The Star Builders

by Arthur Turrell

  To bring ordinary hydrogen isotopes to the extraordinary plasma temperatures required of JET isn’t easy. Conventional heating methods don’t work. You can’t use a fire since there shouldn’t be any oxygen in the chamber, and in any case, fires just don’t burn anywhere near hot enough. More elaborate heating is required.

  The star builders’ first method is to push enormous currents through the hydrogen. Ever notice your phone charger getting hot if you leave it on for a long time? A current going through a wire generates heat. The same effect works in hydrogen gas—push a current hard enough through it, and it heats up. The second method is, somewhat surprisingly, radio waves—not the kind you normally listen to but ones that are specially tuned to a frequency that makes the atoms in the hydrogen gas vibrate. As they vibrate, they crash into each other and heat up. The final method that magnetic confinement star builders use to get to 100 million degrees is to fire beams of atoms into the gas at great speeds (and so great energies). These atoms crash into the increasingly hot gases, dumping their energy as they go and heating what’s already in the chamber even further.

  The amount of gas that they make fiercely hot is very small—if you scooped up all of the particles and weighed them, they’d have a mass of just 0.1 milligram (approximately 35 millionths of an ounce). This makes for a gas density in the chamber that is just one-millionth that of air. The Sun is many, many times more dense than this (which means more collisions); JET gets even hotter than the Sun partly to make up for its much lower density of particles.4

  In an energy-producing experiment, that fraction of a milligram would be made up of equal parts of the two special isotopes of hydrogen that it is easier to make fuse, deuterium and tritium. JET is an experimental reactor, built to explore the physics of nuclear fusion. It was never designed to produce electrical power for the grid. Because of that, most of the time JET uses deuterium alone as it’s easier to handle and much more common than precious tritium. Physicists can still learn a lot by running with deuterium alone because most of the problems with reactors on the front line of net energy gain come down to the plasma, rather than the fusion reactions themselves. And deuterium does fuse with itself, so they can still record nuclear reactions and see some energy being released; it’s just a much less likely reaction than the one between deuterium and tritium.

  Trapping Hot Hydrogen

  As we’ve already seen, the three keys to net energy gain are temperature, density, and—finally—confinement. Once the star builders here have filled JET with hot hydrogen plasma, they need to keep it that way. And this is the great problem facing tokamaks: most machines can only run for a few seconds at a time before they lose control of their plasmas.

  So how do you create a trap for something that is more than 100 million degrees? Most people’s first instinct is that the hot deuterium (and sometimes tritium) is trapped by the metal walls of the reactor chamber itself. But that doesn’t work, not if you want to create a mini-star. If you’ve ever put a hot pan in water, you’ll know why. There’s hissing and fizzing as the heat energy drains straight out of the pan and into the water. This fizzling out is exactly what star builders don’t want to happen to their fusion plasma. If all of the high-energy particles escaped into the wall, they’d transfer their energy over too, effectively cooling the remaining hydrogen and stopping the fusion reactions dead. It’s not good for the walls either; they can reach 300 degrees Celsius (572 degrees Fahrenheit) during a typical experiment, but they wouldn’t survive 100 million degrees Celsius (180 million degrees Farenheit).5

  Star builders need to pull off a much more impressive trick: to stop the plasma from touching anything at all. It sounds impossible: they need an invisible floating box that is impervious to heat. The star builders’ solution sounds impossible too: invisible force fields. But it’s not science fiction, even if it sounds like it.

  The trick is magnetic fields, with which star builders weave a complex web. Everything is controlled by those four fundamental forces of physics—the strong force, the weak force, electromagnetism, and gravity. Electromagnetism is responsible for electrical charge, magnetic fields, and light. The temperatures required for fusion to work create a plasma, which consists of charged particles. This makes magnetic fields a good, if tricky, way of trapping and controlling the plasma.

  The idea of using magnetic fields to confine plasma was one that star builders came up with very early on in their attempts to do fusion. When a charged particle encounters a line of the magnetic field, it gets trapped and follows it, orbiting around the line of the field in loops as it goes. The size of the loop that charged particles gyrate around is named the Larmor radius after the physicist Joseph Larmor. The stronger the magnetic field, the smaller the Larmor radius, and the more tightly bound the particles are to the magnetic field lines. Because the particles are moving anyway, in addition to gyrating, the pattern they trace out is a helix—as if they were on a helter-skelter (an amusement park ride in the UK, with a spiral slide built around a tower). Stronger magnetic fields provide better confinement of plasma partly because charged particles in the plasma are kept closer to the magnetic field lines.

  Regardless of the neatness of trapping particles on magnetic field lines, early machines using this approach had many problems that JET has had to overcome, or at least attempt to overcome.

  Perhaps the first serious magnetic confinement fusion design can be credited to Max Steenbeck. Prior to the Second World War, Steenbeck was the laboratory director of a Siemens-Schuckert plant and a member of the Volkssturm, the Nazi people’s army designed to be the last line of defense during an invasion of Germany. The plans for his confinement device found their way to Imperial College, where they were refined by the son of J. J. Thomson, G. P. Thomson, who filed a patent for a doughnut-shaped, or toroidal, star machine in 1946. This was inspired by the “pinch” effect whereby large currents traveling through a conductor can, via the magnetic force, squish the conductor inward. It’s this effect that causes lightning rods to become misshapen once struck.6

  The simplest version of the toroidal pinch machine consists of an airtight Pyrex doughnut coated with metal that, when current flows through it, creates an electric field in the gas inside. This field provides enough energy to rip the atoms of the gas inside apart into glowing plasma and to push the charged particles around the doughnut’s tube like NASCAR racers. Electric and magnetic fields are so interrelated that they cause one another: the current running around the ring creates a magnetic field that wraps around the inside of the ring at right angles to the current. This same effect is why a magnetic compass gets confused next to a live electricity cable.

  But the really clever trick of the toroidal pinch is how the interplay of magnetic fields and current traps the plasma. The electromagnetic force dictates that whenever there is a current in one direction and a magnetic field in another, charged particles must move in the third and final direction, at right angles to the other two. Imagine that you are a particle going through the doughnut’s ring as part of the current; then the magnetic field is wrapping around you in a circle, pointing clockwise. The direction of the force that’s generated then points back at you from every angle. It’s this force that pinches the plasma’s particles together in the middle of the ring, squishing them up to get them closer to the density and temperature needed for fusion to have a chance. The name “toroidal pinch” arose because the hot fusion fuel is pinched into a torus shape: a doughnut within a doughnut created by an invisible force that can briefly hold the hot fuel in place without its touching anything.7

  Toroidal pinches aren’t the only early magnetic confinement device that star builders tried. Another early design, pioneered by scientists at Lawrence Livermore National Laboratory (where NIF is based), was, essentially, a sausage of magnetic fields, with the magnetic field lines going lengthways along the sausage. Particles would zip up and down. At either end, the magnetic field lines came together in dense bunches, like the tie in a wheat
sheaf, the effect of which was to reflect particles back the other way. These early star traps were optimistically called magnetic mirrors, because they were supposed to reflect the particles. In reality, many particles leaked out of the ends—more like a polished window than a mirror.

  There is a problem with these early pinch and mirror machines that goes beyond mere leakiness though—a problem that means they’re unlikely to ever be good for net energy gain. JET has only partially solved it. Both toroidal and magnetic mirror star machines suffer from debilitating instabilities: catastrophic failures in which the magnetic field gets so rucked up that the confinement is completely lost.

  Dr. Fernanda Rimini spends her days trying to understand the instabilities that plague magnetic star machines. She’s a physicist and one of the session leaders of experiments on JET, responsible for translating the scientific ideas of her colleagues into settings that are feasible for the machine. She has a commanding role in the control room. While we’re chatting in one corner, several people try to ask her questions on this or that technical detail. Despite the importance of her role in the smooth operation of JET, and my happening to catch her at the end of a shift, she’s still full of energy. I don’t think she ever switches off.

  “Heavy metal music for me is the perfect soundtrack to an afternoon spent building with Lego,” she says, when describing how she relaxes in her spare time. She has even re-created in Lego the building that houses JET and a miniature version of the control room. She says that what’s kept her in star building is the curiosity and fun. “And you’re doing work for the future of humankind,” she adds. “It’s an energy source, and it’s important.”

  Fernanda describes her role as halfway between pure physics and engineering, and I can’t help thinking that this is currently apt for the whole of fusion the world over.

  We’re talking in the busy control room, and I struggle to hear her over the background ruckus as months of planning come to fruition next to us. Fernanda explains that if they can control instabilities, they’ll be well on their way to having a working star machine. Right now, instabilities are one of the reasons JET can only operate for a few seconds at a time.

  While instabilities aren’t unheard of elsewhere in nature—take, for example, an out-of-control garden hose or the snowball that grows into an avalanche—plasmas are especially susceptible to them, to the point where the disruptions caused by plasma instabilities frequently stop machines entirely. Instabilities arise in fusion because nature hates extreme conditions in temperature and density as well as magnetic and electric fields existing in a small patch of space or time. Given a chance, nature will even out such inequalities. Because there are such extremes of energy involved in fusion, the reestablishment of balance can happen in abrupt and surprising ways, like a stretched elastic band suddenly being released.

  Professor Jerry Chittenden of Imperial College London has said that attempts to confine the fuel for fusion are subject to Murphy’s Law: anything that can go wrong, will go wrong. Nuclear fusion is flighty and all too easy to stop—one way in which it’s far harder than nuclear fission (and also one reason why it’s safer).

  Those early toroidal pinches suffered two instabilities in particular when run with higher currents. One is the kink instability, a devastating fusion-stopper that JET’s clever design mitigates. It’s the kink instability that causes the breakup of the hoops of plasma thrown out of the Sun’s surface. In toroidal pinches, kink instabilities occur because the magnetic field is stronger on the hole side of the torus’s loop than on the outer side, since the charged particles are closer together on the inside curve. But because the magnetic field is weaker on the outside of the curve, the pinching effect is weaker too. This allows some charged particles to bulge out, weakening the field further. The weaker field permits even more charged particles to bulge out, and so on, until the process runs away with itself, confinement is lost completely, and fusion stops dead. The second problem with toroidal pinches is the sausage instability. This amplifies tiny imperfections in how much the column of plasma is pinched, causing a shape like a chain of sausages to form and to ruin confinement. While JET is less prone to instabilities and the disruptions they cause than earlier magnetic star machines, they’re still a big problem.8

  JET is a tokamak, the most popular and successful type of magnetic confinement fusion machine. Even among tokamaks, JET is famous because it holds the world record for the highest Q (the ratio of power out of the plasma to power put in) ever achieved, 0.67. The machine itself is impressively big: the reaction chamber stands around six meters (approximately twenty feet) tall and three meters (approximately ten feet) wide. It’s shaped like a fat doughnut, with a tube that’s two meters (approximately six and a half feet) in diameter.

  Tokamaks were first developed in the Soviet Union in the early 1960s, the name derived from the Russian terms toroidalnaya kamera and magnitnaya katushka, meaning “toroidal chamber” and “magnetic coil.” The genius of the Russian star builders, and what allowed them to stop the kink instability from being so debilitating, was to introduce a second magnetic field to their toroidal chamber.

  Figure 5.1 shows the basic design elements of a tokamak, including the reactor chamber. Inside, there’s the ring-around-the-doughnut field, known as the poloidal field (one example loop is shown with a solid line and arrows), that was also present in the toroidal pinch. As in the toroidal pinch, this magnetic field can be generated by the particles’ own motion. But there’s a second field too (shown with a dotted line). This is the field that the Russians added. The second field, which goes around the doughnut’s tube like the particles do, is called the toroidal field because it has the shape of a torus. The clever part of the Russians’ design is that the two magnetic fields combine to create an overall magnetic field that twists around the tokamak helically, like a slide wraps around a helter-skelter, but on the inside of the torus.

  Figure 5.1: The basic design elements of the tokamak, the leading magnetic confinement fusion star trap. The reactor is shaped like a doughnut. Two different magnetic fields combine to make a helical field (shown with filled-in arrows) that wraps around the inside of the tube of the doughnut-shaped reactor chamber. Charged particles (electrons and hydrogen nuclei) get trapped on the magnetic field lines, gyrating as they travel along them. Magnetic confinement keeps these particles from hitting the chamber walls—sometimes.9

  Particles are confined to gyrate around the helter-skelter magnetic field lines. Because they move from the inner to the outer part of the doughnut’s tube as they go around it, particles spend as much time on the inner side of the tube as they do on the outside. That stops the bulges endemic to toroidal pinches from developing in the first place, which prevents them from growing, which keeps the plasma flow uninterrupted.

  Russian tokamak scientists suppressed the kinky tendencies of the toroidal pinches that had gone before. Straightaway, their tokamak design provided tenfold better confinement than alternative magnetic schemes of the time. The strategy was so successful that, initially, no one outside of the USSR would believe how good the results were. To settle the question, a team of scientists from Culham was sent to Russia at the height of the Cold War to use Thomson scattering as a more precise check on the temperatures. The Russians were right; tokamaks were confining hot fusion fuel like no other machine before them. Tokamaks soon became the leading magnetic star traps the world over, and they remain so today, and JET is, for now, the most successful and largest of them.

  Fernanda and her colleagues follow a punishing schedule that tries to squeeze every last bit of understanding, and penny of value, out of JET. They begin one shift at 6:30 and go until 14:30, and then run another, with more experiments, from 14:30 to 22:30. By the time a fusion experiment is prepared for the machine, she reckons there have been around six months of preparation, and they aim to run one every half hour.

  Although JET is less prone to instability than some other designs, there is still a whole
menagerie of instabilities just in tokamaks: edge-localized modes, sawtooth oscillations, tearing modes, ballooning modes. And then there are the insidious ways that particles, and their energy, try to escape confinement, like the bizarrely named “banana regime transport” (after the shape the particles trace out). These are driven by energetic collisions that send a particle spinning out of the web of magnetic fields. Each different way for particles to escape presents a different headache for star builders, and there are pages and pages of mathematics that try to understand how they happen, and what scientists can do to mitigate them.

  Today, Fernanda is coordinating experiments that look at how injecting pellets of frozen fuel at just the right moment can stop the instabilities from killing the confinement. They do it by cooling down a portion of the fuel.

  When a tokamak runs into a bad instability, it can cause a catastrophic loss of energy called a disruption. In fast disruptions, the energy in the fuel is violently dumped into the vessel in a few milliseconds.