A few days ago, Lockheed Martin announced progress on a design for a nuclear power-plant that can be built at small scale, maybe tractor-trailer-size, and will produce almost no radioactive waste. Lockheed’s team is saying it expects to have a workable reactor within five years. If that were to happen, the world would literally change: The fossil-fuel business would immediately be halfway to obsolescence because electrical power would soon become (as they used to say in the ‘50s) “too cheap to meter.” Coal would become an antique fuel, used in the odd pizza oven; super-cheap electricity would lead us to retire most gasoline and diesel engines. Carbon pollution, not to mention the politics of oil-producing countries, would become pretty nearly trivial problems.
How can that work? What is fusion power?
Broadly, it is the alternative to fission, the process that today’s nuclear power-plants use. Fission plants amass very heavy atoms (usually of uranium isotopes) and cause them to split, emitting high-energy particles, and thus, a lot of energy. That energy is used to boil water, and the steam is used to turn a generator and make electricity.
Fusion power-plants would work on a different principle: Very light atoms (usually hydrogen isotopes called deuterium and tritium) would be made to collide, combining to form slightly heavier atoms (specifically, helium) and spinning off neutrons. That reaction, too, releases a great deal of energy, and in fact, a similar fusion process powers the sun.
So why would that be any better than fission?
Because a fission reactor’s byproducts are extremely radioactive, many to a degree that will be dangerous for centuries. The products of a fusion reaction are much more benign — a little bit of ordinary helium, plus a lot of neutrons. Something has to soak up the neutrons and then be disposed of, but it’s not going to be anywhere near as toxic as the stuff (like plutonium) that comes out of a fission reaction. Also, a fusion reaction cannot run out of control: If something goes wrong, the plasma quickly cools and the whole thing just stops. There cannot be a meltdown in a fusion plant.
Hasn’t this been on the horizon for years?
Sure. There have been programs at Princeton, Los Alamos, and the University of California for decades, and the Soviet Union and the U.K. had similar projects going. Over the past 40 years, the Princeton Plasma Physics Laboratory has built several large test machines, called tokamaks, after a Russian prototype, that have inched closer to the goal of a working power-plant. (My father spent his career working on those projects at PPPL, retiring in 1999.) Most current research hopes are pinned on a test machine called the International Thermonuclear Experimental Reactor, or ITER, under construction in France. It’ll be switched on around 2020, by which time it will have cost at least $50 billion. A lot of people in the field are concerned about that number. As Dr. Bruno Coppi, a plasma physicist from MIT, noted to me this week, a competing smaller and cheaper approach — whether Lockheed’s or anyone else’s — is almost surely a good thing for the field.
Wasn’t there some big fusion breakthrough back in the ’80s?
You may be thinking of “cold fusion,” a notorious (and very different) process. In 1989, two scientists with an experiment at the University of Utah, Martin Fleischmann and Stanley Pons, announced that they’d come up with a way to produce fusion power at tabletop scale with simple materials. They made the key mistake of calling a press conference before submitting their project for peer review and getting a huge amount of attention upfront. Within days, other scientists pointed out that they’d made fundamental errors, and their work was discredited and their careers effectively ended. Further research on cold fusion — just to double-check that Fleischmann and Pons hadn’t happened upon something after all — has not been fruitful.
Will any of this actually come to pass?
There’s a wry saying among energy researchers: “Fusion power is ten years away … and always will be.” That’s because it faces severe practical hurdles. To get a fusion reaction started, you need to heat the fuel to about 10 million degrees, producing the superheated gas known as plasma. (The operating temperature is many times hotter than that.) At those temperatures, it can be contained only with extremely large and powerful electromagnets, and it tends to worm its way out of confinement and dissipate. So far, tokamaks have been able to produce enormous and very brief single pulses of energy but cannot produce power over time as an electrical-generation plant would. So far, they consume more electricity than they make.
There is also a secondary problem: A fusion reaction produces high-energy neutrons and alpha particles, and the neutrons in particular bore into the metals used in the machinery itself, making them brittle and eventually causing them to crack.
So what did Lockheed figure out?
That is somewhat unclear because the team left a lot of detail out of the press release and has produced no peer-reviewed papers. However, a look at the site reveals some clues, and they suggest that Lockheed is producing a variant of an old design, one that precedes the tokamak era. In the 1950s, several smallish experimental devices called mirror machines were built. The basic design was fairly simple: a big copper coil that created a magnetic field, with a stronger electromagnet at each end. A plasma held in the center of the coil was further confined by the two magnets, which were only somewhat effective. In this new variation, the coil is made not of copper, but of a superconductor — one of the high-tech materials that, kept extremely cold, exhibit zero resistance to electricity. It can thus make a lot more magnetic field while using less juice, increasing the likelihood that you’re building a power plant rather than a power-consumer.
Sounds promising. What’s wrong with that?
Well, two things. The point of those old designs was for the plasma to heat up the casing around the whole coil, and that heat would in turn drive the power plant. You can’t do that, though, if you are simultaneously cooling the whole shell to keep the superconducting magnet working. On top of that, the neutrons streaming out of a fusion reaction also destroy superconductivity, and would thus bring this machine to a halt. These sound like pretty basic problems.
Are they insurmountable?
The Lockheed team is implying, cagily, that it’s got clever solutions to those and various other secondary complexities. The most detailed interview so far suggests that the team is at least aware of those and other challenges that it faces, and further implies some progress toward solving them. For example, shielding the superconducting coil can be managed — but the researchers suggest that the shield would be about ten feet thick, making a reactor that would weigh a couple of hundred tons. You couldn’t put that on a truck, as the initial announcement implies. There’s a lot of promise here, definitely. If I were betting, though, I wouldn’t sell the ExxonMobil stock just yet.