The AMTEC eletric generator and Carnot.

[TDPerk]

In a thread I've lost track of, MSimon and Giorgio (and to a degree myself) had a verbal tussle wherein we quarreled about a device being subject to the Carnot limit.

It--the AMTEC--forces liquid metal (I think it was sodium) through a ceramic membrane of beta alumina. In order to pass through the membrane, the sodium must lose an electron or it won't fit. After the sodium ions pass through the membrane, and the electrons pass through a load, they are recombined on the other side of the membrane.

MSimon I think is partly wrong--the separation of electrons from the sodium and later recombination is not in and of itself subject to the Carnot limit. Where he is completely correct is the heat engine which is producing the pressure differential across the membrane is subject to the Carnot limit. It looked a bit like a Brayton cycle, abjectly, hopelessly a Carnot limit machine.

Giorgio is I think all wrong, I think, in that he was saying the AMTEC is not subject to the Carnot limit, and clearly after reading the whole paper, it is. The word "reversible" is even used to describe the heat engine producing pressure across the membrane.

I reflexively supported Giorgio when MSimon made his blanket claim--if I recall it correctly--that everything is subject to the Carnot limitation. That's how I read what MSimon was writing.

The direct conversion of ions to power which is to be used--we hope--in the Polywell is a prime mover not subject to the Carnot limitation, it is not a heat engine. It is not required that the ion/electrons be hot, and for that matter it would be better if they were cold.

Apologies are sincerely rendered.

[MSimon]

Of course the final product - energetic ions - is not subject to Carnot.

The separation of "hot" ions from"cold" ions is subject to Carnot.

You essentially have a Maxwell's Demon type system. It picks out hot ions and rejects cold ones. That is subject to Carnot.

Any way no apologies required. BTW you need a cold side not only for the energy differential but also to have sodium in a vapor form with low electrical conductivity. Otherwise the sodium shorts out your electrical and temperature differences.

[TDPerk]

MSimon, in the Polywell heat is still an artifact of the ion generation system which is not required for the direct conversion to work. If for example beam collision worked better, there would be net fusion power, and the heat generated would be an unnecessary and unwanted side effect. Non-Carnot power conversion systems exist. They are not especially practical yet, but they exist.

[MSimon]

MSimon, in the Polywell heat is still an artifact of the ion generation system which is not required for the direct conversion to work. If for example beam collision worked better, there would be net fusion power, and the heat generated would be an unnecessary and unwanted side effect. Non-Carnot power conversion systems exist. They are not especially practical yet, but they exist.

I agree. But if you are using heat to do the ion separation then you are subject to Carnot. And the fuel cell as presented is subject to Carnot.

Now it may be on a deeper look even the ions from a Polywell are subject to Carnot. But it matters not since their temperature is so high that the inefficiencies predicted by Carnot are very small. i.e. source temps on the order of billions of deg K vs sinks at less than 1,000 K. i.e:

(1 - 1E3/1E9) = 99.9999% efficient (theoretical limit - roughly)

There are bigger losses to worry about.

[Giorgio]

What I have felt from the start of the discussion is that there has always been confusion on the meaning of "where" you can apply carnot, and what is the meaning of using carnot-efficiency as a reference for a process.
It's obvious that an heat engine is subject to carnot, and I never stated the opposite, but in the Amtec (like TDPerk correctly says), there are different processes involved apart the simple thermal cycle.

As there are different processes involved, and as not all the processes are thermal ones, you have what is called a "complex process system", and you cannot apply the carnot limit to the whole process, but you have to split it into the sub-processes and apply the needed law to each single step. The consequence the whole system is not subject to the Carnot efficiency limit, but you can use the Tc and the Th of the system to "compare" your complex system in respect to a simple carnot system operating between those temperature.

This papers does explains AMTEC in better terms. I will try to get it and post a link to a copy of it for everyone to review.
http://ieeexplore.ieee.org/Xplore/login ... ision=-203

[MSimon]

Any time you have a Maxwell Demon separating hot ions from cold ones you are subject to Carnot.

And my opinion is that the paper is flat out wrong in saying Carnot efficiency can be exceeded for the thermal process (getting hot ions to cross a mechanical barrier).

As to the rest I have no opinion.

If the paper was a dissertation I believe the reviewers failed to look into it deeply enough. If it was an ordinary peer reviewed job I'd say the review was cursory at best. Peer review being a rather slap dash process.

Of course once you have the hot ions in the chemical cell they are no longer limited by Carnot. I would not dispute that. Chemical efficiencies then take over and can exceed Carnot WITHIN THE CELL.

So let me see

As good as I can remember:

Hot side: 1200 C = 1473K
Cold Side: 900 C = 1173K

1173/1473 = .796 which gives Carnot of 20.4%

Then subtract the cell losses. And pumping losses.

One thing to keep in mind is that by extracting the hot ions on the hot side you are directly lowering the temperature on the hot side. i.e. cooling it with no energy extraction. But you do get the hot ions. So that is something.

It would seem to me that there is an optimum temperature to operate the device for maximum energy extraction. Too hot and the ions have more energy than needed to cross the barrier. Too cold and they don't cross. Of course the maximum power point may not be the maximum efficiency point.

[MSimon]

You also need to "collect" the hot ions as soon as they cross the barrier so that they don't go back the other way.

[Giorgio]

And my opinion is that the paper is flat out wrong in saying Carnot efficiency can be exceeded for the thermal process (getting hot ions to cross a mechanical barrier).

We all agree on this point. The paper never says that, nor did I.
Let me get an hold on a copy of it if I can and post a link. It will probably clear all the misunderstanding we are facing.

[MSimon]

And my opinion is that the paper is flat out wrong in saying Carnot efficiency can be exceeded for the thermal process (getting hot ions to cross a mechanical barrier).

We all agree on this point. The paper never says that, nor did I.
Let me get an hold on a copy of it if I can and post a link. It will probably clear all the misunderstanding we are facing.

I'm satisfied that we are now on the same page.

[Giorgio]

I'm satisfied that we are now on the same page.

We have always been on the same page regarding that point.

[MSimon]

I'm satisfied that we are now on the same page.

We have always been on the same page regarding that point.

It wasn't clear to me or I would have dropped the issue long ago. Any way thanks for sticking with it until we resolved it.