That little blurb above? It’s not quite true. Yes, the reaction chamber itself produced more energy than was put into it, but the lasers pumping in that power were only about 1% efficient. Overall, they’re a long way from turning on the tap and pouring out endless energy. But, let’s continue…
Now that we’re past the difference between fission and fusion, there’s one more hurdle to understand: Confinement.
Getting two atoms—specifically the nuclei at the core of those atoms, to stick together—means overcoming a huge amount of electromagnetic force trying to push those nuclei apart. Our friendly neighborhood nuclear furnace, the Sun, achieves this by simply being huge. Gravity in the core of the Sun becomes great enough to push hydrogen together, and bam! Tan lines. Actually, solar power is just fusion power with the reactor kept 93 million miles away. Most wind power, too. Even fossil fuel is just fusion energy that was stored long ago through the photosynthesis of plants and microorganisms. It’s all fusion—well, fusion and turtles—all the way down.
However, those of us who don’t live directly on the Sun lack a massive gravity well close at hand. Somehow we have to come up with other ways to keep atomic nuclei close together so they can be fused. That’s what confinement is all about. There are basically two approaches—magnetic confinement, in which those nuclei are held inside incredibly (incredibly) intense magnetic fields, and inertial confinement in which those nuclei get thrown together quickly.
Truthfully, creating a controlled fusion reaction here on Earth isn’t all that difficult. With fair regularity, your local paper, or 6 PM news, is likely to report the story of a genius child who whipped up a fusion reactor in the family garage. You can watch popular YouTube personality and former Vox writer, Cleo Abram, put one together along with the Queen of S#itty Robots, Simone Giertz, in this video:
What they’re building is a Farnsworth Fusor, designed by Philo Farnsworth, who was also instrumental in designing the first practical television sets and video cameras. This system, from that damned clever Prof. Farnsworth (not the one with a rocket delivery company), uses a kind of inertial confinement known as “electrostatic inertial confinement.” It builds up a huge charged field, hustling the atoms together at the center, where some portion of them can be made to fuse. Dead simple, and can be constructed from materials that were on hand to a basement inventor in the 1960s.
Sandia National Labs did a lot of work with electrostatic confinement over the next couple of decades, but there have always been very good reasons to believe that, no matter how powerful you make that fusor, or how efficiently you build this system, it’s not possible to extract from it more power that it takes to run. It can never be used as a power plant.
Math. Taking the fun out of things since Imhotep.
However, there is another system of inertial confinement that has long shown promise—laser confinement. In it, microscopic spheres filled with frozen hydrogen (technically, deuterium and tritium, which are more easily fused together isotopes of hydrogen) are smacked by lasers that are blasting at them from many directions at once, like, 100+ directions at once, with every laser designed to focus in exactly the same spot.
The idea goes back to 1960, just months after the laser was first demonstrated. In 1974 a now-extinct private company, KMS Fusion, demonstrated the first successful inertial confinement fusion using lasers, but their design proved difficult to scale, and figuring out how to turn a momentary blast of fusion from a single bead of fuel into a continuous stream of power proved too great a challenge. It didn’t help that a high-power laser at that time required a huge amount of space and infrastructure. How big? Each was roughly the size of a football field. That’s also, BTW, the size of the lasers in the current big announcement. You’re never going to get that Mr. Fusion on the back of your Delorean.
Throughout the 1970s and 1980s, the race for fusion power was a sort of alternative, less publicized space race. The Soviet Union built more and more powerful systems based on a form of magnetic confinement designed by a pair of Soviet scientists, the tokamak. The US and Europe poured billions into both magnetic and laser confinement. Everyone made incremental progress.
Only the increments were tiny. With each step, it became clear that the current generation of machines were never going to reach the magic point where power out exceeded power in. Eventually, it became clear that not even the machines still on the drawing boards could hit that mark. Achieving fusion was easy. Achieving practical fusion power turned out to be a problem on par with herding cats using self-driving cars.
In 1990, the US decided to build the National Ignition Facility (NIF) at Lawrence Livermore National Labs in California. It would use a “beamlet” technology to greatly increase the number of lasers hitting the target fuel. Construction actually began in 1997, and it wasn’t until 2009 that it first conducted experiments. By then, it had run up bills of roughly $4 billion—about four times the original estimate. However, the machine got several upgrades along the way (increasing the base number of lasers from 10 to 18, each of which travels just under a mile before being split into 288 “beamlets”).
For the most part, both Congress and researchers were happy, because NIF did what it was designed to do, which was not to produce a controlled, stable reaction like the kind needed for a fusion power plant.
Nope, the NIF was designed for a wholly different purpose. It was built primarily for weapons research. It went on to do dozens of experiments supporting this research over the next decade. And in the middle of this run, back in 2013, the National Ignition Facility announced that it had exceeded that magic breakeven point. Queue press release.
However, that announcement was a big exaggeration. After scientists had a chance to cut through a ton of weasel-wording, it became clear that the actual output was about 0.08% of the input. (Yeah, but the fuel actually absorbed … blah blah blah … not relevant, and not exactly the kind of system that’s going to power the nation.)
After a number of years in which the lasers were used to blast holy hell out of a variety of materials, the NIF got back to planning some fusion experiments in 2016. And this time, the goal of actually reaching true breakeven was on the table.
Last year, the NIF was the first facility of any kind to produce what’s known as a “burning plasma”—one in which more of the heat comes from the fusion of the atoms than it does from the power being input from outside the system. That was a pretty good sign of what was coming next.
Sure enough, on Dec. 5th the researchers put 2.05 megajoules of laser energy into the system, and got back 3.15 megajoules of energy. In the form of an explosion that caused some damage to the equipment, enough so that there’s still a little doubt about those numbers. Somewhere in less than a millisecond of power, a Tic Tac-sized fuel pellet kicked out about the same energy as three sticks of dynamite exploding.
Of course, powering the lasers actually took almost 300 megajoules, but that’s something that we’re all agreeing to ignore for the moment.
So it works. But this doesn’t mean that everyone in the Bay area is going to be dragging an extension cord to Livermore any time soon. From the start, the NIF was never designed to generate power. It can do a momentary blast of power, not the sustained reaction with all the accompanying equipment to capture the output (as heat) and channel it into electricity. That’s the bad news. Over eighty years since the first stab at making a fusion reactor, this is still basic proof-of-concept research.
However, it’s also the good news. The NIF, which was designed thirty years ago, was able to do something that really wasn’t in its script to reach this point. Exceeding breakeven with an inertial confinement fusion reactor is a doable thing. Now we just need to do it again, and better, with a constant stream of fuel pellets being fed into an incredibly well-coordinated system of lasers and a surrounding jacket ready to capture the resulting energy and charge that new electric Delorean. Right now, in the best of conditions, the NIF can do about one such explosion a day. That’s 863,999 short of the ten per second needed for an actual power plant. Minimum.
If that sounds like a difficult improvement, watch this video about how the latest generation of computer chips is made. Because the people who have solved this have already solved a large part of the problems ahead.
In a lot of ways, laser inertial confinement is the simplest form of fusion reactor that might make a viable commercial power plant. That’s because, despite how complicated this all sounds, there are some real advantages—such as the thing doing the heating and the thing doing the confining being the same, and the lack of need for materials an order of magnitude beyond what we can produce today . It just needs incredible precision—like in that chip video. The lasers building a new iPhone chip are actually more precise in their targeting than the ones at the NIF. It’s not 1990 anymore (sorry about that, Smashing Pumpkins fans).
That the NIF made this work, doesn’t mean that there will be facilities based on this tomorrow, or even ten years from now. But it should certainly sound the starting gun for would-be fusion folks looking for that startup cash. The real value of this breakthrough may be found in how investors in uranium stocks are out there begging people to ignore it. Because investors are finally starting to see fusion the way that guy in the meme sees the woman in the red dress
For a long, long time, fusion has seemed to be right around the corner. Now we can see the corner.
In addition to the NIF breakthrough, there is the massive ITER project in which 35 countries are collaborating on the world’s largest and most refined version of that original Soviet tokamak reactor. That system can, theoretically, reach a value of 10x more power out than power in, though it’s not designed to be a practical power plant, and there will certainly be a large number of losses involved that whittle away at that possible output. The origin of this experiment goes all the way back to the 1980s, but it should be cranked up for the first time in about two years.
There are also newcomers like Novatron, which just secured funding for their experimental fusion reactor that uses a new magnet confinement strategy that doesn’t require expensive and tricky to maintain superconducting magnets.
Then there is Washington-based Helion, which uses a device that seems like a merging of Farnsworth’s Fusor with magnetic confinement to produce what looks like a breathtakingly simple approach. Helion is projecting that they will produce electricity from their 100-million-degree device next year. Like… next year.
Novtron and Helion are just two of at least 27 companies currently working toward commercial fusion alongside vast research projects like NIF and ITER. They all seem to be counting on improvements in electronics, materials science, and precision machining to carry them over hurdles that stopped past scientists. Some of these designs, like those of Helion, are actually extensions of proposals from the 1950s that ran into barriers more about the engineering of the time than that darn underlying math. Any of these operations could turn out to be successful in the next few years.
So, yes, all the articles that followed up the NIF announcement with a firm “this doesn’t mean fusion power is coming soon” are right when it comes to the NIF, or ITER, or any of the other big experiments in the works . They also have good reasons to be skeptical of all the companies who are waving their breakthrough designs in front of investors. 84 years of failed approaches and that constant refrain of “20 to 30 years away” has made predicting No Fusion Soon the winning ticket for a long, long time.
But I’m betting against the pack on this one. Someone is going to hit the breakeven with one of these new devices, and they’re going to do it in a sustained way that produces electricity. Soon.