clock menu more-arrow no yes mobile

Filed under:

Nuclear fusion could be the perfect energy source — so why can't we make it work?

Our Sun does fusion power just fine. So why can't we?
Our Sun does fusion power just fine. So why can't we?
NASA/European Space Agency

In theory, it's possible to shoot some energy at hydrogen and get even more energy back. The process is called thermonuclear fusion, and if we could ever get fusion power to work — a big if — we'd never have to worry about our energy problems again.

It's not a completely crazy notion. Nuclear fusion already takes place in the sun's core, after all. And the promise of fusion power has led researchers to try their best for decades upon decades. Occasionally, they even make some advances — as happened this past winter, when a group of scientists got closer to fusion power than they ever had before.

Trouble is, the scientific and technical hurdles ahead are still enormous — in fact, we still don't have a full grasp on what all the hurdles might be. Still, the potential pay-off is so massive that countries have sunk billions and billions of dollars into fusion research.

So here's a guide to how far humanity has come on thermonuclear fusion — and how far we still have to go.

What is thermonuclear fusion?

Thermonuclear fusion is the process that occurs when two atoms combine to make a larger atom, creating a whole lot of energy.

Fusion already happens naturally in stars — including the sun — when intense pressure and heat fuse hydrogen atoms together, generating helium and energy. This process is what powers the sun and makes it so hot and bright. Researchers who work on fusion energy are essentially trying to make tiny stars here on Earth.

Isn't fusion a violation of physics?

No. When two atoms fuse, they lose a bit of their mass, which is released as energy. This is perfectly acceptable according to Einstein's famous E = mc2 equation, which says that mass can turn into pure energy and vice versa. (The E here stands for energy. M is for mass. C is a constant number that is the speed of light in a vacuum.)


An illustration of the ITER machine, which, if all goes well, will be doing fusion by 2027. ITER Organization

Don't our nuclear power plants already do fusion?

No. Nuclear reactors perform fission, which involves splitting atoms apart. Fusion, by contrast, is when atoms merge together. Fusion converts more mass into energy per reaction than fission does.

How can the sun do fusion so easily?

The sun weighs about 333,000 times more than Earth does. That mass creates powerful gravitational forces that produce extreme pressures. This pressure, combined with temperatures up to 27 million degrees Fahrenheit, gets atoms to fuse together.

So how do we do fusion on Earth?

We don't have the technology to recreate the Sun's massive pressures, so researchers have to make up for that by getting hydrogen atoms even hotter than the sun does — in the range of hundreds of millions of degrees Fahrenheit. They heat up the atoms using various tools, including particle beams, electromagnetic fields such as microwaves and radio waves, and lasers.

The temperatures needed are so hot that the hydrogen fuel becomes a plasma, a state of matter that exists when a gas's atoms split into positively and negatively charged particles. (Stars and lightning are plasma, as is the luminous matter inside neon signs.)

Researchers have been producing controlled fusion reactions for decades. These days, the big goal that hasn't happened yet is to make a fusion reactor that produces more energy than it takes in.


Plasma, like lightning, is very difficult to control. Alexander Joe/AFP/Getty Images

Is this the same as cold fusion?

No. Cold fusion is the theoretical fusion of atoms at room temperature. No one has ever done cold fusion — although there have been many false claims over the years. Scientists researching fusion energy are more interested in hot fusion, which they have been doing the 1930s — the challenge now is just how to turn it into useful energy.

How are we trying to do fusion now?

There are many approaches. Here are the two most worth watching.

1) Magnetic Confinement: The basic principle of magnetic confinement is to hold plasma fuel in place with magnets and then heat it up using a combination of microwaves, radio waves, and particles beams. Researchers often do this in a tokamak, a donut-shaped reactor (the weird shape helps keep the plasma in place).

In the 1990s, the European tokamak JET achieved 16 million watts of fusion power for less than a second. On the whole, JET was able to produce 65 percent of the energy that went into the experiment.

More recently, an international group is building the world's largest fusion reactor. This is an even bigger tokamak called ITER. The goal of ITER is to produce 500 million watts of power — in the range of a real power plant — for seconds at a time. The researchers also want to produce ten times more energy than is used by the system.

But ITER is already having problems: the project is falling ever behind schedule while increasing its estimated cost (which has tripled to about $22 billion).


See how tiny that person in blue is compared to this giant fusion reactor? ITER Organization

2) Inertial Confinement: This approach is used by the National Ignition Facility in Livermore, California — and it involves 192 lasers.

The NIF fires the lasers at a tiny gold can, which vaporizes and gives off x-rays. Those x-rays then hit a spherical pellet of hydrogen fuel that's smaller than a peppercorn. The x-rays heat and compress the fuel, which turns into plasma. Then a minuscule portion of that plasma fuses into helium, giving off energy and neutrons for a split second.

In February, 2014, researchers at NIF reported that the fuel pellet made more energy than it absorbed for the first time. The method isn't yet useful for any practical real-world power needs: the experiment's lasers used about 100 times more energy than the fuel pellet produced. Still, it was promising: the results were in line with NIF's computer predictions, a sign that physicists' understanding of plasma is improving.


192 lasers hit this gold can, which has hydrogen fuel inside. NIF

Which approach to fusion has the best chance of success?

If people had to pick one, most would put their money on ITER. That's because NIF only researches fusion power as a side project — its main task is performing studies that help maintain and test the US nuclear arsenal.

However, there's also a good chance that no one will succeed in producing practical fusion power. What scientists are currently doing are research projects that won't be hooked up to the power grid. And getting a machine to do fusion for a split second every once in a while is nothing compared with building an actual power plant that can withstand the trauma of doing fusion all the time.

It's a major engineering challenge, and some say that making a commercial power plant will be even harder than getting fusion to work in the first place.

Why is fusion power so difficult?

One big reason is that it requires working with plasma, which is really tricky. Because plasmas aren't that common on Earth, scientists had very little experience with them until they started studying fusion.

Plasma is difficult to hold: The plasma used in fusion-energy research is hundreds of millions of degrees Fahrenheit. You can't hold it using a solid container, because the container would just melt. Instead, physicists have to corral it using electromagnetic fields or work with it so quickly (in less than a billionth of a second) that holding it isn't an issue.

Plasma is difficult to compress: If you don't compress plasma from all sides perfectly evenly, it will squish out wherever it can. Scientific American explained this well: "Imagine holding a large, squishy balloon. Now squeeze it down to as small as it will go. No matter how evenly you apply pressure, the balloon will always squirt out through a space between your fingers. The same problem applies to plasmas. Anytime scientists tried to clench them down into a tight enough ball to induce fusion, the plasma would find a way to squirt out the sides."


The chamber where fusion takes place at the National Ignition Facility. NIF

Will we ever get fusion power?

The folks associated with ITER say that they'll have plasma in the reactor in 2020 and will be doing fusion by 2027. But the project has been dogged by delays, not to mention a very negative review of its management recently. So take those dates with a giant grain of salt.

More broadly, fusion power research has a very long history of always promising that success is just 20 years away. It also has had its share of wackos, hucksters, and well-meaning, but blindly optimistic scientists. For a good, pessimistic argument of why fusion power will never happen, check out journalist Charles Seife's Slate piece from a few years back.

Sounds dangerous. Is it going to kill us?

No. One of the reasons that people are so jazzed about fusion energy is that it should be pretty darn safe — a lot safer than our current nuclear power plants and absolutely safer than a bomb.

What about nuclear meltdowns like Fukushima?

Not an issue here. First off, plasma needs a very carefully controlled environment in order for fusion to happen. So if something goes wrong with the reactor, the fusion reaction will simply stop. That's why there's no danger of a runaway reaction like a nuclear meltdown.

And unlike fission, fusion power doesn't use require fuel like uranium that produces long-lived, highly radioactive waste. What goes into the fusion system is just hydrogen and sometimes lithium, and what comes out is helium (the stuff that's in party balloons) and some neutrons.


Oli Scarff/Getty Images

So fusion power is completely risk-free?

Nothing's completely risk-free, and fusion is no exception. Here are some of the risks.

Neutron-induced radioactivity: Fusion reactions produce high-energy neutrons, which are not themselves radioactive. However, they strike the walls of the reactor with so much energy that the walls can become radioactive. (However, this radioactivity doesn't last nearly as long as the kind at current nuclear plants.)

Tritium fuel: Tritium is a type of hydrogen that's currently used in many fusion experiments. And it's weakly radioactive. But that's probably not a big problem. The EPA says this: "because [tritium] emits very low energy radiation and leaves the body relatively quickly, for a given amount of activity ingested, tritium is one of the least dangerous radionuclides." And the tritium is used in such small quantities that the risk of environmental contamination is exceptionally low.

Would we run out of fuel?

Not for a thousand generations or so.

Physicists like to use the deuterium and tritium forms of hydrogen, which are easier to fuse than the standard kind.

Deuterium naturally occurs in water in high enough concentrations that there's plenty. And we'd need so little (a few gallons of water could provide the same power as a super tanker's worth of oil) that depleting our water resources isn't really an issue.

Tritium needs to be made by humans. It can be produced by fission reactors or by adding some lithium into a fusion reactor. Although lithium isn't super abundant on land, there's enough in the seas to theoretically support 30,000 years of fusion power.


The fuel for fusion power comes from water. ASP/Getty Images

If the sun's already doing fusion, why not just use solar panels?

Many people say that solar (or any other kind of power) is a better option than fusion. Whether or not fusion research is worth the time and money is quite controversial.

But one of the major hurdles with solar power is that the sun only shines sometimes. If we were to go completely solar, it would require large-scale battery technology that we don't yet have. Fusion power, like today's fossil-fuel and nuclear-power plants, could provide energy 24/7 in any location.

Can I do fusion myself?

Probably, although it involves dangerously high voltage. A few years ago, a 14-year-old built a fusion reactor in his basement. Today, you can get accurate instructions on how to build one for $1,000. Have fun.