TED日本語 - マイケル・ラバーグ: 同期したハンマーの一撃が核融合を成功に導く

Wow, this is bright. It must use a lot of power. Well, flying you all in here must have cost a bit of energy too. So the whole planet needs a lot of energy, and so far we've been running mostly on fossil fuel. We've been burning gas. It's been a good run. It got us to where we are, but we have to stop. We can't do that anymore.

So we are trying different types of energy now, alternative energy, but it proved quite difficult to find something that's as convenient and as cost-effective as oil, gas and coal. My personal favorite is nuclear energy. Now, it's very energy-dense, it produces solid, reliable power, and it doesn't make any CO2.

Now we know of two ways of making nuclear energy: fission and fusion. Now in fission, you take a big nucleus, you break it in part, in two, and it makes lots of energy, and this is how the nuclear reactor today works. It works pretty good. And then there's fusion. Now, I like fusion. Fusion's much better. So you take two small nuclei, you put it together, and you make helium, and that's very nice. It makes lots of energy. This is nature's way of producing energy. The sun and all the stars in the universe run on fusion. Now, a fusion plant would actually be quite cost-effective and it also would be quite safe. It only produces short term radioactive waste, and it can not melt down. Now, the fuel from fusion comes from the ocean. In the ocean, you can extract the fuel for about one thousandth of a cent per kilowatt-hour, so that's very, very cheap. And if the whole planet would run on fusion, we could extract the fuel from the ocean. It would run for billions and billions of years.

Now, if fusion is so great, why don't we have it? Where is it? Well, there's always a bit of a catch. Fusion is really, really hard to do. So the problem is, those two nuclei, they are both positively charged, so they don't want to fuse. They go like this. They go like that. So in order to make them fuse, you have to throw them at each other with great speed, and if they have enough speed, they will go against the repulsion, they will touch, and they will make energy. Now, the particle speed is a measure of the temperature. So the temperature required for fusion is 150 billion degrees C. This is rather warm, and this is why fusion is so hard to do.

Now, I caught my little fusion bug when I did my Ph.D. here at the University of British Columbia, and then I got a big job in a laser printer place making printing for the printing industry. I worked there for 10 years, and I got a little bit bored, and then I was 40, and I got a mid-life crisis, you know, the usual thing: Who am I? What should I do? What should I do? What can I do? And then I was looking at my good work, and what I was doing is I was cutting the forests around here in B.C. and burying you, all of you, in millions of tons of junk mail. Now, that was not very satisfactory. So some people buy a Porsche. Others get a mistress. But I've decided to get my bit to solve global warming and make fusion happen.

Now, so the first thing I did is I looked into the literature and I see, how does fusion work? So the physicists have been working on fusion for a while, and one of the ways they do it is with something called a tokamak. It's a big ring of magnetic coil, superconducting coil, and it makes a magnetic field in a ring like this, and the hot gas in the middle, which is called a plasma, is trapped. The particles go round and round and round the circle at the wall. Then they throw a huge amount of heat in there to try to cook that to fusion temperature. So this is the inside of one of those donuts, and on the right side you can see the fusion plasma in there.

Now, a second way of doing this is by using laser fusion. Now in laser fusion, you have a little ping pong ball, you put the fusion fuel in the center, and you zap that with a whole bunch of laser around it. The lasers are very strong, and it squashes the ping pong ball really, really quick. And if you squeeze something hard enough, it gets hotter, and if it gets really, really fast, and they do that in one billionth of a second, it makes enough energy and enough heat to make fusion. So this is the inside of one such machine. You see the laser beam and the pellet in the center.

Now, most people think that fusion is going nowhere. They always think that the physicists are in their lab and they're working hard, but nothing is happening. That's actually not quite true. This is a curve of the gain in fusion over the last 30 years or so, and you can see that we're making now about 10,000 times more fusion than we used to when we started. That's a pretty good gain. As a matter of fact, it's as fast as the fabled Moore's Law that defined the amount of transistors they can put on a chip. Now, this dot here is called JET, the Joint European Torus. It's a big tokamak donut in Europe, and this machine in 1997 produced 16 megawatts of fusion power with 17 megawatts of heat. Now, you say, that's not much use, but it's actually pretty close, considering we can get about 10,000 times more than we started. The second dot here is the NIF. It's the National Ignition Facility. It's a big laser machine in the U.S., and last month they announced with quite a bit of noise that they had managed to make more fusion energy from the fusion than the energy that they put in the center of the ping pong ball. Now, that's not quite good enough, because the laser to put that energy in was more energy than that, but it was pretty good.

Now this is ITER, pronounced in French: EE-tairh. So this is a big collaboration of different countries that are building a huge magnetic donut in the south of France, and this machine, when it's finished, will produce 500 megawatts of fusion power with only 50 megawatts to make it. So this one is the real one. It's going to work. That's the kind of machine that makes energy.

Now if you look at the graph, you will notice that those two dots are a little bit on the right of the curve. We kind of have fallen off the progress. Actually, the science to make those machines was really in time to produce fusion during that curve. However, there has been a bit of politics going on, and the will to do it was not there, so it drifted to the right. ITER, for example, could have been built in 2000 or 2005, but because it's a big international collaboration, the politics got in and it delayed it a bit. For example, it took them about three years to decide where to put it.

Now, fusion is often criticized for being a little too expensive. Yes, it did cost a billion dollars or two billion dollars a year to make this progress. But you have to compare that to the cost of making Moore's Law. That cost way more than that. The result of Moore's Law is this cell phone here in my pocket. This cell phone, and the Internet behind it, cost about one trillion dollars, just so I can take a selfie and put it on Facebook. Then when my dad sees that, he'll be very proud. We also spend about 650 billion dollars a year in subsidies for oil and gas and renewable energy. Now, we spend one half of a percent of that on fusion. So me, personally, I don't think it's too expensive. I think it's actually been shortchanged, considering it can solve all our energy problems cleanly for the next couple of billions of years.

Now I can say that, but I'm a little bit biased, because I started a fusion company and I don't even have a Facebook account. So when I started this fusion company in 2002, I knew I couldn't fight with the big lads. They had much more resources than me. So I decided I would need to find a solution that is cheaper and faster.

Now magnetic and laser fusion are pretty good machines. They are awesome pieces of technology, wonderful machines, and they have shown that fusion can be done. However, as a power plant, I don't think they're very good. They're way too big, way too complicated, way too expensive, and also, they don't deal very much with the fusion energy. When you make fusion, the energy comes out as neutrons, fast neutrons comes out of the plasma. Those neutrons hit the wall of the machine. It damages it. And also, you have to catch the heat from those neutrons and run some steam to spin a turbine somewhere, and on those machines, it was all a bit of an afterthought. So I decided that surely there is a better way of doing that.

So back to the literature, and I read about the fusion everywhere. One way in particular attracted my attention, and it's called magnetized target fusion, or MTF for short. Now, in MTF, what you want to do is you take a big vat and you fill that with liquid metal, and you spin the liquid metal to open a vortex in the center, a bit like your sink. When you pull the plug on a sink, it makes a vortex. And then you have some pistons driven by pressure that goes on the outside, and this compresses the liquid metal around the plasma, and it compresses it, it gets hotter, like a laser, and then it makes fusion. So it's a bit of a mix between a magnetized fusion and the laser fusion. So those have a couple of very good advantages. The liquid metal absorbs all the neutrons and no neutrons hit the wall, and therefore there's no damage to the machine. The liquid metal gets hot, so you can pump that in a heat exchanger, make some steam, spin a turbine. So that's a very convenient way of doing this part of the process. And finally, all the energy to make the fusion happen comes from steam-powered pistons, which is way cheaper than lasers or superconducting coils.

Now, this was all very good except for the problem that it didn't quite work. (Laughter) There's always a catch. So when you compress that, the plasma cools down faster than the compression speed, so you're trying to compress it, but the plasma cooled down and cooled down and cooled down and then it did absolutely nothing.

So when I saw that, I said, well, this is such a shame, because it's a very, very good idea. So hopefully I can improve on that. So I thought about it for a minute, and I said, okay, how can we make that work better? So then I thought about impact. What about if we use a big hammer and we swing it and we hit the nail like this, in the place of putting the hammer on the nail and pushing and try to put it in? That won't work. So what the idea is is to use the idea of an impact. So we accelerate the pistons with steam, that takes a little bit of time, but then, bang! you hit the piston, and, baff!, all the energy is done instantly, down instantly to the liquid, and that compresses the plasma much faster. So I decided, okay, this is good, let's make that.

So we built this machine in this garage here. We made a small machine that we managed to squeeze a little bit of neutrons out of that, and those are my marketing neutrons, and with those marketing neutrons, then I raised about 50 million dollars, and I hired 65 people. That's my team here. And this is what we want to build. So it's going to be a big machine, about three meters in diameter, liquid lead spinning around, big vortex in the center, put the plasma on the top and on the bottom, piston hits on the side, bang!, it compresses it, and it will make some energy, and the neutron will come out in the liquid metal, going to go in a steam engine and make the turbine, and some of the steam will go back to fire the piston. We're going to run that about one time per second, and it will produce 100 megawatts of electricity.

Okay, we also built this injector, so this injector makes the plasma to start with. It makes the plasma at about a lukewarm temperature of three million degrees C. Unfortunately, it doesn't last quite long enough, so we need to extend the life of the plasma a little bit, but last month it got a lot better, so I think we have the plasma compressing now. Then we built a small sphere, about this big,14 pistons around it, and this will compress the liquid. However, plasma is difficult to compress. When you compress it, it tends to go a little bit crooked like that, so you need the timing of the piston to be very good, and for that we use several control systems, which was not possible in 1970, but we now can do that with nice, new electronics.

So finally, most people think that fusion is in the future and will never happen, but as a matter of fact, fusion is getting very close. We are almost there. The big labs have shown that fusion is doable, and now there are small companies that are thinking about that, and they say, it's not that it can not be done, but it's how to make it cost-effectively. General Fusion is one of those small companies, and hopefully, very soon, somebody, someone, will crack that nut, and perhaps it will be General Fusion.

Thank you very much.

(Applause)