical chemist from Georgia Tech who came to MIT, and he explained the physical chemistry of steelmaking for the world. Before that, it wasn't quite empirical, but it was sort of experiential alchemy — they didn't really know how to control it. John Chipman in the '30s, '40s, and '50s applied the principles of physical chemistry which he had learned on aqueous chemistry — water-based chemistry — and he said, these principles should apply at high temperatures, and he showed that they did. John Chipman, and then Tom King and John Elliott after him, were two of his protégés — Tom King was the department head when I was a student. I knew John Chipman; he had retired in '62, I didn't start here until six years later, but I went to see him. He's a wonderful Southern gentleman, but his name, if you go to Japan or China, he's like a god, because he explained steelmaking — this huge industry — and made it such that it's no longer this sort of alchemy empiricism.


§4. Continuous casting [17:53]

§4.p1

So from Andrew Carnegie's day until about 1970, they would take this steel and they would cast it into ingots. These are the cast iron ingot molds on railroad cars, and there's the hot steel that they're actually getting ready to pour. This is what they call a hot top — it's a ceramic thing that helps feed the liquid into the mold, because the metal shrinks as it solidifies and you want to keep feeding liquid in on the top, and it's obviously hotter at the top. You can see the glowing red here, and it takes a day for this stuff to solidify. You can't move these cars for at least twelve hours or so once you pour it in.

§4.p2

But then some people — again I think it was Austria — decided they could try continuous casting. Continuous casting had been done on lower temperature metals like brasses, copper alloys, and I told you they were doing it on gold alloys, which is not all that different than brass. In continuous casting, you have a ladle with molten metal that you melt somewhere else, and you pour it into something called the tundish, which is just a bathtub holding things, and it drops the metal into a mold. The mold is water-cooled copper, and you solidify a little skin on this. In this yellow band, you actually have molten metal, and the whole thing is pliable, and so you have these rolls. This whole thing stands about seven stories tall, with liquid metal in the center. If you get a breakout, it takes about a week to clean it up, because you've got three or four hundred tons of molten steel hitting the floor of this shop. They do have a stopper in the tundish, but even there you've still got this thing that can be ten feet wide and ten inches thick, and it's 50% liquid on average. If that skin breaks — and they did break, we'd have breakouts once or twice a year in the '70s when I worked at the steel company. They almost never have it today because they've got all kinds of process controls and infrared temperature sensors, and it's too expensive to have a breakout.

§4.p3

When you're sitting there processing 5,000 tons of steel a day through here, you don't want to lose 25,000 tons of production over a week while you clean it up, plus the whole cost of essentially rebuilding six stories' worth. It's not a small job to fix this, but you do have breakouts from time to time, not very many anymore. In fact, the blast furnaces used to need to be relined about every six months. If you go back four hundred years ago they probably had to be relined every three or four months, which might be a season. But now blast furnaces will go for three or four years before they have to be relined, because they've improved the ceramics, the fire brick that's in there. They cost too much — it can cost several hundred million dollars to rebuild a blast furnace, it can cost a hundred million dollars or a couple hundred million to rebuild all the warm parts in this continuous caster. The Japanese had the record for running continuous casters for a year or a year and a half at a time, making a continuous strand of steel that goes for about a hundred and fifty miles. They basically cut it off with a torch down here. What's the torch? Pure oxygen again. You blow pure oxygen on something that's hot like that, it goes right through it like a nice hot knife through butter. It's a cold oxygen knife through hot steel. Pretty exciting to see some of these things, and also to see the size of these things.

§4.p4

So you cut slabs, and the diagram says go to storage. In fact, as we've increased productivity, we don't go to storage anymore and let this thing cool down — we actually go directly to the rolling. We've had to reconfigure the steel plants, and the most efficient steel plants are the ones that have a good layout so the process flows straight through.


§5. AOD refining and stainless steel sensitization [22:29]

§5.p1

There's a process called argon-oxygen decarburization, and it's the way we make all the stainless steel in the world today and most of the nickel-based superalloys for jet engines. Very high quality process. You think it looks like the BOF, the basic oxygen furnace, where you burn the carbon out of the steel. But in this case you take the low-carbon wrought-iron type of steel and you blow oxygen and argon through, such that the oxygen is at a lower partial pressure — you generate carbon monoxide at a lower pressure — and you can get your carbon down below 0.03 percent. You can get it down to 0.01 percent. That means you can make very good quality stainless steel.

§5.p2

The seeds of this were generated down in the basement of Building 8 with one of the graduate students of John Chipman, who was just doing basic science on the chemistry of burning carbon out of steel. He left MIT, and he said, we could make stainless steel this way by blowing argon through. And now this is an AOD vessel. It has dropped the price of stainless steel by a factor of five to ten in the world, and it's better quality stainless steel — it's cleaner. It applies to nickel-based alloys, and now a lot of your cast alloys go through this AOD refining process. It was basic research in the basement of Building 8 here that started it. It was first done when Krivsky [?] left MIT and went to work for a very small steel company that was willing to take a chance. The big steel companies were big and bureaucratic — they wouldn't take any chances.

§5.p3

Student: [inaudible question about carbon levels]

§5.p4

Yep. The lowest carbon you're going to get out of that BOF is 0.05 percent carbon. If you want to go lower you can, but you're going to start losing iron as iron oxide, which is just wasted energy. If you want to go lower, you actually probably start at one-tenth, and then you do the AOD, and the AOD can get you down, in some cases, to an order of magnitude lower in your carbon. For good weldability and corrosion resistance you want to be below 0.03, and you get six times lower than that, at relatively little cost. It also helps you with recovery of your chromium. That was why the price of stainless steel dropped so much. When you just throw the chromium in there, a lot of the chromium would go up into chromium oxide in the slag. Just like the problem of oxidizing the iron, you oxidize the chromium, which is fairly expensive. The nickel you throw in is an alloying element that has been made somewhere else by an electrolytic process at a nickel mill, but you made your stainless steel that way. Now we don't generate as much slag that has a bunch of chromium oxide and iron oxide. We control the carbon chemistry by controlling the vapor pressure of the carbon monoxide by diluting it with argon.

§5.p5

If you really want to save money, you actually start out with a nitrogen-oxygen mixture, but then you don't want the high nitrogen in most cases, so you switch over to argon-oxygen, and that way you don't have to use as much expensive argon. Steel mills are using so much liquid oxygen and liquid argon that Linde will come and build a refrigeration plant to liquefy air right there next door and pipe it right over to your mill — a six-inch line, a lot of liquid oxygen. The only people who use more liquid oxygen in shorter periods of time is the U.S. Air Force sending rockets off. Burning oxygen is potentially very violent. Anyone ever seen the Purdue University charcoal-starting contest on YouTube? If you look up Purdue and charcoal starting, they used to have a contest of who could design the system that would get your charcoal briquettes burning fastest. They basically had to stop it when one person decided to bring a twenty-foot pole with a bottle of liquid oxygen and pour it on the thing, and the carbon — it just goes, it's all gone. Pure oxygen burns things really quick. Plus the fire department in West Lafayette wasn't real happy with the safety of a bunch of people standing around this thing. But you can see it on YouTube.

§5.p6

In steels, in stainless steels, and to a certain extent in some nickel-based superalloys, you don't want a lot of carbon. What happens is — I have a grain structure, I'm going to just kind of draw the grains in here, ideally in your metal — and if you have carbon, carbon diffuses faster than chromium. If you heat up the stainless steels typically in the 800 to 1200 degree Fahrenheit range, which it turns out you will have in the heat-affected zone of a weld — you can't get around it, if you're going to make a fusion weld, somewhere in the heat-affected zone you're going to have that temperature. What happens is the carbon diffuses to the grain boundary, the chromium also diffuses to the grain boundary but from a shorter distance, because its diffusivity is about a hundred times slower. So that little bit of carbon, if you're at 0.05 carbon with 18% chrome, you will precipitate out chromium carbide. I'll call it XY — it could be Cr₂₃C₆, these carbides are really complex — or Cr₃C₇, there are a bunch of different carbides. That means a region of about two microns on either side of the grain boundary will be depleted in chromium. If you go and analyze that, you might find it only has ten percent chromium in the depleted region. There's chromium everywhere, but the carbon has to diffuse from a larger distance, and it's a small atom and it does diffuse from a larger distance, and that denudes — makes the grain boundaries lower-chromium and therefore not as corrosion-resistant.

§5.p7

In fact, John Wulff, my academic grandfather, had a patent in the 1940s to sensitize stainless steel, heat it in this region, get something like this, throw it in nitric acid overnight to make stainless steel powder. The nitric acid doesn't attack the 18% chrome, it does attack the 10% chrome, eats away the grain boundaries, and you get typical intergranular cracking. This cost General Electric a couple billion dollars in the nuclear reactor industry in the '70s from the welds getting sensitization cracking. If you were Westinghouse, you used Inconel — about ten times the price of stainless. If you're building a pressurized water reactor, because of the higher temperatures, they needed Inconel. General Electric had a boiling water reactor, they could use stainless — less expensive material but more susceptible to stress corrosion cracking. You get cracks through the reactor safe end. That means that you're not going to be able to flood the reactor with water if you have a near meltdown. And you will not have a near meltdown — you'll have a real meltdown.

§5.p8

So there's a lot of corrosion chemistry to stainless steels, and a lot of it is controlling the carbon content, which goes back to the steelmaking process. It's all because of AOD refining that has made all this practical today. Any other questions on that?


§6. The productivity revolution in steelmaking [31:37]

§6.p1

This is a chart I put together a few years ago. Back in the 1600s when they had Saugus Ironworks, it probably took about one person-year per ton. The guy had to go chop the trees, make charcoal out of the logs — you basically stack up the logs, cover them with straw and mud, and then you light the inside, and there's not enough oxygen so you burn off all the lignin in the logs, and you end up with — not coke, we call it charcoal, but it's like today's coke made from coal. So that was your carbon for your blast furnace. You then had to go get the iron ore, dig it out of the ground; you had to get the limestone, digging it out of the ground. When it's all manual work, it takes a lot of time. And then you've got to run your blast furnace. So I estimated it probably was at least a person-year per ton. They probably worked sixty-hour weeks — these people were working long hours.

§6.p2

Then Bessemer came, we probably dropped down and got bigger and more efficient over the next 200 years. Bessemer learned how to blow air into a converter. Carnegie came along and invested in things like the basic open hearth. Things kept marching along. After World War Two we had the basic oxygen furnace in the '50s and '60s; we had continuous casting in the '70s. Continuous casting went from having a yield of 65% to 97% in tonnage products — pounds poured versus pounds shipped. You went from 65% to over 95% because you didn't have to cut off the top third of every ingot, because of the piping porosity and things like that. The continuous casting process was a big jump in productivity of saleable product. We used to have to pour 150 million tons of steel to make a hundred million tons that we could sell in the United States.

§6.p3

Mini mills, which we'll talk about in a second, and process technology cut it off here. When I worked for a steel company in 1975, it was 50/50 labor and raw materials. Back here it was labor-intensive. Today it's like 90% raw materials, because we've taken all the labor out. Here's the labor on a log scale, and we're down to less than twenty minutes a ton — person-hour per ton — to produce steel. As Gordon Forward, a graduate of this department who started a mini mill, said: if the Koreans or the Japanese want to sell steel in the United States, all he has to do is produce it within $30 of their cost. Because the raw materials — coke, limestone, and iron ore — they're transported, they're commodity around the world. Everybody's paying the same price pretty much. But it costs $30 a ton to ship a ton of steel across the Pacific Ocean. So Forward said, if I can make it for only $30 a ton more than they make it in their big fancy integrated mills, I could build a smaller mill and I can beat the price. And he did, and we'll talk about that.

§6.p4

The whole story of steel over the last fifty years — there's a book that MIT put together in the '80s, where the beginning sentence is "to live well a nation must produce well." Paul Krugman, who was a faculty member at Sloan School then — he went to Princeton where he won the Nobel Prize in Economics — he says productivity isn't everything, but in the long run it's nearly everything. If you want to talk about quality of life, maybe not over a one-year period, but if you want to talk over a ten- or twenty-year period, productivity is everything. In the old days it might be productivity growth over a hundred years, but today it's over a much shorter time constant.


§7. Productivity vs. competitiveness; overcapacity and Mittal [36:27]

§7.p1

Here's a little cartoon: "all nineteen layers of management agree that we have to cut some of the fat from the first-line employees if we're to stay competitive." So I'll ask you this question, which I asked at the National Institute of Standards and Technology when I was on their visiting committee about ten years ago — and I got kicked off, in part for asking the question. What's the difference between productivity and competitiveness? They didn't know, by the way, so if you don't know, it's okay. This is the National Laboratory that's responsible for productivity. They got up and said, "we're the National Laboratory for competitiveness." This guy gives us an introductory little talk, and at the end I said, well, you mean you're the National Laboratory for productivity, right? "No, we're for competitiveness." I said, do you know the difference? And the answer is, he didn't.

§7.p2

For an economist, productivity is how many hours per ton it takes to make steel, or to make a computer, or to make anything. It's the labor productivity. Competitiveness, on the other hand, has to do with exchange rates and regulations and trade restrictions — all these things that Congress controls or the economy controls, that have nothing to do with technology. You can talk about it as externalities to productivity. I talked about technology and I said there's all these externalities that affect the technical decisions we make. Well, competitiveness is the same thing on an economic scale to productivity. So productivity is how efficiently can I manufacture something. Competitiveness is what's the selling price on a global economy. That's why people fight about whether the Chinese yuan — remember in China they have yuan externally, they do different currencies, different pieces of paper, because they don't want their people being able to do international transactions. They have great control over the exchange rates. If they have artificially low exchange rates then they can be very competitive even though we may be more productive. The Japanese are more productive at producing a ton of steel than the Chinese, but they have lower labor rates in China — but labor is not the big thing.

§7.p3

They're going to keep making steel, they will dump it on the world market. We do have a steel glut, an overcapacity in the world. When you have a productivity curve that has gone like that, what happens to your manufacturing capacity? You've got excess manufacturing capacity. Over the decade of the 1980s, when everybody thought the U.S. steel industry was hitting the toilet — that's what Wall Street thought — they were doubling their productivity. I've got the data that shows we were consuming a hundred million tons of steel each year in the United States. The workforce went from half a million employees to 250,000 employees. What does that say about productivity? The productivity doubled. It went up by four or five percent in a year for ten years. And the only industry that had a higher productivity growth rate in the United States was the mining industry, which has an even lower self-esteem on Wall Street, right?

§7.p4

What was the office productivity — all the great IBMs and Goldman Sachs and everybody? It went down by one percent over ten years. Because they got computers and they didn't know how to use them. Initially, computers actually killed productivity. They're now sort of getting past that after twenty-five years. Nonetheless, some of the most productive industries were mining and steel. So if we look back at this — to live well you must produce well — but everybody thought steel was crap as an investment. It was, because they had a worldwide overcapacity. Steel mills would go bankrupt, they would sell their assets, some third-world nation would buy it, and the next thing you know they're taking the old equipment with their low labor and maybe their cheap iron ore or ready access to metallurgical coal, and they're competing with old ancient equipment that they bought for ten cents on the dollar — and you haven't gotten rid of it. After about ten or fifteen years of this, the steel companies realized they should just scrap their old plants rather than see them show up competing with them again.

§7.p5

These types of lessons we have to learn over and over again. One of the great industries in the United States — or in New England — back a hundred and twenty-five years ago was the shoe industry. This is where they made shoes. I remember as a young assistant professor going to this basement warehouse in Boston, in an old building that had probably been built two hundred years before, and they had all kinds of shoe-making machines, and they were refurbishing them from these old mill buildings in New England and selling them in South America. Rather than selling the old equipment, they should have melted it down, because those shoes ended up coming back being made by lower-cost labor in South America. Same thing happened in the steel mills, and it took the management a little time to realize: selling your old equipment to someone who has an economic advantage of low labor costs just means that you have worse competition that you have to face.

§7.p6

Because of these productivity gains, they had 30, 40 percent overcapacity. The United States was only using a hundred million tons of steel in 1980, in 1990, but the workforce went down by a factor of two. Wall Street and the labor union said, oh, these companies are dying. They were, because no one would invest in them because they had too much overcapacity. It took them until the 1990s to get rid of some of that overcapacity. In the meantime, this guy in India named Mittal started buying up old steel plants around the world. Anybody know what happened to Mittal? They're talking about merging with Thyssen Krupp right now, to become — they are Arcelor Mittal. Arcelor was the French steel company which at one time merged with British Steel. Arcelor Mittal is based in Luxembourg. Mittal is this Indian who realized steel is a big industry, the world needs steel, and the Wall Street bankers — they're so smart — are selling all the equipment for ten cents on the dollar or less. He decided to start buying it and wait for the overcapacity to work its way out of the system, and now he's a multi-billionaire.

§7.p7

Arcelor Mittal is the largest steelmaker in Europe. If they merge with Thyssen Krupp they will become the second-largest steel company in the world after Baosteel, which is the Chinese steel company. It turns out, since 1964 when Bethlehem Steel started to build Burns Harbor, no private company has invested in a new integrated steel mill. Bethlehem Steel was the last company in the world to do it. Cost them five billion dollars, they almost went bankrupt, but by 1974 they had the most profitable year they ever had because they had a new fancy modern high-productivity steel mill, as opposed to these 1910 four-hearth versions they were working with. Burns Harbor is still probably the most efficient plant in the United States, and it's grown — people have continued to invest in it while closing other mills. In the meantime, Mittal, understanding the economics and the need and the overcapacity, became a very rich person, only because there'd been plenty of steel plants built in the world since 1964, but they've all been built by countries.


§8. Government-built steel and the American denial [46:05]

§8.p1

Take the Koreans. In the 1990s, I was the POSCO Professor. POSCO was the Korean steel company. The president of Korea took one of his colonels and said, "Mr. Park, I want you to go and start a steel company." Park went and talked to U.S. Steel. U.S. Steel sold them the technology, and they built POSCO Steel in Korea, which by 1997 was the world's largest steel company. From the early 1970s, in twenty-five years they became the largest steel company in the world, until the Chinese passed them by a long shot. But that was at the backing of the government of Korea to build what was, at that time, probably a ten- or fifteen-billion-dollar plant. They built a couple of them in Korea — Korea doesn't need but a couple. Saudi Arabia's got lots of natural gas, so what do they do with their natural gas? They build a gas reduction facility. They don't use blast furnaces — they use all the natural gas that they can't get out of the country any other economical way, they import iron ore, and they make steel. Because steel is fundamental to a growing economy. That's what the Japanese did after World War Two — they built steel companies and then they built shipyards, they became the largest shipbuilder in the world. Other people have passed them by, but people are all using that model.

§8.p2

The road to industrialization is: first make steel, then start making ships if you're close to the ocean, and then you start making automobiles and other things. That's what the Koreans did, but they did it after the Japanese. You didn't start seeing Hyundais on the road until after you saw Toyotas, right? Koreans are just copying the Japanese, and the Americans are sitting there like deer in the headlights, watching all this happen and denying that there was anything to it. I remember the month I left Bethlehem Steel, Edgar Speer, who was the president of U.S. Steel — he was also president of American Iron and Steel Institute, which is a collection of all the companies — he was on the front page of Iron Age magazine, and in the article he said the Japanese are not using anything we don't know about. What an idiot. They were using the best technology in the world, and we knew about it.

§8.p3

That's sort of like General Motors. Toyota wasn't using anything that General Motors didn't know about. General Motors had the best technology in the world for making automobiles. It's right there in the research laboratory — it's not on the road, it's right there in the research laboratory. They spent billions of dollars. In the 1980s you could have purchased Toyota on the Tokyo Stock Exchange — if the law would allow you — for less than the production equipment budget of General Motors for that decade. General Motors invested 50 billion dollars in new plants and equipment, decided their employees were dirt, worse than dirt, and they had practically a bounty to put in a robot to get rid of a person. In the meantime, what were the Japanese doing? They were developing the Toyota Production System and treating people like they were empowered to make decisions, like intelligent human beings. In Detroit we treated people like dirt. They weren't allowed to make any decisions. If they thought something was going wrong — shut up, I'm the management, do as I say. So when General Motors went bankrupt I said, they earned it. And they did.

§8.p4

So that's enough for today. Actually, it's enough until the eleventh and twelfth. If you have some questions — a couple of my trips got cancelled, so I will be around some of the next couple of weeks, but I'm going the next two days.

Cases referenced

  • Bethlehem Steel Burns Harbor §2.p6

    The last integrated steel mill built in the world by a private company (1964, finished c. 1972). Five million tons per year, two blast furnaces, ~$5 billion. Burns Harbor's relining and "government job" culture used to introduce the corruption-and-arrogance theme; in §7.p7 it returns as the case showing that high-productivity assets did exist in U.S. steel and were retained while others closed.

  • Massabi/Mesabi Range iron ore depletion §2.p4

    Mesabi was pure iron ore consumed during WWII and just after; postwar transition to sintered iron oxide at Bethlehem produced 5,000 lb/day of cyanide emissions until fixed. Used to introduce the Air Quality Act of 1972 and the pre-EPA regulatory environment.

  • Continuous casting and steel industry capacity collapse §4.p2

    Continuous casting raised yield from 65% to 97% (pounds poured vs. shipped) — eliminated the cut-off-the-top-of-every-ingot loss. Combined with BOF and mini mills, drove the productivity gains of the 1980s that halved the workforce and produced 30–40% overcapacity. Tom's "we used to have to pour 150 million tons to ship 100 million" is the canonical framing figure.

  • Continuous caster breakout §4.p2

    Breakouts happened once or twice a year in the '70s, take a week to clean up — 300–400 tons of molten steel hitting the shop floor, 25,000 tons of lost production, hundreds of millions to rebuild. Today rare due to process controls and IR sensors.

  • Argon oxygen decarbonization (AOD) process development §5.p1

    Born in basement of Building 8 from a John Chipman graduate student (Krivsky [?]) doing basic carbon-burnout chemistry; commercialized at a small steel company because the majors wouldn't risk it. Dropped stainless steel price by 5–10×, enabled <0.03% C for weldability and corrosion resistance, recovered chromium from slag.

  • General Electric boiling water reactor stress corrosion cracking §5.p7

    Sensitization of stainless steel welds in BWR safe ends cost GE "a couple billion dollars" in the 1970s. Westinghouse PWRs avoided it by using Inconel (10× the cost) for higher temperatures. The teaching consequence of the John Wulff sensitization patent and the broader carbon-control rationale.

  • John Wulff stainless steel sensitization patent (1940s) §5.p7

    Wulff's 1940s patent: deliberately sensitize stainless steel, immerse in nitric acid overnight, dissolve chromium-depleted grain boundaries to produce stainless steel powder. Used pedagogically to show that the same intergranular attack mechanism that makes BWR welds fail is exploited industrially for powder production.

  • Saugus Ironworks §6.p1

    ~1600s. One person-year per ton of steel produced — the baseline anchor for Tom's productivity chart. Manual charcoaling, manual ore mining, manual limestone, 60-hour weeks.

  • Steel mini mills rise (1970s–1990s) §6.p3

    Gordon Forward's $30/ton transoceanic shipping argument: if mini mills can produce within $30 of Japanese/Korean cost, they win on freight alone. Mini mills + process technology pushed labor to <20 person-minutes/ton.

  • MIT Commission on Industrial Productivity — Steel Industry Study §6.p4

    The 1980s MIT-authored book whose opening sentence is "to live well a nation must produce well." Frames the productivity-is-everything thesis Tom uses through §6–§8.

  • NIST competitiveness vs. productivity misunderstanding §7.p1

    Tom's visiting committee anecdote: NIST staff claimed to be "the National Laboratory for competitiveness" and couldn't distinguish productivity (labor hours/ton, internal) from competitiveness (exchange rates, regulations, external). Tom got kicked off the committee in part for asking. Used to develop the externalities-of-productivity framing.

  • U.S. Steel industry decline and management failure §7.p3

    Workforce halved 1980–1990 while output held at 100 million tons/year; Wall Street read it as death, Tom reads it as a 4–5%/year productivity gain doubling output per worker. Edgar Speer's *Iron Age* claim ("the Japanese are not using anything we don't know about") used as the canonical exemplar of denial.

  • New England shoe industry decline §7.p5

    Boston basement warehouse refurbishing 200-year-old shoe-making machines to ship to South America — same pattern as the 1980s steel-equipment sales. The shoes came back from South American labor; "they should have melted it down." Used to make the don't-sell-your-old-equipment-to-your-future-competitor point.

  • Lakshmi Mittal steel mill acquisition strategy §7.p6

    Mittal bought distressed steel mills worldwide at ~10¢ on the dollar while Wall Street dumped them; became multi-billionaire as overcapacity worked out. Arcelor Mittal (Luxembourg-based) is the largest European steelmaker; the Thyssen Krupp merger discussion (2017) would make them #2 worldwide after Baosteel.

  • POSCO (Korea) government-built steel mill surpassing Nippon Steel §8.p1

    Korean president's colonel Park sent to start a steel company; U.S. Steel sold the technology; POSCO became the world's largest by 1997. Tom held the POSCO Professorship at MIT in the 1990s. Used to make the case that since 1964, every new integrated mill in the world has been government-built (POSCO, Saudi gas-reduction, China).

  • Saudi Arabia gas-reduction steel facility §8.p1

    Saudi natural gas as steel reductant — sidesteps coal-and-coke entirely. Imported iron ore + domestic natural gas + sovereign capital = integrated steelmaking. Part of the road-to-industrialization sequence (steel → ships → autos).

  • Japanese steel industry dominance 1970s §8.p1

    Post-WWII Japan built steel, then shipyards, then autos — the canonical industrialization sequence Koreans then copied with a 25-year lag. The Edgar Speer *Iron Age* denial line is the U.S. response.

  • General Motors bankruptcy §2.p7

    "They earned it." Parallel to the steel-company bankruptcies of the 1980s. GM's $50 billion plant-and-equipment budget in the 1980s could have bought Toyota outright on the Tokyo exchange. GM bountied robots-for-people while Toyota developed the Toyota Production System.

Layer 2 — cleanup edit
p1 00:00

This will be, this is my sixth lecture plus the introductory lecture, and we're gonna change things now because of the students' schedules and stuff to do, you do six lecture modules. Which in future years, but now we're gonna do, we're starting now to make those six lecture modules and we're going to take the old modules, we're gonna break them up from twelve to six, okay. And in any case, so there's no class after today until next class. Next live class is Tuesday October 11th, and that should be Dr. Belmar [Bellman?], okay. And then I think on the 12th I may just do a recitation, okay, so you can bring questions and ask questions and whatever so far as that goes.

p2 01:01

But to finish up your twelve lecture modules, I did lecture this stuff back in 2016 in the fall, and if you do those four lectures — 6, 7, 10, and 12 — I looked at them briefly this morning, there's not going to be too much overlap. There will be new materials in there. And if you look at spring 2016, I did something called Structural Materials Selection where I didn't spend as much time on externalities and economics and things like that. If you look at 11 cent— 11 and 12, so that would be your next six lectures to complete this module for this semester. It's a little weird because you're on twelve module, twelve lecture modules, and we're in this transition where we're gonna go to six lecture modules, okay. And it doesn't really matter, you're not going to be quizzed on it anyway, so who cares, okay.

p3 01:53

The next lecture, because Dr. Bellman's out of town, I'm out of town, some of my travel got changed, but it's just as easy to— he'll finish up some things on the 11th and maybe I'll do something on the 12th, we'll let you know about that. And then we're done with the live lectures. I told you we'd be done before Halloween and before it starts getting too dark out early in the morning. Even though we're not doing anything live for the next week and a half, I would encourage you to watch some of the modules just as if it was being done live. There's some — much was, I don't know, I understand why a couple of times the thing was dark, okay. I got to clean up some of these old things, and so I did look at some of these this morning. I didn't watch the whole thing but I think those six lectures will be— there's not a whole lot of redundancy, okay, I tried to cut it out.

p4 02:50

Any questions on that? Hopefully you got your, and Brian's supposed to be checking on this yesterday, everybody's, you know, a little half-page write-up on what your topic is going to be. I responded to people who sent them directly to me, Brian's going to check through them and we'll get that straightened out this week, and you can start writing your papers, which are due I think October 31st or something like that. The first draft has to be turned in, and then you have your little summaries to write up. I think a summary is due sometime around mid-October. But if you watch these six modules you'll be done with this module, and you may even be done with Dr. Bellman's module, okay. So you may be done with a couple of modules. Any questions on the assignments? Nope.

p5 03:45

So we were talking about melting of metals, and specifically one of the problems, a serious problem, was how do you melt steel, which melts at 1600 degrees centigrade, and you can't get but about a thousand degrees centigrade out of a normal fire burning in air. If you take hydrocarbons and burn them in air, you're going to get 2,000 degrees Fahrenheit, a thousand degrees C. And so you can, with a little bit of help, you can melt cast iron, you certainly can melt copper alloys, but you can't melt steel. And before Henry Bessemer, the way they made steel, they would take cast iron, which is like four or five percent carbon, and they would blow air through it and the carbon would burn off as carbon monoxide, and it would leave a lower carbon version of iron, and I told you that's wrought iron, which is malleable, can be forged, as opposed to cast iron which is not malleable. You try to forge it, it'll crack, okay. It's just brittle, so far as that goes.

p6 04:55

And Henry Bessemer came along with his converter, which was a better way to blow oxygen through the iron. They called it puddling iron — if you want to learn about puddling you can go look it up on Wikipedia, but guys would just stand there blowing compressed air through a bath of steel, and it was a very slow process. Andrew Carnegie came along and the railroads were being built around the world and he started making steel in a big way. If you want to learn about how he got into this business, there were a couple of shady deals in Pittsburgh, okay, and he ended up owning a steel company which is now U.S. Steel as it's come down through the last 150 years. And he became the richest man in modern history, even richer than Bill Gates or Warren Buffett or Carlos Slim or any of the guys today that are worth a lot of money. In constant dollars he was the richest.

p7 05:58

And we talked about how Carnegie actually built what they called integrated steel mills, where you start out with coal and you have a coke oven where you put the coal into a rarefied atmosphere, just like we made charcoal basically. You didn't have a lot of oxygen, you burn off all the organics out of the coal, and you end up with something that structurally has some structural integrity, basically a graphite, okay, which is called coke, okay. It's not the kind you drink, it's not the kind you shoot, it's basically this black lump of stuff. And the coke ovens in Pittsburgh were one of the major reasons for the Air Quality Act of 1972, okay. The coke ovens burning off all the organics in the coal just created a terrible atmosphere from a human health perspective. Just going off in the air, okay. And that was acceptable back then.

p8 07:00

Back when I had worked for Bethlehem Steel, they had changed their iron ore charge from something that was basically just sort of a rock that came out of what's called the Mesabi Range in Minnesota. It was virtually pure iron ore, okay, it was wonderful material, but we used it all during World War Two and just after. And so they went to sintered iron oxide, which they got from a bunch of rock in Minnesota but they had to do some processing. And apparently when they started putting the sintered iron oxide in with the coke and the limestone in the blast furnace in Bethlehem Pennsylvania, they found they were producing five thousand pounds a day of cyanide going up the stack, okay. Didn't bother to tell anybody because that upset some people, but it didn't matter legally, they had no duty to clean up, they're just— okay. But they actually did fix it, they realized it was something that should be fixed, and they figured out what the chemistry was that was giving them— and let's face it, cyanide is just carbon-nitrogen, right? And so if you start burning carbon in air with nitrogen, with a lot of oxygen, you can get a chemistry that gives you CN, so you've got cyanide. But people living in Bethlehem just had to breathe the air, you know, that was the different rules they had back in the old days, okay.

p9 08:28

And because of some of those types of things, eventually Congress came up with the Air Quality Act of 1972, which is actually one of the beginning laws for the Environmental Protection Agency and stuff, okay. So it's just like the labor unions. I mean, whether you're pro-labor or not, if you go back and look at what the management was doing to the workers before unions, back a hundred years ago, I mean, they basically hired Pinkerton guards when they had a problem. These guys just, you know, fire at will, okay, and they would murder them, okay. There are a number of stories at Alcoa and US Steel of Pinkertons being hired to come in and shoot down a little of the rowdy labor force. They were all immigrants anyway, you know what the president thinks of immigrants, right? Oh sorry. Again, it was sort of a different standard, okay.

p10 09:33

And anyway, so integrated steel mills got to the point where they would consume tens of square miles and they would cost about five billion dollars — today would be about twenty billion. I don't think I mentioned it the other day, but the last integrated steel mill to be built in the world by a private company was Bethlehem Steel Burns Harbor in Indiana, in 1964, finished about nineteen— about eight or ten years later. And it could produce five million tons per year, which was five percent of the U.S. steelmaking capacity at the time. You know, it had two blast furnaces so if one went down you didn't shut down the whole mill, because it can take three months to reline a blast furnace. In fact, that's when I lived in Bethlehem Pennsylvania, boy they relined the blast furnace, and that's when everybody got new grills for their patio. And I mean everything, you know, everything went— it was a huge project, cost a couple hundred million dollars, and they were building all kinds of things in the shops, many of which actually went to the steel mill but most of which went to the people's homes. [laughter] Yeah, extra stuff is, you know, they call it a government job, okay, if you've ever heard that in some of these things, okay.

p11 10:53

There was, without getting into it, there was a lot of nepotism — not necessarily family nepotism but people scratching each other's backs. When you got into management in the steel companies, they were pretty corrupt, okay. In the 1930s, Bethlehem Steel had six of the top ten paid executives in the United States working for Bethlehem Steel, and Eugene Grace, who ran Bethlehem Steel through the '30s and the '40s, got paid a bonus on every ton produced. Even though during the Depression Bethlehem was losing money, he got paid a bonus for every ton produced. Whether they were making money or losing money, he made a fortune. And people just kind of accepted this as the way, you know, what happened with management, okay. And it's sort of the same type of thing, a lot of these companies like Bethlehem went bankrupt a few years later in the '80s. And it's just like I said, when General Motors went bankrupt and someone said, well what do you think about General Motors' bankruptcy, my comment was, well, they earned it, okay. And the steel companies earned the bankruptcies they went through, okay. There's a lot of arrogance and other things, okay.

p12 12:10

So the blast furnaces — we talked about before, you've got this big tall furnace which might be ten stories tall nowadays. You blow fuel in, you pour coal in the top — the fuel you put in here could be oil or natural gas, just speed up the process — you have all your coke, limestone, and iron ore, and those three components go in the top and they work their way down. And every now and then, about once or twice a day, you tap the heat, which means you drain off some of the liquid cast iron at the bottom of the blast furnace. You have to preheat the air, so you have these two big brick ovens, one of which is being preheated by the exhaust gas and the other one is preheating the air that's going into the furnace and being blasted into the furnace. And about twice a day you would switch over your ovens, so the one that has cooled off now is going to get heated up and the other one that's now hot from the exhaust gases will be preheating the air, and you just switch it over.

p13 13:18

And we did the same thing with open hearth, which we don't have anymore in the world, but that's how Andrew Carnegie made his steel, after the Bessemer converter. In open hearths, over about a hundred years we used open hearth, and you have this big brickwork below us, same type of thing in glass front melting furnaces. We have the same thing — instead of a bed of steel you got a bed of glass — and in order to get the temperatures you have to preheat the air coming in. You can't get more than 2,000 degrees with normal fuels, hydrocarbon fuels, whether it's natural gas or oil or coal or whatever. You're only gonna get about 2,000 degrees. Take my welding course if you want to learn why, okay.

p14 14:02

And then in the late '50s, early '60s, some Austrians decided, well, I can go back to the puddling process — instead of blowing in air, I could blow in pure oxygen. Because in Germany they developed the Linde process for liquefying air and distilling it, and so they had lots of excess oxygen. They had lots of nitrogen, and they also— you got 1% argon, and you basically would distill the air into those three components. And you had lots of liquid oxygen, and you can blow that down through a water-cooled copper lance, and essentially in this vessel that might be four or five stories tall, you base it with just five or ten feet of liquid cast iron in the bottom. You blow this in and in twenty minutes you can burn off all that carbon. And when it's going, it's really exciting, okay. But it's the same thing, you got a flux of limestone and sand and oxygen lance. But the volume inside here is about twenty or thirty times the volume of the steel, and it foams up as a big froth, okay. And sometimes people commit suicide by diving into it. It's happened — when I was at Bethlehem Steel — you can tell because the body floats. Steel is heavier than water, so you float right on top. You see this depression in the shape of a person.

p15 15:38

Anyway, why doesn't oxygen— because there's carbon, and the carbon burns to carbon monoxide, which is a reducing gas. If you keep blowing forever, after you've burned off all the carbon, you'll start burning off the iron and form iron oxide. How do you tell? They know from experience how long you blow — how many tons of liquid oxygen do you put in to how many tons of cast iron, you knew how many tons of carbon you had, and from experience they can get down— they don't blow below about 0.05 percent carbon in the steel, because at that point you get to burning off iron oxide and now you're going back to where you started from, and that's not very efficient, right?

p16 16:21

Now the person who worked all this out was a guy named John Chipman. If you go right across the hallway, you know through the courtyard here, you hit the Chipman Room. John Chipman was a physical chemist from Georgia Tech who came to MIT, and he explained the physical chemistry of steelmaking for the world. Before that it was almost— it wasn't quite empirical, but it was sort of experiential alchemy, they didn't really know how to control it. But John Chipman in the '30s and '40s and '50s defined, on the principles of physical chemistry which he had learned on aqueous chemistry — you know, water-based chemistry, but he said, well, these principles should apply at high temperatures, and he showed that they did. And John Chipman and then Tom King and John Elliott after him, which were two of his protégés — and Tom King was the department head when I was a student, okay. But John Chipman, I knew John Chipman, he was retired, he'd retired in '62, I didn't start here until six years later, but I went to see John Chipman. He's a wonderful Southern gentleman, but his name, if you go to Japan or China, he's like a god, okay, because he explained steelmaking, this huge industry, and made it such that it's no longer this sort of alchemy empiricism, as far as that goes.

p17 17:53

So from Andrew Carnegie's day until about 1970, they would take this steel and they would cast it into ingots. And these are the cast iron ingot molds on railroad cars, and there's the hot steel that they're actually getting ready to pour. Actually this is what they call a hot top — it's actually a ceramic thing and it helps feed the liquid into the mold, because the metal shrinks as it solidifies and you want to keep feeding liquid in on the top, and it's obviously hotter at the top. You can see the glowing red here, and it takes about — it takes a day for this stuff to solidify, okay. You can't move these cars for at least twelve hours or so once you pour it in.

p18 18:38

But then some people — again I think it was Austria — decided they could try continuous casting. And continuous casting had been done on lower temperature metals like brasses, copper alloys, and I told you they were doing it on gold alloys, which is not all that different than brass. But continuous casting, you have a ladle with molten metal that you melt somewhere else, and you pour it into something called the tundish, which is just a bathtub holding things, and it drops the metal into a mold. The mold is water-cooled copper, and you will solidify a little skin on this. And this yellow band, you actually have molten metal, and the whole thing is pliable, and so you have these rolls. This whole thing stands about seven stories tall, you have liquid metal in the center. If you get a breakout, it takes about a week to clean it up, because you got three or four hundred tons of molten steel hitting the floor, okay, of this shop. They do have a stopper in the tundish, but even there you've still got this thing that can be ten feet wide and ten inches thick, and it's 50% liquid, okay, on average. And so if that skin breaks — and they did break, I mean we'd have breakouts once a year or twice a year or whatever in the '70s when I worked at the steel company. They almost never have it today because they got all kinds of process controls and infrared temperature sensors, and it's too expensive to have a breakout, okay.

p19 20:11

When you're sitting there processing 5,000 tons of steel a day through here, you don't want to lose 25,000 tons of production over a week while you clean it up, and the whole cost of essentially rebuilding six stories worth of— it's not a small job to fix this, but you do have breakouts, okay, from time to time, not very many anymore. And in fact the blast furnaces used to need to be relined about every six months. If you go back four hundred years ago they probably had to be relined every three or four months, which might be a season. But now blast furnaces will go for three or four years before they have to be relined, because they've improved the ceramics, the fire brick that's in there. Again, they cost too much — you know, it can cost several hundred million dollars to rebuild a blast furnace, it can cost a hundred million dollars or a couple hundred million dollars to rebuild all the warm parts in this continuous caster. So I think the Japanese— used to be the Japanese had the record for running continuous casters for a year or a year and a half at a time, and making a continuous strand of steel that goes for about a hundred and fifty miles. And they basically cut it off with a torch down here. What's the torch? Pure oxygen again. Don't, ee, you'll— guess the thing's very hot, just come here, blow pure oxygen on something that's hot like that, that's right through it like a nice hot knife through butter. It's just a cold oxygen knife through hot steel. Pretty exciting to see some of these things, and also to see the size of these things.

p20 22:01

So you cut slabs and it says go to storage. In fact, as we've increased productivity, we don't go to storage anymore and let this thing cool down, we actually go directly to the rolling, okay. And so we've had to reconfigure the steel plants and things like that, and the most efficient steel plants are the ones that have a good layout so the process flows straight through, okay.

p21 22:29

There's a process called argon-oxygen decarburization, and it's the way we make all the stainless steel in the world today and most of the nickel-based superalloys for jet engines and stuff. Very high quality process. You think it looks like the BOF, the basic oxygen furnace, where you burn the carbon out of the steel. But in this case you basically take the low carbon wrought iron type of steel and you blow oxygen and argon through there, such that the oxygen— is the lower pressure, you generate carbon monoxide at a lower pressure, and you can get your carbon down below 0.03 percent. You can get it down to 0.01 percent, so far as that goes. And that means you can make very good quality stainless steel.

p22 23:20

The seeds of this were generated down in the basement of Building 8 with one of the graduate students of John Chipman, and he was just doing basic science on the chemistry of burning carbon out of steel. And he left MIT, and he said, you know, we could make stainless steel this way by blowing argon through. And now this is an AOD vessel. It has dropped the price of stainless steel by a factor of five to ten in the world, and it's better quality stainless steel, it's cleaner stainless steel. It applies to nickel-based alloys, it applies now— a lot of your cast alloys, cast metals, go through this AOD refining situation. And it was basic research in the basement of Building 8 here that started it. But it was first done when Chriskey [Krivsky?] left MIT and went to work for a very small little steel company that was willing to take a chance. The big steel companies were big and bureaucratic, they wouldn't take any chances.

p23 24:26

Student: [inaudible question]

p24 24:30

Yep, yep. Lower than 0.05 — I mean, the lowest carbon you're going to get out of that BOF is 0.05 percent carbon. If you want to go lower you can go lower, but you're going to start losing iron as iron oxide, which is just wasted energy. If you want to go lower, you actually probably start at one-tenth, okay, and then you do the AOD, and the AOD can get you down, in some cases, to an order of magnitude lower in your carbon. For good weldability and corrosion resistance you want to be below 0.03, and you get six times lower than that, okay, at relatively little cost compared to the big cost. It also helps you with recovery of your chromium. That was why the price of stainless steel dropped so much. In order to get down to this low— when you just throw the chromium in there, a lot of the chromium would go up into chromium oxide in the slag. And just like the problem of oxidizing the iron, you oxidize the chromium, which is fairly expensive. The nickel you throw in is an alloying element that has been made somewhere else by an electrolytic process at a nickel mill, but you made your stainless steel that way. And now we don't generate as much slag that has a bunch of chromium oxide and iron oxide. We basically control the carbon chemistry by controlling the vapor pressure of the carbon monoxide by diluting it with argon.

p25 26:07

And if you really want to save money, you actually start out with a nitrogen-oxygen mixture, but then you don't want the high nitrogen in most cases, so you switch over to argon-oxygen, and that way you don't have to use that expensive argon. But in fact steel mills are using so much liquid oxygen and liquid argon that Linde will come and build a refrigeration plant to liquefy air right there next door and pipe it right over to your mill. You know, like a six-inch line, something — a lot of liquid oxygen. The only people who use more liquid oxygen in shorter periods of time is the US Air Force sending rockets off, okay. But burning oxygen is a very violent — potentially violent — anyone ever seen the Purdue University charcoal-starting contest on YouTube? If you look up Purdue and, you know, charcoal starting, they used to have a contest of who could design the system, you know, that would get your charcoal briquettes burning fastest. And they basically had to stop it when one person decided to bring a twenty-foot pole with a bottle of liquid oxygen and pour it on the thing, and the carbon — it just goes, it's all gone, okay. Pure oxygen burns things really quick. Plus the fire department in West Lafayette wasn't real happy with the safety of a bunch of people standing around this thing, okay. But you can see it on YouTube.

p26 27:42

In steels, in stainless steels, and actually to a certain extent in some nickel-based superalloys, you don't want a lot of carbon. Because what happens is, I have a grain structure — I'm going to just kind of draw the grains in here like this, okay, ideally in your metal — and if you have carbon, carbon diffuses faster than chromium, and if you heat up the stainless steels typically in the 800 to 1200 degree Fahrenheit range, which it turns out you will have that in the heat-affected zone of a weld, okay. I mean, you can't get around it — if you're gonna make a fusion weld, somewhere in the heat-affected zone you're gonna have that temperature. What happens is the carbon diffuses to the grain boundary, the chromium diffuses to the grain boundary but from a shorter distance because its diffusivity is about a hundred times slower. So that little bit of carbon, if you're at 0.05 carbon with 18% chrome, or whatever the amount of chrome is in your stainless steel, you will precipitate out chromium carbide. I'm gonna call it X Y, it could be Cr₂₃C₆, okay. I mean these carbides are really complex — or Cr₃C₇, I mean there are a bunch of different carbides. But that means that a region of about two microns on either side of the grain boundary will be depleted in chromium. If you go and analyze that, you might find it only has ten percent chromium in the depleted region. Because there's chromium everywhere, but the carbon has to diffuse from a larger distance, but it's a small atom and it does diffuse from a larger distance, and that denudes, or makes the grain boundaries lower chromium and therefore not as corrosion resistant.

p27 29:45

In fact, John Wulff, my academic grandfather, had a patent in the 1940s to cost— sensitize stainless steel, heat it in this region, get something like this, throw it in nitric acid overnight to make stainless steel powder. Because the nitric acid doesn't attack the 18% chrome, it does attack the 10% chrome, eats away the grain boundaries, and you get cracking — typical intergranular cracking along the grain boundaries. Cost General Electric a couple billion dollars in the nuclear reactor industry in the '70s from the welds getting sensitization cracking, okay. Yeah, they could— well, if you were Westinghouse, you used Inconel, as she said, you know, that was wonderful, about ten times the price of stainless. But if you're building a pressurized water reactor, they needed, because of the higher temperatures, they needed Inconel. General Electric had a boiling water reactor, they could use stainless — less expensive material but more susceptible to what is called stress corrosion cracking, okay. And you get cracks through here in the reactor safe end. You know, that means that you're not gonna be able to flood the reactor with water if you have a near meltdown. And you will not have a near meltdown, you'll have a real meltdown.

p28 31:10

So there's a lot of corrosion chemistry, obviously, to stainless steels, and a lot of it is controlling the carbon content, which goes back to the steelmaking process. And it's all because of AOD refining that has made all this practical today, okay. Any other questions on that?

p29 31:37

We also have— well okay, look what's going on here, okay. So if I go to— this is a chart I put together a few years ago. Back in the 1600s when they had Saugus Ironworks, it probably took about one person-year per ton. The guy had to go chop the trees, make charcoal out of the logs — you basically stack up the logs, cover them with straw and mud, and then you light the inside, and there's not enough oxygen so you burn off all the lignin in the logs, and you end up with — not coke, we call it charcoal, but it's like today's coke made from coal, okay. So that was your carbon for your blast furnace. You then had to go get the iron ore, dig it out of the ground, you had to get the limestone, digging it out of the ground. And when it's all manual work, it takes a lot of time, okay. And then you gotta run your blast furnace. So I estimated it probably was at least a person-year per ton. They probably worked sixty-hour weeks, these people were working long hours.

p30 32:46

Then Bessemer came, we probably dropped down and got bigger and more efficient over the next 200 years, and Bessemer came along and learned how to blow air into a converter. Carnegie came along and learned about — or invested in — things like the basic open hearth. And things kept on marching along. But then after World War Two we had the basic oxygen furnace in the '50s and '60s, we had continuous casting in the '70s. Continuous casting went from having a yield of 65% to 97% in tonnage products — pounds poured versus pounds shipped. You went from 65% to over 95% because you didn't have to cut off the top third of every ingot, okay, because of the piping porosity and things like that. It gets more complex but nonetheless the continuous casting process was a big jump in productivity of saleable product. We used to have to produce 150 million tons of steel — pour 150 million tons of steel — to make a hundred million tons that we could sell in the United States.

p31 33:55

Mini mills, which we'll talk about in a second, and process technology was to cut off here. And I mentioned, I think, that when I worked for a steel company in 1975, it was 50/50 labor and raw materials. Back here it was labor-intensive. Today it's like 90% raw materials, okay, because we've taken all the labor out, okay. Here's the labor on a log scale, and we're down to less than twenty minutes a ton, person-hour per ton to produce steel. As Gordon Forward, a graduate of this department who started a mini mill, and we're gonna talk about the mini mills in a second — said if the Koreans or the Japanese want to sell steel in the United States, all he has to do is produce it within $30 of their cost. Because the cost of the raw materials — coke, limestone, and iron ore — they're transported, they're commodity around the world, okay. Everybody's paying the same price pretty much. But it costs $30 a ton to ship a ton of steel across the Pacific Ocean. So Gordon Forward said if I can make it for only $30 a ton more than they make it in their big fancy integrated mills, I could build a smaller mill and I can beat the price. And he did, and we'll talk about that.

p32 35:35

But basically, in all of this, the whole story of steel over the last fifty years — there's a book that MIT put together in the '80s, we started writing it, where the beginning sentence is "to live well a nation must produce well." And Paul Krugman, who was a faculty member at Sloan School then, he went to Stanford [Princeton] where he won the Nobel Prize in Economics, he says productivity isn't everything but in the long run it's nearly everything, okay. If you want to talk about quality of life, maybe not over a one-year period, but if you want to talk over a ten or twenty year period, productivity is everything, okay. In the old days it might be productivity growth over a hundred years, okay, but today it's over a much shorter time constant.

p33 36:27

So here's a little cartoon: "all nineteen layers of management agree that we have to cut some of the fat from the first-line employees if we're to stay competitive," okay. So I'll ask you this question, which I asked at the National Institute of Standards and Technology when I was on their visiting committee about ten years ago, and I got kicked off before— asking the question and other things. What's the difference between productivity and competitiveness? They didn't know, by the way, so if you don't know, it's okay. And this is the National Laboratory that's responsible for productivity. And what they did is they got up and said, "we're the National— NIST is the National Laboratory for competitiveness." And I says — this guy gives us an introductory little talk at the end — I said, well, you mean you're the National Laboratory for productivity, right? "No, we're for competitiveness." I said, do you know the difference? And the answer is, he didn't.

p34 37:33

Well, for an economist, productivity is how many hours per ton it takes to make steel, or to make a computer, okay, or to make anything. It's the labor productivity. Competitiveness, on the other hand, has to do with exchange rates and regulations and trade, you know, restrictions, all these things that Congress sort of controls or the economy controls, that have nothing to do with technology. You can talk about it as externalities to productivity. You know, I talked about technology and I said there's all these externalities that affect the technical decisions we make. Well, competitiveness is sort of the same thing on an economic scale to productivity. So productivity is how efficiently can I manufacture something. Competitiveness is what's the selling price on a global economy. And you know, that's why people fight about whether the Chinese dollar is, or Chinese renminbi — yuan, they have, remember in China they have yuan externally, okay, they do different currencies, okay, different pieces of paper, because they don't want their people being able to do international— anyway, they have great control over the exchange rates. And if they have artificially low exchange rates then they can be very competitive even though we may be more productive. The Japanese are more productive at producing a ton of steel than the Chinese, but they have lower labor rates in China, but labor is not the big thing.

p35 39:26

What it is, they're going to keep making steel, they will dump it on the world market if we have a steel— and one of the things is, we do have a steel glut, an overcapacity in the world. When you have a productivity that has gone like that, okay, what happens to your manufacturing capacity? You've got excess manufacturing capacity. Over the decade of the 1980s when everybody thought the U.S. steel industry was hitting the toilet — well that's what Wall Street thought, but they were doubling their productivity. I've got the data that shows we were consuming a hundred million tons of steel each year in the United States. The workforce went from half a million employees to 250,000 employees. What does that say about the productivity? The productivity doubled. It went up by four or five percent in a year for ten years. And the only industry that had a higher productivity growth rate in the United States was the mining industry, which has an even lower self-esteem on Wall Street, right?

p36 40:35

What was the office productivity, all the great, you know, IBMs and, you know, Goldman Sachs and everybody? It went down by one percent over ten years, okay. Because they got computers and they didn't know how to use them. I mean initially computers actually killed productivity, okay. They're actually now sort of getting past that after twenty-five years. But nonetheless some of the most productive industries were mining and steel. And so if we look back at this thing, to live well you must produce well, but everybody thought steel was crap as an investment. It was, because they had a worldwide overcapacity. And it turns out steel mills will go bankrupt, they would sell their assets, some third world nation would buy it, and the next thing you know they're taking the old equipment with their low labor and maybe their cheap iron ore or ready access to coal — the metallurgical coal — and they're competing with old ancient equipment that they bought for ten cents on the dollar, and you haven't gotten rid of it. So after about ten or fifteen years of this, the steel companies realized they basically would just scrap their old plants rather than see them show up competing with them again.

p37 42:01

But these types of lessons we have to learn over and over again. One of the great industries in the United States, or in New England, back a hundred and twenty-five years ago was the shoe industry, okay. This is where they made shoes. I remember as a young assistant professor going to this basement warehouse in Boston, in his old building that had probably been built two hundred years before, and they had all kinds of shoe-making machines and they were refurbishing them from these old mill buildings in New England and they were selling them in South America, okay. Well, in a lot of ways they should have — rather than selling the old equipment — they should have melted it down, because those shoes ended up coming back being made by lower-cost labor in South America, okay. Same thing happened in the steel mills, and it took the management a little time to realize, selling your old equipment to someone who has an economic advantage of low labor costs just means that you have worse competition that you have to face.

p38 43:08

But because of these productivity gains that I showed you in that slide, they had like 30, 40 percent overcapacity. I mean the United States was only using a hundred million tons of steel in 1980, in 1990, but the workforce went down by a factor of two. Wall Street and the labor union says, oh these companies are dying. They were, because no one would invest in them because they had too much overcapacity. Yeah, it took them until the 1990s to get rid of some of that overcapacity. In the meantime this guy in India named Mittal, Mittal, started buying up old steel plants around the world. And now, anybody know what happened to Mittal? Yeah, well they're talking about merging with Thyssen Krupp right now, okay, to become— they are Arcelor Mittal. So Arcelor was the French steel company which at one time I think merged with British Steel. Arcelor Mittal is in— who's based in Luxembourg. Mittal is this Indian who realized steel is a big industry in the world, needs steel, and people are selling— the Wall Street bankers, they're so smart, they're selling all the equipment for ten cents on the dollar or less. And he decided to start buying it and wait for the overcapacity to work its way out of the system, and now he's a multi-billionaire, okay.

p39 44:46

And Arcelor Mittal is the largest steelmaker in Europe. If they merge with Thyssen Krupp they will become the second-largest steel company in the world after Baosteel, which is the Chinese steel company that just produces steel. It turns out, since 1964 when Bethlehem Steel started to build Burns Harbor, no private company has invested in a new integrated steel mill. Bethlehem Steel was the last company in the world to do it. Cost them five billion dollars, they almost went bankrupt, but by 1974 they had the most profitable year they ever had because they had a new fancy modern high-productivity steel mill, as opposed to these 1910 four-hearth versions they were working with. And Burns Harbor is still one of the most— probably the most efficient plant in the United States, and it's grown and people have continued to invest in it while closing other mills, okay. But in the meantime Mittal, understanding the economics and the need and the overcapacity, basically became a very rich person, only because there'd been plenty of steel plants built in the world since 1964, but they've all been built by countries.

p40 46:05

You'll take the Koreans — when I was, in the 1990s, I was the POSCO Professor, okay. POSCO was the Korean steel company. It turns out the president of Korea took one of his colonels and said, "Mr. Park, I want you to go and start a steel company." And so Park went and talked to U.S. Steel. U.S. Steel sold them the technology, and they built POSCO Steel in Korea, which by 1997 was the world's largest steel company. From the early 1970s, in twenty-five years they became the largest steel company in the world, until the Chinese passed them by a long shot. But that was at the backing of the government of Korea to build what was, at that time, probably a ten or fifteen billion dollar plant, and they built a couple of them, okay, in Korea. But Korea doesn't need but a couple of them. Saudi Arabia's got lots of natural gas, so what do they do with their natural gas? They build a gas reduction facility. They don't use blast furnaces, they use all the natural gas that they can't get out of the country any other economical way, they import iron ore, and they make steel. Because steel is fundamental to a growing economy. That's what the Japanese did after World War Two — they built steel companies and then they built shipyards, they became the largest shipbuilder in the world. Other people have passed them by, but people are all using that model.

p41 47:48

The road to industrialization is first make steel, then start making ships if you're close to the ocean, and then, you know, you start making automobiles and other things, okay. That's what the Koreans did, but they did it after the Japanese. You didn't start seeing Hyundais on the road until after you saw Toyotas on the road, right? Koreans are just copying the Japanese, and the Americans are sitting there like deer in the headlights, okay, watching all this happen and denying that there was anything to it. I remember the month that I left Bethlehem Steel, Edgar Speer, who was the president of US Steel — he was also president of American Iron and Steel Institute, which is a collection of all the companies — he was on the front page of Iron Age magazine, and in the article he said the Japanese are not using anything we don't know about. What an idiot, okay. They were using the best technology in the world, and we knew about it.

p42 48:54

That's sort of like General Motors, you know, Toyota wasn't using anything that General Motors didn't know about, okay. General Motors had the best technology in the world for making automobiles. It's right there in the research laboratory, it's not on the road, it's right there in the research laboratory. They spent billions of dollars. In the 1980s you could have purchased Toyota on the Tokyo Stock Exchange — if the law would allow you — for less than the production equipment budget of General Motors for that decade. General Motors invested 50 billion dollars in new plants and equipment, decided their employees were dirt, worse than dirt, and they had practically a bounty to put in a robot to get rid of a person, okay. In the meantime what were the Japanese doing? They were developing the Toyota Production System and treating people like they were empowered to make decisions, like intelligent human beings. In Detroit we treated people like dirt. They weren't allowed to make any decisions. If they thought something was going wrong — shut up, I'm the management, do as I say, okay. So when General Motors went bankrupt I said, they earned it. And they did, okay.

p43 50:13

So that's enough for today. Actually it's enough for until the eleventh and twelfth, okay. If you have some questions— it turns out a couple of my trips got cancelled, so I will be around some of the next couple of weeks, but I'm going the next two days.