ARS Technica by Chris Lee – Jun 22, 2015
Once upon a time, I worked at a research institute that was, for the most part, devoted to nuclear fusion. Although I was never involved myself, two things impressed me about the research. The first was the pure difficulty facing researchers: there are material, plasma physics, and control system issues that are enormously challenging. And, yet, progress is made—I am now, and will forever be, impressed by the achievements that I saw during my short stay among the fusion researchers.
On the other hand, I was also impressed (and not in a good way) by how locked in certain decisions were. This is a natural consequence of doing science that has an enormous infrastructure—once a facility is built, some decisions simply cannot be unmade. These constraints, understandable as they are, make me fear for the eventual success of tokamak fusion. I would rather researchers were given more money so that they had more flexibility in terms of repurposing facilities to try new avenues of research.
Given these thoughts, you would think that I would be enthusiastic about alternative fusion schemes. Yet, because I am vaguely aware of the challenges, and how robustly they are being addressed, I tend to greet alternative fusion schemes with some skepticism. So, it is with some interest and a bit of trepidation that I started examining new research on the Polywell fusion concept, where a group claims to have achieved grand new things in terms of plasma confinement.
Fusion, fission—it’s all the same, right?
Fusion, the energetic process that powers the Sun, is really a golden dream for unlimited and almost clean energy. Nuclear reactors generating electricity today operate on the principle of fission, or splitting elements apart. There are some elements—usually very heavy ones—that are inherently unstable. These elements fall apart to form different, more stable elements. A fission reactor coaxes these elements apart (usually by banging neutrons into them), and in doing so the elements release energy, which the reactor captures and uses.
But generally only elements above a certain mass are amenable to fission (these are called, appropriately enough, “fissile elements”). At the other end of the periodic table, lighter elements can be combined to heavier elements, and in doing so, become more stable. The very lightest element, hydrogen, can have its nuclei forced together to make a helium nucleus. This releases energy because the helium nucleus can exist in a more stable, lower energy ground state.
To put it in perspective, creating helium from isotopes of hydrogen will release on the order of a few MeV of energy (a million electron volts), while a chemical reaction releases about a single eV of energy.
The kicker, though, is that it requires a great deal of energy to force the hydrogen nuclei close enough together for them to start fusing. Methods to achieve this can be divided into two classes: the “we crush solidified hydrogen using big-ass lasers” class, or the “we crush plasmas with big-ass magnetic fields” class.
As much as I love me some lasers, today we are only going to discuss magnetic fields. The basic idea is that, once the electrons are stripped from the nuclei, the hydrogen takes the form of a plasma. A plasma is basically a gas consisting of electrons zipping around at very high speeds and much more slowly drifting ions (positively charged nuclei). Like a gas, plasma has a pressure that is related to its temperature and density. Additionally, because a plasma consists of charged particles, it can be contained and compressed by magnetic fields.
The reason I mention this is that one can differentiate between the natural pressure of the plasma and the pressure applied by a magnetic field. The ratio of these two, called beta, is an important parameter in fusion reactors. Essentially, the efficiency of fusion goes up rapidly as beta approaches one, so magnetic compression aims to get beta as close to that level as possible.
On the face of it, this seems simple enough: for the most efficient reaction, add more magnetic field, get more beta. Unfortunately, two things happen. First, the magnetic confinement tends to leak, so as you squeeze the plasma harder, it escapes. Second, as you squeeze the plasma, it becomes more turbulent and generates massive currents and fields that may destroy your apparatus. There are no simple solutions to either of these problems. There are, however, complicated solutions that have allowed tokamak scientists to reach beta values as high as 0.4, though 0.1 is a more typical value, and the ITER fusion megaproject is expected to operate at a beta of 0.03.
One line of thought, which was recently pushed by Lockheed Martin engineers, is that getting beta very close to unity may well allow for smaller, cheaper fusion reactors. That’s where the Polywell reactor comes in.
Polywhat?
The Polywell reactor arranges magnetic fields differently to that of a standard, traditional tokamak. In a tokamak, the fields set up a plasma that has a donut shape. The plasma is free to circulate around the donut but is squeezed to a thin, high pressure stream.
The Polywell concept, instead, tries to create a magnetic box to confine the plasma in place, which reduces the turbulence and solves many control problems. However, a true box is simply not possible. This is because the force applied by a magnetic field depends on the direction of motion of a charged particle, which causes the electrons and ions to travel in a corkscrew motion around magnetic field lines. At each corner of the magnetic field box, the magnetic field lines point outward away from the center of the box, so the plasma can spiral out of the box. The upshot is that the harder you squeeze, the quicker the plasma leaks out, leaving you with a low beta plasma.
To help overcome this, additional high energy electrons are injected into the plasma. The electrons create a large negative potential that draws the ions to the center of the box, slowing their escape. Nevertheless, even with the electrostatic draw slowing ion escape, the magnetic field still wins in the end, because the electrons are also driven to spiral along magnetic field lines.
But researchers quickly realized that if the plasma was dense enough—in other words, if it had a high beta—it would exclude the magnetic field lines, creating a sharp boundary between the plasma and the magnetic field. The sharp boundary acts like a mirror for charged particles, vastly slowing their rate of ion escape. This unfortunately creates a chicken and egg scenario: if you have a high beta plasma, a Polywell design will keep it confined at high beta. But first, you must have a high beta plasma.
Cheating for the sake of physics
This is where the researchers chose to do the intelligent thing and cheat. Typically, in magnetic confinement fusion, you initiate the plasma from a gas. Here, however, the researchers used a sheet of plastic. Essentially, they sent vast amounts of current through a sheet of plastic, generating a large amount of plasma in a short time. This was then injected into the confinement area and the box turned on. Because the plasma was initiated from a solid, the total number of ions and electrons is much larger than can be achieved in a similar time with a gas. This meant that the researchers started with a much denser plasma and, hopefully, a higher beta.
Unfortunately, the experiment was not refined enough to directly measure beta, so the researchers make no claims other than that beta is high. What is important is that they show that the magnetic field is partially excluded from the plasma. They don’t compare directly to gas injection systems, but they do show, through analysis of the X-ray emission, that the density of electrons remains rather high over the period in which the magnetic field is on. That is actually a complicated statement, because the magnetic field is pulsed and the plasma is very short-lived; but during the short lifetime of the plasma, the magnetic field is rather constant.
The researchers also showed that, by changing the power used to generate the plasma, the magnetic field boundary progressively sharpens as the plasma density grows. These two together are a good indication that, indeed, a high beta plasma will be contained reasonably well by a Polywell configuration.
It should be noted that this is a long way from anything remotely close to fusion. First of all, although they claim a beta close to unity, the sharpening of the magnetic field boundary was very small—the researchers estimate this to be on the order of 10 percent. I would guess that they need an order of magnitude higher density to actually get a boundary sharp enough to be useful. And, since they are already starting from a solid target, I am not sure how they can manage that. I guess if they can increase the power of the plasma injector and use larger targets, they could get there.
What I fear, though, is that research like this will be used to under-fund ITER. My honest opinion is that both should be funded. It is not yet clear that the Polywell concept will ever go beyond a concept, while tokamak designs, and their relatives like the stellarator, are well-developed. Although we can’t predict the economic success of tokamak reactors, I am pretty confident that ITER will meet its science and engineering goals.