§1. Why we don't build structures from the vapor phase [00:03]
The processing rate matters. For structural materials — where you have to produce thousands of cubic meters, large volumes — you just can't start with the vapor phase, for two reasons. The first is that it's too slow. You can't get enough atoms there fast enough. Even at room temperature there's about a thousandfold difference between the density of a gas and the density of a condensed phase. The typical number for methane, liquid natural gas, is 800. At room temperature there's an 800-fold volume change. Whether it's 800 or a thousand, there's about a thousandfold change, and therefore there's at least a thousandfold difference in processing rate compared to pouring a liquid as opposed to depositing a vapor. And in fact vapors are usually deposited at far less than 100% concentration in air. They're at very low partial pressure, which could be 10⁻³ to 10⁻⁶ of an atmosphere. So if you have 1/1000 the density and a partial pressure that's 1/1,000 to 1/1,000,000, you can have a processing rate that's a million times less, or a billion times less. It's much slower.
Vapor-phase processing — you can put coatings on structural materials to improve corrosion resistance or reflectivity or other optical properties; you can change some properties. But as a basic structural material to build a bridge that's going to go across a river, you're going to pour liquid. That's one reason. The second reason is energy cost. To get something into the vapor phase takes ten times the energy of getting something into the liquid phase. You have to break ten times as many bonds, because you have to break every bond, whereas in a liquid you only have to break about 10% of the bonds you'd have to break from a solid. So the energy cost is tremendous and the processing rate is slow, and therefore vapor-phase processing is something we do for functional materials that have extremely high value added — semiconductors — that are a million times as valuable per pound or per unit volume.
§2. Energy content as the dominant cost driver [02:45]
Another way to look at this comes out of a book on material selection by Mike Ashby, who used to be at Harvard and then retired a few years ago from the University of Cambridge. This is 1980 dollars on one axis, but the energy content in gigajoules per ton — he's got various structural materials. Gravel: all you have to do is crush the rock. Not a lot of energy per pound. Aluminum: aluminum oxide is very stable, and that's what you mine out of the ground, and it takes a lot of energy to make aluminum metal from the oxide.
Copper at that time was 100, rising to about 500. That's because we've used up most of the good copper ores in the world. The really good copper ores came out of the Belgian Congo — what's now Zaire, or whatever the name is, central Africa, right where they're having all the wars all the time. They used to have 6% copper ores. The largest copper mine still in operation in the United States is the Bingham mine right outside Salt Lake City. If you've ever flown the right way into Salt Lake City, you look down at the world's deepest pit. The Bingham mine is now less than half a percent copper in the ore. We mine gold out of ores even less than that, in parts per million, but gold is 10,000 times as valuable per atom as copper. So there's a lot of energy content in the copper just in the mining. In aluminum it's basically getting the oxygen out of the aluminum oxide. Plastics — the energy content is basically what you would get if you just burned it as a barrel of oil, plus the processing of the plastic.
Steel is around 50, and remember steel is the high-volume material in the world, other than gravel and cement. Cement is about 2 billion tons a year; steel's about a billion tons per year. Back in 1980, at $4.40 per gigajoule, steel would be about $220 of energy cost per ton, and that was about right back then. The ore might cost you $100, the energy might cost you $200, you might buy the finished steel at $400 a ton because there was some other processing to make it into shape. Aluminum at $4.40 times 300 back in 1980 — you'd have $1,400 of energy cost, and at the time the cheapest aluminum was about a dollar a pound. So steel might be 20 or 30 cents a pound, aluminum a dollar. Aluminum's got a third the density, 40% of the density, so you're getting more volume per pound, but it's still a lot more expensive. Now aluminum's up around two bucks a pound and steel up around 50 cents a pound, depending on the type.
Energy costs have become the dominant factor. When I went to work for a steel company in the mid '70s, it was about 45% ore and raw materials, 45% energy, and 10% profit. Today it's about 90% energy cost. Raw materials really aren't that significant anymore. That's partly because the only industry over the last 30 years that has exceeded the productivity of the manufacturing sector is the mining sector. If you look at productivity growth over the last twenty or thirty years in the service sector — doctors, insurance companies, Wall Street — what's office productivity been? It's been around 0%. No productivity growth whatsoever. In the medical business it's going down. In the financial sector maybe some people are getting rich, but the productivity growth is not there. They always say, well, we've got computers now, we're more efficient at screwing people in the financial industry.
In manufacturing it's been about 4% productivity growth. In the mining industry it's been about 6%. Average that over twenty years and you find that ore costs about a third as much in real dollars as it did twenty years ago — 1.06 to the twentieth power. In manufacturing, productivity has also gone up dramatically. Before the Chinese started blowing away the metals market about four or five years ago, if you went back to 2000, a ton of steel cost about half the price in real dollars of what it cost in 1980, with inflation taken out. That's because of the productivity increases.
On the kind of plot I showed you — pounds per year used versus dollars per pound — stone is very low cost because it has very low energy content. Cement, steel — these have all benefited from better mining costs. Things at the high end have gotten more expensive in a sense, because they have more energy content, because energy cost has become the big push in the cost of materials over the last twenty or thirty years. Steel is 95 pounds out of every 100 pounds of metal; aluminum is less than 2%; copper just over 1%; stainless steels are about 10% of that 95%.
§3. Why there are no all-aluminum Tauruses [10:35]
Student: [question about an article on gasoline prices and aluminum cars]
Yes — but you have to wait for the new design. That $4-a-gallon article was written originally with numbers good from the mid '90s. The break-even point right now might be $6 a gallon. In fact gas has been $6 or $8 a gallon in Norway for years. But the United States is the market where everybody dumps their cars, and we have gasoline prices that are less than $3 a gallon. That's why people make steel cars. You don't see all-aluminum Tauruses. You don't see an all-aluminum Toyota Corolla or Camry. The reason is you can't make a $20,000 car out of aluminum. Aluminum is just too expensive to make a low-priced car. But all-aluminum Audis — if I want to make a $40,000, $50,000, $60,000 car, sure, I can make it out of aluminum.
The body in white, which is the basic metal structure for a unibody car, only costs about $1,000 or $2,000 in steel. But if it costs $4,000 in aluminum, and you're trying to sell the car for $20,000, you can't afford an extra $2,000 of aluminum. If you're selling an Audi for $40,000, who cares about a lousy $2,000? The person buying an Audi for $40,000 is pretty much insensitive to whether it's 40 or 42. But the person buying a Taurus at $20,000 cares.
Student: Do they make Audis out of aluminum?
Yes, that's a big deal. Audis are almost all aluminum. Audi and Alcoa had a big thing back in the '90s. I used to say the break-even point was $3 a pound. I gave my talk at a metallurgical conference, and Peter Bridenbaugh, who was a graduate of this department, at the time was senior executive VP of Alcoa — he had the longest title, research, development, safety, technology, environment, and something else. Peter was giving a talk in the same session. I was talking about how you're not going to make aluminum Tauruses because it's too expensive, and Peter was coming in to give a talk about all-aluminum vehicles and the work they were doing with Audi. He came up to me afterward and said, "It's not $3 a pound, Tom — it's $4 a pound is the transition." He admitted it was even higher than what I had estimated. He used the $4 number. That's a mid-'90s number.
So in the material selection lectures, you have to be careful when you start talking about prices of things. I can talk about a ship or a railroad where 20 cents a pound is the value of a pound saved over the life of the vehicle — a commercial freighter, cargo carrier — versus a spacecraft where it's $20,000 a pound. They're two different beasts. But Wall Street doesn't understand that difference. They can't tell the difference between 20 cents and $20,000. You would hope they could, but they don't understand it in a technological sense.
§4. World casting production and apples-versus-oranges in handbooks [15:08]
This is from Modern Casting — pounds of cast metal, world casting production in 2004, 2005. Gray iron castings, 40 million tons. Ductile iron, which is just a better grade, 20 million tons. Malleable iron, which is expensive — requires a furnace to heat-treat for long hours — has dropped to a million tons. Steel, 9 million tons. Aluminum, 12 million tons. There's a problem with this graph. It's from the Metals Handbook 10th edition, and their source is Modern Casting. Anybody know what the problem is?
I told you we make a billion tons of steel, and they say steel castings output is 9 or 10 million tons. They say 12 million tons of aluminum — they say aluminum's more than steel. What's the problem?
Student: [inaudible]
The only steel they're counting is where the final product is a cast shape of steel. They don't care whether you made a plate or an I-beam out of steel. These are the casters; they only count steels that were cast as a piece of steel. But in aluminum, where everything is cast — that's 12 million tons of aluminum produced in the world, much of which ended up in Audis as sheet metal. What's the largest use of aluminum in the world, the largest single fraction? 40% of all aluminum production goes into beverage cans. All of that started out as a casting, and all those beverage cans are part of this 12 million tons. The steel number should be a billion tons, not 10 million, because they're comparing apples and oranges. The aluminum number is everything that was cast for any application, even sheet metal. So be careful — just because it's in a handbook, just because it comes from a reliable source, doesn't mean it's reliable.
§5. Two cautionary tales about citation [18:12]
Back in the late '70s I was giving a talk at a welding conference, and I decided I wanted to know how many welders there were in the United States. How would you estimate that?
Student: [suggestion involving college graduates]
There are lots of people who weld who never graduated from any college. College graduates don't correlate with the professions. I went to the number of pounds of electrodes that were sold. Someone has to be consuming these electrodes. I divided knowing how much metal a guy could lay down in a day, and I came up with a number like 800,000 people full-time welding. Except not everybody's a full-time welder. So I said, maybe about half a million are full-time welders. You go to a shipyard or John Deere and they've got guys who are welders — they weld all day. But if some farmer or auto mechanic welds two hours a week, you have to count those too. So I figured about 2 million welders total — probably half a million full-time and a million and a half part-time. Some people weld one hour a month, but are you going to include those?
I get up and give this talk about 2 million welders. About 1990, twelve years later, they come up with a new edition of the welding handbook, and I'm flipping through it and it says there are about 2 million welders in the United States, and the citation is the US Bureau of Labor Statistics. I thought, hey, they got the same number I did. So I called up the American Welding Society and said, where did you get this Bureau of Labor Statistics reference? They said: you. The Bureau of Labor Statistics had called up, and the only number they had ever heard was mine, so they gave it to them. But they didn't reference me in the welding handbook. They referenced the government — they're more reliable.
The other reference: Clayton Christensen up at Harvard Business School. He's been chair of the faculty; some people say he's the greatest business school professor in the country right now. He wrote a book called The Innovator's Dilemma fifteen, twenty years ago. I've known Clayton since he was a graduate student. He lives up the street from me; I was in a meeting with him just yesterday morning. Very nice guy. Back when he was writing The Innovator's Dilemma, my daughter — now in her mid-30s — was a college student, and she needed a job one summer, and Clayton gave her a job doing research for his book. One of the case examples in his book is the steel industry, and how the big integrated producers didn't innovate and the little electric-furnace steel makers did innovate. He asked Rebecca to go find out what it cost to build an integrated big steel mill, and she came home one night and said, "Dad, I can't find it anywhere. Do you know?" I said, sure, Rebecca, I've got a slide in my office.
I bring in a copy of the slide. The slide was made in Saginaw, Michigan, in the airport when I was visiting General Motors with one of my graduate students, and we were talking about what his thesis was going to be, and I jotted down what it cost to build an integrated steel plant versus a mini mill — people were talking about micro mills back then. I pulled the numbers out of the air. Well, not completely — it was my best estimate: $15 billion to build an integrated steel plant in the mid '90s. I know what it cost to build a steel plant. The last steel plant ever built by a company in the world was the Bethlehem Steel Burns Harbor plant in 1965. It cost $6 billion. Burns Harbor, Indiana. You take inflation and so it's $15 billion to build one in the mid '90s — today it's probably $20, $25 billion. My number's not way off. $15 billion order of magnitude is the right type of number. So I gave the slide to Rebecca, and if you look in Clayton's book, I'm referenced for the cost of a steel plant. I just sort of pulled it out of the air — well, I had a basis for it. But now it's referenced in this great management text. So be careful when you see numbers. I only tell my students about it, because — hey, Clayton wrote it in his book, it must be the gospel, right? Clayton doesn't even know this story. All he wanted was a reference. I'm the professor of metallurgy at MIT — who would know better? In fact, I'll bet you I'm right.
§6. Why nobody builds new integrated steel mills anymore [24:16]
Let me finish that story. We've had new steel mills built since 1965. POSCO, the world's largest steel company, in Pohang, Korea, built a steel mill in the 1970s. But POSCO Steel was not bankrolled by financial people — it was bankrolled by the Korean government. There have been probably a dozen steel mills built in the world since 1965 on greenfield sites, but not a single one has been built by a company that was risking shareholder money. It was countries who built a steel mill because they saw steel as vital to the growth of their nation.
Why? Once you get to the $10, $15, $20 billion range, very few companies are willing to risk that type of investment. If you're in the aircraft business, Boeing has to — when they come out with a Dreamliner, that's a $20 billion investment, supposed to be $10 billion. Intel, a new fab, over $1 billion. There are only a few companies in the world profitable enough to take the risk of a $1 billion investment. Usually it takes a nation to do that.
So one of the things I'm trying to get across today is the magnitude of some of these industries in structural materials. It's not a small business. These are businesses where running a blast furnace can be a value of $100,000 an hour. You guys are in the Navy — what costs the Navy $100,000 an hour to run? I was going to say a carrier.
Student: [reactor reference]
The reactor — 30%, well, okay 100%, but 30% is basically full tilt for a reactor.
§7. The carrier as economic unit [26:10]
Student: [About something costing] $100,000 an hour now. You guys are in the Navy — what costs the Navy $100,000 an hour to run?
Student: A carrier.
One carrier — the reactor at full tilt is maybe thirty percent of operating cost. But if you have 5,000 personnel on a carrier and you start adding in the operations, running that reactor might be fifty percent of the cost of running the ship. I was just figuring: you've got a $15 billion investment in that ship and you've got to amortize it over forty years. A kid in high school could run through those numbers if you taught him how. You've got eleven or twelve carriers now, so you're talking a million dollars an hour or $2 million an hour just for the carriers — without the battle groups that go with them. It takes a nation to run a carrier.
§8. Iron-making: the three ingredients [27:45]
It takes a nation to build a big steel mill nowadays. Now I want to talk about how we do casting. Since ninety-five percent of all casting is basically making steel — I'm not trying to emphasize steel too much, but let's face it, there's a lot of steel in the world — I'm going to talk about steel, because you're going to be involved in buying lots of steel, or working with lots of steel whether you buy it or not.
Steel is made from three things: iron ore, limestone, and coal. At Saugus Iron Works they were cutting trees, not coal — making charcoal. But you need an energy source, and traditionally that was coal. In the bogs in Saugus they could get iron ore, and there's limestone all over New England. The limestone is the flux, the iron ore produces the iron, and the coal goes through coke ovens.
Coal ordinarily has all kinds of volatile components, so you do the same thing with coal that you do with wood for charcoal: you burn off all the volatiles in a reduced-oxygen environment. That's a coke oven. If you want to see a carcinogen factory, it's a coke oven. When I moved to Bethlehem, Pennsylvania, you could smell the coke ovens all over town — the sweet smell of sulfur. There's a whole area of chemistry dealing with this — coal tar. If you go back to 1900 you would have taken courses at MIT in coal tar chemistry. That's the volatiles that come off at 2,000°F out the stack. What you're left with is carbon similar to charcoal but from coal — we call it coke.
You take that, and you crush the limestone — you don't have to do much more than that. Today we have to take iron ore from different parts of the world, grind it and sinter it and make it into pellets. They used to be golf-ball size; now they're smaller than that to get faster reactivity — marble-size pellets of iron ore. In the old days, anyone ever heard of the Mesabi Range in Minnesota? The iron ore was so pure all you had to do was crush it up like limestone and stick it in your blast furnace. But we ran out of the Mesabi Range right after World War II — we just mined it into extinction. There's still lots of iron ore up in Minnesota, and the Great Lakes were huge shipping routes going from Minnesota to Pittsburgh and Chicago.
§9. The blast furnace and the three steelmaking processes [31:10]
You put those three things into a blast furnace. This is about forty stories tall, and today a good blast furnace costs you two, three, four, $5 billion depending on the size. A big blast furnace will have 4,000–5,000 cubic meter internal volume. It will produce 5 million tons a year — 10,000 to 15,000 tons a day of cast iron. The blast furnace is 100% ore; it's all virgin iron going to the steel refinery.
Steel is made by three processes: electric furnace, open hearth, and basic oxygen furnace. They had to add some flux into all of these. The electric furnace can start with pig iron, but that's an expensive way to do it. Thirty-five years ago when I worked at Bethlehem Steel, we used to figure $180 a ton for the cast iron coming out of the blast furnace. So 10,000 tons a day at $200 a ton — that's $20 million a day, about a million dollars an hour for a big blast furnace. The small ones we had at Bethlehem Steel, which were built in 1911 — they were very proud of their ancient technology — were only worth $100,000 an hour. A big blast furnace today costs more than a carrier to run.
The electric furnace can take anywhere from 0 to 100% scrap; typically they take 100% scrap. The open hearth used to take 70 to 80% virgin cast iron from the blast furnace and couldn't take more than 20 or 30% scrap. The basic oxygen furnace can take about 30% scrap and has to take 70% virgin ore. Before 1880 we didn't have a steel industry in the world. Andrew Carnegie became the Bill Gates of his day — he was the richest man in the world.
§10. Recycling and the postwar steel economy [34:43]
In the United States we use 100 million tons of steel a year. Some of that steel lasts 100 years before it's recycled; some lasts thirty years. In the first half of the 20th century we were putting in maybe only 50 million tons a year; in the second half we were certainly putting in 100 million tons. Take 50 million tons times 100 years — that's 5 billion tons of steel. That has to be recycled, and in fact 70 to 80% of all the steel we produce is recycled. Can you imagine the size of the landfills if you couldn't recycle steel? Well, you can imagine, because concrete can't be recycled — and we're putting 2 billion tons of cement into the environment worldwide in a year. You can crush it up and make aggregate for roads, but you can't pave the whole world.
Student: [comment about breakwaters and artificial reefs]
They call it an artificial reef for the environmentalists, because otherwise they're not going to allow you to dump trash in the ocean. But it actually isn't particularly harmful — concrete started out as limestone and ends up as calcium hydroxide. It's not a terrible thing. We don't recycle a lot of concrete, but we're using twice as much of that as we do steel.
In the 1960s and 70s the cost of scrap dropped dramatically. One of the reasons people say the Japanese attacked Pearl Harbor was that we were starting to put restrictions on the amount of scrap steel we would sell to Japan. We were the consumer of steel for the first half of the 20th century; we were producing the scrap that was coming back after twenty or thirty years of use. We had all the steel scrap compared to most of the rest of the world, and we were withholding it. This wasn't the only reason for Pearl Harbor, but it was one — read the historians. We put restrictions on the amount of scrap they could buy, and that was fueling their industry.
The steel industry is so huge it actually controlled a lot of the world's commerce. In 1962, US Steel wanted to raise the price of steel. None of you were born in '62, but I was. There was a big battle between the president of the United States and the president of US Steel, and Kennedy forced the US Steel industry to roll back its prices, because he said raising the price of steel would be inflationary to the world economy. It's the same thing as oil prices today. He could control US Steel; he can't control OPEC. You could pump oil out of the ground for a nickel a barrel back in 1960, and we had plenty in Texas and Saudi Arabia had even more. But then when oil became scarce, energy became the big thing, and not steel.
§11. Bethlehem Steel, 1975: the fully depreciated plant [38:45]
After World War II the United States controlled 75% of the world's steel production. Why? Because we bombed out all the other capacity in all the other countries. We didn't have a war on our turf — there was not a single steel mill bombed in the United States. We built a lot of capacity in World War II, but it's not what we built, it's what we destroyed.
In 1975 when I went to work for Bethlehem Steel, the managers who had been hired in 1945 right after the war were now the leaders of Bethlehem Steel and US Steel, and they still had the mindset that whatever they did, they were the king. They now in 1975 controlled 25% of the world steel production, and everybody else was making steel in modern plants they had built since World War II. Bethlehem Steel was very proud that our coke ovens were built in 1910 and our blast furnaces were built in 1911.
I went to a seminar as a new hire where the Vice President of Finance said it cost us nothing to make steel because our plants were fully depreciated. I was stupid enough to raise my hand — I was only 24 years old — and say "Excuse me sir, I don't understand why it doesn't cost us anything to make steel just because our plants are fully depreciated." He said, "That's because you don't understand finance." That was the answer he gave me. I later took an accounting course, and I learned that it's true I didn't understand finance — but neither did he. The fact that your plant is depreciated has nothing to do with what it costs to make something in it. That's an accounting game; it has nothing to do with the real cost. Our costs were double the Japanese cost, because we were building in 1911 plants that we were holding together with duct tape and a prayer. The Japanese had brand new mills, no more than ten or fifteen years old, the most productive in the world. And we were wondering why they could ship steel to the United States and sell it for less than we could make it.
§12. Gordon Forward and the mini-mill calculation [41:10]
What's the shipping cost? It's been about the same for the last thirty years. What's the shipping cost to get a ton of cargo across the Pacific or the Atlantic? You guys are in the business of ships and you don't know?
Student: Ten cents.
No — $30 a ton. Gordon Forward, who started one of the electric-furnace steel companies, started a mini-mill in 1975. Everybody said: how can you compete with the Japanese and their big integrated mills, $5 billion plants? Well, first of all, it only cost him $250 million to build his mini-mill. Sure, it only produced one-fifth or one-tenth as much, but the capital cost was less than one-tenth — it was about 5%. It couldn't produce the full range of products that the integrated mills could. But he used to say, "If we can keep our labor cost below $30 a ton, we don't care if the Koreans pay their employees nothing, we can compete in the American market." Because it cost the Koreans $30 to ship their steel over here, and their only competitive advantage was labor cost. [Forward] is a graduate of MIT, this department; he's retired up in Vancouver now, Canadian by background.
§13. Cupolas and Mao's Great Leap Forward [43:17]
Steel today is a trillion-dollar worldwide industry — a billion tons times $1,000 a ton; I did that math in my head. Before I get to the three furnaces, let's talk about cupolas.
Student: [question about adding scrap]
You can add scrap in an electric furnace, but you can't in an open hearth furnace or a basic oxygen furnace — and I haven't told you yet what these furnaces are. Out of the blast furnace you get pig iron, or cast iron. It's actually carried over — you don't solidify it, that would be a tremendous waste of energy — so you carry it in a hot metal car, about 100 tons at a time, a big railroad car with lots of bogeys. It only goes about half a mile through the plant over to the steel-making side. The iron-making part of the plant might have a thousand employees; the steel-making part another thousand. The cast iron has lousy mechanical properties — today maybe $300 a ton value for the molten iron. By the time it's steel as a casting with no shape yet, you're talking about $400 to $600 a ton. Give it shape as an I-beam or a sheet with particular properties and you're up to $800 to $1,000 a ton selling price.
On the other side, they take the hot steel and continuously cast it — that's where I started out, I was going to talk about casting. They continuously cast it, or they ingot-cast it, or they make pig iron which you can sell to someone to make cast iron parts. From ingots you can do great big hot forgings, you can do ingot breakdown to make blooms and shapes and plate; continuous casting makes slabs. We're actually going to talk about all this, not just for steel but for all metals — but I'm going to dwell on steel because it's the 900-pound gorilla, 90% of the metals.
I've been walking through steel mills for 35 or 40 years and it still amazes me every time I go through a mill, to see the scale of things. I'm sure it amazed each of you to walk through a shipyard the first time. But shipyards are just a small version of a steel mill. Integrated steel mills are big — in fact they're the biggest things in the world. The bigger steel mills are dealing with 350-ton castings — 700,000-pound castings, the largest castings made in the world. A shipyard may be dealing with a 100-ton casting. You have some of the biggest cranes in a shipyard, because you actually lift whole sections that weigh a lot more than 300 tons. But in terms of a fabricated hunk of metal, steel mills outdo shipyards.
[Tom turns to the cupola diagram.] If you look up "cupola" in the dictionary, it's the little house with a weather vane on top of a barn. But a cupola is also a melting furnace for cast iron. There's nothing really fancy about it — it's not all that different from Saugus Iron Works. You put your charge — iron ore, limestone, coal — in the top, and you have a flame at the bottom where you blow air in. These are called tuyères, t-u-y-è-r-e-s. Later we're going to talk about a blast furnace. A cupola is nothing more than a small blast furnace. Saugus Iron Works is nothing more than a small blast furnace. But we call them cupolas when you're just making cast iron.
In the 1960s, Mao Tse-tung in China decided he was going to industrialize China, and as part of the Great Leap Forward, all the communes were supposed to have their own cast iron cupola. So instead of making pottery and clay utensils, they were going to make cast iron. Mao Tse-tung was leading China right into the 16th century. By 1975, ten years later when he passed away and they opened up China, all of a sudden between '75 and '85 we were flooded with little cast iron trinkets, because every commune now had to try to figure out some way to get currency. I have a cast iron Christmas tree stand that was made in China. This was the beginning of Made in China — the modern Made in China. China had probably half of the world's cast iron capacity in these little inefficient, environmentally unsound cupolas — maybe sort of Saugus Iron Works fancy. Now they've gotten rid of a lot of that and gone upscale. There's apparently one town in China that makes 90% of all the world's neckties. They learned to specialize.
So this is three different versions of a cupola. This is just an air-cooled wall. This one is a shell wall — it has an inner cavity that's water-cooled; if you want to really pump stuff through there, you don't want to melt the outside walls. And this one is even less sophisticated — a single shell, and they just flood the outside. If you came up to it, you'd just see a curtain of water — they're pumping Niagara Falls around the thing to keep the walls from melting. Productivity goes up as you cool the walls.
§14. The Detroit blast furnace they kept running [51:58]
They do this same type of wall cooling on the huge steel blast furnaces — not on the modern ones, but I had one in Detroit that was built in 1925. The walls were starting to buckle — it was an old riveted structure — but it was producing $100,000 of product per hour. They knew it was falling apart, but no one wanted to call the shot to shut it down. They would send welders up there while the thing was still hot, welding the shell back together. The shell was buckled, thinned down to one tenth of its original thickness. They were flooding the outside with water to keep the whole thing cool. At $100,000 an hour, who's going to shut it down? Not the outside consultants, not the internal staff — no one was going to take that responsibility. So they kept pushing it until it blew up one night and caused a half-a-billion-dollar loss. Hey, the insurance company paid for it.
§15. The Mississippi flood and federal flood insurance [52:49]
So they kept pushing it — nobody outside, nobody internal was going to take responsibility for shutting it down at $100,000 an hour — until it blew up one night and caused a half-billion-dollar loss. The insurance company paid for it.
Student: [reaction]
I'm just telling you the way it really is out there in the world. Funny thing about insurance companies — all those houses that got flooded this last Mississippi, Missouri River, whichever river, a million acres of land. The government actually sold those people insurance, nobody else would. If you're in the lowlands, no insurance company is stupid enough to insure you. The only one stupid enough to insure you is the US government.
You've heard the Arctic ice cap is melting, Greenland is losing ice, the water level is going to rise. You can't get insurance on the coast from anybody except the US government. But those people who want to have beachfront property — your taxes are going to pay for them to have those beautiful days in the summer, and on the days in the winter when they have to evacuate, we will rebuild their house for them. Your taxes, my taxes. Aren't we wonderful people.
§16. Who has authority to shut a plant down? [54:59]
Now you know why I call this course "Materials Processing for Poets." You're not going to know how to process materials when you're done, but you're all going to be managers anyway. And you're going to be the ones where some chief comes to you and says, I've got a crack in the wall of the reactor — do you think I should have shut it down?
I'm not kidding. They don't know, and they don't have the authority to shut it down. They do if they think it's safety-related — only safety-related. Maybe in the Navy, but commercially, not everything is safety, or they rationalize it away that it's not, because no one really wants to take responsibility for making the expensive decisions. They know that no matter what decision they make, someone's going to second-guess them afterwards and tell them they were wrong. You lost the profitability of the company because you made the wrong decision.
Everybody can do Monday-morning quarterbacking, and the only thing that will save you is safety. I did it because of safety — that's the only defense you have. And it's not all that different even on Wall Street. People have to make billion-dollar decisions, and most people don't understand what they're doing well enough to make an intelligent decision. So a lot of times there is no decision, which is the worst decision.
§17. Energy crises and diversification [56:46]
Student: [earlier question about energy crisis]
The problem with energy crises is the same story investors have had for centuries: you don't put all your eggs in one basket. The reason we had an energy crisis in 1973 is that we were almost totally dependent on oil. We had no capacity to immediately switch from oil to natural gas to coal in most of our big energy-consuming industries. When the oil supply got disrupted, all of a sudden we had a crisis — there wasn't enough energy available, because the only energy we could use was oil.
We had another energy crisis in 1978 or thereabouts, nowhere near as bad. The reason it wasn't as bad is that the managers who'd had the government come in and tell them, you've got to shut down your plant except in the middle of the night because we don't have enough electricity — and they were losing hundreds of thousands of dollars an hour because they couldn't run at capacity — said, I'm not going to let this happen again. They went out and bought insurance, in the sense that they could run their furnaces on oil or gas, and some of them on coal. They diversified. So the next energy crisis, in terms of actual restraint on oil supply, may have even been worse, but it wasn't felt anywhere near as bad, because industry had the ability to flip the switch and change to natural gas or coal. You don't have to have everyone have flexibility — you only have to have some flexibility. In 1973 we had no flexibility.
People have been predicting for fifty years that we're running out of oil in the next twenty years. Why are we not running out of oil? There's plenty of oil still in the world, but not at $2 a barrel. We've used up all the $2-a-barrel oil. The only $2-a-barrel oil that's left is in Saudi Arabia — and actually it costs them about $2.50 to pump it out of the ground now. I heard this years ago, in the 1980s, from a guy from AT&T who was working over in Saudi Arabia: they don't even meter the gas. When you fill up your car, it's a flat rate — two bucks. You need five gallons, two bucks. You need twenty gallons, two bucks. In Libya I read it was ten cents a gallon, whatever. They don't care, the price is nothing. The inconvenience is having to get out in the sun to fill up your car. To them oil is cheaper than water.
§18. Tar sands, Alberta, and Venezuelan heavy oil [60:17]
So — energy crisis? We've got as much heavy oil in the form of tar sand in North America, including the Canadians, as Saudi Arabia has in reserves. 200, 250 years. Now, you can't get it for $2 a barrel. When I first worked on a case up in Alberta, it was about $30 a barrel, and oil was going for about $30 a barrel, and they were barely able to make do. Now, because of environmental concerns, it costs them $50 a barrel, but it's selling for $100, so they're making out.
Anybody been to Alberta? Richest province in Canada. Money flows like water — billions and billions of dollars of oil money. All the other provinces are going bankrupt; Alberta is not. The most American province in Canada — because all the Americans have moved up there to get jobs. It wasn't like that twenty years ago when I was first up there, because when it cost you $30 a barrel and the price of oil was dropping at times to $15 a barrel, you could lose your shirt. But now oil is well above $50 and staying there, and they make lots of money.
There's a trillion tons of heavy oil in the Orinoco River valley of Venezuela. Lots of oil in Venezuela. That's the light stuff. After World War II they started shipping oil to the United States — and there's a good little story I can tell you about a shipment that came up here right after World War II and started firing up the Boston Edison plant, which was the energy generator back then.
§19. John Wolff and the vanadium-in-Venezuelan-oil story [62:38]
John Wolff was my thesis adviser when I was a freshman. John had the office that I'm in now. He told me this story when I was a freshman. He got a call one morning from Boston Edison. They needed a consult, because overnight all the stainless steel in their whole plant had just rotted away to nothing. And he said, Well, I know the answer, but it will cost you $5,000. $5,000 in the late 1940s was like two or three years' salary — kind of like $200,000 or $300,000 today. He was going to extort a fair amount of money out of them, because he knew the answer. He'd been looking at impurity elements in oil for ten or fifteen years.
John was German — W-O-L-F-F. During the war, he wasn't involved in things like the Manhattan Project. But he knew there were lots of things going on in the buildings we're in right now that had to do with people who were interested in uranium. John Chipman — the Chipman Room is just across the courtyard — John Chipman was working on a crucible material to hold molten uranium. No one had had a use for uranium before. John Wolff didn't know exactly what this was about, but at a cocktail party of faculty one evening he said, You know, there's a lot of uranium in the oil in the Balkans. It was one of the impurity elements in the oils he was studying. He said there were two Secret Service agents in his office the next morning asking him what he knew about uranium. So there were people at MIT who knew about the Manhattan Project — but John didn't, because he was German.
Anyway, after World War II, when Boston Edison's stainless steel went to pot, they said, If you know the answer, we'll pay your fee — we need to know. They'd just lost a million-dollar plant; what's $5,000? So once the agreement was signed, he said: Last week you got your first shipment of Venezuelan oil. They said yes. He said: Venezuelan oil contains vanadium, and vanadium forms an oxide that melts off the protective chromium oxide of stainless steel. They had just oxidized their stainless steel, because the vanadium impurity in the oil ate it up overnight.
Today the American Petroleum Institute has specs on getting the vanadium out. There is really only one alloy — a nickel-chrome alloy — that can survive vanadium. I actually got hired briefly on a job up in the tar sands area of Alberta, because they have vanadium in their oil too. They have processes now to get rid of the vanadium, but back then no one really knew about it, because the US always had the nice crude from Texas that didn't have this problem. John Wolff knew about it at the right time and found the right client who had a big enough problem that he could put a couple of kids through college on that one answer. You've got to be at the right spot at the right time.
§20. Cupolas and cast iron [66:23]
[Tom holds up a cast-iron part.] This is a piece of cast iron. This part was in my office when I showed up Friday night after my travels. I need it back — I have to do some tests on it — but it's part of a big tractor-trailer. It actually holds the king pin, the big six-inch-diameter pin that holds the trailer to the tractor on the fifth wheel. This was the lockdown to keep the thing from bouncing out, and it turns out it did bounce out in one case. We don't have to get into the accident that caused. But there are complex shapes for which making something out of cast iron is fine.
A modern cupola: you pour the material in at the top, you have different types of ceramic on each side, you have to have expansion joints. It's really hot down at the bottom where all the fuel is melting and the stuff is coming down from the top. You have these blasts coming in — these are tuyeres — just like the bellows in the blacksmith shop. You heat things up by blowing air in. Every now and then you open the plug — just a big plug of sand — and the cast iron comes out. You cast your pigs, your little brackets, your Christmas tree stands, whatever you're making.
That's cupola technology. It's still used in bigger cupolas in lots of foundries all over the country. People who make brake calipers, engine blocks for cars — they're typically melting in cupolas. The American Foundrymen's Society has gotten money out of the Department of Energy for more efficient cupolas that use less energy, so you can pay Americans more to compete against the Chinese who are making cast iron parts and paying their people nothing.
§21. The blast furnace: anatomy and scale [68:48]
This is a blast furnace — not a cupola. It has what they call the bell at the top, which is basically a valve. You pour your stuff in, drop the bell, everything falls in. You have to have abrasion-resistant walls. You have erosion all the way along. You have thermal shock from thermal expansion. The difference from a cupola is that this might be twenty stories tall in a steel mill. It's not a small furnace. This furnace could cost you $2 billion just to rebuild.
Student: How thick is the steel shell?
Maybe one inch thick. There's an internal pressure of a few PSI — five PSI. The blast pressure: you're blowing air in pretty fast, basically sonic velocity, to get the fastest burning rate. You've got a mixture of coke pellets, sinter pellets (iron ore), and lumps of limestone. The limestone starts to melt, and down at the bottom you get slag — a molten glass of slag floating on top of the cast iron. The carbon is mixed in. Down at the very bottom you get this pool of iron. On one side you take off the slag — there's a little tap to get rid of the molten limestone and silica glass that came from the iron ore rock — and a slightly lower tap for the iron. The whole depth of the molten stuff may only be three or four feet. You tap it off every forty-five minutes to an hour, take a couple hundred tons off.
In a big modern Japanese or Korean mill — 5,000 cubic meters, 10,000 tons a day, maybe 15,000 tons a day. One of these can produce four to five million tons of steel a year. That's half a percent of the world total out of one blast furnace. POSCO, the world's largest steel company, produces 30 million tons at one site — three percent of the world's production at one steel mill. Much of it goes into Korean shipbuilding, which is maybe ten or fifteen miles away. They may have five blast furnaces producing 25 million tons of steel a year. It only takes five of them.
Lots of refractory in there. Three or four feet of ceramic refractory brick lining the thing. You may spend $100 million on ceramic brick relining; it takes six months. The furnace will run for five years or more before you have to shut it down and completely rebuild it. The rebuild can cost you a billion dollars. On the smaller, older ones — and probably even the newer ones — they have injectors that can spray more ceramic on the inside walls, because you hate to lose half a year's production at a couple hundred thousand an hour in downtime. They have all kinds of ways to keep these things going. The ceramic gets eroded away over time.
The temperature gradient across the wall: 2,400°F on the inside, a couple hundred degrees on the outside. Sometimes they water-cool the outside steel shell — just a sheet of water coming down, a waterfall. They weld rods on the inside so the ceramic brick can be attached to the shell. It's a pretty complex structure. A multi-billion-dollar facility.
§22. Blast furnace stoves and air preheat [73:41]
Here you can see the temperature difference going up the furnace. The gas comes off — it takes about 600 pounds of coal for every ton of cast iron, so a 600-to-2,000-pound ratio. With all that coal burning, you have two stacks coming up alongside, and the gas comes down. Most of this is coke oven gas — well, coke oven gas is one thing, blast furnace gas is another. The blast furnace gas goes through one of these regenerators, and the air blast comes in through the other, hot regenerator to preheat the air.
You have this on ships when you're burning fuel in steam boilers — you preheat the air to get more efficient burning. You can't get these temperatures unless you have a blast. The reason it's called a blast furnace is that you're blasting air in, just like the old bellows in the blacksmith shop. This is a big bellows. The brickwork in the stoves: you have two of them because one is being heated by the exhaust gas — cooling the exhaust gas and heating the brick — while the other is heating up the incoming air. Every three or four hours you switch over. It's just energy conservation.
These stoves are as big as the blast furnace itself. That's why even though the furnace is a $2 billion piece of equipment that costs a billion dollars to rebuild, the whole facility for a blast furnace operation is probably $5 billion today. You have to have all your ore handling equipment, raw materials handling, and all these blast stoves. It's a lot bigger than a football field — probably ten football fields' worth of real estate. Not a small facility.
§23. ASTM specs and A36 structural steel [76:04]
Are you guys familiar with ASTM specs? There's a six-volume set for steels, a five-volume set for all other metals. Back in 1998 — the last time I bought a set for myself; you can get the current year at the MIT Library — this was Section 1 of ASTM. There are thirty or forty different sections. All the ASTM spec books together would take more than the width of this room. Plastics, anything. ASTM is down in Philadelphia, and they make a fortune off selling specs. They have committees.
If you look in here under structural steel, reinforcing, pressure vessel, railway — I picked out A36, garden-variety ASTM A36 structural steel: plates, I-beams, bars. This is what a civil engineer uses to build bridges and buildings. 36 ksi yield. Garden variety. And it says, in Section 6, Process: the steel shall be made by one or more of the following processes: open hearth, basic oxygen, or electric furnace. Those are the three we talked about. So for garden-variety A36 you can use any steel mill in the world. But there are other specs in here — high-strength steels, like the Navy HSLA-60 — with tighter requirements.
Now look at the spec on cast irons. It's not going to tell you anything about open hearth, because open hearth, electric furnace, and BOF — basic oxygen furnace — are steel-making processes. What's the difference between steel and cast iron? Cast iron melts at around 2,000°F, and it's got 3 or 4% carbon. In some forms, like ductile iron, it's almost as good as cast steel, but not quite, in terms of mechanical properties. The bulk of cast iron: you hit it hard with a sledgehammer and it fractures.
Have you ever seen the water main guys repairing a leak? They get down in the hole, uncover this big cast iron water main, and when they need to cut it out, they don't have to saw it — they just take a sledgehammer and hit it. It just shatters. That's what carries our water, and that's why you have big floods in Cambridge. Some of those cast iron pipes are 100 years old. So they've gotten a fair amount of service.
Now, the basic open hearth — we haven't built one in the world since the early 1970s. But for 100 years before that — well, 90 years — this was the way steel was made. It took 24 hours to make 300 tons of steel.
§24. Basic Open Hearth and the basic oxygen process [79:26]
Those cast iron pipes are 100 years old, so they've gotten a fair amount of service. Now the basic Open Hearth — we have not built one in the world since the early '70s, but for a hundred years before that, or 90 years before that, this was the way steel was made. It took 24 hours to make 300 tons of steel. It took forever. This is the size of a football field, and you can tap it into ladles. You had big brickwork on either side the same size as this to preheat the incoming air, to recover the heat from the outgoing air, and you switched over every 6 to 12 hours in your brickwork. So you had three football fields for one furnace to make 300 tons of steel.
And then, in a little place in Austria — actually I think it's 1958 — they decided to try pure oxygen and burning off the carbon. Steel is basically burning the carbon off the blast furnace iron. Blast furnace iron has 3 or 4% carbon. You've got to get it down to a tenth of that value to have steel. You've got to burn off the carbon. In an Open Hearth you just use air flowing over the furnace to slowly burn off the carbon. Here you blow a supersonic jet of liquid oxygen into a molten steel bath, and you burn off — not in 24 hours but in 10 minutes — all the carbon. So now every 45 minutes for the whole cycle you're producing 300 tons of steel, as opposed to 3 or 400 tons in a day. Pretty big productivity increase. It all occurred between 1965 and 1980.
So the steel industry had this tremendous new high-productivity process. They had to have a source of liquid oxygen. Companies like Linde will build a liquid oxygen plant right there on the steel plant grounds, because about 15% or 20% of all the liquid oxygen in the world goes to making steel — burning the carbon out of the cast iron. And the plant is quite a bit smaller. You have all this stuff to take care of the gas coming off. Here's your vessel, might only be 20 or 30 feet tall. You have ingots, ladles, your scrap. You throw in some scrap, you put in molten iron, you put it all in the vessel. You only fill up the bottom four or five feet of the vessel — a person might be 20 to 30 feet tall standing next to it, and you only fill up about the bottom 20% — and here's your lance for oxygen. But when you blow that oxygen in there, this thing fills up with a foam of molten steel. It's really exciting. You want to see fireworks, you want to go through a steel plant. 2700, 2800, 3000 Fahrenheit foamed steel, 300 tons. It's neat, but it's big scale.
§25. The electric arc furnace and the scrap economy [83:30]
And here's an electric furnace, which has been around for 120 years ever since Edison and Westinghouse. It's nothing more than three carbon electrodes — about three-and-a-half feet diameter carbon electrodes — three-phase electricity. The big ones today will run about 150 MVA, 150 megawatts (actually MVA if you're an electrical engineer, because it's reactive power). Big electric arcs. 200, 300, 400 volts at 50,000 amps. You can melt 100 tons of steel in half an hour, 40 minutes. You can use 100% scrap.
Remember we put 50 billion tons of steel in the environment, and we're getting back each year about 70 or 80 million tons in this country, and in the world we're getting back about 500 or 600 million tons a year of scrap. Well, we don't need to make everything in blast furnaces anymore. If I went back to just 40 or 50 years ago, 80% of my steel was made from ore, because the world was on a growth spurt after World War II as it was industrializing and rebuilding. They needed more and more steel. Around the 1970s and 1980s the market flattened, partly because of the energy crisis, partly because the world couldn't afford more steel. We were getting so productive that we could make so much in the plants we had that we didn't need new plants. That's why since 1965 no company has been able to justify building a new plant — only countries who are doing it for political-economic reasons, not for business reasons. The countries would build excess capacity, and all these other countries that had enough capacity, like the United States, were being competed against by countries who wanted to own their own steel mills. They didn't want to just buy from the United States anymore.
And so all of a sudden the US steel industry was operating at 50, 60% of capacity, when they needed to — just like the airlines, the airlines can't make money unless they're operating at 80, 85% of seats filled. Well, the US steel industry couldn't make money at 60% of capacity, when whole countries like Korea or Japan or China or Romania were building steel plants. The American steel industry couldn't afford to replace their 1911 blast furnaces. They didn't have the profitability. President Kennedy helped ensure that in 1962 when he made them roll back prices. The whole steel industry is a very complex story, but from the 1980s and 1990s Wall Street would say this is a dying industry — I'm interested in silicon. Do you know how many bridges have been built out of silicon? Some of us knew they would still be building steel cars, steel bridges, steel ships. But no one would believe that on Wall Street.
They didn't believe that at the Navy either. They thought they were going to build all-composite ships in the future, and then they built some and they found they didn't work so well. They found the modulus of steel had a certain advantage when you made big things. That's the old story of the Visby, the Swedish corvette, the world's largest composite ship — you can only take it out in good weather. Don't take it out in a storm. That's good as long as you have an agreement with the other side that they'll only do battle in good weather, right.
§26. Why electric furnaces tolerate more scrap, and the residuals problem [87:45]
Why can the electric furnace tolerate more scrap? The Open Hearth and the basic oxygen furnace need molten metal to get the reaction going. They've got to build up a froth from that liquid. The liquid steel is going to react with the iron. If you put cold scrap or even warm scrap in and you blew pure oxygen on it, you'd just be blowing your oxygen away. There's nothing to react until you get a liquid. You can throw 30% scrap into the basic Open Hearth or the basic oxygen furnace, but you can't put in so much that you'll freeze all your liquid. You have to have molten cast iron so that when you blow the oxygen in, the oxygen will react with it. The electric furnace, all your heat's coming from electricity — you can put cold stuff in, 100% scrap.
The problem is scrap is not pure. It's got alloy steel, it's got carbon steel. The conventional wisdom was you could not make all grades of steel in an electric furnace. From a basic oxygen furnace you're getting out virgin iron, and you can add alloying elements to it. So the basic oxygen furnace had fewer non-metallic inclusions and didn't have residual elements. If I get a small piece of steel and people ask, was this made in a basic Open Hearth or electric furnace, all I have to do is look at how much nickel and chrome is in there — because when you're using scrap you've always got a little stainless steel mixed in, could be stainless steel screws in something. You always have some residual elements, and a steel metallurgist can look at the composition of a little bolt and tell you whether it was made by electric furnace or basic oxygen furnace.
When you're remelting you can't get all the elements out. Some of them you can burn away and others you can't. Anything more noble than iron — because if you start burning away the iron you no longer have anything of value, right? If you go back to ore, to iron oxide, you haven't gained much. As a result, elements like bismuth they won't let near a steel mill. At the 10-part-per-million level, bismuth will go to the grain boundaries of the steel and embrittle it. It'll be junk. And you can't refine it out of the steel. You can blow oxygen in until you've burned all the iron away to nothing before the bismuth will burn away. That's just the fundamentals of the chemistry of the process.
However, they're going to replace lead in solders, and a lot of them are using bismuth. Well, you know what that's going to do to a lot of the world's steel? It's going to make all the scrap useless. We're going to have to landfill it. We don't have enough of those bismuth alloys replacing lead alloys yet, but in the next 10 or 20 years they're going to have to segregate the scrap more carefully than they do today.
§27. Mexican aluminum scrap and roadside lead [91:21]
Give you an example. In making aluminum, lead is just as bad as bismuth is to steel. It doesn't alloy with the aluminum, it goes to the grain boundaries. You try to roll the aluminum and it'll just crumble to pieces when you try to hot-roll it, because the lead's at the grain boundaries, it melts, and the aluminum just falls apart into pieces. A number of years ago, Alcoa Tennessee, which makes a lot of can stock, was having all kinds of problems with Mexican aluminum scrap. They traced it down — they had too much lead in the Mexican scrap, and they thought the Mexicans were slugging, because they buy scrap by weight. They thought they were throwing in lead weights, because lead is cheaper than aluminum, just to get the price up. And they couldn't find it.
Finally they did a more extensive study. At the time — this was in the early '90s — the Mexicans were still using leaded gasoline, and they don't have five-cent deposit laws in Mexico. People would just throw their cans along the side of the highway. Other poor Mexicans would come along and collect the cans. But if you plotted the lead concentration from the side of the highway, it was very high right at the edge of the highway and within 200 feet on either side it dropped to very low values. It was the lead deposits from the auto exhaust that was getting on the cans on the side of the highway and causing enough contamination. Alcoa quit buying Mexican aluminum scrap, because they couldn't make good-quality aluminum by using the recycled material from Mexico. Took them a while to figure that one out. They knew what the problem was, but they finally tracked down the source.
§28. Historical shares: Open Hearth, BOP, electric, and the rise of mini-mills [94:15]
Here is the history of total raw steel produced. This came out of The Making, Shaping and Treating of Steel, a US Steel publication. Now it's an American Iron and Steel Institute publication, but back when US Steel made money they had ten editions going back like 50 years on how to make steel. The last edition the American Iron and Steel Institute published, because the steel companies couldn't afford to do this kind of public-relations publishing anymore. Back in 1955, 90% of all the steel was basic Open Hearth. That was what Andrew Carnegie made his money on, 24 hours for 300 tons. The basic Open Hearth — the "basic" means it's a basic slag rather than an acidic slag, and it makes a cleaner steel — dropped from 1960 to 1980. In 20 years, the basic Open Hearth shut down. About 1970, when it was clearly a dead process, US Steel built the world's largest basic Open Hearth, a 450-ton furnace in Pittsburgh. Clever management. The rest of the world had seen the trend, but not US Steel's management.
The basic oxygen process, which uses a supersonic oxygen jet and makes everything in 20 minutes, took everything over in that same amount of time. The electric furnace was growing, because scrap became more and more available. We had been putting so much steel into the environment, we were getting more and more scrap back. A whole series of new mills called mini-mills replaced the integrated mills, starting in the 1970s, because some of these guys like Gordon Forward or Ken Iverson at Nucor realized you could buy scrap at $100 a ton, you could buy pig iron at $200 a ton. Both end up as the raw material for steel. Which one did you use? The $100-a-ton product or the $200-a-ton product? Duh. That was the growth of the mini-mills. From 1975 to today, about 50% of the steel made in the United States comes from electric furnace mills.
The conventional wisdom at the steel companies — and I was one of these employees back then — was that you couldn't make the same quality steel. All they could ever make was garbage steel. They might be able to make kids' little red wagons, they could make concrete reinforcing bar, but they couldn't make automotive sheet, they couldn't sell General Motors. And therefore they were never going to be important. That's why Clayton Christensen, in his book Innovator's Dilemma, defends the steel industry as these great paragons of managerial wisdom who were torn between integrated production and electric furnace production. He presents it as if it was a real question that someone with a brain would have to think about. But in fact it was obvious to anyone.
§29. Kim Clark, Armco, and ingot vs. continuous casting [98:08]
I remember another person I knew when I was a graduate student, a guy Kim Clark — became the dean of Harvard Business School. Kim and I, our wives were best friends. His son knocked my oldest son's teeth out twice in six weeks on the playground. I remember going to a meeting with Kim once — this was when I didn't have money to feed my family, and Kim had just gotten back from Disney World with his family — and he was down there lecturing to the American Iron and Steel Institute, the presidents of all the AISI companies. This was back around 1980. I said, well Kim, what are you telling them about? Kim had studied labor economics at Harvard. Kim helped me build a Pinewood Derby track for Scouts — it's the only track of the four that doesn't go straight. Kim doesn't quite know which end of a screwdriver to use. He's not mechanically inclined, he's not technically inclined.
And he was explaining to them about this dilemma: should they put in ingot casting versus continuous casting, which is something I haven't brought up yet. Ingot casting versus continuous casting was a process in 1970 that was absolutely obvious. You could go from 65% yield to 95% yield on your steel. Does it take a genius to know which one you would rather have? But for Armco Steel, it's a dilemma. For a businessman it's a dilemma. Armco Steel decided to put in ingot casting, and three years later, after they built it — takes about three years to build the plant and invested a couple hundred million dollars — they decided they were going to have to put in continuous casting. Anyone with a brain would have known before, but Kim said he was defending this dilemma. He was telling these managers how difficult this decision was. Well, of course you tell them that — they're paying your consulting fee, and you're not going to tell them they're idiots. My problem in my profession is I will tell them they're idiots. But in fact they were idiots. Kim became Dean, and here I am lecturing to you, but at least I have my integrity. I'd rather be lecturing, believe me, than sitting in those meetings.
§30. BOP process variations and the Sparrows Point pulpit [100:45]
This is the trend in the industry — it's all BOP and electric furnace, but you'll still see it in the specs, because the specs were written 80 years ago. There are lots of variations on the BOP process. The bath level is way down here. You can blow hydrocarbons and oxygen in the bottom. This comes out of the US Steel book, so they like to talk about the Q-BOP process — bottom blowing, rather than all top blowing. You can get a froth from the bottom rather than from the top. It's a little more efficient but it's a more complex process. The problem with the Q-BOP furnace, you have to blow at all times, even when you're not blowing oxygen. You have to have a flame coming in here so the steel doesn't leak out the bottom — you have to have enough pressure to keep it from leaking out.
Here are the different size vessels. From 1964, Allegheny put in an 18-ton vessel. McLouth put in a 60-ton vessel in '54. These were sort of experimental things back then. By '73, Bethlehem put in a 300-ton vessel, and these are all on the same scale. It doesn't get a lot bigger because volume goes as the cube of the dimension. The Making, Shaping and Treating of Steel gives you the times for making. Charging — putting the hot metal and the scrap in the furnace — 5 or 10 minutes. Refining, blowing, about 15 or 20 minutes. Sampling: you have to sit there and wait, because you're changing the chemistry so quickly. You have to take a sample and get a full chemical analysis — that will take 4 to 15 minutes. And it's worth a fortune. This facility is turning out steel at a value of $4 million an hour. A few minutes is worth a lot of money. If you got a faster way to do the chemical analysis, or if you could do it in situ — a guy over here in Building 12 was working for years trying to take the chemical analysis machine and put it right down that lance into the 3,000-degree furnace. Never worked, but he was trying to figure out how to do it. Tapping — pouring the steel — 48 minutes, pouring the slag off. Put all this together: about 45 minutes to make 300 tons of steel. Pretty productive.
And the guy doing it has, if he's lucky, a high school diploma. In 1975 I got thrown out of the BOP shop at the Sparrows Point steel plant, right next to the shipyard, when the guy found out I was from MIT and had a doctoral degree. He threw me out of the pulpit. They were casting a special heat of steel for an LNG carrier, and I had developed the steel and gone down there. When he found out that I actually had a degree that he didn't have, he threw me out of the pulpit. I couldn't be there while he melted the steel. And the steel came in off-chemistry. He was one of the highest-paid employees at Bethlehem Steel in the mid '70s. He was making probably $100,000 a year as an hourly employee, but he was responsible for $100,000 an hour worth of product, and getting it right. I still wonder if he got it wrong just to spite me. I hadn't said anything. I'll tell you what I did. There was no one there, it was a slow time after the oil embargo. We came in, there was one chair in the pulpit and there were three of us, and I went and sat in the chair. When he came back in I stood right up immediately and said hello to him. But he was very upset that I had been sitting in his chair.
And then I asked him — I was looking at his process sheet, the plan for how they're going to make the steel — I said, are you going to shoot for the high end of the range or the low end of the range? He said, you got one of those degrees like Sully? Sully was an MIT metallurgist, head of the Melt Shop Engineers. I had a hard hat — should have brought it with me, still have it in my garage at home — said "T.W. Eagar Research." White. They could tell the difference. That was white for management, so the crane operators could make a better hit if they were dropping something. He says, you like Sully, you go to MIT? And my boss, who had also gone to MIT, breaks out laughing. The next thing we know, the guy walks out of the shop, and in comes one of Sully's engineers and said, you guys better get out of here. So I got thrown out of the melt shop because I asked a question and I was sitting in his chair. You may have met some people like that in your day too.
I don't think I'll be back — I'm not here the rest of this week. I'm going to visit the Army and see how they blow up vehicles in Afghanistan.