I never quite understood the math behind power densities in a fusion reactor.
In the sun, isn't energy production occurring at something like 100-1000 W/m3? So, if you want to build a multiple MW fusion plant, shouldn't these plants be ridiculously huge compared to, say, a wind turbine rated at a couple of MW?
Is the density of the plasma so much higher in a fusion reactor?
Also, something else I never grokked, how do you get the power out? The plasma heats up, but how do you turn that into useful electrical energy?
Nevertheless of course I hope it does work as advertised... someday.
Edit: thanks everyone for the thoughtful, insightful replies!
The density is actually lower than the sun in magnetic confinement fusion (MCF) devices because we can’t squeeze plasma together as hard as the sun’s mass can. Inertial confinement fusion (ICF) can squeeze harder than MCF devices, but has serious unaddressed engineering issues.
The trick is in higher temperature plasma. The sun fuses protium (lone protons). We don’t have the confinement necessary on Earth to do this, so we fuse deuterium (1p+1n) and tritium (1p+2n). This reaction is more energetically favorable and is achievable on Earth. Coupled with giant microwave ovens and clever geometry and electromagnetic tricks, we can make plasmas much hotter (faster moving particles) than the sun can.
Once a plasma is fusing, it emits a lot of heat (alpha heating and fast neutrons). A plasma that requires no external heating (no microwave ovens) is said to be “ignited”. We don’t necessarily need or want ignition to have a successful reactor, but it’s a cool thought.
The major trouble with fusion reactors is keeping particles in the bottle long enough to fuse. Since they’re leaving anyway they have to go somewhere. You can tune vessel geometry and magnetic fields to have designated strike points where most of the plasma will exit confinement. These are called divertors. Run some coolant through your divertors and you have a heat source that can boil water and spin a turbine.
Here my knowledge gets shaky because I know that the fastest particles coming out of a D+T reaction are neutrons (they weigh much less than an alpha particle). Since neutrons are electrically neutral I think they are much less likely to become thermalized (they are not likely to bump into another particle on their way out). I’m not sure how neutron thermalization happens in reactor simulations, but I’m under the impression that it does.
The trick is that deuterium (and tritium) are vastly more reactive than ordinary protons. The fusion of the latter is extremely slow, because it involves the weak nuclear interaction to convert a proton to a neutron.
lithium blankets traps neutrons. Sadly, the reactor chamber becomes radioactive (but a short period radioactivity that only requires to store the contaminated chamber for decades)
Project Rho has lay-person accessible detail on type of fusion reactions.[1] Well worth a read. The takeaway is, if we can't do D-T fusion, then don't even think about any of the other kinds. They're orders of magnitude harder.
The temperature at the center of the sun is about 15million kelvin. The plasma temperature of a fusion reactor is around ten times higher. Also, a rector uses deuterium–tritium (D–T) fusion, and the sun uses hydrogen fusion. D-T requires less confinement and is more energetically favorable. If DT is like setting off a fire cracker then hydrogen fusion is like igniting a damp log.
I chuckled/was staggered when I heard that the power density in the Sun's core is approximately the same as a compost heap! (albeit a rather large one...)
Even at the center of the sun's core, the power density doesn't go above 300 W/m^3. At the core boundary it's < 10 W/m^3, and it's < 2 W/m^3 overall.
Terrestrial reactors try for much higher reaction rates. But most designs still have power density issues that would make them uneconomical even if they could make net power. Net power is actually a very low bar, corresponding to an EROI of 1.
>The plasma heats up, but how do you turn that into useful electrical energy?
I appreciate that many people are commenting 'you couple the plasma to a working fluid', but I think the original comment was more along the line of how you couple a confined plasma to a working fluid. By definition the plasma is in a hard vacuum, magnetically bottled. What, then, is the coupling method? Thermal photons escaping confinement? I genuinely have no idea myself, but would really like to know.
You're right, I was curious about capturing the fusion products in some way and still being able to extract useful energy.
I looked up the effects of neutron radiation on materials[]. Sounds like a hell of an engineering challenge to come up with a robust way of getting that energy out!
Radiation damage to materials occurs as a result of the interaction of a [neutron] with a lattice atom in the material. The collision causes a massive transfer of kinetic energy to the lattice atom, which is displaced from its lattice site, becoming what is known as the primary knock-on atom (PKA). [...] The magnitude of the damage is such that a single 1 MeV neutron creating a PKA in an iron lattice produces approximately 1,100 Frenkel pairs.
Yep. And anything touched by the neutron flux would become radioactive if it captures neutrons. For example steel, normal steel contains carbon, carbon captures neutrons, so the steel becomes radioactive, and also brittle. And then you need wiring and insulation and coolant and pumps and all that.
- operates in very high magnetic fields, which means high forces
No such material is known so far.
It is a bit like some engineer from 1700 said: "Well, you could just build a more efficient and compact and light steam engine, and connect it to a machine which has flapping wings, and then you have a transport vehicle which can carry people across the Atlantic ocean at supersonic speed, and at little cost."
Not such a bad analogy, as the blanket will get hot, and the heat will be used to turn water into steam, which will power generating plant. Just like in the 18th and 19th centuries.
It would have had to have been an unusually well informed and far-sighted engineer though, as Savery's engines first worked in 1698 and the first for-sale commercial engine, Newcomen's, wouldn't happen until 1712. And they were incredibly inefficient: 0.01% to 0.1%.
Do you know how far along are we in understanding how to build an effective blanket that can withstand the neutron flux while maintaining its physical integrity?
It's not a solved problem (since really testing a wall design requires a working high power source of fusion neutrons), and it's not clear it can be solved in a sufficiently practical way. Indeed, the ability of the first wall to withstand the energy flowing through it will likely cause DT fusion reactors to have low volumetric power density, with sad effects on their economics.
SPARC's approach is for the blanket to be a molten salt, and also function as a coolant. That will surround an inner wall which gets replaced annually. They tested joints in the REBCO superconductors, so the reactor can be opened up on hinges.
You're correct, that's a good ballpark for the average heat generation rate in the sun. However, 99% of heat generation in the sun occurs in the core (approximately the center 25% by radius) and this area is much more energy dense. Further, fusion reactors could achieve an even higher volumetric heat generation rate then the sun's core does.
To collect energy, heat would be transferred to a working fluid (e.g., molten salt) by exposing that fluid to the hot plasma. Then the working fluid would be used to boil water and spin a turbine.
There's something that feels so archaic about using a nuclear reactor to... boil water and spin a turbine. It's disappointing that we haven't figured out a better way to convert to electrical energy than what we were using in the 1800s.
We've had solid-state thermoelectric generators for a while that convert heat flux directly to electricity, they're just really inefficient (only 5-8% efficent). [0] I can't find anything suggesting a fundamental limit to the efficiency, but it would take a lot to catch up to steam-powered generators at 33%-48% efficiency [1].
I'm not well versed in physics, but when I think about the fact that electric motors haven't changed all that much it sort of makes sense.
Inducing a current by moving something in a magnetic field, or vice versa, may seem like simple designs, but perhaps that's why we haven't replaced them yet. Rotational motion seems easy to service and efficient.
Same with water being particularly easy to turn into a gas using heat, as well as being super plentiful on Earth. There are just a lot of things going for a steam turbine
> It's disappointing that we haven't figured out a better way to convert to electrical energy than what we were using in the 1800s.
We haven't found a better way to convert heat to electrical energy. We've done great things with solar, kinetic, gravity, etc.
I think the better analysis is to chart the efficiency of the conversion over time. I couldn't easily find a chart showing this, but I assume gigawatt scale turbines operating at 50%+ efficiency are modern engineering marvels compared to the earliest 7kW prototype made by Charles Parsons in 1884.
The reason that steam still drives turbines (as opposed to having the salt drive them directly) is that it’s hard to engineer blade materials that will withstand temperatures much higher than superheated steam.
I recommend the book "5 Equations that changed the world", it gives you the idea how modern world works and how that is direct influence of math. For example how electricity is being generated using dynamos.
The sun "burns" ordinary hydrogen. The first step in this proton-proton-chain is the reaction of two protons to a deuteron, which requires conversion of one of them into a neutron. This is an interaction of the weak nuclear force, which has very small range, so it is slow. Reactions considered for terrestrial fusion don't require the weak force, and consequently won't run on pure hydrogen.
as far as i am aware, solar power is the only prevalent power source that does not derive from spinning a dynamo. Besides wind (and perhaps there, too) almost all of the force used to spin dynamos includes some kind of boiling (eg: dams happen to use the force of the water after the "boil"/evaporation, nuclear, coal, natural gas, etc are all steam-based thermo-mechanical power plants.)
Note that gas powerplants are combined cycle power plants, they contain both gas turbine and steam turbine, so part of their spinning power (from gas turbine) is not based of boiling.
Dumb question: Is it possible to make a photovoltaic cell that works with infrared or longer wavelengths? ... Googling says there is research and some claimed results.
In theory, one could have "generate heat, and have the apparatus surrounded by a vacuum, surrounded by infrared photovoltaics" be the new universal backend for power generation. I have no idea about the associated efficiencies, of course.
Certainly the only major electrical generation method, but MHD and thermocouples can be the appropriate solution sometimes, and if you count storage systems as well as primary production, there are also fuel cells & batteries.
It's not just that the fusion only occurs in the core, but the fusion rate is also very low; it's only because the sun is so incredibly massive that this slow reaction rate can self-sustain. With a much smaller fusion reactor, the reaction rate needs to be much higher to keep itself going and to create enough power to run the electromagnets that keep it confined.
Exactly: there is no prospect or intent to actually, ever make a Tokamak that produces useful power. Besides the power density problem and the really absolutely enormous expense and already-sunk cost, the necessary neutrons would destroy the most expensive part of the reactor in short order, and make the rest radioactive.
The whole research program gets its funding as what amounts to a jobs program to keep high-neutron-flux physicists employed and available to draw upon for weapons work. That is one reason why any fusion process that does not emit neutrons is not given any of the research funds: weapons work doesn't need high-alpha-flux physicists. (Secondarily, they have papers that purport to show e.g. p-B fusion could never work.)
If we ever do get practical fusion, it won't be in a Tokamak, it probably won't be on the Earth's surface, and it certainly won't help resolve global climate disruption.
The money being spent on Tokamaks, on the other hand, absolutely could help a great deal with global climate disruption. But not while also maintaining the all-important high-neutron-flux population.
In the sun, isn't energy production occurring at something like 100-1000 W/m3? So, if you want to build a multiple MW fusion plant, shouldn't these plants be ridiculously huge compared to, say, a wind turbine rated at a couple of MW?
Is the density of the plasma so much higher in a fusion reactor?
Also, something else I never grokked, how do you get the power out? The plasma heats up, but how do you turn that into useful electrical energy?
Nevertheless of course I hope it does work as advertised... someday.
Edit: thanks everyone for the thoughtful, insightful replies!