So... what if it didn't work?

[MSimon]

jmc,

I think a continuously operating Polywell is a must. Seconds to minutes of operation. I estimate the cost at $5 to $20 million. The majority for power supplies capable of continuous operation.

However, if that shows a good result the supplies can be used to scale up to 100 MW.

[cuddihy]

Similarly I think requesting $100 million dollars when you still don't really know what's going on inside your little machine would be a big mistake. It would be setting the project up for a fall. 100 million dollars may not be much for solving the world's energy problems but it is not good value for money with regards to creating some great big new machine from out of the blue whose underlying principles of operation are still not properly understood.

If such a $100 million dollar machine was indeed built, it would probably offer new insight into the behaviour of polywell systems, it would almost certainly not achieve its objectives of net power, by failing its objectives it could result in the whole programme being abandoned in disillusionment even if in principle a prudent well thought out approach could have actually eventually produced a successful fusion device.

I don't know, jmc, the history of tokamoks seems to indicate that lower-than-billed performance is a sure ticket to ever bigger budgets, from JET on up to ITER.

But no, actually I agree with simon that, regardless of the "estimated timescale" necessary for a pulse machine to display continuous operation features, you HAVE to do a small-scale REAL continuous machine prior to going up in size, because you don't know what problems continuous operation is going to present and better to deal with them on a small scale rather than big & expensive.

[jmc]

MSimon:

I think they built WB-4 on a budget of less than 5 million and I'm reliably informed that it could go for up to 30 minutes.

Another thing would be just to build a load of cheap different sized machines all operating in pulsed mode, just to get some kind of handle on how these things empirically scale with size.

Then there is the fabled dodecahral configuration which everyone says will perform massively better than the spherical one but noone has ever built or tested. I think there are plenty of useful experiments still left to do in pulsed operation, although continuous ops are the obvious next step towards a reactor.

Anyhow I could live with spending 5 to 20 million dollars on a new machine, even in the presence of scant evidence I think if your asking for 100 million you should have some reasonably well proven scaling laws that this is a viable reactor concept though. I don't think you can say "we don't have any evidence, or proven theory this can work, but I have a hunch it might, please give us 100 million dollars"

cuddily:
JET was actually quite successful, it was built on time and on budget. It achieved 65% fusion power/heating power the highest ratio ever demonstrated. Their planning a campaign soon to get 100% non-inductive current drive for five seconds, it achieved record tokamak plasma temperatures of 40keV. In addition to this inspite of the fact that its gone over its rated lifetime JET still operates pretty well, I would even go as far as to say it was the success of JET which enabled the ITER project to be justified. ITER is in many ways just a larger version of JET. I would, however be very surprised if ITER met as many of its targets as successfully as JET did.

[MSimon]

jmc,

The demand now is political. Science and engineering have taken a back seat. It is now a case of what ever funds are reasonably required will be delivered. $200 million is down in the political noise level.

You are thinking VC money (your path is correct for that). But we are no longer in a VC environment on this.

This will not be a case of baby steps one at a time. It will be a lot of parallel steps. Manhattan Project like.

[Roger]

you HAVE to do a small-scale REAL continuous machine prior to going up in size, because you don't know what problems continuous operation is going to present and better to deal with them on a small scale rather than big & expensive.

Maybe,

Scaling can be dealt with by doubling the prior 30cm coils to 60cm, or go to Dr. Nebels 100MW/160cm size. Actual scaling is a monstrous issue. But can be done with fairly cheaply, a double size WB-7 might be built for what, under 10 mill. (60cm coils).

I feel scaling is a separate issue, we also need the LN2 cooled WB-7x. Under 20 mill.

Then there is the Dodec. Same over all size as WB-7, for maybe 6 mill?

Build 3 machines to deal with these 3 issues, is there any reason they should not be done during the same time frame ? Can some of this work be done at a "known" University.


Should the Navy Fund Dr Kulcinski at U. of Wis @ Mad. to build one? When should leading academic institutions be brought in ? "cause were going to need them. AS long as Polywells dont fizz out.

[dweigert]

For some strange reason, the symmetry of having it done at the University of Chicago appeals to me. Too bad there are no more squash courts to take over...

[MSimon]

Build 3 machines to deal with these 3 issues, is there any reason they should not be done during the same time frame ? Can some of this work be done at a "known" University.

That would make sense.

[MSimon]

For some strange reason, the symmetry of having it done at the University of Chicago appeals to me. Too bad there are no more squash courts to take over...

Worse: the tennis courts that were built where the squash courts used to be are gone too.

I played on the tennis courts in '62. Visited the area in June of '07.

Here is a link to my blog on the subject:

http://powerandcontrol.blogspot.com/200 ... usion.html

Sadly the Victor Zaveduk link no longer works. It was great.

[TallDave]

So... what if it didn't work?

There's a few options given here:

http://en.wikipedia.org/wiki/Suicide#Ritual_suicide

I like seppuku better than the ghat.

[TallDave]

Then there is the fabled dodecahral configuration which everyone says will perform massively better than the spherical one but noone has ever built or tested.

Well, you just get 3-5x better cusp confinement. It's not THAT great.

[TallDave]

For 100 million I'd need some pretty convincing evidence, frankly I can't really envisage how 6 months of data on one machine could possibly present sufficient evidence to justify such a radical increase in funding

The relative increase over prior funding really isn't that important; the comparison to other similar projects seems more relevant.

$100M isn't a lot by the standards of the fusion community. The Livermore magnetic mirror project spent 3x that and never even fired it up. And that was in 1980s dollars.

If they think there's some real chance a 1.5M device could yield net power, the risk/reward ratio is so high it would be silly not to build it.

[Shubedobedubopbopbedo]

Seem to be some tumbleweeds blowing around in here.

[MSimon]

Maybe the neutron count will be so high that they can use it as a super generator to enrich Thorium for molten salt reactors?
Would it make more U-233 fuel and generate net power that way?

Discussed at length at NASA spaceflight and/or here.

A hybrid reactor with a BFR as an electrically controllable neutron source and a U238 blanket.

The question was proliferation. But the principles are the same.

You have a reactor that can't go critical.

[KitemanSA]

Maybe the neutron count will be so high that they can use it as a super generator to enrich Thorium for molten salt reactors?
Would it make more U-233 fuel and generate net power that way?

The primary problem with thorium based molten salt reactor is the need to reprocess, which has the first step of chemically removing the bred U233 which is the actual fuel (thorium being fertile but not fissile). U233 is a potential bomb making material too; and being able to CHEMICALLY seperate it is SCARY! Therefore, to be safer, there should be no reprocessing. However, without reprocessing, there needs to be an external source of neutrons to overcome the poisoning of the protactinium (the intermediary between Th233 and U233).

Has anyone seen a study where a polywell or other fusor is coupled to an UN-reprocessed MSR?