We Will Know In Two Years

[Art Carlson]

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.

The only temperature problem I know of in a tokamak is that cyclotron radiation kicks in at some temperature to become worse than bremsstrahlung, but I don't remember when that happens. The details can get complicated, but a high-beta device like polywell should have much less an issue with cyclotron radiation.

What is a "thermal tail problem"? Do you mean the extra kick (Is it a factor of 2, or is it less?) you get by running at the resonance of the p-B11 cross section? (Factor of 2 was probably for the CBFR. Polywell should be less because it is isotropic.) If you want to claim that as a credit, then I insist on booking a big debit for the power to maintain that non-Maxwellian distribution.

[MSimon]

Art. Here is the kick:

[TallDave]

I had understood p-B11 temps to be challenging for tokamaks, but I'm not clear on why exactly. Possibly it was just the cyclotron radiation issue, or maybe just high temps in general relative to their difficulty in IEC.

As I understand and recollect it, the thermal tail problem is the side reactions you get because the distribution is Maxwellian, which make the process somewhat less aneutronic. Someone did a post on this a while back, pointing out the conventional calculations for aneutronic fusion assumed such a distribution.

[chrismb]

toks envision making 10% more tritium than they use with a Lithium blanket neutron to tritium converter.

I thought the breeding factor was more like 1.02 or 1.03, but I wouldn't want to argue about it.

Just to clarify, as DT says, it comes down to the 'tune' of the beryllium breeding. Some endothermic neutron breeding WILL be required to maintain enough tritium, if tritium breeding is to be exploited within a fusion reactor. Quite obviously, each and every lone reaction consumes a lone triton and emits a lone neutron. That lone neutron has a lone reaction with lithium which MUST be caught 100% AND result in a tritium if the T were to be 'maintained'. It is self-evident that that process can never be 100%, so one would simply assess how many neutrons DON'T produce T with the Li, then insert a controlled, tuned amount of Be so as to endothermically breed enough neutrons necessary for the purpose. Hence, a 'well-oiled' speculated tokamak of the 22nd century would have a 1.000 tritium breeding factor overall.

To my mind, this whole tritium breeding business presents ITER in a bad light. Having asked the question directly to ITER press office, the reply I got was that no tritium breeding programmes had been planned. There is still 'talk' about it and as I understood the reply, the Japanese and Russians are particularly keen to forward their proposals for breeding modules. However, that ITER has not yet nailed down this matter, yet T breeding should be a fundamental requirement of a 'proof-of-principle' DT-burning machine, (not our Dan Tibbets - he wouldn't be a sustainable source of fuel!!) suggests to me a degree of lax-ness of intellect over the actual objective of ITER.

Personally, I think it is easy to critique it - ITER is an all-out engineering solution to problems that would exist for a working machine. Why not just make a working machine out of bog-standard materials, (e.g. stainless steel rather than beryllium), forget the breeding, forget the SC magnets, forget it all and just build a go-darned machine that actually pumps out net-power neutrons on demand! Prove it can be done first, before spending vast wealth on materials and material development!! No-one would question funding materials' engineering after that! (What is left would need to be disposed of more carefully, but once done it would secure an infinite amount of funding for the improvement of those materials.)

As I say, I think ITER is actually missing some real thinking in what its objective *could most usefully be*. It is a somewhat blind 'do everything' mass of an idea which has now gained an uncontrollable momentum.

[Art Carlson]

I had understood p-B11 temps to be challenging for tokamaks, but I'm not clear on why exactly. Possibly it was just the cyclotron radiation issue, or maybe just high temps in general relative to their difficulty in IEC.

I've heard it said before (also by Bussard?) that tokamaks have trouble reaching high temperature, but I've never heard it from the tokamak community. The optimum temperature for a pressure-limited, homogeneous D-T plasma is 15 keV, and the optimum maximum temperature for a D-T tokamak with a realistic temperature profile is around 30 keV, so nobody was much interested in higher temperatures. There is a well-known and well-understood beta limit in tokamaks, and a well-known but not so well understood density limit, so there is in a sense a lower limit on temperature, but in practice no upper limit except, under some conditions, available heating power.

As I understand and recollect it, the thermal tail problem is the side reactions you get because the distribution is Maxwellian, which make the process somewhat less aneutronic. Someone did a post on this a while back, pointing out the conventional calculations for aneutronic fusion assumed such a distribution.

OK, if that's what you mean. The thermal tail puts a limit of how clean p-B11 can be (very clean vs. super clean), but not on the feasibility of net power.

[Art Carlson]

toks envision making 10% more tritium than they use with a Lithium blanket neutron to tritium converter.

I thought the breeding factor was more like 1.02 or 1.03, but I wouldn't want to argue about it.

Just to clarify, as DT says, it comes down to the 'tune' of the beryllium breeding. Some endothermic neutron breeding WILL be required to maintain enough tritium, if tritium breeding is to be exploited within a fusion reactor. Quite obviously, each and every lone reaction consumes a lone triton and emits a lone neutron. That lone neutron has a lone reaction with lithium which MUST be caught 100% AND result in a tritium if the T were to be 'maintained'. It is self-evident that that process can never be 100%, so one would simply assess how many neutrons DON'T produce T with the Li, then insert a controlled, tuned amount of Be so as to endothermically breed enough neutrons necessary for the purpose. Hence, a 'well-oiled' speculated tokamak of the 22nd century would have a 1.000 tritium breeding factor overall.

The Li6+n->He4+T reaction is, of course, a 1:1 proposition., but there is also the Li7+n->He4+T+n reaction, which gives you a tritium without costing you a neutron. Since 92.5% of natural lithium is Li7, this reaction generally puts you in the plus. You don't get the 2X that MSimon would like because the reaction is endothermic. If you need some extra umph, you can isotopically tailor your lithium or add something that with a large (n,2n) cross section, like Be or lead. A lithium-lead mixture is sometimes proposed as a blanket fluid. If you have cojones, you can also add uranium.

To my mind, this whole tritium breeding business presents ITER in a bad light. Having asked the question directly to ITER press office, the reply I got was that no tritium breeding programmes had been planned. There is still 'talk' about it and as I understood the reply, the Japanese and Russians are particularly keen to forward their proposals for breeding modules. However, that ITER has not yet nailed down this matter, yet T breeding should be a fundamental requirement of a 'proof-of-principle' DT-burning machine, (not our Dan Tibbets - he wouldn't be a sustainable source of fuel!!) suggests to me a degree of lax-ness of intellect over the actual objective of ITER.

ITER is "build-to-cost". If they don't give the money you need to do the job right, then you have to decide what is least painful to leave out. They don't have power turbines, either, although nobody thinks it will be a big problem when the time comes to add them.

Personally, I think it is easy to critique it - ITER is an all-out engineering solution to problems that would exist for a working machine. Why not just make a working machine out of bog-standard materials, (e.g. stainless steel rather than beryllium), forget the breeding, forget the SC magnets, forget it all and just build a go-darned machine that actually pumps out net-power neutrons on demand! Prove it can be done first, before spending vast wealth on materials and material development!! No-one would question funding materials' engineering after that! (What is left would need to be disposed of more carefully, but once done it would secure an infinite amount of funding for the improvement of those materials.)

One line of argument is that we should be doing more materials development because it takes a long time and otherwise we will some day know how to build a working tokamak power reactor but we won't know what to build it out of. That would delay the glorious day of fusion power by a couple valuable decades.

As I say, I think ITER is actually missing some real thinking in what its objective *could most usefully be*. It is a somewhat blind 'do everything' mass of an idea which has now gained an uncontrollable momentum.

I thought at the time (I'm not sure what I think now) that ITER should have been made 50% more expensive so that it could do enough to be the last step before a prototype reactor. As it is, you will need a "DEMO" to test things like tritium breeding and power production before you can actually build a prototype. That will also cost a decade or more.

I'm not quite clear what you're advocating. Should ITER do less or do more than planned? Both paths have been suggested, but I think the compromise taken is not out of line.

[Art Carlson]

MSimon posted a graph of the p-B11 fusion cross section above. To answer the question of how much a polywell gains by being mono-energetic over being Maxwellian, one would have to find <sigma*v>, once averaged over a monoenergetic but isotropic distribution, and once over a Maxwellian distribution, comparing the two results at constant plasma pressure (not forgetting the electrons). Who feels up to that?

[chrismb]

I'm not quite clear what you're advocating. Should ITER do less or do more than planned? Both paths have been suggested, but I think the compromise taken is not out of line.

Clearly, I'm advocating that it should do less - it should singularly focus on kicking out over-unity neutrons continuously. As you say, there HAS to be [at least] a 2nd round anyway because of the limited scope of this one, so why go a kinda 50:50 when the first isn't guaranteed to work when you could do a 10:90 and PROVE that it can work in the first 10%, thereafter guaranteeing future funding.

If you need to prove a device works and it requires "function A" and after that it requires functions "B, C, D and E", then why on earth would you build a prototype to test out ONLY the functions A B and C? Either test the critical function A, or build one that does A, B, C, D AND E! I have no problem with contemplating a project for all functions, double the budget and go straight to DEMO, sure, if the plan is there then do it, but ITER is a half-way house that I believe will end up costing 30E9 euros and STILL NOT demonstrate a grid-power-producing reactor can be built.

If you want to get a man on the moon, would you try to launch the first test rocket with a prototype lunar lander attached to it? 'Course you bloomin' wouldn't - you'd see if you could get a rocket into space first!! If you can't do that, why worry about building a lunar lander?? Conversely, if you think you know enough to go straight for the moon, then just do it! ITER has a deep ambivalence about its objectives that is not immediately obvious unless you stop and say to yourself "hey - I'm paying for that and I'll be paying for DEMO aswell - isn't there a better/cheaper way to do ALL of this?" I would go so far as to suggest the big part of the budget on materials is JUST a political manoeuvre so that all the partners can feel they've got research work they can do in their own countries. It's not necessary at this stage, is a distraction to the principle objective of showing a continuous 'burning' plasma is possible, will ultimately lengthen the overall timeline to the completion of DEMO , and finally will make funding DEMO much harder because there WON'T be an outcome from ITER that will entirely prove that DEMO should be built - there will STILL be arguments over viability even if ITER hits the objectives of its currently-planned-for experiments.

[chrismb]

The Li6+n->He4+T reaction is, of course, a 1:1 proposition.

There are plenty of other Li6+n outcomes, and so also Li7+n. If Li6+ a thermal neutron gives the T then I guess that will dominate, but up at around the 14MeV neutron energy there are plenty of other outcomes, all around the 0.1barn cross-section, by the look of it. 2n+Li5, 2n+p+He4, Li7+hv. there's even a p+6He [->p+(e-)+Li6] outcome with a cross-section peaking at ~0.04 barn at 4MeV. Do you have a link for a review on likely ratios of 14MeV neutrons into a natural lithium mix, and/or with other metals? I guess there must be something, given the importance of this for viability of DT burning.

[Art Carlson]

Do you have a link for a review on likely ratios of 14MeV neutrons into a natural lithium mix, and/or with other metals?

Sorry, I don't. All I've seen as a plasma physicist with an interest in reactor design is what I said above: the simple Li6 and Li7 reactions, and occasionally neutron multipliers. I know/believe that detailed, 3-D neutronics calculations using realistic materials and geometries have been carried out, but I've never looked at them myself.

On the other point, I don't find your argument convincing, that either A or A+B+C+D+E should be funded but nothing in between, but I don't think it's fruitful to argue about it. The ITER design, good or bad, is cast in concrete and nothing will change it. If we ever get to examine evidence that the polywell has better-than-cusp confinement, then the argument might get interesting again.

[chrismb]

On the other point, I don't find your argument convincing, that either A or A+B+C+D+E should be funded but nothing in between

But I have presented an argument. What argument successfully argues for mealy-mindedness and whimpish half-measures?

Evidently, the absence of an argument must be intrinsically assumed to be less convincing than an argument that is actually presented. Therefore your reply is even less convincing!

Arguing on cost or time would be a start, but for ITER even this cannot be claimed. Destined to be forever over time, over cost and lacking in a certain 'élan', ITER's objectives are a luke-warm agglomoration of International committee-like musings of intellectuals and has all the hallmarks of lacking visionary leadership necessary for such an undertaking for this possibly most important of all undertakings of mankind in the 3rd millenium; the goal of viable fusion energy.

Could that lack of leadership stem from the fact that those in charge know they will be dead before the end of the project, and that's if it even goes to plan? I would not be so cynical, there are good scientists at work on the project. But when has any committee (let alone one of scientists) ever come up with a good idea? Good ideas come from individuals, and tokamak needs some good ideas. It's still a sputtering unit-seconds squirt of uncontrollable neutrons with a flat-top inductive ramp that can only go one way before the primary saturates. Self-sustaining banana orbit are still a fantasy. ELM control is still a fantasy. A proper understanding of H-mode is still a fantasy. Divertors that can take the 50MW/m2 that would be needed in DEMO are a fantasy and exceed the performance of all substances known to man. ITER alone is going to exhaust a few year's supply of beryllium in one go, yet we're going to make hundreds of these power stations? The SC magnets use liquid helium which is a finite resource with less that 70 years of know reserves (at current consumption, imagine if we REALLY started building lots of SC tokamaks). The magnetic field in JET takes 1 GJ to initiate, yet with the production of 20MJ of neutrons in one pulse at its best this is claimed to be a net-energy output??

Yet still, those on the research industry gravy train shrug and grin and say 'well, we've got to try to see if we can do it first'. What? Try what? A, B, C, D and/or E??? What is it, the REAL objective of ITER? What will ITER do that JET and DEMO couldn't do together?

Generating innovation in a decades old project with decades to go is almost impossible - moribund outlooks are guaranteed by those old-and-bold scientists who tell the young'uns that they've tried everything before, and the young'uns believe it and grow up to say the same to the next generation. And so they are destined forever more to fiddle around the edges (literally!) with 1950's ideas trying to make them work by ever more layers of complexity on ever more layers of complexity in the design.

[Roger]

So it appears we're talking about a lot more than just 3 neutrons from four test runs of one machine.

By memory, detectors neutron counts......5

but something like 106k fusions? ... no?, in the best run of WB-6?

[IntLibber]

I had understood p-B11 temps to be challenging for tokamaks, but I'm not clear on why exactly. Possibly it was just the cyclotron radiation issue, or maybe just high temps in general relative to their difficulty in IEC.

As I understand and recollect it, the thermal tail problem is the side reactions you get because the distribution is Maxwellian, which make the process somewhat less aneutronic. Someone did a post on this a while back, pointing out the conventional calculations for aneutronic fusion assumed such a distribution.

The problem for tokamaks is they acheive fusion by temperature and pressure, with no consideration for velocity. All the plasma in the tokamak is residing at the same high potential all the time. Polywell causes ions and electrons to be constantly moving out of the lowest potential up into these arcs so they are constantly passing through the zero potential well. They are meeting at a point of lowest kEv potential, so they are not suffering losses at that point, the bremstrahlung would be occuring if the point at which the particles met in the center was at a very high potential.

Essentially a polywell is like having a ton of small particle accelerators all shooting beams at the center at the same proper velocity to fuse.

[KitemanSA]

So it appears we're talking about a lot more than just 3 neutrons from four test runs of one machine.

By memory, detectors neutron counts......5

but something like 106k fusions? ... no?, in the best run of WB-6?

Dang I wish we had the WB7 data!

[MSimon]

And so they are destined forever more to fiddle around the edges (literally!)

That got a big LULZ out of me.

What is really amusing is that the ELM problem has been rather well known for a long time yet even proposed solutions are an ITER afterthought for a design that has been in the works for a decade. i.e. we think we can fit some SC leaks in the current design.

I still like what Nick Krall (a fan of Bussard's work and considered one of the top plasma experts in the world) had to say on the subject:

"We spent $15 billion dollars studying tokamaks and what we learned about them is that they are no darn good."