tokamak versus fusor efficiencies.

[Chris Bradley]

I've been keeping my eye out for information to enumerate the actual energy input into a tokamak pulse, so as to get somewhere towards a realistic comparison with fusor efficiencies for Steven S.'s table.

I've had a recent piece of information off someone who has told me that the total energy needed to energise the magnetic coils in JET is approx 1GJ.

(I'm not sure anyone knows an exact figure, because it gets lost in all the other 'noise' of the huge energy expenditure in driving all the other systems. The 'out-of-the-wall' rated power consumption per pulse is 10GJ, by JET's press releases.)

Now, 1GJ of magnetic energy is some lump of energy! Considering that the total energy output for JET's best DT pulse was 22MJ of neutrons, it kinda puts the idea of 'break-even' into context.

So this piqued my interest and I went looking for info on ITER as well. It turns out that the energy required to energise ITER's magnetic coils to max is quoted as ~40 - 46GJ.

When discussing the energy/power Q for tokamaks, all the talk I have ever seen has been directed to the instantaneous ohmic/RF/NBI power input into the plasma compared with the neutron stream out. This compares up towards unity, which is widely reported.

But let's talk energy. In JET's best ever DT pulse, this looks to have been somewhat over 50MJ worth of ohmic/RF/NBI energy for 22MJ of neutrons out, giving a Q of about 0.4. But what about this whopping magnetic field generation? 1GJ? This puts it at 22MJ/1GJ (not even counting the other power inputs), or Q=0.02.

And what about using DT instead of DD in a fusor. Well, there is an established empirical (and theoretical) relationship used in these experiments to relate the two - a factor of 200. So, JET's equivalent best pulse, using an extrapolation to DD, is an efficiency of 1E-4, putting it about 4 orders of magnitude better than a fusor.

So in JET it would take about 60 seconds of continuous output at its 16MW best-seen transient peak output, and for ITER it is about 80 seconds of its rated output. Then tie in the fact that if the neutron output were to be converted back into electrical energy for the magnetic fields, say at 40%, that means ITER has to run at its 500MW rating for at least 3 minutes before it pays back the energy expenditure of generating the magnetic field.

The argument is, of course, that once the magnetic field energy is energised, so a continuous operation would not then need a repeat of that energy. That may well be the case, but one thing it shows is that pulsed operation of less than 3 minutes at a time per pulse is non-viable, and tokamaks have only been shown to work as pulsed devices to date. Continuous operation is still only a theoretical possibility - as far as I understand it - and requires an as yet only-theoretical thermally-driven bootstrap current to take over from the initiating primary-circuit plasma current that currently can only be operated for a matter of seconds.

So I reckon the best tokamak efficiency has been show to be, DD equivalent, as Q=1E-4.

If anyone has further info, or corrections to my understanding, I would be interested to hear.

best regards,

Chris MB.

[Wilfried Heil]

The plasma in a Tokamak is held together by the magnetic field of a toroidal current, which flows through the plasma itself. This current is induced by a changing external magnetic field. When the plasma ignites, it becomes the single loop secondary winding of a gigantic transformer. To keep this current alive, an external poloidal magnetic field is slowly ramped up, until the structural limits of the magnets are reached. This is then also the end of the run. Thus the Tokamak is doomed to single shot intermittent runs, unless the plasma current is constantly excited by additional means.

It is possible to build a machine with a different topology - the Stellarator - that can contain a plasma indefinitely. The downside is that it has a larger surface area of the plasma, which amounts to higher losses, while a Tokamak can be nearly spherical and relatively compact.

"Wendelstein 7-X" - Germany
>http://www.ipp.mpg.de/ippcms/eng/for/be ... index.html
"LHD Large Helical Device" - Japan
>http://www.lhd.nifs.ac.jp/en/
The purpose of any of these machines is not to reach breakeven between fusion energy output and all the external power sources needed to sustain it, but to study the containment of the plasma.

The goal that has been set for ITER is more modest than that for a power reactor. The hopes are that it will be possible to reach and define the as yet unknown engineering conditions for _ignition_ (and self-sustained burning), in which case there will be a breakeven between the energy deposited in the plasma by charged fusion products and the energy which is lost from the plasma at the same time, under otherwise static conditions.

How much energy has to be spent to set up the magnetic fields is irrelevant, at least for a research reactor. This can always be improved upon later, or built in a completely different way. JET uses water cooled copper coils.

I very much doubt that a future power producing reactor will be of the Tokamak type, but rather of some other design which can operate continuously. It will then be the task of engineering to build such a machine, after it becomes clear that the conditions needed for ignition exist and it’s parameters can be properly defined.

Another outcome, which is also possible, would be that even for a plasma of the size in ITER, the losses could still be higher that the fusion heating. But if it works - then the successor to ITER might well be a Stellarator.

[Steven Sesselmann]

Chris,

Thanks for putting in the effort to bring us those figures. It seems to be consistent with the figures quoted in the report by M. Keilhacker et al.

It just makes you wonder how the efficiency of a fusor would compare, if the same amount of money was spent building one.

I think it is quite likely that a $10,000 table top D+D fusor will soon break the 1e-8 barrier (Carl, we are waiting.., which would be only 4 orders of magnitude lower than the worlds biggest multi million dollar effort.

More importantly, the fusor is a continuous mode reactor, and does not rely on any theoretical bootstrap current to work.

I think it would be a bit unfair to list JET in the fusor efficiency table with a Q of 1e-4 but it could get mention it in the text below the table.

Steven

[Chris Bradley]

Wilfried Heil wrote:
> How much energy has to be spent to set up the magnetic fields is irrelevant, at least for a research reactor. This can always be improved upon later, or built in a completely different way.

Not sure I see any evidence for the 'can always'. But some progress has been made, granted.

Perhaps, though, one could argue that the electron conduction losses in a fusor are irrelevant also. Many hopefuls who arrive here maintain the idea that electron losses 'can always' be eliminated as well!!

>
> I very much doubt that a future power producing reactor will be of the Tokamak type, but rather of some other design which can operate continuously.

Not sure I understand you. The current aim IS to get the tokamak running continuously, through bootstrapping banana orbits arising from thermal transport and trapping ions within the plasma.

This IS the great white hope for tokamaks. The aim of ITER to produce a burning plasma (that is self-sustaining, as opposed to 'ignition' which is different) is only part of it, other aspects are to achieve 'steady-state' conditions, which fundamentally means these self-sustaining toroidal currents.

Anyway, point being there *are* magnetic energy demands when setting up the fields to confine the plasma and I am suggesting it is legitimate to include those if comparing with all the ionising losses associated when setting up a fusor's electrical field, so 1E^-4 is my suggested equivalent best efficiency (wrt DD/fusor) demonstrated to date by a tokamak, whether or not it is intended to be an experimental device.

best regards,

Chris MB.

[Chris Bradley]

Ok. Took a bit of digging around because data on LHD fusion shots seems very thin on the ground.

Total stored magnetic energy = 1.63GJ (ref: http://www.nifs.ac.jp/report/nifs051.html)
Total input of heating energy during a pulse = can't find that specifically, but what would it matter relative to 1.6GJ?

Total neutron output for DD per pulse = 2.4E15
(ref table 1, http://www.jspf.or.jp/PFR/PDF/pfr2008_03-S1024.pdf - note, I'm pretty sure they've got the power indeces mixed up between those two bits of data, remember the x200 factor)

LHD EFFICIENCY = [2.4E15 x 2.5E6MeV x 1.6E-19J/ev]/1.63GJ = 5.9E-7

Actually, I've just realised an error in my conversion to DD outputs for JET. The 200 factor is a rate for neutrons, not for energy. So because the DT neutrons are 6 times more energetic than the DD neutrons, so the 'DD-equivalent' efficiency for JET should have been 1.7E-5.

Sorry about that!

Again, anything I am misunderstanding or erroneous data, please correct me.

best regards,

Chris MB.

[Wilfried Heil]

>Total neutron output for DD per pulse = 2.4E15
>(ref table 1, http://www.jspf.or.jp/PFR/PDF/pfr2008_03-S1024.pdf - note, I'm pretty sure they've got the power indeces mixed up between those two bits of data, remember the x200 factor)

Nope, that article contains a lot of typos and is a mess to read, but table 1 is correct.

LHD is intended to produce 2.4x10^17 neutrons from D+D fusion and 4.3x10^15 neutrons from D+T fusion per 10s shot. The fusion fuel used is deuterium only, the DT neutrons come from the fusion of some of the tritium which is produced during the fusion run, hence the lower number.

[Chris Bradley]

Blimey! So if the x200 reactivity DT:DD is applied as a prediction to them running DT, then they'd end up with 5e18 neutrons in a pulse, or 107MJ? JET has only managed 22MJ in a DT pulse.

[Richard Hull]

I enjoyed seeing the mathematical efforts poured into the above comparisons. I also appreciate the efforts of all the dedicated physicists, engineers and technicans from our own humble IECF group and that of the ITER effort. Yet, for all that, I remain bemused over discussions of energies that do not have a continuous time running possibility of feeding the energy hungry bull dog in anything like the next 50-100 years, barring the luckey donkey.

Fusion is no where near the point where it is just a matter of engineering refinements. Fission was reduced to mere engineering the instant the fission based CP-1 pile went active under the bleachers in Chicago only 4 years after the discovery of fission, itself.

Only 15 years would separate the first ever research pile and the first fission electrical power going to the grid to millions of people. The fusion power clock has been ticking away since d-d fusion was discovered in 1934 by Oliphant and Harteck. After 75 years, still no fusion power and little possibility of same for the next 50 to 100 years.

The magnitude of spectacular claims for fusion energy must be considerably watered down by the paucity of results obtained, considering the billions spent.

Richard Hull

[Retric]

The problem with your assessment is you can easily recover the energy stored in the magnetic field. They are not doing that because the value of the energy is less than the cost to build a system to recover it. Just like they are avoiding building a steam engine to recover the energy produced with ITER.

JET can do ~30 pulses a day and ITER is expected to do 10 min pulses with a high Q value 5 - 20 and do steady state operations at a lower efficiency level. But assuming a similar recovery time ITER would only ever operate for around 5 hours a day so it's power output is not that big a deal. The next fusion system should focus on increasing efficiency at every level.

[Chris Bradley]

Jkirby wrote:
> The problem with your assessment is you can easily recover the energy stored in the magnetic field.

This would be a problem with the interpretation of my assessment than the assessment itself?

I am, simply, trying to evaluate an honest-to-goodness, now, real engineering calculation of the efficiencies of current tokamaks wrt a fusor. Energy in versus energy out.

Besides, would the 'easily recover' mean 100% efficient recovery? The heat in a fusor can be easily recovered as well!

"Tokamaks are only experimental"?.... fusor's can barely be classified as 'experimental', better they're described as mostly cobbled-together PRE-experiments!

Today, as we speak, there are fusion devices kicking around this planet. If the question is 'what are the proven efficiencies of these devices' there is a conversation that can be usefully had, as you are saying, but it's just not an answer to THAT question! The answer is a simple, neat, dimensionless number for each device.

best regards,

Chris MB.

[Retric]

I would say 95% or more, having a few grams of plasma cool down inside the system has little impact of the magnetic field. The magnetic field is basically just electricity so recovering that energy has more to do with storing electricity and how efficiently you want to do that vs any complex issue. The only problem is to restart the system you can't have start over and rebuild the magnetic field so the best solution would be a high efficiency Flywheel which they are probably already using to reduce the grid demand on start up.

PS: When you start talking about losses on that level you need to have a lot of heat show up somewhere and dumping that much heat as soon as the system fails would melt stuff.

[Chris Bradley]

Wilfried Heil wrote:
> LHD produces 2.4x10^17 neutrons from D+D fusion and 4.3x10^15 neutrons from D+T fusion per 10s shot. The fusion fuel used is deuterium only, the DT neutrons come from the fusion of some of the tritium which is produced during the fusion run, hence the lower number.
>
> They normally use a mix of light hydrogen and deuterium, in order to keep the fusion rates low.

Wilfred,

When you posted this, my immediate reaction was to, simply, ask them the question. So this is 'FYI' wrt the discussion in hand:



>>>>>>>>>>

----- Original Message -----
From: "広報室" <nifs-kouhou@nifs.ac.jp>
To: <chris_macdonaldbradley@yahoo.co.uk>
Sent: Friday, January 09, 2009 12:32 AM
Subject: National Institute for Fusion Science


> Dear Chris Macdonald-Bradley,
>
> Hello. This is the Public Relations Office in the National Institute for
> Fusion Science.
> First of all, we thank you for your inquiry the other day.
> Attached is our answer to your question. We hope that they will be of your
> help.
>
> ___________________________________________________________________________
> Thank you for having interest in our experiment.
> Your question is ‘what is the highest number of neutrons ever emitted by
> LHD during an actual experimental pulse?’. Answer is that we have never
> emitted any neutron by LHD during actual experiments. The reason is that we
> have never used deuterium gas in LHD experiment. We only used hydrogen,
> helium, neon and argon gases in LHD from the start of LHD project. Paper
> which you found is described about the activation by neutron in future
> experiment in LHD. Now we are planning the experiments using deuterium gas
> in LHD to get more powerful plasma. In these future experiments, several
> times 10 to the 17th neutrons per pulse are expected.
> If you have further questions, please ask us without hesitation.
>
> Yours sincerely.
> National Institute for Fusion Science

>>>>>>>>>>>>


So the answer is that LHD cannot yet appear on Steven's table. It is, as yet, not a fusion experiment but a plasma containment experiment.

best regards,

Chris MB.

[Dan Tibbets]

I just read this thread. A couple of items.

I believe the record for the highest Fusor output was by by Robert Hirsch back in the late 1960's. Uing DT fuel he got ~ 1 trillion neutrons per second.
http://www.brian-mcdermott.com/fusionproof.htm

Do I understand the comments correctly. Once Tokamacks' reach ignition, the reaction becomes runaway- power increasing untill the magnets cannot contain the plasma, and the machine shuts down ( ie no longterm steady state possible, only runs limited by the reserve capacity of the magnets)? The power out cannot be throttable by fuel feed, magnet 'detuneing', etc?


Dan Tibbets

[Chris Bradley]

The discussion above didn't quite explain the actual issues/fields in a tokamak.

There are two magnetic fields generated.

Firstly, a poloidal field generated by the current being induced into the plasma. The plasma forms the secondary of a huge transformer. The core of that transformer is initially saturated one way. Then the power is reversed and it swings all the way to saturation in the opposite magnetic direction. During that ramp up of magnetic field, the plasma current is driven in one direction. This can, clearly, only last for a finite time - the time it takes the core to reach opposite saturation. If it were then reversed again (as it would in an AC transformer) the plasma current would also reverse and the whole thing would disintegrate in a puff of instability. So tokamak pulses to date are only as long as the induction pulse is ramping up the core to saturation in one direction.

The hope is that once the plasma initiates various thermal transportation within the torus so it will self-drive this current, so this saturation pluse will be just a 'starter motor'. Unfortunately, this starting action is as far as tokamaks have got to, the rest is still mostly theory.

This really should be kept in mind. There has been 50 years of tokamak development but it's like having spent 50 years building a car whose engine you can crank over but it doesn't actually start!!

Secondly there is a toroidal field. This is by a big solenoid all the way around the torus (which will be the bit which is super-conducting in ITER and is the case in a few other machines). This field 'confines' the ions to magnetic surfaces that are nested within the shape of the outer.

There are additional poloidal coils in most machines aswell, but these are mostly just to 'tune up' the poloidal fields that the inductive pulse generates and shape the plasma's cross-section to that desired.

The first induction field takes up a few MW to drive it (and the inductive 'pulse' lasts for 10 seconds or so). The second, the generation and creation of the toroidal field, takes 1GJ in JET and more in others. It is this toroidal field energy I have been toying with as something which really hammers the idea that these machines have got anywhere near over-unity efficiency before, though some folks would like you to believe that they have (as they don't count the energy put into these fields).

Once the thing runs continuously, sure, this is just a 'start up' investment of energy, but my thread here is trying to establish and compare current, today's, real genuine fusion machines' performances, warts'n'all, not just speculations on the future. (like - "I've built a really fast car, but it only cranks over at the moment" - this doesn't add up to an engineering reality to me and I can legitimately ask 'what is the fastest car today' and the answer cannot be a car due to be built or finished off tomorrow.)

[Wilfried Heil]

Chris -
the fusion data are _future_ operating parameters, from an engineering white paper which estimates the probable activation from those future runs.

Dan -
a Tokamak's plasma confinement time is limited, because the poloidal field (from axial coils in parallel with the plasma) can't be ramped up indefinitely. The resulting helical field stabilizes the plasma, but is induced only as long as the magnetic field changes. This is independent of whether the plasma shows fusion or not.

It may be possible to sustain this field externally in an inefficient way by "helicity injection" or by making use of thermal currents which would then draw a lot of power from the plasma.

So the Tokamak is a neat and simple design for short plasma experiments of a few seconds duration each, but may not be the best choice to begin with if continuous operation is desired.

[Richard Hull]

Chris Came up with a brilliant analogy that I will latch onto now as it is so majestic and yet simple in concept. This relates to the engine analogy of the last 50 years of fusion research. (it is actually 60 years if you include Project Sherwood, the rather secretive study of fusion reactor systems starting in the late 40's)

I would like to note that the first engines were one lungers (single cylinder jobs). All were different in concept as no one claimed to know how to make an engine. Stellarators, accelerators, mirror machines, torus types, etc.

When the one lungers wouldn't start, a tiny bit of money flooded in and they figured two cylinder machines might have a better chance of starting. They never did, of course.

By the late 60's, it was apparent that only hot fusion would work in a 4 or 6 cylinder multimillion dollar fusion motor. Still no go.

By the early 70's it was settled that only a tokamak or its derivative would work in a 8 or 10 cylinder fusion motor. At least 4 different 10 cylinder, near billion dollar each 10-15 year motor construction efforts dawdled into the late 90's. Some 2-3 second attempted starts turned over, but tended to blow the "mains" or "drop a valve" during the brief, abortive turnovers. Turbo's were added in the form of massive magnetic systems and, currently, the ultimate multi-billion dollar inverted V-16 turbo charged motor is now on the dynamometer as ITER and almost ready to attmept to be started, It has the best valves and roller bearing "mains". We are forewarned, however, that if this thing starts and runs for a decent test run, that "this is not and will not be a production engine." It is sort of a "motorama", auto show, eye-popper. We will have to wait many years for a real production motor, so get used to it. AND this assumes this V-16 starts and runs with no real issues like all the other motors that never started or started but crapped out just as, supposedly, useful RPM and horsepower developed shortly after crank over.

Fusion, it appears, never had a tin lizzie, ( model T ford), that actually did what a tin lizzie could do.... Actually supply motive power, transport goods,etc., albeit rather poorly. Fusion never even had a model A ford. Instead, Fusion appears to need to be born big, really big!

To many folks way of thinking, if you can't make a little tin lizzie, you should not be trying to make a Ferrari. It just seems reasonable.

I really hope I live to have to eat my words, but I fear I will not make it to fusion power on the mains. It will probably reside where it has always resided; in the minds.

Richard Hull