The Truth About Nuclear Fusion Q Factors
backreaction.blogspot.com
backreaction.blogspot.com
Years ago I worked for a brilliant physicist who believed (usually correctly) that any factor less than an order of magnitude was just an "engineering problem".
It think the summary on fusion is:
- it's the most amazing energy source we have lined up in the mid term
-ots almost certainly going to work
- we are making meaningful but slower progress than we would like
- it's dramatically underfunded
But i do not understand the general approach to plasma-vessel construction, in particular, why it is so passive. Its either "bend" to orbital shape dicated by the plasma physics properties when constructing (Wendelstein, Stellerators etc.).
Or increase the containment vessels magnetic field strength (scaling up the reactor, high-temperature-superconductors) until it is capable to hold all the chaotic increasing fluctuation inside no matter what.
My Question: Is there not a more active, not brute force approach for containing the plasma? Sort of "Whack-The-mole" countering the escape-events as they happen in a Tokamak?
If i understood correctly, the system adds to itself constantly energy (its a reactor after all) and any chaotic butterfly wing flapping on the plasma current can shape up to a tornado breaching the containment and ending fusion.
I think General Fusion's approach is interesting, they're compressing the plasma with a liquid metal wall.
control of normalized plasma pressure: https://papers.nips.cc/paper/2019/hash/7876acb66640bad41f1e1...
plasma profile transport modeling: https://iopscience.iop.org/article/10.1088/1741-4326/abe08d/...
hybrid dynamical modeling of gross plasma quantities: https://arxiv.org/abs/2006.12682
uncertainty quantification for plasma dynamics: https://arxiv.org/abs/2011.09588
It's still early days for this work and for us but we're looking at pushing reinforcement learning in methods and engineering to solve this problem
Have you considered antagonistic training? One AI tries to destabilize the proces, the other trains not against a simulation, but against the destabilizing input and a succes-metric?
- the dataset is a mess. The experiments that have been conducted on the tokamak that we have access to were done for very many different reasons and under many different configurations of the machine so there is not a clear method for disambiguating what dynamical changes are due to differences in the system vs underlying dynamical truths
- the simulators available are very slow and not that accurate
- the physics is hard enough that it's not possible to develop a controller in closed form (obviously)
This implies that we need a version of reinforcement learning or model-predictive control that is substantially more robust and sample-efficient than currently exists. We're working on that but obviously it's an open research problem.
Papers I would recommend from our collaboration on control of normalized plasma pressure: https://papers.nips.cc/paper/2019/hash/7876acb66640bad41f1e1...
plasma profile transport modeling: https://iopscience.iop.org/article/10.1088/1741-4326/abe08d/...
hybrid dynamical modeling of gross plasma quantities: https://arxiv.org/abs/2006.12682
uncertainty quantification for plasma dynamics: https://arxiv.org/abs/2011.09588
It's still early days for this work and for us but we're looking at pushing reinforcement learning in methods and engineering to solve this problem.
I don't understand why you're saying the Wendelstein follows the path of the plasma, it's not like plasma has a shape it wants to be in, you just sort of push it the way you want it to go, and wherever it goes you try and push it the other way again, to make it go in a circle.
What other shape than the Stellerator would someone choose if they wanted to use even more fine control over the plasma?
W7-X is not a tokamak nor a nuclear machine. It's designed to run all day and no one on the planet doubts it. The upgrade to get it to 30 minute pulses is primarily to the heat dissipation subsystem. The ECRH subsystem of W7-X is eyewateringly beautiful.
In fact, she might even stand to gain if funding gets diverted from fusion research towards other projects.
I find the claim almost trivially true: we care about electricity production, so that should be the main goal we're tracking.
The plasma physics for making a better (cheaper) reactor has been pushed in smaller machines, as shown by better triple product values and empirical scaling laws. Triple product is a measure of how well the plasma is confined. That's the metric to focus on (when accounting for the fact that the science machines are small with scaling laws). In that regard, we have made excellent progress.
A high Q machine (such as ITER) is valuable because it lets us study burning plasmas. It's an expensive pill to swallow but an essential step towards making a real reactor. Qtotal = 1 is a nothingburher. Qplasma = 1 is also a nothingburger, but at least going past that barrier grants new knowledge.
I believe that’s why they’re being made: Q being a function of both physical size and magnetic field strength, and the new superconductors being useful for increasing the latter so the former can be reduced.
But I have no idea if any of them are currently attempting Q_total > 1 or not: too much hopeful PR for me to know what stage they’re really at.
If the gain of the system overall is greater than Q-plasma, you’ve invented some other energy source in there somewhere and should throw away your reactor and focus on that thing. :D
Edit: one might think Q-plasma > 1 is just an internal milestone, not a public goal, but it’s not. Fusion startups are all going to be forced to make a public fuss over Q-plasma > 1 because the amount of funding you need to get from there to Q-total > 1 is MASSIVE and that milestone will be an investor aphrodisiac.
The Europeans are doing quite a bit of preliminary design scoping and engineering for DEMO, a 'demonstration' reactor which is supposed to come after ITER and generate net electricity, but there's not been a site selected, for example. Similarly, Chinese researchers are working on their CFETR, 'Chinese Fusion Engineering Test Reactor'.
I’m not talking about that category.
I’m talking about private sector stuff by groups which claim developments in superconductors means they can do it privately for much less, e.g.: https://arstechnica.com/science/2021/09/mit-backed-fusion-st...
I lack the knowledge to tell real projects in this sector from snake oil.
Considering even the best solar panels have under 50% efficiency, increasing the total solar radiation to earth would be very bad for the climate.
It'd be interesting to compare the energy directly released as usable heat from burning hydrocarbons, to the indirect increase in energy absorption over the life of the released greenhouse gases.
Not sure, but I reckon the cumulative greenhouse effect would be massively more than the usable energy released during combustion.
Citation needed.
The radiation to earth varies already, sometimes we're farther away from the sun and the output of the sun isn't constant either. The earth seems to handle that just fine.
That sounds remarkably similar to "how can there be global warming if we still have a cold season and a hot season".
I don't say space-based solar arrays are necessarily a problem, but arguing that earth handles variations, as if the net total doesn't matter, is a little blinkered.
Yet, this also happens with nuclear, coal, gas, oil...
The only solution is to capture the energy that is already being received from the sun in order to prevent it from heating up the atmosphere on site.
E.g. solar panels on the desert and wind turbines
Or stop screwing with it entirely (it worked fine without our help for quite a while) and build nuclear reactors everywhere, which scale wonderfully and work even when it's dark, cloudy, or shrouded in smoke.
Specifically, one of the killer engineering problems of space based solar is getting the power back to Earth. Tom Murphy wrote the canonical post on this: https://dothemath.ucsd.edu/2012/03/space-based-solar-power/ Depending on what frequency you're using and how big the orbital antenna is, the receiving rectenna array is going to be a couple kilometers across. If you use a beam power density that isn't going to cook birds flying across the antenna, then you're not receiving too much more power than regular solar irradiation. The atmosphere eats some of the power, the receiving antenna eats some power, and then the rectified power has to be converted to AC, losing a few percent more power. Maybe 50% total transmission loss. For the cost of putting a million tons into orbit, you might want to just build more ground solar.
Building transcontinental and transoceanic UHV power transmission lines circling the northern hemisphere, providing the same function is probably cheaper, but may be more politically difficult.
But that's moot. We have a cornucopia of viable storage methods in active development--viable meaning capable of being scaled up to global scale in 20 years or so.
Do you have more info on this? I’m skeptical of any claims of scaling to global scale that quickly.
At least a couple of teams are quite far along with hot rock energy storage. Good for the one week to one month timescale.
Various other battery technologies for sub-day timescales - lithium for grid frequency stabilisation, flow batteries.
I don't know of any active trials of ammonia energy storage, one of the most scalable candidates for seasonal storage, but ammonia engineering is very old and very widespread, so there don't appear to be roadblocks to scaling up rapidly once required.
The YouTube channels "Just Have a Think"[1] and "Undecided with Matt Ferrell"[2] focus on this stuff, and do some research.
Edit: you should be sceptical; well done! Scaling really does take forever. That's why the only viable technologies are those already lying around all over the place, being used for other things.
https://arpa-e.energy.gov/sites/default/files/2021-03/07%20D...
2.4 GHz is one of the best frequencies to use because losses through storms and what not are quite low. However, we now have bluetooth and wifi devices which operate at this frequency now. 1000 km from the receiver, the incident power is stronger than the minimum sensitivity of a bluetooth receiver.
[0]https://www.researchgate.net/publication/348442155_Microwave...
Fusion research is applied. It is justified by the useful end goal, being an energy source, not by pure science. A pure plasma physics program would look very different (and likely not be able to justify the current fusion budget.)
So, simultaneously advocating fusion research, while suggesting solar will win, is internally inconsistent.
So, with super heated steam of say 600 degrees C, I'd make that a maximum of about 66% theoretical, so 50% actual sounds pretty good.
If you could build a plasma heat engine, you could get near 100% but I've no idea what that would technically look like.
They probably mean 90% of a theoretical maximum.
For example, one of the results that stood out to me in thermodynamics class was that the maximum possible efficiency of an internal combustion engine (I forget which cycle) was a function of the compression ratio. In other words you could never reach 100% conversion of chemical energy to mechanical energy and the theoretical "efficiency" was a function that increased with compression ratio but could never reach 100 percent. So lets assume an engine with a given compression ratio could have a theoretical efficiency of 50 percent, but a real-world design only achieved 45 percent. Someone might say 45/50 is 0.9 or 90 percent of the theoretical limit. That's a measure of how good the design was vs what's theoretically possible, but it has little to do with the actual efficiency of the energy conversion.
IIRC the best heat-to-mechanical energy conversion devices are rockets with a theoretical efficiency of 50% but it's been half my life since I studied thermodynamics, so maybe I'm off on some of this.
I'd rather spend a trillion on fusion and get it done Manhattan project style than continue droning children in Afghanistan or funding $3.5T in pork barrel bullshit here at home. Just drown it in money completely until it's done. Put the very best people on it and let them have at it for real. We've done it before, several times, it worked. There's no reason why it wouldn't work this time. If there is any 21st century "Moon landing" type goal, this is it, and the United States is still best positioned to accomplish something of this magnitude. For how long that will be true I do not know, but it is true now.
And "late" is better than never.
It is absurd to say that fusion research's current level of funding is a waste.
https://www.theguardian.com/environment/2021/oct/06/fossil-f...
ITER is (among other things) supposed to be a physics test, but won't come until 2035. So SPARC's goal is to 'leapfrog' that by 10 years.
You're right that there's a lot of work to be done between Q_plasma = 10 and generating energy. A lot of this work on the materials side, to engineer materials that can last for years in a fusion reactor environment.
But this does not say that they are going to put power on the grid in 2025. They say net energy from fusion by 2025 which may just mean Q-plasma > 1, it does not say (which kind of proves the point of the article). They don't have any estimates of an on the grid fusion plant.
https://cfs.energy/news-and-media/cfs-commercial-fusion-powe...
[0]https://www.psfc.mit.edu/sparc/faq [1]https://arxiv.org/pdf/1409.3540.pdf
Edit: There has been Edison, and there has been Musk. If you're not either of them, and you claim other people are doing their engineering wrong ... consider that you might be mistaken.
Make a big song and dance about it, you look like a charlatan.
The game changer here is not very new reactor designs, but cheap high performance high-temperature superconductors that enable very high field and therefore much smaller size.
Other fusion work, like ICF or the LLNL laser fusion or Lockheed or General Fusion all seem to be built on designs that are far less understood and less Q performance so far.
And so far, the company, CFS, has been achieving their milestones for the performance of their magnets. The HTS magnets and coils is the main ingredient that the startup is optimizing for. Sometime in the last year, they implied that their Qplasma was much better than they minimally hoped for. I think they achieved Qplasma > 10, but Dennis was keeping the details proprietary.
From all the talks I remember, they are targeting power on the grid with the ARC reactor by 2035.
Edit:
This is the best recent video I've seen, about SPARC and ARC. Jumped to 2:25 for Dennis Whyte.
https://www.youtube.com/watch?v=bHJyoqDO0zw
It's targeted to MechE students and gives a lot of details about other aspects of a potential ARC reactor design, e.g. how to get the heat out.
In short, from my basic understanding, there are quite a few additional major challenges involved with fusion power generation. e.g. After creating net-positive fusion, the heat must be efficiently extracted without stopping the reaction. Also tritium must be continually extracted from the FLiBe (fluorine-lithium-beryllium) bath that stops and collects neutrons and extracts the heat.
Since I'm a total novice in all this, I don't know if these require incremental innovations or major advances. But from the video, the problems mentioned seem to be more tractable than achieving fusion ignition or Q>10.
Abdou's team (the fusion engineering guy at UCLA) rejected molten salt blankets for this reason, among others, after trying really hard to get them to work in studies.
One big problem with fusion is the low power density. I harp on that a lot, but it's been known to be a very serious problem for decades. ARC's power density is 40x worse than a PWR's reactor vessel. It's difficult to see how fusion can beat fission given this. I suspect the optimistic numbers for fusion come from using a way too cheery cost estimation methodology, something that would predict fission is far cheaper than it actually turned out to be.
I was always somewhat concerned about the abundance of critical raw materials for ARC or other fusion projects. I had assumed that the rare-earth elements in the REBCO tape, e.g. yttrium, would be constrained. I didn't suspect that beryllium could be a limiting resource. But I wonder if the lack of supply is related to true scarcity or just to a lack of a profitable market currently.
These are all questions I'd want to ask domain experts in mining and fusion.
But I agree that fission would be a better solution for baseload, at least for the next 10-20 years. If only newer modular designs were actually approved...
I have a suspicion that this is being funded at all because the magnet technology would be useful in non-fusion contexts (hybrid electric aircraft, superconducting generators in wind turbines.)
In contrast, IIRC a single ARC reactor of that design would have 90 tonnes of beryllium (although that could be reduced by half if the secondary loop used a different molten salt.)
This energy should be considered by fusion scientists, because it may end up invalidating certain topologies that can achieve a Qplasma > 1, but can't achieve a Qtotal > 1. For instance from the article, it seems like pulsed lasers are relatively inefficient at turning input energy into laser energy, which might mean they would need a QPlasma > 100 before QTotal aproaches 1. I don't know which of those inefficiencies are deemed to be fundamental and which are potentially improveable, but it could suggest the whole line of research is not worth perusing.
They get to uncork some champagne, but they still need to get Q up to ~10 before handing it off to the powerplant engineers.
That is not to say there should be no money spent on researching fusion or the standard model (or beyond). Just that the current system does a poor job of allocating limited funds by putting most of the monies in one basket until well after other ideas/technologies should be given time of day and ample funding.