Did you like Michael's article?
One of the main challenges facing fusion today is keeping a high-temperature, high-density plasma in the middle of a vacuum chamber and holding it there even as it reaches temperatures that are ten times hotter than in the core of the sun. This may sound impossible. But to me, it’s what I want to help the fusion community solve.
Reaching these high temperatures is not the problem. With the push of a button, a fusion reactor can heat a plasma up to 150 million degrees Celsius in temperature in less than one second. The problem is figuring out how to efficiently control this high-energy plasma to keep it from touching the walls of the tokamak except for the divertor, which is used as the exhaust point of the reactor.
My focus is on measuring the temperature and density of fast and turbulent transport processes with the help of a new diagnostic which I developed during my PhD thesis.
The dotted lines represent magnetic flux surfaces. Although only eight are shown in this picture, in reality there is an infinite number of these magnetic layers sitting within each other (we often use the term “nested”) just like layers in an onion. Since the plasma consists of charged particles, we can control it with these magnetic fields. The particles can move rapidly along one magnetic flux surface (layer), but only very slowly from one layer to another – in other words, they are magnetically confined.
Theconfined region of the plasma is where the magnetic flux surfaces are “closed” like a ring. Further out are the “open” layers that touch the divertor. We call this region the scrape-off layer. The last closed magnetic flux surface, called the separatrix (the purple line in Figure 1), separates the confined region of the plasma from the scrape-off layer.
With a size of only a few centimetres, in contrast to a total plasma radius of 2 metres in the ITER tokamak, the plasma edge region is very small compared to the overall plasma size - like an eggshell compared to the egg. Despite the fact that the desired fusion reactions do not occur in this region, this area plays a key role in realizing fusion on earth, because the processes that do occur within it are of major importance for 1) the overall behaviour of the plasma and 2) the power exhaust process.
The first important role of the plasma edge region is to set the boundary for the hot core plasma in the confined region (the yellow, orange and red regions in Figure 1). It surrounds the burning plasma and acts like a shield between it and the reactor walls.
In the plasma edge region, the gradients (changes in a given direction) of temperature and density are very steep – i.e. the properties change fast over the radius. The combination of the width of this layer and its gradient defines the boundary to the plasma core. We have found that the higher the core boundary temperature is, the better the conditions will be for fusion to occur in the plasma core.
Think of the plasma edge region as the insulated outer walls of a heated house. With no insulation and open windows, the heat flows right through the walls. In this case, the temperature gradients in the outer wall region are low. If we close the windows and insulate the walls, the temperature in the wall region – from just inside the house to just outside the house – changes from high to low over a very short distance. The temperature gradient is now steep. So we can say that the higher the temperature difference at constant heating, the more efficient the thermal insulation of the house.
Back in a tokamak, the insulating properties of the plasma edge region can be explained by transport barriers which can build up just inside the separatrix in the confined region. The better this barrier, the higher the temperature difference and thus the higher temperature and density will be at the plasma core – which makes it more likely that fusion will occur successfully.
The second role of the plasma edge region is to moderate the power distribution in the scrape-off layer and the divertor heat loads. This is still a key nuclear fusion research challenge because the amount of power we are talking about is enormous!
The steady-state operation of a nuclear fusion device is the ultimate goal for an efficient fusion power plant. That means keeping plasma parameters like temperature and density, as well as the fusion power produced, constant over time. To sustain and control the fusion process, continuous external heating is necessary.
For ITER the foreseen heating power is 50 megawatts - the power equivalent of 500 car engines or 25.000 hair dryers. Moreover the desired goal for ITER is to use this 50 megawatts heating power to generate 500 megawatts of power released by the resulting fusion, which corresponds to a power amplification factor of Q = 10.
Because we do not want the machine to continuously heat up, the challenge becomes removing the 550 megawatts (50 megawatts heating plus 500 megawatts fusion power) from the tokamak chamber. This means: What goes in has to go out.
But how can we remove power from a nuclear fusion device?
80% of the fusion power (400 megawatts) is carried by the fusion neutrons. As they are electrically neutral, they do not interact with the magnetic field cage and are therefore distributed evenly over all wall components. The same holds true for the generated photons (light emitted by the plasma), which carry a total power of 40 megawatts to the wall. These two parts are not the problem. Heat exchangers placed behind the first wall use them to generate electrical power.
The challenge is handling the remaining 110 megawatts of power, carried by charged particles that slowly move from the hot core to the relatively-cooler plasma edge. Physicists call this process diffusion. Once the particles cross the separatrix and reach the scrape-off layer, they are transported to the divertor. Due to diffusion, some charged particles reach more outward layers before touching the divertor. The speed of this diffusive outward motion influences the width of the blue area in Figure 1, which is a belt roughly three centimetres wide surrounding the divertor. If all the power in the scrape-off layer was deposited in this tiny region, we would end up with a power density (amount of power per area) of 100 megawatts/square metre. No material can withstand these power loads, as it is higher than if we placed the material directly on the surface of the sun!
This means we have to minimise these power loads somehow. To manage this, we combine two strategies. First we must get rid of some of the power in the scrape-off layer by additional radiation (plasma has to emit additional light, which is again homogeneously (equally) distributed over all the machine components). Second we must increase the area in the divertor reached by the particles.
Heat-carrying filaments, which occur in plasma, can help us do this. Filaments are structures several metres in length, which stretch along the field lines. In the cross-section, they appear as a small dot of only a few millimetres in diameter. Comparable to flares in the sun [see Figure 2], they can carry temperature and density radially outwards, speeding up the slower diffusive outward transport of heated plasma particles. In contrast to diffusion, where heat is transferred between particles, heat transported by the filaments is due to the movement of the particles themselves. We call this process “convection”. It is the same physical principal as when you use your convection oven, which uses a fan to move the hot air particles around.
To be able to control these filaments through plasma shaping, fuelling and heating, we need good experimental access to the underlying physics. In other words: we need to conduct experiments to better understand the physics of these filaments and how we can control or at least influence them.
The filaments are turbulence-driven and can move radially outwards with velocities of several kilometres per second. Diagnostic instruments that measure parameters like the temperature and density of plasmas therefore have to deliver a sufficient temporal and spatial resolution (be very quick and able to record all points in the diagnosed area).
In fact, they record one million frames per second to catch these filaments! This is so fast that watching two seconds of a recorded plasma discharge at the standard television framerate of 24 frames per second would take around 22.5 hours - the same as watching all the Star Wars movies back-to-back. This would take too long, so we use atomised and statistic-based analysis tools to catch the occurrence of filaments.
Someday, we hope that these powerful diagnostics, combined with enhanced theoretical models and improvements in materials, will help us control even the hottest of plasmas and harvest the heat energy created by fusion reactions in future devices like ITER and DEMO.
Controlled fusion for electricity production was promised by scientists a long time ago. But now I can see how all the parts are coming together and how the remaining problems are being identified and addressed one after another. I find it so interesting and rewarding to get to work on solving one of the pieces needed to realise the promise and make fusion happen within my generation – the ITER generation.