Just to reiterate, Wendelstein 7-X is not about fusion. 7-X is a research project about controlling a plasma. It is not meant to evaluate whether energy production is viable, neither is it setup to do so.
> 7-X is a research project about controlling a plasma.
Which is the primary impediment to developing viable fusion reactors. So yes, it is absolutely about fusion. We're not trying to contain plasma because we think it'd be a neat thing to do. We're trying to contain plasma because it's necessary for fusion.
We already know that energy production is viable if we can solve the containment problem. Saying this project "isn't about fusion" is like saying "guns don't kill people, bullets do".
> We already know that energy production is viable if we can solve the containment problem.
No. That's only a small part of the problems. What is probably a far more difficult problem is the fact that any nuclear fusion which is remotely viable in this century will use a Deuterium-Tritium reaction. And tritium is an isotope which does not occur in nature - it is unstable. Therefore tritium has to be bred, using heavy water reactors. And lots, lots, of uranium. The idea is now to breed tritium, using neutrons from nuclear fusion, and further nuclear chemistry which is supposed to happen in a breeding blanket. But this runs into an additional problem, the deuterium-tritium reaction has no neutron surplus, as uranium fission has. And in addition to the very difficult task of exfiltrating traces of Tritium, one essentially needs a breeder technology (similar to Monju in Japan), and there are huge, unsolved material sciences problems, such as constructing a large structures which withstands very strong magnetic forces, very high temperatures, and a very high neutron flux, and this of course without any elements like carbon which will become radioactively activated.
Oh, and I forgot, the whole process is also unlikely to be clean, because of the breeder technology needed.
And all these questions are so extremely difficult, that today's large projects, like ITER, do not even address them.
More on these issues for example in the article by Michael Moyer which is contained in the page below (Michael Moyer, "Fusion's false Dawn", Scientific American 302, 50 - 57 (2010)
doi:10.1038/scientificamerican0310-50 )
Tritium can be bred in any sort of fission reactor. The US has recently used light water power reactors to breed tritium for supplying the US nuclear weapons stockpile.
Lithium-7 plus a high energy neutron yields tritium, helium, and a lower energy neutron. I don't know what a practically engineered system can do, but a 14.1 MeV neutron from D-T fusion formally has enough energy to transmute 5 lithium-7 nuclei into tritium and helium (each transmutation is endothermic to the tune of 2.466 MeV).
I agree about the huge unsolved material science problems. I just don't think that tritium availability lies on the critical path of unsolved problems for industrial fusion power.
| Tritium can be bred in any sort of fission reactor.
Tritium deliberately made in fission reactors would cost maybe $200M/kg.
A DT fusion reactor consumes 55kg of T per GW(th)-year. That would cost $11B/GW(th)-year if the T came from deliberate production in fission reactors, which is far too expensive.
Even producing the startup tritium for a fusion reactor using fission reactors is problematic.
There's a serious loop in the development schedule for DT reactors. They absolutely need good tritium breeding to make enough tritium, but such blankets cannot be developed without access to the large amounts of T that they themselves are needed to produce. Running around this loop is going to be laborious and painful, even after all the plasma physics issues are solved.
M.E. Sawan and M.A. Abdou, "Physics and Technology Conditions for Attaining Tritium Self-Sufficiency for the D-T Fuel Cycle", Presented at the Seventh International Symposium on Fusion Nuclear Technology, May
22-27, 2005, Tokyo, Japan; to be published in
Fusion Engineering and Design.
Abstract + conclusions:
"There is no practical external source of tritium for fusion energy development beyond ITER and all subsequent fusion systems have to breed their own tritium. To ensure tritium self-sufficiency, the calculated achievable tritium breeding ratio (TBR) should be equal to or greater than the required TBR. The potential of achieving tritium self-sufficiency depends on many system physics and technology parameters.
[...] It is clear from the above discussion that both
the required and achievable TBR values depend on many system physics and technology parameters. Many of these parameters are not yet well defined. In addition, the
rapidly decreasing tritium resources imply that the time window for the availability of tritium to supply fuel for the DT physics devices is closing rapidly. It is, therefore, necessary to establish without delay an extensive R&D program to determine the “phase-space” of plasma, nuclear, material, and technological conditions in which tritium self-sufficiency can be attained.
[...] Tritium self-sufficiency in DT fusion
systems cannot be assured unless specific plasma and technology conditions are met.
[...]
Page 32 of that presentation, "Tritium Supply Calculation Assumptions", says it is assumed that the following will NOT happen:
• Restarting idle CANDU’s
• Processing moderator from non-OPG CANDU’s (Quebec, New Brunswick)
• Building more CANDU’s
• Irradiating Li targets in commercial reactors (including CANDU’s)
• Obtaining tritium from weapons programs of “nuclear superpowers”
• Premature shutdown of CANDU reactors
But assumptions 3 and 4 have already been invalidated. China has added CANDU reactors since the presentation and is building a few more abroad for other countries. The United States has started irradiating lithium targets in commercial reactors.
So in my judgment, closing the tritium cycle still isn't on the critical path to fusion industrialization. There are other crucial problems to solve before "where will we get enough tritium for these fusion reactors?" becomes a crucial issue.
"without any elements like carbon which will become radioactively activated"
Carbon has very low neutron activation. Silicon carbide has been proposed as a material for fusion reactors for this reason. But otherwise your points are well taken.
I think that undersells it. While it is true that the 7-x is about proving the concepts of the Stellerator, you can use the plasma stream from such a device as part of an energy system.
Electric current; net momentum difference between positive and negative charges within the plasma. I.e. nuclei moving clockwise, and electrons counter-clockwise within the tokamak, or vice-versa. The plasma may have a high temperature (and thus induce fusion) without such current.
No. 7-X hasn't performed any fusion at all and while they later want to use plasmas that are closer in composition to what you'd find in a fusion reactor, very little fusion will take place — intentionally, to avoid generating significant amounts of neutron radiation.
It's perhaps important to point out that fusion is not difficult; it's very easy in fact. The hard part for power generation is plasma physics, i.e. being able to contain and control a plasma effectively. That's what 7-X is designed to research. If you can control a sustained plasma, fusion power is relatively easy.
Yes it has. It hasn’t broken energy breakeven but fusion is definitely happening in that plasma and, for that matter, plasmas at lower temperatures, lower densities and with lower confinement times. Getting particles to fuse is pretty trivial.
He was an inventor of a variety of television technology, and I believe the designs were very similar to his fusor. Unfortunately, the design cannot be made "exothermic" IIRC, and the research ruined him financially.
I am no plasma physicist, but the concern I hear about fusion is that whenever we scale a system up we find the plasma is behaving in new ways, resulting in new loss mechanisms.
Identifying and understanding those new behaviours sounds more like science than engineering to me. (That said, scientists frequently do engineering and engineers frequently do science; I've been on both sides of it).
It explicitly has not been designed to improve fusion efficiency, but to do plasma-handling with minimal fusion happening. This way, it needs a lot less shielding than a fusion-centric experiment would need, and is thus cheaper to build and easier to maintain.
that's not correct. It's definitely about fusion and to test a different concept as the tomahawk, the problem is that you need a self sustaining fusion and that happens at 100M°C. Of course there're lots of other experiments and interesting stuff happening, but it's also to find a viable concept to generate power (that's also why they get lots of european funding). There's a good interview with a scientist and a director from wendelstein 7-X but it's in german https://alternativlos.org/36/
Because the article is only about the mentioned achievement. The project page of Wendelstein 7-X (https://www.ipp.mpg.de/w7x) details what the project is about, though.
What does "At an ion temperature of about 40 million degrees" mean?
How does "ion temperature" relate to "regular" temperature?
Is it similar to how we can cool down things to fractions above absolute zero, by using lasers to reduce the movement of atoms? i.e. is the "ion temperature" a measure of how fast are they moving, as opposed to the every-day meaning of temperature as "how hot something is"?
I realize that "how hot something is" depends on how fast the atoms of said "something" are moving, but since they specify "ion temperature" I'm guessing it has a very specific meaning.
I searched for "ion temperature" but only found about Electron temperature [0].
Other replies have explained that the ions and electrons can be at different temperatures, and that is of course true. I can offer some more explanation why. While you have linked temperature to some average kinetic energy, this is not a strictly accurate notion of temperature. Temperature only means something in the context of thermodynamic equilibrium; for any system out of equilibrium, often a temperature cannot be defined, even as an average kinetic energy. The two concepts are linked to each other, in the sense that an average kinetic energy at the particle scale is a significant (and sometimes only, but not always) component of the internal energy of the system. But temperature is first and foremost an equilibrium concept.
A plasma is composed of both ions (partially ionized atoms, or if they are fully ionized, bare nuclei) and electrons. Together, they constitute two co-located but separate fluids, whose motions may be distinct; indeed, because electrons are so much lighter than ions, they respond to forces much more readily. In equilibrium, a single fluid must have sufficiently rapid interactions so that the particle distributions are driven to a Maxwellian. In that case, for a single fluid, a temperature is well defined. In the case of a two-fluid system, self-interactions (such as interactions between ions and themselves, and electrons and themselves) may be sufficient to establish two separate equilibria corresponding to each fluid; hence, the electron vs. ion temperature. Only in the case that ion-electron interactions are sufficiently rapid would those equilibria be driven together to a single-temperature fluid, in which case, Te = Ti.
This picture becomes rapidly more complicated at very high temperature (usually, around 0.1 keV or about a million K), at which point the photons being exchanged by hot charged particles become dynamically important, and a third temperature, the radiation temperature, may emerge. At this point, the plasma must be described using three temperatures - if, and only if, you are in a situation lucky enough for equilibrium to manifest. (Fortunately, equilibrium is not usually that hard to access.) In many cases, such as when the system undergoes a strong shock, the system may be driven very far from equilibrium, but only temporarily. In others, some underlying energetic process may continuously drive the system away from equilibrium, resulting in a metastable state; this is the case in stellar atmospheres, in which NLTE (non-local thermodynamic equilibrium) processes matter a great deal. Usually, physicists resort to kinetic theory to try to understand such situations.
"For this reason, the ion temperature may be very different from (usually lower than) the electron temperature."
So if the ion temperature of the plasma in the Wendelstein 7-X is about 40 million degrees...the electron temperature could be even higher?
What does it even mean to have a temperature of 40 million degrees? AFAIU, the Sun has temperatures much lower than that. That's utterly mind blowing...
The surface temperature of the sun is approximately 5,778 K (5,505 °C, 9,941 °F) [2].
I can guess what the insane pressure means ("estimated at 265 billion bar (3.84 trillion psi or 26.5 petapascals (PPa))"), but I don't know what millions Kelvin means. It should be kinetic energy. So maybe it means the ions vibrate very fast?
if it were a solid, vibration would be a reasonable guess, but unbound particles, the primary form of kinetic energy is basic translational movement, like a baseball's got most of it's kinetic energy in moving from the mound to the batter and a bit of it in how it's spinning.
Keep in mind the mean free path of translational movement, even at those very high pressures, is quite large relative to the size of a proton (even more so an electron).
High energy density systems are quite hard to develop an intuition for.
While average kinetic energy is one way to think about temperature, think about it instead in terms of the Maxwellian particle distribution [1]. In this sense, the Maxwellian is a parameter of this distribution, and as it increases, the location of the distribution changes. A high temperature corresponds to, in this hypothetical system, more particles with higher velocities in the system. (This technically only applies to certain systems in which the assumptions hold. For example, in systems in which other degrees of freedom matter, such as vibration/rotation of molecules, the interpretation is harder. At high temperatures, these degrees of freedom are destroyed, though others, like ionization, can matter significantly.) Of course, these systems are highly collisional, in the sense that each particle will probably go only a short distance before Coulomb scattering off another particle. So the particles are not really vibrating; the energy is in translational motion, but in the mean, there is no real directionality to it, so the particles won't typically stream out of the plasma.
This can be readily understood in the context of achieving controlled fusion. Nuclei are positively charged, and therefore exert forces that tend to repel them away from other nuclei. The nuclear scale, at which nuclear reactions must occur, is very small, and the 1/r* Coulomb potential is quite large at those distances. As a result, it is to be expected that an individual fusion reaction should only occur when the kinetic energy of an incoming nucleus is sufficiently strong to overcome the Coulomb barrier. (Practically, quantum mechanics effectively reduces the barrier through tunneling, but nevertheless it is quite high.) As a result, you need a sufficiently large number of nuclei with high speed to have an appreciable fusion rate, i.e. your temperature must be high enough. (Hence, "thermonuclear" fusion.) This has to be contrasted with your confinement quality, in the sense that you have to keep your fuel at that temperature and at a sufficient density (so that your collision rate is high enough) long enough for fusion burn to consume an appreciable fraction of your fuel.
In context, the temperature seems a simpler quantity to discuss rather than the pressure. It is easier to conceive of this pressure as a momentum flux, and in stars, gravitational confinement demands that the plasma pressure balance the crushing momentum flux of the weight of the star. Such static pressures are simply outside of our ability to intuit, and result in quite counterintuitive properties of matter.
Ion temperature is the important metric because it is ions that do the fusion, not electrons.
The sun has immense gravity and mass that produce very high pressures which permit fusion at 'low' temperatures. Maintaining high plasma pressure is difficult on Earth, so higher temperatures are used.
I also think that if you click on the flag at the top right to get the German translation, then let Chrome translate this to English, it reads a bit better.
> The objective of fusion research is to develop a power plant favourable to climate and environment. Like the sun, it is to derive energy from fusion of atomic nuclei
While plasma containment seems like an essential step towards fusion-based energy, is fusion-based energy a desirable goal?
Right now, fission-based nuclear energy is only used as a heat source for steam-driven turbines. It's a fantastically complex and expensive way to make heat.
Is fusion energy on Earth going to do something similar? And if so, is this source of heat actually going to be cheaper than using the sun as a distant fusion reactor? (Which is what drives both solar and wind energy on Earth.) It seems quite doubtful that fusion could ever be cheaper than solar or wind energy, the tech-to-energy ratio just seems massive in comparison.
So if it's not Earth where fusion energy will be useful, how about in space? Here on Earth, the Carnot cycle is used to convert heat into useful mechanical energy, then finally electrical energy, which is usually the true goal. But the Carnot cycle requires dissipating massive amounts of heat; just as much energy is dissipated as heat as gets turned into electricity. And heat dissipation isn't such an easy thing in space, is it?
I feel like I'm missing some rather large puzzle pieces here. Without them, fusion does not really seem like an engineering goal towards a good energy source, just engineering research into better containment of difficult materials. Could somebody help enlighten me?
There are other potential ways to extract electricity from a plasma. Since plasma is high-energy charged particles, you can extract voltage directly from their kinetic energy. https://en.m.wikipedia.org/wiki/Direct_energy_conversion
Aha, that's exactly the term I was missing! Thank you so much.
In particular, the gamma radiation energy capture in that Wikipedia article addresses another somewhat undesirable aspect of fusion in a really great way.
The hope is that fusion will, indeed, be less complex and expensive than fission. This is for a few reasons:
1. The fuel is much more common and easier to handle - uranium and plutonium are a pain to extract/transmute, handle, and package.
2. The radioactive contamination is limited to induced radioactivity in the structural materials of the reactor core; there aren't the safety concerns of radioactive neutron moderators and the like that you see in a fission reactor. Preventing the release of those materials in the case of accident is one of the big drivers of fission plant costs.
3. The core itself is fail-off, again reducing the cost and complexity of safety measures.
4. They are not dual-use (civilian/military) to the extent of the fission supply chain.
5. By comparison to the rather indirect method of using solar panels and wind turbines to extract energy from solar radiation, steam turbines are a simple and efficient way of extracting energy from a fusion reaction. The Sun is enormous, but a fusion reactor would be right here and we can extract energy from one with well-understood, efficient technology.
Such things are often claimed. But every single statement here is false:
ad 1) Nuclear fusion as we know it does not rely only on deuterium, it needs tritium. Tritium needs to be produced by uranium fission, and after having a hypothetical fusion power plant running, by a breeding process. The technology of that breeding process is today completely unsolved.
2) Fusion reactors, if they aver will exist. will be large and will have a very large neutron flux. The will have breeding blankets which need to be several meters in size. Anything within these structures will be activated by neutrons. That means common materials such as carbon steel can't be used, because the carbon would become highly radioactive. Also, the neutron radiation weakens all known materials so that the structures will probably need to be periodically replaced. This alone will probably lead to a higher amount of nuclear waste than uranium fission plants.
3) It is correct that the fusion reaction will stop to operate when the core is damaged. But this does not mean that the plant is safe then. It will contain substantial amounts of tritium and other radionuclides. By the way, you may have seen nice photos of scientists in white coats which stand in the vacuum vessel of a a fusion power plant. Except that, if there were actually fusion to happen in these devices, nobody could enter them before they have been decontaminated in a time-consuming and expensive process. Consequently, in experiments such as ITER, no fusion happens at all.
In addition, a fusion plant will be very complex, and especially it will work with extremely large magnetic fields which have by themselves a tremendous energy content. If there is any electrical or structural failure, it is likely that the whole structure just blows apart. And it will not safe to be standing nearby.
4) As said, tritium is a key ingredient in the fusion technology that is aimed for, and it also happens to be a key ingredient of thermonuclear weapons. In addition, the required breeding technology is also highly relevant for military uses, because if you can breed tritium from lithium, you can also breed weapons-grade uranium and plutonium.
5. Steam turbines are not physically simpler than wind turbines. Apart from that wind power converters, as well as photovoltaic energy converters, can come pretty close to the physically possible optimum, there is another problem: Large steam turbines need cooling for thermodynamical reasons. Lots of cooling. The amount of cooling required for today's uranium fission plants is already a problem. For one, they cannot be built in arbitrary places - they need to be at rivers or at the coast. As the fukushiuma plant has shown, having a nuclear plant at the seaside can be a bit risky. But cooling by rivers also has costs, heating up a river to much messes up its ecology. Also, France often has energy problems in the summer, and the reason is that the rivers are too hot to cool its numerous nuclear plants.
Here another article on further problems of nuclear fusion - see the article of Michael Moyer.
As a closing remark, I am not saying that fusion power plants are impossible. Sometimes, technological break-troughs happen, similar to the advances which made airplanes and aviation possible. But today, we don't even know what is required to make it possible. We are in a similar position like people from 1800 were in respect to aviation. And also, we have more urgent problems to solve - we need clean energy within the next decade, not in the next century. And fortunately, we do have the technology for that, we just need to put it in march.
1. No. The current bulk of research focuses on this pathway because d-t is both energetically favorable, and occurs at lower temps than other energetically favorable pathways. Tritium is a key component of fusion as it is currently studied in many, but not all cases and is the most likely first reaction. If you had said that the bulk of current research focuses on d-t, I’d agree, but it’s definotely not “nuclear fusion as we know it.” D-3He is a thing, not to mention stellar fusion which we certainly won’t create in a power plant, but we do know about it.
2. No, although the kinds of reactors being dreamed of in the “short” long term do fit this bill. Again, the much more difficult to manage (not to mention fuel) D-3He reaction is aneutronic, and doesn’t require a breeding blanket. Another difficult reaction would be p-7Li, although the fuel would be easier to come by. I’m not claiming that we’re remotely capable of harnessing these reactions, but it’s important to point out that they exist. A high neutron flux, and the need for Trtium are not inherent to fusion in general.
3. A bigger issue would be sputtering from the neutrons and He infiltration of the vessel. Othereise, assuming d-t fusion... this is accurate.
4. We can already breed tritium, and plutonium... this is a reality we have to live with. Fusion won’t change that either way. We can also engineer plagues, and a lot of other nasty things, and every year it all becomes easier to do “out of the garage” so to speak. If we’re going to survive, we need a better solution than unspilling the milk. Maybe embracing fission energy and greatly raising quality of life around the globe would be a good first step?
5. Agreed, although it is worth mentioning that there are ways to harness energy without a turbine, they are also complex. Fusion won’t save us, it won’t save our kids or grandkids. We need fission, solar, and wind, and peobsbly a century or more of research into aneutronic fusion if we want to survive well.
You point out that alternative reactions like D-3He are theoretically possible.
But my argument is not that D-T (Deuterium-Tritium) fusion is theoretically impossible. It is practically impossible. While it is possible to generate very small amounts of Tritium in commercial reactors, a fusion power plant would need much much larger quantities before it could even be started. And while the D-D reaction is also theoretically possible, it is even more difficult to achieve practically.
This is similar to aviation in 1800: While it was theoretically possible to build a flying machine, there was, for example, no known engine which would have been able to power an airplane. And for somebody to say "we just go ahead and will solve the technical problem of having a low-weight power source within 20 years" would have been delusional. You can't summon technological break-throughs like a mythical ghost in a bottle.
So to sum up, your argument against that todays nuclear fusion research, which is dominantly based on the D-T reaction, will not lead to any energy reaction in foreseeable time, is that there are more exotic, more difficult, end less well understood reactions which are also theoretically possible but not even close to being practically usable in any way.
Just a layman. But wind and solar have big limitations. They are available only in a limited amount only under certain conditions and their variability isn't under our control. And if you think energy storage would solve all this, batteries aren't cheap. We would like to have an infinite source of energy that we can control on demand.
> They are available only in a limited amount only under certain conditions and their variability isn't under our control.
If one connects wind turbines to an electrical grid in a larger area, the variability averages out.
Battery storage is being discussed. But what is probably more economical is that a significant proportion of energy usage is somewhat flexible in time. Take hot water generation. It is expensive to store the amount of electricity to bring 100 liters of water to boil, but it is very easy to store 100 liters of hot water for a few days. Heating and cooling make a very big proportion in modern energy use.
And this isn't a new idea at all - all of human history, our energy management was around making use of a surplus of energy, like food, for times when less was available, and using storage.
Not an expert by any means but I think the benefits of fusion power lie in the future: Namely as a power plant for interstellar exploration. The energy received by the sun might be enough to power infrastructure on earth but planets and space ships further away might need their own power source.
Fortunately the ITER project does wast amount of R&D and basic engineering that solves problems common to both design. If the problems in stellators are solved and it proves to be superior to tokamak in practice there is no need to start from scratch.
There's also quite a bit of talk going on between scientists on both projects. If I remember correctly, the microwave heating design of Wendelstein 7-X also solved a few problems for ITER and in general it's more of a friendly competition.
I view it as less of a competition than as projects working on different parts of the same problem - ITER seems to be working more on the problems of managing a fusion reaction (materials, procedures, shielding, etc.), while using a better-studied containment design with lower technical risk. Meanwhile, Wendelstein is ignoring all those other challenges in order to study an alternate containment design in isolation.
Not really. Wendelstein is not for practical usage of fusion. It's only there to make a point that the Tokomak design is inferior, and all the money should have went to the germans instead.
But thanksfully both teams are working together. We will see if something better will come out of it. Even if I'm extremely sceptical on both.
Whenever you're shown a new gosh-wow fusion design, ask "What is the power density of this design?"
The power density of a PWR fission reactor core is 100 MW/m^3.
The power density of ITER (gross fusion power/volume inside the cryostat) is maybe 0.05 MW/m^3.
There are fundamental engineering limits to fusion that guarantee the power density will suck compared to fission. So my default take on any fusion design is that it will never be competitive with fission, never mind the things that are cheaper than fission.
So where are they along the project plan to useful fusion energy? Since that is not the end goal of this project, how far along the project plan for this device are they?
I'm really well past the phase of being excited about milestones like "record for a stellarator". Put it in context, show where we are on the plan. Perhaps it's behind, but still moving forward - that's fine. Confirming things work as predicted is awesome and tells me they should be moving forward at a good pace right? So where are they on that gantt chart?
Nowhere close, but as you rightly stated, that isn’t the goal of this experimental device. For what it is, which is an exploration of the dynamics of containing high temperature plasmas, it is a resounding success. What you’re asking is sort of like someone asking Babbage how close he is to a Pentium 90. Babbage still did some incredible and foundational things, and it was still a world away from playing Doom.
There are a lot of issues outstanding for fusion to be a viable power source, and this reactor is simply exploring the dynamics of one of those issues. Putting aside hype and desperation, there is no chart yet. For a fusion plant to work and compete with any other plant, you need it to have good uptime, relatively low maintenance burdens in terms of downtime and costs, good output, a viable source of fuel, a fusion cycle that is safe and sustainable, and more.
As of today the d-t cycle isn’t sustainable because it requires tritium, produces a lot of energetic neutrons that along with helium will undermine the reactor. The problem is that more viable fusion cycles occur at significantly higher temperatures and it’s a struggle to control plasma at d-t temps. The lack of control both quenches the reaction, and exposes parts of the reactor to plasma which erodes the material. Neutron bombardment causes the metal to become brittle, and helium infiltrates and undermines it too. So a reactor would be down for maintenance a lot, which would make it difficult to work as an economically viable source of power.
As of today the tritium for the reaction has to be bred in fission reactors. There are theoretical plans to use a “blanket” impregnated with an isotope of Lithium to breed tritium within the fusion reactor, but so far no one has gotten it to sustainably work near necessary levels. As a result a fusion reactor today would require fission reactors, and it might make sense to ask why we don’t just stick to the far more mature and efficient fission technology.
There are lots of other issues that aren’t as big or obvious as what I’ve described, but ITER and desperation hype aside, there are more unsolved problems than solved. So where this reactor is concerned, be excited for what it is away from the context of “fusion any day now!” hype. Appreciate the slow, but steady progress in the context of incremental research.
If you want clean power in your lifetime, support fission plants.
Mostly agreed, but: "If you want clean power in your lifetime, support fission plants."
Not anti nuclear or anything, but you really want a large share of wind+solar. Less hard to handle waste and cheaper. Once we are approaching 80% wind+solar we can discuss how to best deal with the difficult remaining 20% (load shifting, expanded transmission grids, storage, CCX and Carbon2Gas gas plants and nuclear are all options that are possibly going to play a role).
They're only cheap and low-waste if you don't have to worry about storage, and that needs to be solved well before the 80% point. Battery manufacture is pretty dirty, and even if the environmental issues were not a concern, it's not clear that enough raw materials (rare earths, etc.) exist in the world to move the entire power grid to intermittent sources plus battery storage. And so far, no alternative storage sources have proven themselves broadly viable (pumped hydro is probably closest but requires fairly specific climate and geography).
Personally I'm with Bill Gates on this one: we basically need an innovation miracle to solve our energy and climate problems, but probably just one, and it can be in any of four or five different fields (renewables + storage, fission, fusion, biofuels, carbon capture, etc.). Success in any one of those spaces is far from guaranteed, so we shouldn't put all our eggs in one basket, and should be aggressively pursuing all of them in the hopes that at least one will pan out.
Could you please point me to reputable studies that back up your claim that storage will be needed before very high penetration?
Even for extremely high carbon reductions (95% on 1990) the best models I've seen consider transmission lines and sector coupling considerably cheaper. And these models make very aggressive assumptions on the fall of battery prices. E.g.:
Edit:
The thing to look at is figure 11. on page 16. It gives a policy trade off. Given political limits on the amount of transmission grid extension (x axis) what is the cost of the economically optimal mix of technologies (y axis). Left graph shows sector coupling, right graph shows sector coupling. The second graph almost completely eliminates the need for battery storage.
Nuclear is not in this mix, partly because this is exploring policy constraints on transmission capacity in the context of Europe (and nuclear is pretty much a non-starter politically) but it clearly shows that the idea that large amounts of battery are inevitable is outdated.
I'll leave you with a quote from the conclusions:
"The All-Flex-Central scenario with optimal transmission costs just 13% more than today’s system [even when excluding health benefits]"
You're right that arrangements that mostly avoid storage are possible, but might not be cost efficient in a way that leads to their eventual adoption, and any mass adoption of renewables at high percentages of overall capacity requires some kind of major investment beyond the cost of generation itself, to deal with the intermittency, be it storage or transmission (and the latter is really only helpful if by "renewables" you mostly mean wind -- the intermittency of solar is obviously highly correlated over broad geographic areas). Still, I think the more interesting question isn't "could it be done" -- it's pretty clear that it could -- but "could it be done in a manner that's cost-competitive with non-renewable technologies," and there's pretty abundant research at this point that shows that any advantages renewables have in terms of levelized cost of ownership get more than wiped out once these kinds of investments are factored in; see, e.g., https://www.lazard.com/perspective/levelized-cost-of-energy-... From the executive summary:
> Despite the sustained and growing cost-competitiveness of certain Alternative Energy technologies, advanced economies will require diverse generation fleets to meet baseload generation needs for the foreseeable future
with specific numbers in the actual report.
It's easy enough to focus on Europe or the Americas where the rich have the comparable luxury of being able to overpay for generation if the political will is there, but the overwhelming majority of new generation capacity brought online globally in the next century won't be in these places, it will be in Asia, Africa, and South America, as these countries work towards developmental parity with the West. We need to come up with solutions that aren't just possible, but cheap, and deployable in places that are starting from basically nothing in terms of modern grid infrastructure, and where hyper-inefficient diesel generation is currently the norm, to the extent that electricity is broadly available at all (as is the case in large parts of Africa in particular today).
Again: this isn't to say that renewables can't work, or that any other particular technology will or won't be the solution. We just don't know yet, both in terms of feasibility and economics.
The study I linked to looks at nothing but the cost of dealing with intermittency and comes to the conclusion that it's barely more expensive than the current system.
Also, once you go into the developing world context a lot of things change. What is difficult and expensive in the European context is to guarantee the last few hours of electricity a year. If you can take an area that has a few hours of electricity a week to that point, you have a massive win already. Plus you can build sector coupling in from the beginning.
In a green field scenario you _really_ want wind/solar as a decentral bootstrapping technology that scales down, that can get you a really long way.
Bangladesh has >5 million solar home systems installed. That's a Norway or a Denmark population running 100% on decentral off-grid solar.
Edit:
An important point here is to be aware of time horizons. A study looking at a 2025 or a 2030 system looks very different from one that looks at a 2050 system. If the foreseeable future means the next ten years, then sure.
> We need to come up with solutions that aren't just possible, but cheap
Renewable energy sources are already cheaper than traditional nuclear energy, in many parts of the world already cheaper than fossil sources, and rapidly getting even cheaper:
And at the same time, Westinghouse had to file for chaper 11 bankruptcy because ... well, nuclear power does not seem to be so economical at all.
As an interesting side note, both nuclear power as well as renewable energy sources have the cost structure that almost all investments are up-front, while the relative amount of running costs is very small. Because in a market system, the marginal cost of production per unit determines the market price, and the market price is therefore close to zero, both technologies have the problem that they actually need subsidies and incentives to be created. In other words, while renewable power sources definitively need incentives, nuclear power also can't exist without huge subsidies.
> Renewable energy sources are already cheaper than traditional nuclear energy
Again: only when looking at the cost of generation alone. It's more expensive if you factor in the storage and/or long-distance transmission necessary for high utilization (even the study I'm responding to says so, though by a smaller amount than I've seen elsewhere -- ~13%).
> well, nuclear power does not seem to be so economical at all.
Nuclear power isn't economical under the current regulatory environment and set of political realities. There's no fundamental reason that that need be the case. More people die from wind and solar per year than nuclear (mostly installers and technicians falling off of things), and obviously both are dwarfed by orders of magnitude by coal once externalities are factored in. If we were as risk-averse with those sources as we are with nuclear, they would be expensive too. Coupled with the slowness of construction eliminating economies of scale or effective market competition, and you don't have a great situation.
(but wind energy was not the primary cause, the main cause was bad planning)
What happens at large scale is that the fluctuations induced by variable wind speeds smooth out over larger regions. And having a large, interconnected grid is usually much cheaper than battery storage.
What would also help is diversification. In Scotland, there was a fascinating project to generate electricity from wave power, the Pelamis wave power converter. It had working 500 kW installation but was scrapped then.
> If we were as risk-averse with those sources as we are with nuclear, they would be expensive too.
A wind power plant blowing up will not cause half of Europe to be contaminated with huge costs to agriculture, like it happened in 1986. You must also not forgot the extreme health costs of uranium mining.
There are papers claiming that the various 100% renewable work by Jacobson, Breuer etc. suffer from a number of wildly optimistic assumptions. See e.g. doi:10.1016/j.rser.2017.03.114 . Also see work by Christopher Clack et al. (one of which caused the infamous Jacobson lawsuit, FWIW), and Jesse Jenkins et al.
Which I find considerably more convincing, given that it addresses the concerns raised in detail and gives a much wider overview of the literature and ideas around.
In my view the response article actually demonstrates that Heard et.al. had an ideological motivated conclusion they wanted to get to and chose evaluation criteria to fit their desired outcome.
> Then you are no doubt aware of the response article:
Yes, I am. I'm not particularly convinced by it, it seems to largely repeat the same claims made previously. Anyway, there's a preliminary informal response to that at http://4thgeneration.energy/response-to-brown/
> In my view the response article actually demonstrates that Heard et.al. had an ideological motivated conclusion they wanted to get to and chose evaluation criteria to fit their desired outcome.
Clearly Heard is a nuclear advocate, yes, but AFAICT the criteria they chose aren't unreasonable. OTOH it's hard to argue Jacobson, Breuer, Brown et al. aren't pushing an ideology either, since they claim to be motivated by decarbonizing the energy supply, yet they are excluding one of the very few sources which, historically, has provided large-scale low-carbon energy.
To make Fusion work, there is a lot of unproven technology required.
For renewables not.
And the storage is a problem yes, but not one that requires miracles. Just investing.
(And there are much more technologys around than just batteries. Air-pressure storages for example. Or hydrogen. Or different H2 based ones like Methan, etc.
or artifical lakes for water pump storage, etc.)
But to make Fusion work any time soon, there are indeed miracles required. If it works one day, nice. But I would not bet on it, to solve any problems we have today.
> To make Fusion work, there is a lot of unproven technology required.
For the critical issue of tritium breeding, it is more like non-existing technology. What is meant to bridge the gap between today's plasma physics experiments and working power plants includes a large amount of fairy dust. It is not a valid comparison to compare that against the advantages and disadvantages of the existing renewable sources. Even the Pelamis wave power plant (which was discontinued for whatever reasons) is far far closer to large-scale energy generation than any nuclear fusion process could be in the next 50 years.
I think this is incorrect. Renewables + Storage is possible now. It's only a matter of price. At marginal cost right now solar+wind are cheaper than alternatives, at total system costs it's within a factor of two (or at break even for optimistic studies) of the current system and falling.
Fusion is not possible now. As climate change is happening now, any delay in transitioning to clean and available solutions is purely ideology driven.
This too is a political will problem. Designs for proliferation-resistant gen IV nuclear plants have existed for a long time, and many have been tested. We don't use them because nobody builds nuclear plants any more out of overblown fears of meltdowns (which, incidentally, also can't occur in most gen IV designs).
The "can't melt down" property is usually referred to as "passive safety": https://en.wikipedia.org/wiki/Passive_nuclear_safety . These tend to have components that cause all of the nuclear fuel to passively flow out of the reactor and into a cooling chamber in the event of coolant failure. In liquid-fueled designs, this is accomplished via a "freeze plug," essentially a cork made of a low-melting-point material at the bottom of the reaction chamber that's kept solid by active cooling and rapidly melts in the event of a power failure such that the reaction chamber drains. Equivalent mechanisms exist for pebble bed reactors, though, where all the pebbles fall into a cooling chamber. In either case, production reactors have been build that include these features and the physics are very well understood.
Proliferation resistance is trickier, and not all gen IV designs focus here. There are a couple of areas of attention here. The first, and probably more mature development-wise, are alternative fuel cycles like the Thorium fuel cycle that don't produce easily usable fissile material out the other end (and do produce a bunch of U232, which in addition to being non-fissile is also difficult to steal because it's super-dangerous to handle). Secondly, there are designs that breed and then immediately burn fuel in situ without reprocessing, such that there's no point during which the fissile material exists outside the reactor to be stolen. I don't think any of this class have actually been built yet, but the Traveling Wave design is probably furthest along, and TerraPower is building one of those in China with a target completion date of ~2025.
Two important caveats though: these are "proilferation-resistant" in the sense that fuel would be hard for non-state actors to steal; state actors are another concern as that article points out, but they also don't really need to breed fuel, and can just enrich uranium directly without that much difficulty, as Iran and North Korea have both demonstrated, if they're willing to pay for it. At this point the physics are very well-understood, so this is a problem in need of diplomatic solutions more than technological ones. And second, the focus here is on material for fission weapons. I don't think any nuclear technology has good defenses against using material for dirty bombs.
All of that said, I'd still stand by my original point that newer designs are well-understood and dramatically better in these respects than most currently operating plants, and are only not being built (at least in the US and Europe) for political reasons.
I am extremely skeptical about such claims. One such alternative design was the THTR design, which was implemented in the Jülich AVR reactor. It was also claimed to be passively safe.
But there were incidents when the plant was basically out of control. Worse, the design is based on graphite spheres which contain the fuel. That is only safe as long as the spheres in the hot reactor do not come into contact with air, which will cause them to burn, or with water, which will form hydrogenium-oxygen mixtures. Burning graphite was both a main ingredient in the Windscale fire, and in the Chernobyl disaster. There were also important mechanical problems with the spheres. In retrospect, these claims for passive security were unwarranted and the plant was dangerous. With the experience from the AVR, one can also say, that the nuclear industry is not transparent at all about safety problems. Also, it is not only very hard to make such plants inherently safe, if this also very expensive. Unfortunately, this conflicts with the goal of every company, which is to make a profit which is as large as possible.
The AVR was also not a Gen IV reactor (about which the parent/gp claims were made), and was built over 40 years ago. In fact construction began in 1961, and it was commissioned in ‘69, so really it was tech from 57 years ago that was put into practice 49 years ago.
"All of that said, I'd still stand by my original point that newer designs are well-understood and dramatically better in these respects than most currently operating plants, and are only not being built (at least in the US and Europe) for political reasons."
Well your original point was written a bit more absolute...
With this I go along. And as I said, I also prefer fission as the short term solution. But only as a transition on the way to fully renewable (or allmost full, I am for pragmatism).
As there are just many problem involved with nuclear power, like danger and waste and the need for uranium, etc. that you would not have with solar energy. So this is the goal for me, fission power only as a way to get there or where there are not really other options. (submarines, spacemissions, etc.)
I’m not in the US, but I do remember. The thing is I remember other disasters relating to energy, including yearly deaths from pollution. It’s not as though gas, oil, and coal production or use are somehow safe, free from catastrophic failure, and subsequent deaths.
Solar thermal that stores energy in molten salt sounds like a cool technology that can drastically reduce the need for batteries. You could probably set up large solar thermal plants in the Sahara and power Europe via HVDC lines without needing many batteries.
Now look at some of the figures in this slide deck:
https://www.svensktnaringsliv.se/Bilder_och_dokument/mattias...
(in particular, slide 3, showing rate of added generation/year for various countries, comparing nuclear to wind+solar+geo+bio energy, compared to the needed rate of addition of carbon-neutral energy to hit 2 deg C warming goals.) These data are also sourced from the BP statistical review and the World Bank.
If you trust that information, evaluate for yourself whether or not we can reach 80% penetration of wind and solar in time to avoid anything short of catastrophic warming.
If your response is to extrapolate current trends in cost reductions for wind and particularly solar, please bear in mind that there are no indefinite exponential growth processes in the real world where transfers of matter and energy are concerned, only processes that show logistic growth. And we can't necessarily predict when we'll reach the plateau phase of logistic growth.
What I conclude from that, personally, is that we shouldn't gamble on continued trends in solar and wind adoption to get us to decarbonization.
Adding nuclear is more expensive than wind/solar _right now_. Not in a hypothetical future but as of this very moment. We know because the UK government has decided to subsidise a new nuclear power plant as a low carbon must run for the future. That is on track for generating electricity in 2025. So I'm not exactly buying the "added capacity per year" stats there as terribly relevant for now.
China is ramping up investments in renewables (including crucial grid strengthening measures) and hasn't approved a nuclear power station in two years (as of beginning of this year, I haven't heard if it has resumed building them now).
Again, I'm not ruling out that nuclear has a role to play. But most of the lobbying for nuclear seems to be based in an instinctive dislike of wind+solar rather than in facts. Nuclear is terribly expensive and there really is a waste problem (even if CO2 is clearly the more severe problem).
> Adding nuclear is more expensive than wind/solar _right now_
Indeed, wind and solar can produce power very cheaply nowadays, which is awesome. We should definitely build more of them. OTOH, due to their variable nature, their value to the grid reduces as their penetration increases, so it's a race of declining costs vs declining value due to increasing penetration.
Nuclear, being somewhat dispatchable, doesn't suffer from this.
Given the magnitude of the climate crisis, IMHO we should build about every low-carbon source we can, as fast as we can. Including wind, solar, and yes, nuclear.
> China is ramping up investments in renewables (including crucial grid strengthening measures) and hasn't approved a nuclear power station in two years (as of beginning of this year, I haven't heard if it has resumed building them now).
Additional investment in transmission lines and sector coupling is politically more viable and cheaper, and already that is stalled. So I'd rather push for that. (Plus we avoid a massive scale waste problem). Let existing ones run as long as possible, and if China or other countries want to go for the nuclear option, great! I'm simply objecting to the idea that that's the only/easiest/optimal option.
The variability of renewables is of course a well studied problem [1] and exactly why, after a certain point, we need transmission capacity (or, more expensive, local storage). The variability averages out on large scales.
Honestly, I don’t even like nuclear power. I just see it’s track record in expanding availability of carbon neutral energy sources. To me, going all in on solar and wind and potentially running into scaling issues down the line after abandoning nuclear energy seems more risky than the waste issue in nuclear power. Also, as I noted in another reply, I think we need a carbon tax to correct our accounting for the different energy sources. After that, if we start lots of different experiments and find that there really are no issues scaling wind and solar, I will be overjoyed at being proven wrong.
A global carbon tax in 1990 would have avoided a lot of the mess we find ourselves in. As it is we don't have the luxury of implementing one. Even something as weak as the Paris agreement hinges on one election in the US. A globally enforceable carbon tax is much much less feasible than universal cheap fusion power.
That said, we are in the middle of the experiment you are talking about. Renewables are cheaper now than new nuclear, despite nuclear having received a massively larger amount of subsidies over the years, there are still many parts of the world that are building nuclear so it's not like we're losing the technology, but really, there doesn't currently seem to be a good reason for building new nuclear.
Let me put it another way, a relatively small island like Ireland is going to be capable of going 100% wind/solar relatively soon. They are not strongly connected to the EU grid. Smaller disconnected islands already have gone 100% renewable. So we will have plenty [1] of test cases of increasing size as we ramp up renewables. Conveniently, electricity is also much more expensive on small islands, so it's also economical to test every technology you want to bring to maturity there first.
If we find that beyond a certain size it doesn't work we'll have a decade of warning before we hit that wall with the big continental size grids.
You make good points. I just don’t think we have the luxury of leaving ourselves in a position where burning fossil fuels might remain more economical than carbon-neutral alternatives. A carbon tax seems the only way to get around that. I agree that a carbon tax remains unlikely, which is why I spend most of my days in a depressed and anxious haze for the future :)
But maybe a group like the Citizen’s Climate Lobby can pull something off (they suggest using the revenue of the tax as a monthly dividend for all citizens. Never underestimate the power of a monthly bribe in generating consensus in the American public).
My bet is on pursuing everything that gets us closer to decarbonization, but it appears there is the most potential in the immediate future for fission. The challenge there is of course economics, which is why I think we need a carbon tax to add the future impact of climate change into our accounting for various technologies.
There is a chart here from the same site (which would have been useful for them to have included in the main article) which shows how this device theoretically compares with others and with the end goal:
Why is it unclear? AFAICT, it lies outside the shaded region (just below the bottom left part). The scale is in the chart is a little unclear, but all the data is in the article.
Can we stop with the "fusion will always be x years away" meme? Fusion is mostly a number of invested dollars away. A big reason why we don't have fusion reactors yet is because nobody invests sufficient amounts into the necessary research. Check this graph http://imgur.com/sjH5r from the old Slashdot interview
> You might say that we’re not a certain number of years away from a working fusion power plant, but rather about $80-billion away (in worldwide funding).
That's a surprisingly small number for the impact nuclear fusion could make. I wonder why we haven't seen any billionaires taking up the challenge, even though this might be more impactful than space tourism or people on mars.
For context, that's over twice the size of the Gates Foundation. You have to compare that to all other possible investments, including energy startups that could pay off a lot sooner.
Note that SpaceX is set up as a business, not a charity; Musk may talk a lot about Mars but that's not really where the money goes.
The money funding SpaceX could hardly be allocated towards sustainable colonization of Mars in a more efficient manner. If you can think of one, please email Elon.
Sure, SpaceX has good business plan. That's the point. Building rockets only for a future Mars voyage and not making any money along the way would be a bad way to do it.
By contrast, fusion will only be a scientific research project for many years and many billions.
Time/budget predictions are hard, especially for cutting edge research. So predicting required funding until fusion is profitable sounds unreliable too.
The bottom like is that fusion energy is now engineering problem and not basic science problems. Engineering risks are more manageable and predictable.
It's like Manhattan project, you can do it if you want it bad enough.
Fusion has fundamental engineering obstacles, not just unsolved problems.
The basic problem is that heat transfer limits guarantee that fusion reactors will be big and expensive (meaning: power density an order of magnitude worse than a fission reactor core, and probably worse than that.)
The core of a PWR fission reactor that one can build today has a power density of 100 MW per cubic meter. The power density of ITER (dividing the gross fusion power by the volume inside the cryostat) is 0.05 MW per cubice meter, 2000x worse. Other fusion designs aren't quite as ludicrous, but will still be much worse than fission.
The problem is that heat in the PWR has to get out of fuel rods that are 1 cm in diameter. The heat in a fusion reactor has to get out of a plasma vessel that is meters in diameter. The ratio of surface area/volume is orders of magnitude higher for the fission reactor.
that spot on. IMO Its mostly a funding issue. My bold prediction is that we will have net energy from fusion within a decade of the point where market Begins to price in the ultimate demise of fossil fuel based energy. sooner the FF projects start getting counted as stranded resources because of EVs & renewable quicker the money starts to flow into fusion research.
the other rapid development scenario would be some sort of fusion arms race between asia (ch/kr/jp) and the rest of developed world.
You have to admit that there is something to this meme, It seems like Fusion suffers from a lack of focus and direction. Ive heard all sorts of horror stories about physicists and engineers going off on fusion design tangents that pull funding away from more viable designs as far as energy-in/energy-out.
I understand that there is a level of experimentation and R&D that is guaranteed not to generate viable results but at some point you need to trim the fat, get everyone moving in one direction etc. This grand project has been going on for over 50 years at this point, It seems like we should at least have one design that everyone should be focusing on. If they had real, defined commercial goals (even if they miss the goals) it might be easier to secure funding because your average politician would see the benefit to them.
I'm not a nuclear engineer so take that all with a grain of salt.
If you spare half an hour to Google, you can find out what we want collectively;
1) Find global spending on fusion research and note it down.
2) Pick some other stuff to compare it to. Like global annual military budget. Annual profits of the largest companies and what they sell to make it. Or how much do we spend on 'entertainment'. Or popularity of Kardashians(given even I know that f#cking name...) or whatever.
You'll be quick to realize that the fusion is sitting in the corner, waiting for us to "want to have it".
It makes me feel depressed to see on every fusion article in 2018, there is a paragraph allocated to explain how fusion is different from fission and how it is not... basically not a bomb.
The topic feels home in HN, but I do not know if you ever brought up fusion in a talk with family or even among your young friend circle. I certainly have engaged in that experiment myself. People do not even know what the heck is fusion, let alone lobbying to fund it with tax money. In this state, I say, the advancement on fusion is rather stellar, I congratulate wholeheartedly anyone putting sweat and money in it to bring things to this level.
I've been following fusion research on and off since I've been a little kid. It's always been far in the future. Back then it seemed like I could actually see a fusion reactor when I'm an old man. Now, 25 years later, the rate of progress generally feels like the date has shifted to "not in my lifetime". It's kind of a weird feeling.
Edit: Addendum: I vividly remember images with visualization results for numerical simulations on the reactor vessel shape for Wendelstein 7-X. That was back in 1992 or so. This just shows how long this experiment was in the making.
