We Will Know In Two Years

[tombo]

And 5X or 10X is a much more comfortable margin.

Dream on!

This whole scheme is waaay too close to the hairy edge to expect safety factors like that.

[MSimon]

And I forgot to mention that alphas leaving through a cusp greatly simplifies direct conversion.

[icarus]

What a lot of people don't realise is that once a mass market is identified, and the know-how is in place (intellectual property), the total cost of any mass produced item asymptotes down to about exactly it's raw material cost.

If you want to know how much a washing-machine, car, fridge, etc cost to produce then compact it down into a ball of stuff and figure out what the weight cost of the component materials are ... add in a little for energy to drive the factory robots and shipping transport to store and you're done.

If you can mass produce a 10kW fusion reactor out of a hunk of steel, copper, plastic, etc that costs less than ~ $30,000 RAW material cost, I reckon you'd be in business.

[Art Carlson]

How about comparing the high-end hopes for polywells to the high-end hopes for tokamaks when tokamaks were in the same stage as polywells are now, i.e. having produced three neutrons? It's easy to be optimistic as long as you are also ignorant.

True. OTOH, we can say something the tokamakkers could not at that point: it will not cost $20B+ to find out if Polywell works (where "work" means "produce energy at a competitive cost").

Actually, come to think of it, they still can't say that about tokamaks, billions of dollars and neutrons later.

I don't see any difference. In the early days of tokamaks, before "anomolous transport" became a household word, the estimates of development cost and time were very low, like they are now in some people's minds for polywells. If transport in tokamaks had turned out to be much better than it is, cheap tokamaks would have been quickly in production. If it had turned out to be much worse, the line would have been given up before much money had been spent. That's all the same for polywells. If the transport proves to be marginal, then much research will have to be done in expensive machines for decades before we have that answer.

Also, did anyone ever seriously envision burning p-B11 in a tok? My 1950s text seems to rule that out.

Sure they did. And they rejected it for the same reasons that apply to polywells. (Since Rick hasn't provided us with a publication yet, we don't know whether his idea to get around the problem is "serious", and if it is, whether it might also be applicable to tokamaks.)

[Art Carlson]

The other half of the point I was trying to make is that toroidal magnetic systems and IEC don't mix very well because of the confinement properties of the toroid. The advantages Polywell supposedly has would largely evaporate, in exchange for a lot more trouble than tokamak currently has.

Art, you seemed to more or less agree with this assessment. Is that right?

In the privacy of my own home, I tend to think the polywell works like a cusp confinement device, not like IEC. But beyond that, I don't see any way of combining the strengths of tokamaks with those of polywells, so I think that means we agree.

[Art Carlson]

Now I do not know who you are, but from where I stand Dr. Nebel is the only one working on a real Polywell fusion device, that makes him the expert, he has the data and that trumps theory. I hope you don't mind if I take his word over yours.

Feel free. But if you are a scientist, you know that lots of things can go wrong with experiments, from faulty instruments to faulty interpretation, so it's not a good idea believe experimental data blindly. Unfortunately, that is what we have to do with Dr. Nebel's data. None of us has ever seen it. We don't even know what instruments he was using. At least I have laid my theory on the table for your inspection.

[Art Carlson]

Just to beat a dead horse: toks envision making 10% more tritium than they use with a Lithium blanket neutron to tritium converter. As Art points out that is cutting it rather fine for an unproven total design. It would not be hard to come in with 10% less tritium than they use and there goes the whole shooting match.

For an experimental design you really want to come in at 2X what you hope for in all the critical design parameters. And 5X or 10X is a much more comfortable margin.

I thought the breeding factor was more like 1.02 or 1.03, but I wouldn't want to argue about it. The engineering designs are rather detailed, and the neutronics calculations are well developed, so people don't seem to worry much about this. I would agree it is one of many problems that probably won't but could rear its head later. Right now there are certainly bigger headaches - transport, tritium accumulation, ablation and forces from disruptions, ....

One reason you would like a bigger margin, aside from just sleeping better, is to reduce the time to build up new reserves. You can find enough tritium reserves to fuel your first few reactors, but the ones after that will have to get their tritium from the excess produced by the first generation. If the breeding ratio is too low, that could be a bottleneck.

[Art Carlson]

How about comparing the high-end hopes for polywells to the high-end hopes for tokamaks when tokamaks were in the same stage as polywells are now, i.e. having produced three neutrons?

Art,
Isn't this just a bid disingenuous? You keep says "3 neutrons" but it was actually ~3 out of the expected ~3, every time they did it. That is a whole different situation. Your statement suggests they got no data. The reality is the data match expectations across the board. I like the sound of the second way a whole lot better!

I'm sure you do, but that's not the way it was. Check the papers. I believe there were actually 6 or 7 neutrons involved from 3 or 4 shots. It's not like 3 neutrons were predicted by any theory, and it's not like there were enough experiments done to make an empirical prediction. They just happened to see a few neutrons under the last conditions they tried before their funding ran out.

[Art Carlson]

Just to beat a dead horse: toks envision making 10% more tritium than they use with a Lithium blanket neutron to tritium converter. As Art points out that is cutting it rather fine for an unproven total design. It would not be hard to come in with 10% less tritium than they use and there goes the whole shooting match.

For an experimental design you really want to come in at 2X what you hope for in all the critical design parameters. And 5X or 10X is a much more comfortable margin.

The advantage for Polywell is that if it works it scales nicely and it is not too large to begin with so R^3 scaling and B^4 scaling is not too hard to come by. Going to 1 m coils (from .3 m coils) gives you about a 30X power boost. And going to 1 T (from .1T) gives you about a 1E4 power boost. If it produces 1 mW in the current set up you are at 300 W for the improved situation. Going from there to 10 T and you are in the 3 MW range. Once that is accomplished (if it can be) then you are in the "unlimited funds for further development" regime.

The gist of this and some of the following posts is simply the advantage of high beta. For a given field strength, a beta =1 machine has 100 times the power density of a beta = 0.1 machine. Since costs usually scale a little more slowly than the volume, the CoE in the compact machine will not be 100 times less, but it might go in that direction. But it's not like we have a slew of machines to choose from. I would love nothing more than having two fusion concepts with adequate confinement, so I could pick and choose according to other criteria. Transport, transport, transport!

[TallDave]

I don't see any difference. In the early days of tokamaks, before "anomolous transport" became a household word, the estimates of development cost and time were very low, like they are now in some people's minds for polywells.
...
If the transport proves to be marginal, then much research will have to be done in expensive machines for decades before we have that answer.

Well, one large difference is that we know the problems tokamaks ran into that made them so expensive and the Polywell concept is in some sense a response to those problems. Hence, it's much more likely that we'll find out relatively cheaply and quickly if Polywells can work. It's second-mover advantage.

Even if results are marginal, it's hard to imagine the concept won't be abandoned long before anyone starts talking about a $20B science project that still can't promise anything close to commercial viability.

Sure they did. And they rejected it for the same reasons that apply to polywells.

Hrm? They share the brem problem, but with a tokamak you also have both a thermal tail problem and a temperature problem for an aneutronic reactor. Those are much more difficult to solve in a tokamak, since in IEC temp is just voltage and the distribution is much closer to monoenergetic.

[TallDave]

I'm sure you do, but that's not the way it was. Check the papers. I believe there were actually 6 or 7 neutrons involved from 3 or 4 shots. It's not like 3 neutrons were predicted by any theory, and it's not like there were enough experiments done to make an empirical prediction. They just happened to see a few neutrons under the last conditions they tried before their funding ran out.

4 shots, 12 neutrons according to Valencia. He also apparently got neutron counts from WB-4.

WB-4 produced fusions in DD under a short-pulsed-mode drive in December 2003, at about 1E6 fus/sec at 12 kV drive energy
and 10 kV well depth.

He also mentions getting fusions from MPG-1, MPG-2 and PZLx-1. So it appears we're talking about a lot more than just 3 neutrons from four test runs of one machine.

[TallDave]

You may have missed it but at high fields there is no first wall problem. Dr. N. first pointed that out about a week or two ago. And I did some calculations and came up with .35T as the transition point for a machine with 2 m diameter coils. Above that the alpha gyroradius is smaller than the coils.

For p-B11, yes. I had assumed the post was talking about a 2GW D-D machine. Which I'm not sure really made sense, but I was even more wrong than that...

..becase then I read it over again, and slapped myself for confusing second-generation "2G" with two gigawatt, and deleted that part of my comment (but not before getting caught )

[MSimon]

You may have missed it but at high fields there is no first wall problem. Dr. N. first pointed that out about a week or two ago. And I did some calculations and came up with .35T as the transition point for a machine with 2 m diameter coils. Above that the alpha gyroradius is smaller than the coils.

For p-B11, yes. I had assumed the post was talking about a 2GW D-D machine. Which I'm not sure really made sense, but I was even more wrong than that...

..becase then I read it over again, and slapped myself for confusing second-generation "2G" with two gigawatt, and deleted that part of my comment (but not before getting caught :oops: )

Any reaction that emits alphas will be helped by high fields. Of course it will do nothing for neutrons. That can be handled in two ways. Minimize the thermalization of the neutrons and let them pass through the coils to be thermalized outside the machine. Or maximize the thermalization and let the neutrons get absorbed by B10 in the coil structure.

Studies will need to be done.

In any case the first wall problem is eliminated or greatly reduced.

[D Tibbets]

Several comments in respose to reading through this thread.

The history of research into the Polywell concept has streatched over ~ 20 years, not just a few. Compared to Tokamacs this is perhaps 1/3 to 1/2 the time period, depending when you decide it started- with research in Russia, or when it gained popularity in the West during the 1970's. Polywell research has suffered in comparison due to the much smaller bugets and the lack of open communication concerning details. Converesly Polywell research and expectations have benifitted from the huge amount of research done in various electrostatic and magnetic scheems (nothing is in isolation).

I've read in layman sources that breading tritium in Tokamacs with the lithium blanket might barely produce a surplus of tritium, but if more than a tiny fraction of the neutrons are lost to other structures than the lithium blanket the trituim will be in a negative balance. There is a proposed workaround using beryllium to boost tritium production, though at increased cost and complexity. Any fusion method that could fuse D-D or other fuels would have a major advantage over D-T machines. Of course you would still need to figure out what to do with the waste tritium from D-D fusion. Either reburn it or start a buisness selling Beta voltaic batteries.

P-B11 fusion, if it worked might resolve most of the first wall problems, but even D-D fusion in a Polywell could potentially have a significantly decreased problem- at least in terms of heat loading on the magnets. Presumably, the tritium and He3 fusion ash would leave through cusps like the alphas from P-B11 fusion. Direct conversion would be more difficult due to the increased spread in the kinetic energy, but even with thermal conversion the charged particles only hitting the chamber walls instead of a significant portion hitting the magnets eases the thermal handling engeenering. Of course there would be the rebounding neutralized ash products and spalled atoms to contend with.
I could see direct conversion of a wide range of charged ash products being desireable, even if the efficiency suffered alot from having to accomidate the various species, especially if the conversion grid was also used to direct the ions into collecting ports- pumps at speeds too slow to cause significant spalling and befor neutralization.

[EDIT]
Additionally, I might ( will ) add that R. Bussard reluctantly said that second generation Polywells MIGHT be made compact enouth to fit on a locamotive or semi, but no smaller. If you haven't already you should listen to the below link. Bussard expands on what he said in the Google talk.
http://archived.thespaceshow.com/detail.asp?q=709


Dan Tibbets

[D Tibbets]

Opps... sorry. The previous link is to an interesting interview with Tom Ligon in which Dr Bussard joins in. But it is not the interview I was thinking of.

A PDF transcript of that interview I was thinking of is at-
http://www.askmar.com/Robert%20Bussard/ ... erview.pdf

The original website seems to be defunt.


Dan Tibbets