US Senators: What Will ITER Really Cost?

[kcdodd]

I don't know how many times I can say that I think there is nothing physically wrong with Tokamaks. I have nothing against the idea of them, except that they are big, expensive, and ultimatly I don't think anyone will ever build one for commercial purposes. That is why continuously comparing everything to them is pointless. It doesn't matter if tokamak (a) got better results than non-tokamak (b) if the design of (a) cannot be scaled to compete economically with say solar panels. Even so, although the scaling currently looks promising for at least a really expensive proof of principle for man-made fusion energy, so far it has not been proven that tokamaks can generate net power at any price level. I have not seen any scheme so far that shows a clear way to scale to a practical result, which also includes polywell. I am not a cheerleader for anyone, and I'm not going to play your game.

[Joseph Chikva]

That is why continuously comparing everything to them is pointless.

Agreed - pointless to compare. But:

I have to agree that the "no toroidal stars" argument against the tokomak and other toroidal devices is weak. Unfavorable field curvature is a much stronger argument.

My answer was only that there is not really proven more favorable curvature (confinement concept) after which we (you and me) have started arguing.
To speak about weakness of tokamak field configuration is a big blef. As TOKAMAKs really proved that they are stable enough for providing minutes order plasma lifetime.
The argument that spherically symmetric field is better is bluff as there was possibility to prove but that still is not proven.
The reasonings about advantage of convex fields against convex is wrong and I easily explained you others why. Get it please. I think that is not difficult to understand.

Best regards,
Joseph

[kcdodd]

Defining curvature as k = n dot ( (b dot grad)b ), where n is unit vector pointing out of plasma, and b is field aligned unit vector. I would claim that all things being equal, k > 0 is better than k < 0. Are you saying that you disagree with that statement? Why? You have to point to some reason that k < 0 is better than or equal to k > 0, and not simply that conferment is merely possible with k < 0 only. If you only want to keep talking about k < 0 being just ok enough, then I'm not interested. I have never claimed that k > 0 is mandatory for fusion. Merely that it would be favorable if end-losses were not an issue.

[Joseph Chikva]

If you only want to keep talking about k < 0 being just ok enough, then I'm not interested.

Does not matter in what are you interested.
I am only claiming that TOKAMAK having field similar to concave at other similar conditions provides better confinement than Stellarator with "minimum-B".
So, such principle is less significant. And confinement issue is much complex than keeping of certain rather simple principle.
My advise to not call any field configuration "bad" or "good" without experimental proof only on base of popular in last century principle that does not work.

[KitemanSA]

Carter,
I have Joe's posts blocked, so I have been reding your side of the conversation. What strikes me most is that you sound like you are trying to reason with a petulant child. It seldom works.

[Joseph Chikva]

Carter,
I have Joe's posts blocked, so I have been reding your side of the conversation. What strikes me most is that you sound like you are trying to reason with a petulant child. It seldom works.

I see here quite many petulant children believing that tens thousands experts around the world are idiots and only they know the way how to make humanity happy.
And how? They believe that only they know well known since last century but never proven experimentaly reasonings about advantage of convex fields.
Who is child? Me or you?

[kcdodd]

Of course it matters to me what I am interested in. I am not interested in throwing around vague words. If you can't give me physical principles, then this is a waste of time. And I do not think tokamak experts are idiots. I wonder how different the conversation would be though if one of them were here instead of you.

[Joseph Chikva]

Of course it matters to me what I am interested in. I am not interested in throwing around vague words. If you can't give me physical principles, then this is a waste of time. And I do not think tokamak experts are idiots. I wonder how different the conversation would be though if one of them were here instead of you.

What physical principles do you want?
I can inform you only that there is not any universal way how define "good" or "bad" field configuration. I see that you do not think tokamak experts are idiots. But the most people here think that desision makers (not tokamak experts but people of e.g. Fusion Scince Office of DOE whose responsibility is to know all promising fusion concepts and not only TOKAMAKs) are idiots throwing money for shit, while they know how to reach total happiness with order of magnitude less money.
But actually those people know nothing, their knowledge of plasma physics is very eclectic and they present well known for a long time and not proven experimentally reasoning about "good" field configuration as something new capable to make revolution.
In fact USPO expert said to Dr. Nebel that Polywell has not any novelty and is very similar to Russian Galatea that is forgotten for a very long time.
What vague words am I throwing if all said by me are only obvious and well seen facts?

[kcdodd]

How would I know what physical principles you are thinking of. One reason for k < 0 as bad is the basic Rayleigh–Taylor instability. Now, how is that not bad?

[Joseph Chikva]

How would I know what physical principles you are thinking of. One reason for k < 0 as bad is the basic Rayleigh–Taylor instability. Now, how is that not bad?

Very good to speak with person who at least understands that instabilities of various types always present in plasma.
The talk is not about to prevent of all types of instabilities but to avoit their destructive development till enough quantity of ions will fuse.
That is why any fusion machine has finite confinement time (as absolutely stable plasma would be confine for infinite time period).

The talk is about only what mechanism of mitigation of instabilities has that or another machine.
Recall that ELM is another type of instability that is a feature of H-mode (high confinement mode). And so on.
Nevertheless TOKAMAKs provide minutes order lifetime of plasma.
The answer is to run at beta<0.4 (Troyon limit), to have low aspect ratio, to use H-mode, etc.

Now please you answer what mechanism for mitigation of electron-ion two stream instability Polywell having "ggod field configuration" has?
The answer is - none. As there are three ways for mitigation of such types of instability and none is posible for polyweel design.
Despite this people here talk about posibility to run machine at beta=1, scaling law B^4R^3 and are going to increase B in 125 times (from 0.8 T to 10 T).
Do you expect too that even running acceptable long time (I am not so sure as we can not find any experimental results of past experiments) at 0.8 T Polywell with the same success will run with number density highe in 15625 times?

[kcdodd]

So, your answer is that it IS bad, but that there are workarounds to bad things. That is fine, but different than your claim that there is no bad field curvature. Tokamak's simply have to live with it.

As to a polywell, I spent quite a bit of time several years ago modeling a polywell type scenario. I could never convince myself that a polywell could even be created at high densities (greater than say 10^17 m-3 or so). Too low for practical fusion. Now, as to streaming instabilities, I don't know. You would have to solve for the equilibrium first before you would know what would happen from there, and I could never make an equilibrium in the typical polywell sense in simulation. However now we are not talking about curvature anymore. Just because an experiment cannot take advantage of favorable curvature for completely unrelated reasons doesn't mean all machines with that curvature are dead ends. And who knows, perhaps those working on polywell can find a workaround in the same way that H-mode was found essentially by accident in the 80's. How many tokamaks where built before that? I would guess many dozens. Even now it is not really fully understood, but without it ITER would be costing $40b instead of $20b. So, the moral is that polywell could probably actually work if you simply made it big enough and ran it at lower densities. But do you really want a $40b dollar polywell? No, of course not. And disregarding good field curvature because of an unpredictable workaround is just silly.

[Joseph Chikva]

So, your answer is that it IS bad, but that there are workarounds to bad things. That is fine, but different than your claim that there is no bad field curvature. Tokamak's simply have to live with it.

As to a polywell, I spent quite a bit of time several years ago modeling a polywell type scenario. I could never convince myself that a polywell could even be created at high densities (greater than say 10^17 m-3 or so). Too low for practical fusion. Now, as to streaming instabilities, I don't know. You would have to solve for the equilibrium first before you would know what would happen from there, and I could never make an equilibrium in the typical polywell sense in simulation. However now we are not talking about curvature anymore. Just because an experiment cannot take advantage of favorable curvature for completely unrelated reasons doesn't mean all machines with that curvature are dead ends. And who knows, perhaps those working on polywell can find a workaround in the same way that H-mode was found essentially by accident in the 80's. How many tokamaks where built before that? I would guess many dozens. Even now it is not really fully understood, but without it ITER would be costing $40b instead of $20b. So, the moral is that polywell could probably actually work if you simply made it big enough and ran it at lower densities. But do you really want a $40b dollar polywell? No, of course not. But ignoring basic physics in the hope of finding a workaround by accident is no way to plan things. And disregarding good field curvature because of an unpredictable workaround is just silly.

Bad or good are scinfific words, my friend. The better word is "acceptable results".
10^17 m^-3 density means that Polywell will have in million times lower power density than TOKAMAK. And so Polywell never will not compact and therefore cheep.
Very good that you know hystory how H-mode was discovered. But TOKAMAK even without H-mode has resource to run with seconds order lifetime. Because keeping some condition there is the mechanism to mitigate instabilities development.
I am claiming that there in Polywell is not the mitigation mechanism of at least electron-ion two-stream
And if anybody ignores basic physics, that is not me. That is those Polywell fans speaking about posibility to run at beta=1 including developers (at least Dr. Nebel).

[kcdodd]

10^17 is just a number where it became hard to keep the ion and electrons separated. The fields in the cusps were such that ions would simply stream out. But I don't claim what I did was comprehensive or conclusive.

[JoeP]

"I am really do not understand what is better."

[MSimon]

There seems to be a problem in engineering thinking.

Thought experiment:

1. You build a machine it takes up 1 m^3 and delivers 1 KW. You can build them in 1KW increments.

2. Your friend builds a better machine. It delivers 100 KW per m^3. Its smallest size is 10,000 m^3.

Which will you buy?

Energy density isn't everything.

And of course the steam plant vs direct conversion throws a little wrench into the economics. And scalability.