So the Artificial Sun Boils a Kettle — How a Hundred Million Degrees Becomes Electricity

There is really only one way to make electricity

Think of power stations. Coal, gas, oil, nuclear, heat from underground. What they burn differs every time. What happens next is identical: boil water into steam and let the steam turn a very large windmill.

That windmill is a turbine. At the end of its shaft sits a generator where a magnet and a coil pass each other and electricity appears. A bicycle dynamo does exactly this, only smaller.

Fusion is no exception. Make a plasma at a hundred million degrees and what you do with it is boil water. The gap between the nickname and the kettle feels deflating — and the whole of the engineering lives inside that gap.

Boiling water in five steps

The previous piece ended with helium and a neutron, and only the neutron leaving. It strikes the wall of the machine. It damages that wall, and it also makes it hot. That heat is the heat we are after.

1. Pipes run densely through the wall with water inside them. The water takes up the heat of a wall being struck by neutrons.
2. But this water does not boil. It is held at around a hundred and fifty atmospheres, and under that pressure water stays liquid well past three hundred degrees.
3. The hot water goes to a large vessel called a steam generator. Thousands of thin tubes pass through it, and around those tubes sits entirely different water.
4. The hot water inside the tubes heats the water outside them. That outer water is not pressurised, so it simply boils, and steam rises.
5. The steam goes to the turbine: ranks of hard metal blades along a shaft, struck in passing, turning it. At the end of the shaft is the generator.

The wall has a second job. Put lithium in it, and the arriving neutron strikes the lithium and turns it into tritium. Tritium barely exists in nature and has to be made somewhere, and here it is made on the spot. One neutron from one joining boils the water and makes the next fuel.

Why two separate waters?

It sounds roundabout. Why not send the water heated at the wall straight to the turbine, instead of using it to heat a second lot of water?

Because the inner water has been circling through a place where neutrons pour down. It cannot be let out towards the building where the generator stands. So the two waters never touch. They hand over heat through a tube wall and the inner water goes back to the wall again, sealed, going round for ever.

A nuclear station is built exactly this way — the machine of the seventh piece. Building a fusion plant means a new front half bolted to a back half that has been running for decades.

And Carnot's limit applies here unchanged. No machine, however well made, turns heat entirely into work; some must be thrown away. The cooling tower beside a station, or the sea running past it, is where it goes. A fusion plant is no different: more than half the heat collected will never become electricity.

The order of it is hard to hold on the page. Follow it yourself in the board below — from the neutron striking the wall, through water and steam and turbine, to the socket.

Tool A Vessel for a Hundred Million Degrees — Why Fusion Is So Hard Six animated steps from fuel to electricity, a tank of nuclei whose temperature and density you raise yourself, and the Sun and KSTAR running side by side. Why heat alone is not enough, and why we must run hotter than the Sun. (Korean only)

A little further in

Electricity in, electricity out

A hundred million degrees does not arrive by itself. There are three ways to heat it.

1. Drive a current through the plasma itself; resistance heats it. Like an electric fire — except that resistance falls as it gets hotter, so past a point it stops working well.
2. Fire radio waves into it, much as a microwave oven heats food.
3. Shoot in a beam of very fast atoms from outside — throwing hot things in to warm the whole.

All three eat electricity. Which is why fusion talk always comes with a multiple attached: how many times what went in came back out.

The international machine being built in southern France by many countries together aims at ten. Fifty million watts in, five hundred million out.

There is a reason it is so much bigger than KSTAR. Confinement improves with size, and not just in proportion: with the square of it. KSTAR is about ten metres across, the French machine about thirty. Three times the size, nine times the grip. It also explains why the palm-sized reactor of the films is so hard. Making it small is the last thing that will be solved.

The American laser result shows the difference well. Two megajoules were delivered to the target and three came back. That was the first time fuel gave out more than it received, and it was a real result.

But firing that laser drew something like three hundred megajoules from the mains. As power generation it is short by about a hundredfold. That does not make it meaningless: a threshold was crossed for the first time, and the remaining hundredfold is a question for engineering rather than physics.

Ignition — where it starts burning by itself

There is one hope here. The previous piece said helium carries charge, so the magnets keep it inside. That helium goes on heating the plasma — a stove burning within the fire.

The more fusion happens, the larger that stove; the larger the stove, the less heating is needed from outside. When the inner stove alone keeps the fire going, that is ignition.

It is like lighting a bonfire. At first you keep feeding paper and twigs. Then the logs themselves catch and you can stop. The moment the match can be taken away is ignition.

No magnetic machine has reached it. The laser experiment above was reported as first across that threshold, and the two ways of holding the fuel are different enough to make a direct comparison awkward.

Is nothing left over?

Clean energy, people say. That is half right.

Take the right half first. There is no high-level waste of the spent-fuel kind that a fission station produces. And as the previous piece said, any slip in conditions puts the reaction out, so there is no runaway to build against.

But it is not zero. Walls and structures struck by neutrons for years become radioactive themselves, and dismantling a machine at the end of its life yields low and intermediate level waste. Tritium, one of the fuels, is a radioactive gas, and keeping it in and recovering it is no small task.

How is it different from a hydrogen bomb?

It is the same reaction. Joining hydrogen, in both cases. The manner is the exact opposite.

What the previous piece described as a weakness — that any slip puts it out — becomes the safety feature here. Being hard to light and being easy to extinguish are the same property.

What dangerous reactions have in common is that they happen readily at room temperature. Uranium splits at room temperature the moment a neutron strikes it, which is why, once too much of it starts, it is hard to stop. Fusion is the reverse: it barely happens below a hundred million degrees, so the moment anything goes wrong and the temperature drops, it simply stops. Break the machine and outside air rushes in and puts the fire out. It is like building a box underwater and lighting a fire inside: crack the box and the water ends it.

And the helium-3 on the Moon?

A question that keeps coming. The facts first: helium-3 really has accumulated in the lunar soil, carried there by the wind from the Sun over billions of years. Earth has almost none, because Earth's magnetic field turns that wind aside.

But neither KSTAR nor the international machine in France uses helium-3. They use deuterium and tritium. So why does the Moon keep coming up? Because what comes out is different.

Skipping the kettle. That is the real reason people look at the Moon: it goes around Carnot's limit — around having to throw half the heat away — rather than through it.

The sums are hard, though. The helium-3 route needs a far higher temperature, several times the hundred million of the present route. The concentration in lunar soil is punishingly low, so an enormous quantity of ground would have to be dug and heated. And side reactions mean neutrons do not vanish entirely. It is honest to file it as a candidate for a long time hence.

For what it is worth, what anyone is actually after at the lunar south pole is ice rather than helium-3. Water. Water settles drinking, breathing and fuel all at once.

What you would study to do this

The people in the first piece light the fire. The people here take electricity out of it. They are different seats.

1. Someone who designs how heat is taken off the wall — mechanical engineering and thermal fluids, meeting two conditions that fight each other: do not melt, and absorb everything.
2. Someone who works with steam generators and turbines — power plant engineering, much the same work as in today's thermal and nuclear stations.
3. Someone who contains and recovers tritium — chemical engineering and radiation protection. Stopping leaks is both the safety and the fuel supply.
4. Someone who actually runs the plant — operations. Reading gauges, working valves, maintaining plant. Trade certificates in electrical work and energy management lead straight here.
5. Someone who inspects and regulates — the principle that builders and watchers are kept apart applies here too.

The fourth line is the unexpected seat. If a fusion plant is ever built, most of the people inside it day to day will not be physicists but people who have run a power station. Experience being earned right now in thermal and nuclear plants carries over intact. It is a path that runs unbroken from the electrical work, instrumentation and pipefitting taught at a technical high school.

The place to visit is the Korea Institute of Fusion Energy in Yuseong-gu, Daejeon, which runs tours on set days. KSTAR itself is visible through glass.

So when do we use it?

Honestly: fusion is tomorrow's answer, not today's. But which tomorrow has to be split in two. Making electricity from fusion for the first time, and a power station selling that electricity, are different events.

The person responsible for KSTAR puts the first in the 2030s — fusion electricity used in daily life once, as a demonstration, with the outline of a pilot plant inside the 2040s. The second, a commercial plant selling power, is generally placed beyond the 2050s internationally. A child at school now will be an adult before that one runs.

Today's answer lies elsewhere — using less, solar and wind, and the nuclear plants already turning. Fusion is what gets built in the meantime.

So the work at KSTAR is not building a power station. It is closer to setting down answers for the people who will build one: what to make the wall from, how to cool the magnets, how to catch a plasma coming apart. Only when those answers are gathered does the first plant stand.

One last thing. Plasma is hard to hold because the motion of a fluid and the forces of electricity and magnetism are tangled together, and the arithmetic is brutal. Lately, artificial intelligence has begun to solve exactly this kind of problem, and work already exists on AI catching a plasma as it starts to come apart in real time. AI eats enormous amounts of electricity; fusion is a candidate to make it; and fusion needs AI to be held. Each needs the other.

The question that remainsThis is a technology its builders will not live to use. If it should be built anyway, what is the reason?