Fusion — Can We Build a Sun on the Ground?

Not a fire

The fire we know is something joining with oxygen from the air. A candle does it; so does firewood. But there is almost no oxygen in the Sun. With nothing to join, it cannot burn in the way we mean.

What happens in the Sun happens much further in. The tiny cores at the centre of atoms crash together and merge into one, and heat comes out at the moment they merge. This is fusion — the word simply means joining.

How small is the thing we are joining?

Small is easy to say and hard to see. Blow one atom up to the size of a football pitch and the nucleus at its centre is a single grain of rice. All the rest is empty space.

It is the grain of rice we are trying to join, not the pitch. Picture two pitches side by side and the job of making the two grains at their centres meet squarely, and the shape of the difficulty starts to show.

If atoms and nuclei are not yet solid in your hands, open the board below first. You can zoom from the football pitch down to the grain of rice yourself, and add and remove protons, neutrons and electrons one at a time to watch hydrogen become deuterium. The deuterium and tritium of this piece are made there.

Tool From the Atom to the Periodic Table Blow an atom up to a football pitch and zoom to the grain of rice at its centre, change protons, neutrons and electrons one at a time, count each element's hands to see why water is H2O, then fold the line into the table and sing through all 118. (Korean only)

Energy from splitting, energy from joining

Here is an odd thing. A nuclear power station gets its energy by splitting nuclei apart. The Sun gets its energy by putting them together. If both give out energy, which one is right?

Picture it as a valley with iron at the very bottom and a slope on either side. Heavy uranium sits up one slope and rolls down by splitting; light hydrogen sits up the other and rolls down by joining. Whichever side it comes from, it arrives at iron, and the distance rolled is the energy.

The Sun joins hydrogen, the lightest element there is: the commonest material in the universe giving the largest return. Do it on Earth and the fuel question all but disappears.

So why is there no power station yet?

Every nucleus carries positive charge, and like the same poles of two magnets, they push each other away — harder the closer they get. To beat that push you must throw them very fast, and making particles fast is exactly what heating them means.

The temperature needed is around a hundred million degrees. And there the real problem appears: nothing on Earth can hold something that hot. Steel and even tungsten melt long before.

A container made of magnets

The answer was to keep it off the walls. To do that the gas has first to be turned into plasma — something a magnet can take hold of. Three steps.

1. Put a strong electric field through the gas. Electrons are pulled off the atoms.
2. With the electrons loose, what remains and the electrons drift apart, each carrying charge. That state is plasma.
3. Anything charged can be handled by a magnet. What sat unmoved beside the strongest magnet now does as it is told.

So a magnetic field is wrapped into a doughnut and the plasma held inside it, floating rather than touching. The design is called a tokamak, and KSTAR in Daejeon is one: a doughnut about ten metres across.

It is a little off to say the magnets push the plasma away from the wall. They do not push it away; they skewer it. Switch on the field and countless invisible skewers appear, and each charged particle is threaded on one, spinning round it and moving only along it. But a skewer has an end, and at the end the particle slips off. So the skewers are bent round into a doughnut. With no end to reach, the particle goes round for ever and never gets out.

This explains those strange headlines about holding a hundred million degrees for a number of seconds. Reaching the temperature is not the hard part. Holding it steady is.

And once they join?

Two nuclei join and out come one helium and one neutron. Put them on the scales, though, and something is odd: everything that comes out weighs a shade less than the two that went in. That missing weight is the energy.

And here the two part company.

That small parting is the key to everything that follows. The one that stays keeps the fire alive; the one that leaves batters the wall and wears the machine out. And it is at that battered wall that the electricity gets made. Which is the next piece.

For the whole sequence in one go, there is a one-minute film at the top of the tool page, running from the Sun's fifteen million degrees to the boiling of water.

Reading about it rarely shows you where the difficulty sits. The board below lets you try it: fire two nuclei at each other, raising the temperature until they stick, then hold a hundred million degrees with magnets and watch what heat alone fails to do.

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

Why hotter than the Sun

The centre of the Sun is around fifteen million degrees — far below the hundred million we aim for. It sounds backwards that the Sun is cooler.

The difference is weight. The Sun crushes its own centre with its own mass, packing nuclei so tightly that they meet often without moving especially fast. Earth has no such gravity, so what cannot be made up in density has to be made up in temperature.

By the arithmetic, the Sun should not burn either

And here something strange turns up. Work out how fast nuclei move at the temperature inside the Sun and it is nowhere near fast enough to beat the push between them — not nearly, by a long way. By classical physics the Sun ought not to be burning.

It is burning.

The calculation was done in the 1930s. Climbing the wall head-on would take three billion degrees, and the number was crushing: no such temperature exists anywhere.

The answer is that a nucleus is not the kind of ball we picture. It is not a speck sitting at a definite spot but something smeared out, like a wave. So it is faintly present even beyond a hill it has no hope of climbing, and once in a great while it appears on the far side without having gone over at all. This is quantum tunnelling.

Redone with quantum mechanics, the answer came out at a hundred million. There was a way through rather than over, so the top of the wall was never needed. Three billion becoming one hundred million is what saved fusion. A hundred million is enormous; three billion was not a number anyone could even attempt.

The fuel is in seawater

The fuels are deuterium and tritium. Deuterium is taken from seawater; tritium barely exists in nature and is bred from lithium. A gram of fuel is said to hold as much energy as eight tonnes of oil.

Being in seawater makes it sound common. It is not especially common: about one hydrogen in six thousand four hundred. Collect every scrap of deuterium in a two-litre bottle of water and you have seven hundredths of a gram. But a gram was eight tonnes of oil, so seven hundredths is not a small quantity either. It is thinly spread, and the sea is wide.

Two consequences follow. The reaction does not sustain itself, so any slip in conditions simply puts it out — there is no runaway to worry about. But the neutrons thrown out keep striking the vessel wall, and what to build that wall from is a large problem of its own.

Where seconds make the news

Fusion has three conditions: how hot, how dense, and how long you can hold it. The first two are largely in hand; the fight now is over the third.

The reason holding is hard lies in the doughnut shape itself. Stand the skewers straight and the field is even everywhere. Bend them into a ring and the inside differs from the outside: skewers crowded on the inner side, sparse on the outer. A plasma is like a balloon squeezed hard, and squeeze a balloon harder on one side and it bulges out on the other — the plasma escapes wherever the grip is weakest.

So the skewers are not simply bent round but twisted as they go. Wound into a helix, each particle passes alternately along the inside and the outside and the difference cancels. The idea dates from the 1960s, and a machine built this way is called a tokamak. Sixty years after the principle, the record is still 48 seconds, because even so the plasma keeps finding new ways out.

KSTAR set a world record holding a hundred million degrees for thirty seconds, then stretched it to forty-eight. Replacing the component that takes the plasma's heat head-on, from carbon to tungsten, did much of that. The next target is three hundred seconds — five minutes. Past five minutes the question stops being whether it works and becomes whether it can be kept running.

A far larger machine is being built in southern France by many countries together, Korea among them. What KSTAR learns goes there.

Why the magnets must be superconducting

One more thing about the magnets. An ordinary electromagnet is a copper coil with current through it. But current meets resistance and makes heat, and a magnet as strong as fusion needs would melt its own coil. It can hold for seconds, not for a working day.

A power station runs all day. So the coils must be superconducting, with no resistance at all, and superconductivity only happens at 269 below. Which leaves a hundred million degrees and minus 269 a few metres apart inside one machine — the hottest place on Earth pressed against the coldest.

And how is the temperature measured? You cannot put a thermometer into a hundred million degrees. The answer is the method used for stars. Hot things glow, and the colour of the glow depends on how hot. The light from the plasma is collected and its colour read. A sun built on the ground, measured with the ruler made for stars.

What you would study to do this

A fusion institute is not staffed only by physicists. One machine needs several different trainings.

1. Someone who works out how the plasma moves — physics.
2. Someone who builds magnets strong enough to hold it — superconductivity and electrical engineering.
3. Someone who designs the doughnut vessel and its cooling — mechanical engineering.
4. Someone who finds a wall that survives the neutrons — materials engineering.
5. Someone who tends and mends the cooling plant that keeps those magnets at 269 below — cryogenic plant and mechanical maintenance.
6. Someone who catches the plasma in real time as it tries to break up — control and computing.

So there is room here for someone who is not especially strong at mathematics but likes building things, or wonders how materials behave. Not one path, but several arriving at the same place.

The fifth line is the unexpected seat. The plant that keeps the magnets at 269 below is a very large refrigerator made of pipework, valves and gauges, and it is open to people who never read physics at university. Let that refrigerator stop and the magnets lose superconductivity and the whole experiment stops with them. Making a hundred million degrees begins with keeping a fridge running.

The place itself is the Korea Institute of Fusion Energy in Yuseong-gu, Daejeon. Open Daejeon on the map and it is there.

One question is still unanswered. Having made a hundred million degrees, how exactly is that heat turned into electricity? The answer is a little deflating. It is the next piece.

The question that remainsIf the hardest part of copying the Sun was not the fire but the container — what else are we still failing to build a container for?