5 ms·

As an uneducated observer to nuclear physics, I could not tell the significance of this achievement. Did we finally learn how to extract more energy from a fusi

As an uneducated observer to nuclear physics, I could not tell the significance of this achievement. Did we finally learn how to extract more energy from a fusion rector than we supplied to operate it? Could anyone more in the know here explain in simpler and more practical terms please?

lisper 5y ago
> Did we finally learn how to extract more energy from a fusion rector than we supplied to operate it? No. > Could anyone more in the know here explain in simpler and more practical terms please? It's impossible to tell because the story is incoherent. The central claim is that they "release[d] a record 59 megajoules of sustained fusion energy" but this makes no sense. One can talk about sustained power (over a period of time) but "sustained energy" is a category error because energy is just power integrated over time. You can get 59 megajoules out of your wall socket if you wait long enough. They apparently did something that had never been done before, but there's no way to tell exactly what that was from what is written in this story. (This kind of obfuscation is not unusual in fusion research. A cynic might argue that this is because if they were clear about the actual state of things their funding would cease.)
- SamBam 5y ago
  I don't believe that this is entirely mis-stated. We already know that fusion power plants are going to operate by igniting plasma in short bursts -- a few seconds, maybe 10s at most -- and generating a huge amount of power during that single burst. You then ignite it again and again. The question is how much total energy you can extract from each of those bursts. By that metric, the total energy is what's important, not the power. Producing 10 quadrillion watts of power isn't that useful if it's only for 100 trillionths of a second. (Both numbers pulled from an actual recent result last August. [1]) "59 megajoules of sustained fusion energy" means a single burst produced that much. That's significantly more than the paltry 1.3MJ from that other result I just linked to. Yes, of course you could get that much from a wall socket over the course of hours, but we know we haven't been able to sustain a fusion reaction for a few seconds, so that's the timeframe we're talking about for a single burst. 1. https://www.livescience.com/fusion-experiment-record-breaking-energy.html https://www.livescience.com/fusion-experiment-record-breakin... EDIT: And, at the bottom of the article it says that the reaction ran at 11MW, so plugging that in it sounds like, indeed, it ran for 5 seconds.
  - robbomacrae 5y ago
    Your description made me wonder if it would be possible to make an ICE to extract the power. I then found a paper called Fusion Internal Combustion Engine from 2010 [0]. Does anyone know if there was any merit to this idea and if anything else has come of it since? 0: https://www.researchgate.net/publication/215544178_Fusion_Internal_Combustion_Engine https://www.researchgate.net/publication/215544178_Fusion_In...
    - lisper 5y ago
      A few elementary calculations reveal how feasible this might be. A typical ICE produces a few hundred horsepower using 4-8 cylinders rotating at a few thousand RPM. 1 HP = 745 watts. Figure out how much energy is released per cylinder on each cycle, and compare that to the energy released in a typical fusion ignition. Also note the cycle time of an ICE rotating at a few thousand RPMs, and compare that to the cycle time of a current state-of-the-art fusion reactor. (Hint: the former is measured in milliseconds, the latter currently in months if not years.)
      - robbomacrae 5y ago
        I'm a layman and so its possible I am missing some limiting factors but I do feel as though this rebuttal does not take into account the possible ranges of values that can be configured when tweaking things like scale and operating speed. For example the Wärtsilä-Sulzer RTA96-C operates at 15-102 RPM generating 100,000 HP (or 74.5 megawatts for your comparison) [0]. Nor does it take into account the difference between a prototype investigation being constantly modified for experimentation and analysis and a production system built to purpose. 0: https://en.wikipedia.org/wiki/W%C3%A4rtsil%C3%A4-Sulzer_RTA96-C https://en.wikipedia.org/wiki/W%C3%A4rtsil%C3%A4-Sulzer_RTA9...
        lisper 5y ago
        Yes, I probably should have used numbers from a large diesel rather than an automobile engine. Wow, the Wärtsilä-Sulzer RTA96-C operates at speeds as low as 15 RPM! That is just mind-boggling. I found this video: https://www.youtube.com/watch?v=jXHvY-zY9hA https://www.youtube.com/watch?v=jXHvY-zY9hA Still, I think you will find that even this doesn't help all that much. A cycle time measured in seconds rather than milliseconds is still orders of magnitude away from what can presently be achieved. > the difference between a prototype investigation being constantly modified for experimentation and analysis and a production system built to purpose. AFAICT, no one has ever built an ICE that is within even an order of magnitude of realistic operating parameters of a fusion ICE. It's a whole 'nuther level of engineering challenge beyond just getting the fusion itself to work. I'm not saying it's impossible, but I'll give you long odds against seeing it happen in any of our lifetimes.
- arlort 5y ago
  They expand under the article (not the clearest visual design though, I thought it was an ad or some "related" section) The record is for energy generated. They generated the most energy, however they did not generate the most power (unlike the previous record which generated both the most energy and most power) because they were focusing on sustained generation The sentence "59 megajoules of sustained fusion-energy", where fusion is the source of the energy, doesn't make sense I think they meant it as "59 megajoules of sustained-fusion energy" where sustained fusion is the focus of the experiment (not sure I managed to get through what I meant and it's worded awfully on the website) I hope / expect the actual paper/technical reports which will come out will be worded more clearly
- Rygian 5y ago
  > You can get 59 megajoules out of your wall socket if you wait long enough. Assuming a French wall socket, 240 V at 13 amps, and a pure resistive load, that will take $ units Currency exchange rates from FloatRates (USD base) on 2020-11-15 3677 units, 109 prefixes, 114 nonlinear units You have: 59 MJ You want: (240 V * 13 A) * s * 18910.256 / 5.2881356e-05 You have: 18910.256 s You want: hms 5 hr + 15 min + 10.256 sec which helps put the figure in perspective.
mattalex 5y ago
> Did we finally learn how to extract more energy from a fusion rector than we supplied to operate it? No. The issue is that most people only report the gain over the plasma (i.e. how much energy was put into generating the plasma) rather than the full amount of energy put into the process (i.e. superconductors, magnets, generating the deuterium/tritium, maintaining the sun-like heat, etc). If you add this to the computation, you end up having to have a "fusion-gain" of around 50x to break even. The reason people report the plasma efficiency instead of the actual operating efficiency is to get funding and hype. Don't get me wrong, this result is still impressive, but it's still orders of magnitude off of the required efficiency.
- Yajirobe 5y ago
  50x is not orders of magnitude off. It's one order of magnitude
arlort 5y ago
> Results fully in line with predictions, strengthening the case for ITER This is the main take away for me, JET is not a standalone project, it's part of the whole ITER project which is supposed to go like this: + JET as a scaled down model provides testing and data for + ITER which I believe is a full scale model and is supposed to generate net gain (heat in vs heat out, not net electricity) and provide information to + DEMO which is supposed to produce net electricity (though not at market rate costs) So the fact that it worked as predicted is a good sign (or at least as good as we can get) that ITER will work which will be a good sign for DEMO etc Also not an expert though
regularfry 5y ago
The energy generated is a small red herring here. For a long time one of the hardest problems in fusion reactor design is what the hell you make it out of. The big win here is that they replaced the walls of the reactor with a new alloy, and it worked according to what theory predicted, which gives them the green light for using that material in ITER. To simplify a little (ok, a lot) there are two big materials problems inside the reactor. The first is the walls: you need something that's going to survive the temperatures, not disturb the reaction, and not get too radioactive in the process. They previously used carbon, which isn't great: it gets radioactive because it absorbs tritium, which is in the fuel. This experiment used a beryllium alloy, which doesn't absorb nearly as much, and worked, validating the material choice for ITER. The second problem is to do with the exhaust. You need to get hot plasma out of the chamber without disturbing the ongoing reaction, and with a tokamak that means ridiculously energetic particles hitting a solid divertor. Again, the problem here is what materials you might come up with that stand a chance of surviving useful operational periods. ITER is currently planned to use beryllium walls and a tungsten divertor, but I don't know what JET's divertor is made of at the moment to know whether this experiment will have informed whether tungsten is a good enough choice. What all this means is that there's one less thing on the "ITER might fail because of..." list.

lisper 5y ago

> Did we finally learn how to extract more energy from a fusion rector than we supplied to operate it? No. > Could anyone more in the know here explain in simpler and more practical terms please? It's impossible to tell because the story is incoherent. The central claim is that they "release[d] a record 59 megajoules of sustained fusion energy" but this makes no sense. One can talk about sustained power (over a period of time) but "sustained energy" is a category error because energy is just power integrated over time. You can get 59 megajoules out of your wall socket if you wait long enough. They apparently did something that had never been done before, but there's no way to tell exactly what that was from what is written in this story. (This kind of obfuscation is not unusual in fusion research. A cynic might argue that this is because if they were clear about the actual state of things their funding would cease.)

SamBam 5y ago
I don't believe that this is entirely mis-stated. We already know that fusion power plants are going to operate by igniting plasma in short bursts -- a few seconds, maybe 10s at most -- and generating a huge amount of power during that single burst. You then ignite it again and again. The question is how much total energy you can extract from each of those bursts. By that metric, the total energy is what's important, not the power. Producing 10 quadrillion watts of power isn't that useful if it's only for 100 trillionths of a second. (Both numbers pulled from an actual recent result last August. [1]) "59 megajoules of sustained fusion energy" means a single burst produced that much. That's significantly more than the paltry 1.3MJ from that other result I just linked to. Yes, of course you could get that much from a wall socket over the course of hours, but we know we haven't been able to sustain a fusion reaction for a few seconds, so that's the timeframe we're talking about for a single burst. 1. https://www.livescience.com/fusion-experiment-record-breaking-energy.html https://www.livescience.com/fusion-experiment-record-breakin... EDIT: And, at the bottom of the article it says that the reaction ran at 11MW, so plugging that in it sounds like, indeed, it ran for 5 seconds.
- robbomacrae 5y ago
  Your description made me wonder if it would be possible to make an ICE to extract the power. I then found a paper called Fusion Internal Combustion Engine from 2010 [0]. Does anyone know if there was any merit to this idea and if anything else has come of it since? 0: https://www.researchgate.net/publication/215544178_Fusion_Internal_Combustion_Engine https://www.researchgate.net/publication/215544178_Fusion_In...
  - lisper 5y ago
    A few elementary calculations reveal how feasible this might be. A typical ICE produces a few hundred horsepower using 4-8 cylinders rotating at a few thousand RPM. 1 HP = 745 watts. Figure out how much energy is released per cylinder on each cycle, and compare that to the energy released in a typical fusion ignition. Also note the cycle time of an ICE rotating at a few thousand RPMs, and compare that to the cycle time of a current state-of-the-art fusion reactor. (Hint: the former is measured in milliseconds, the latter currently in months if not years.)
    - robbomacrae 5y ago
      I'm a layman and so its possible I am missing some limiting factors but I do feel as though this rebuttal does not take into account the possible ranges of values that can be configured when tweaking things like scale and operating speed. For example the Wärtsilä-Sulzer RTA96-C operates at 15-102 RPM generating 100,000 HP (or 74.5 megawatts for your comparison) [0]. Nor does it take into account the difference between a prototype investigation being constantly modified for experimentation and analysis and a production system built to purpose. 0: https://en.wikipedia.org/wiki/W%C3%A4rtsil%C3%A4-Sulzer_RTA96-C https://en.wikipedia.org/wiki/W%C3%A4rtsil%C3%A4-Sulzer_RTA9...
      - lisper 5y ago
        Yes, I probably should have used numbers from a large diesel rather than an automobile engine. Wow, the Wärtsilä-Sulzer RTA96-C operates at speeds as low as 15 RPM! That is just mind-boggling. I found this video: https://www.youtube.com/watch?v=jXHvY-zY9hA https://www.youtube.com/watch?v=jXHvY-zY9hA Still, I think you will find that even this doesn't help all that much. A cycle time measured in seconds rather than milliseconds is still orders of magnitude away from what can presently be achieved. > the difference between a prototype investigation being constantly modified for experimentation and analysis and a production system built to purpose. AFAICT, no one has ever built an ICE that is within even an order of magnitude of realistic operating parameters of a fusion ICE. It's a whole 'nuther level of engineering challenge beyond just getting the fusion itself to work. I'm not saying it's impossible, but I'll give you long odds against seeing it happen in any of our lifetimes.
arlort 5y ago
They expand under the article (not the clearest visual design though, I thought it was an ad or some "related" section) The record is for energy generated. They generated the most energy, however they did not generate the most power (unlike the previous record which generated both the most energy and most power) because they were focusing on sustained generation The sentence "59 megajoules of sustained fusion-energy", where fusion is the source of the energy, doesn't make sense I think they meant it as "59 megajoules of sustained-fusion energy" where sustained fusion is the focus of the experiment (not sure I managed to get through what I meant and it's worded awfully on the website) I hope / expect the actual paper/technical reports which will come out will be worded more clearly
Rygian 5y ago
> You can get 59 megajoules out of your wall socket if you wait long enough. Assuming a French wall socket, 240 V at 13 amps, and a pure resistive load, that will take $ units Currency exchange rates from FloatRates (USD base) on 2020-11-15 3677 units, 109 prefixes, 114 nonlinear units You have: 59 MJ You want: (240 V * 13 A) * s * 18910.256 / 5.2881356e-05 You have: 18910.256 s You want: hms 5 hr + 15 min + 10.256 sec which helps put the figure in perspective.

mattalex 5y ago

> Did we finally learn how to extract more energy from a fusion rector than we supplied to operate it? No. The issue is that most people only report the gain over the plasma (i.e. how much energy was put into generating the plasma) rather than the full amount of energy put into the process (i.e. superconductors, magnets, generating the deuterium/tritium, maintaining the sun-like heat, etc). If you add this to the computation, you end up having to have a "fusion-gain" of around 50x to break even. The reason people report the plasma efficiency instead of the actual operating efficiency is to get funding and hype. Don't get me wrong, this result is still impressive, but it's still orders of magnitude off of the required efficiency.

Yajirobe 5y ago
50x is not orders of magnitude off. It's one order of magnitude

arlort 5y ago

> Results fully in line with predictions, strengthening the case for ITER This is the main take away for me, JET is not a standalone project, it's part of the whole ITER project which is supposed to go like this: + JET as a scaled down model provides testing and data for + ITER which I believe is a full scale model and is supposed to generate net gain (heat in vs heat out, not net electricity) and provide information to + DEMO which is supposed to produce net electricity (though not at market rate costs) So the fact that it worked as predicted is a good sign (or at least as good as we can get) that ITER will work which will be a good sign for DEMO etc Also not an expert though

regularfry 5y ago

The energy generated is a small red herring here. For a long time one of the hardest problems in fusion reactor design is what the hell you make it out of. The big win here is that they replaced the walls of the reactor with a new alloy, and it worked according to what theory predicted, which gives them the green light for using that material in ITER. To simplify a little (ok, a lot) there are two big materials problems inside the reactor. The first is the walls: you need something that's going to survive the temperatures, not disturb the reaction, and not get too radioactive in the process. They previously used carbon, which isn't great: it gets radioactive because it absorbs tritium, which is in the fuel. This experiment used a beryllium alloy, which doesn't absorb nearly as much, and worked, validating the material choice for ITER. The second problem is to do with the exhaust. You need to get hot plasma out of the chamber without disturbing the ongoing reaction, and with a tokamak that means ridiculously energetic particles hitting a solid divertor. Again, the problem here is what materials you might come up with that stand a chance of surviving useful operational periods. ITER is currently planned to use beryllium walls and a tungsten divertor, but I don't know what JET's divertor is made of at the moment to know whether this experiment will have informed whether tungsten is a good enough choice. What all this means is that there's one less thing on the "ITER might fail because of..." list.