Do we need breakeven to consider fusion power sources?

It may be difficult to separate "theory" from "application," but let''s see if this helps facilitate the discussion.
Post Reply
Calmarius
Posts: 17
Joined: Tue Apr 08, 2014 4:19 pm
Real name: Dávid C

Do we need breakeven to consider fusion power sources?

Post by Calmarius »

I did some back of the envelope calculations.

So the fusion gain factor is

G = Q_fusion / Q_in,

G: the factor
Q_fusion: fusion power released
Q_in: input power.

In a practical power plant the setup would probably turn everything into heat and feed into a heat engine. But probably better designs possible (eg. direct conversion with decelerators?)
Both the input power and the fusion power will be turned into heat. So the total heat generated is:

Q_fusion + Q_in = Q_in*(1 + G)

The efficiencies of real heat engines (Chabal-Novikov efficiency):

eta = 1 - sqrt(T_c / T_h)

This is much smaller efficiency than of the (impossible to build) Carnot engine.
And the efficiencies of the current power plants corresponds well to this value.

So in order to make it self sustaining, we need to feed back the electric power into the device:

Q_in = eta * Q_in * (1 + G)

If we solve for G, we have:

G = (1 - eta) / eta

Or for eta, we have:

eta = 1 / (1 + G).

I don't know what was the highest gain factor confirmed. So if you have exact numbers please let me know.

On breakeven the eta required is 0.5. Which means the hot side of the heat engine should be at 1200K. Look like it's doable.

If we consider that all parts of the heat engine is made of tungsten and calculate with 3000K on the hot side, G=0.5 would be enough to be sustainable.

On the forums I read that one would need a gain factor of 5-10 to make fusion power practical... So am I missing something?
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: Do we need breakeven to consider fusion power sources?

Post by Chris Bradley »

'Sub-breakeven' devices have been discussed here and elsewhere at length, both for fusion devices, accelerators and fission piles.

In the case of deuterium fusion, the neutron emissions of a sub-breakeven reactor could be used to make fertile material fissionable and could be used to 'burn up' nuclear waste and long-lived actinides with a highly 'profitable' and practicable thermal output as well as helping eliminate radioactive fission waste.

Ultimately, though, you have to get within reasonable sight of breakeven at least. Let us know when you make one that can do better than 0.1% fusion energy output to energy input .....
Calmarius
Posts: 17
Joined: Tue Apr 08, 2014 4:19 pm
Real name: Dávid C

Re: Do we need breakeven to consider fusion power sources?

Post by Calmarius »

JET claims they reached the gain factor of 0.6. Unless they are lying I would say they are almost there...

It seems there is a confusion about the breakeven, and I'm a bit confused too. When do we reach the breakeven? When fusion_power_released == power_used OR when electric_power_produced_by_the_reactor == electric_power_consumed?
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: Do we need breakeven to consider fusion power sources?

Post by Chris Bradley »

Well, they achieved 18MJ of fusion energy for ~10 seconds for 22MJ input into the plasma, but it also required 1GJ of magnetic energy put into the confinement fields first. They don't seem to trouble themselves with that little detail.

I'd suggest that to make what you are looking for worthwhile one would need to demonstrate a continuous fusion power output and one which approaches a reasonable fraction of breakeven whilst taking into account all energy inputs, would you not agree?

FWIW, ITER will require 41GJ of magnetic energy before it can begin to confine a fusing plasma. At a design output of 500MW, it has to run for at least 2 minutes before it has even generated the energy required to generate those magnetic fields.

Also FWIW, one tonne of TNT is around 4.1GJ of energy, so the magnetic field in ITER is equivalent to 10 tonnes of TNT. Now imagine that field collapsing all at once and giving up its magnetic energy to the confinement vessel. .... I wait with baited breath to see if ITER can survive its first plasma collapse!...
Calmarius wrote:It seems there is a confusion about the breakeven, and I'm a bit confused too. When do we reach the breakeven? When fusion_power_released == power_used OR when electric_power_produced_by_the_reactor == electric_power_consumed?
The reality is that practical fusion power is so far off it makes little difference! The convention is when the input energy into the plasma equals the fusion energy generated. But that is just a 'target' so far. Reality, as you allude to, is that it needs to be Q=10 (fusion energy created is 10 times the input energy), so breakeven at Q=1 is only a milestone and is no practical result per se. It's a bit like hitting 100mph in a car or celebrating the year 2000 - it's just a number with little practical bearing.
Calmarius
Posts: 17
Joined: Tue Apr 08, 2014 4:19 pm
Real name: Dávid C

Re: Do we need breakeven to consider fusion power sources?

Post by Calmarius »

Well, they achieved 18MJ of fusion energy for ~10 seconds for 22MJ input into the plasma, but it also required 1GJ of magnetic energy put into the confinement fields first. They don't seem to trouble themselves with that little detail.
I thought they were thorough enough to measure the input power vs output power in watts. So they are basically lying.
Also FWIW, one tonne of TNT is around 4.1GJ of energy, so the magnetic field in ITER is equivalent to 10 tonnes of TNT. Now imagine that field collapsing all at once and giving up its magnetic energy to the confinement vessel. .... I wait with baited breath to see if ITER can survive its first plasma collapse!...
What would happen? Would it explode? It would be the most expensive fireworks in human history lol...

Valid point anyway. Seeing many fusors in youtube quietly glowing I forgot about the fact that tokamaks are still struggling with plasma instabilities.
User avatar
Dave Xanatos
Posts: 108
Joined: Tue Apr 22, 2014 1:59 pm
Real name: Dave Xanatos
Location: Western Massachusetts, USA
Contact:

Re: Do we need breakeven to consider fusion power sources?

Post by Dave Xanatos »

Isn't what we are really looking for called Ignition? Correct me if I'm wrong (and I know you will, and I probably am! :) ) - but as I understand it, ignition is the point at which the usable, probably electrical, energy harvested from the output is sufficient to power the reactor itself. In other words, self-sustaining as long as the fusion fuel is being continually fed, much like an ICE which, once going, will power itself as long as fuel is available.

Wow... what a day that'll be.
It would take decades of work, by thousands of scientists, in a particle accelerator powered by dump trucks of flaming grant money! - Professor Farnsworth/FUTURAMA
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: Do we need breakeven to consider fusion power sources?

Post by Chris Bradley »

Calmarius wrote:I thought they were thorough enough to measure the input power vs output power in watts. So they are basically lying.
Not sure who is 'lying', I've not seen any JET scientists claim anything not essentially true. Possibly disingenuous, but not false. What do you think you've seen that is a 'lie'?

Dave Xanatos wrote:Isn't what we are really looking for called Ignition?
'Ignition' is the point at which the other, thermal products from the fusion are sufficient to sustain the plasma temperature so that there is no need for an external power input into the plasma (not counting sustaining the magnetic field and ancillaries). This is like setting a dry piece of wood alight, it will continue to burn under its own exothermic output. For DT fusion, this is higher than the fusion power by a factor of 5, because only 20% of the fusion energy (alpha particles) is 'thermal' and the remaining 80% is the neutron flux that will heat an external thermal working blanket.

'Burning' is where external energy is pumped in and it results in more energy out than in, but it'd stop if you take the external energy source away. This is a bit like putting a piece of wet wood into a gas burner - it will add to the heat, but will fizzle out and not keep burning by itself.

Burning is a form of 'energy amplifying' in which the process requires an energy input, but overall is a net positive energy process. You might be interested in looking up Rubbia's 'energy amplifier', which is what the term has come to commonly reference, which uses neutron spallation with fertile fuel. It is being researched academically and commercially as a viable, sub-critical (therefore intrinsically 'safe') reactor.
prestonbarrows
Posts: 211
Joined: Sun Jun 24, 2012 1:27 am
Real name:

Re: Do we need breakeven to consider fusion power sources?

Post by prestonbarrows »

The hand-wavy detail you guys are missing is that JET and most other existing systems use resistive external coils to produce the vacuum field. This means that basically all of those gigajoules of energy input into the coils wind up as waste heat. In a legitimate power-producing tokamak, it's usually assumed you would have superconducting coils to drive the vacuum fields. These still require stupid amounts of energy to spool up, but basically none of it is lost as heat and the vacuum field just coasts without the need for further input power. The energy will stay stored in the magnetic field, like a taught spring, as long as they are kept cool. It could concieveably be recovered when the machine is powered down also. Think of a giant near-ideal inductor. This is why most numbers quoted in today's experimental reactors omit the power needed for the external coils.

This is similar to charging a large capacitor bank or flywheel energy storage or pumping water up into a hydroelectric storage reservoir. The input energy is not 'lost' just transferred into potential energy waiting to be used later. The only losses come from series resistance in the conductors or frictional losses as the case may be.

Most of the workhorse experimental tokamaks today are inductively driven. They essentially are set up as a single-turn stepdown transformer (with the plasma in the torus as the secondary) to drive the huge toroidal currents through the plasma required for confinement. Since you can't keep ramping up the current through the primary side of the transformer forever, there is an inherent run time limit on such devices. Again, a legitimate power-producing tokamak will likely need non-inductive current drive mechanisms such as neutral beam injectors to achieve anything more than a few seconds of run time.

Addressing these issues, among others, is the main goal for the upcoming ITER project if they can sort out their bureaucratic issues first...
User avatar
Richard Hull
Moderator
Posts: 14992
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Do we need breakeven to consider fusion power sources?

Post by Richard Hull »

Big, Big science like ITER, being multi-national, will suffer mightily as the science is at the mercy of governmental budgets, bureaucratic oversight and even the science becomes politically involved and entangled in a self-inflicted choke hold.

Will there even continue to be a stable world able to continue this long term work? Giant fusion budgets will not be the highest priority in a crumbling political and economic world infrastructure, should this become a reality in the near future. Thus, it isn't just the science and engineering of fusion involved here.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: Do we need breakeven to consider fusion power sources?

Post by Chris Bradley »

prestonbarrows wrote:The hand-wavy detail you guys are missing is that
No handwavy or even detail missing. You seem to have missed the point that the OP was discussing an experiment as 'nearly there'. Yeah, sure, if they can achieve stability for more than a couple of minutes it might start making sense to argue that the magnetic field energy is recoverable.

So the only minor matters (as you appear to be suggesting they are all readily solvable) are that tokamaks have not yet been driven by a bootstrap current, which they would require, have not yet been driven by a superconducting magnet, which seems like it would be required, have not yet had their magnetic field energy recovered, which they might not require so long as they can demonstrate continuous stability, and that's all before the fact that no one has yet even shown more energy out than in for a tokamak, let alone Q=10.

I would not regard these matters as 'handwavy' in the context of someone initially believing that the experiment is 'almost there'.

So maybe we can have this conversation again when all the 'if's' and 'should be's' you've mentioned are ever turned into a plausible level of practicability, or maybe I'll be dead by then. It is no handwavy thing to mention that a thing is not 'almost there' if you are not likely to live long enough see it, within several decades.

On the contrary, the energy expense of the magnetic field energy is the thing that others handwavy at, just as you have done. Rather than be specific about how the magnetic energy would be recovered by the coils, you flap your hand a little that someone, somewhere, some time in the future will no doubt do it. Well, if it was that easy, why don't the experiments recover that energy now?

I find it ironic that the thread discussion is specific about how much energy is required for a JET plasma run, yet you suggest that discussion is handwavy when all you do is make vague suggestions towards how these things might be solved in the future with absolutely no guarantee that they are tractable problems.

These issues are ones that tokamak experiments have been trying to overcome for decades. Decades. In fact, scientists now responsible for these experiments are two generations younger than the ones that started it. As a point of discussion, how long do you run an experiment to try to see if something will work before deciding it won't work? Under what circumstances would 'tokamak' ever be deemed a failed experiment? I'm not saying they can't make it work, I'm just wondering if there are any circumstances anyone will decide it has 'failed' and stop funding it. Is it possible that anyone will ever regard it as 'failed', or will it just go on forever?
prestonbarrows
Posts: 211
Joined: Sun Jun 24, 2012 1:27 am
Real name:

Re: Do we need breakeven to consider fusion power sources?

Post by prestonbarrows »

There are few, if any, people in the field who would say we are 'nearly there' in regards to a feasible fusion reactor connected to the grid. The scheduled timeline for ITER is on the order of 15 years before full scale tests are being run, and 30 years for DEMO. (and this is the sunny polished version they are selling their funding sources)
Wikipedia wrote:Handwaving is a pejorative label applied to the action of displaying the appearance of doing something, when actually doing little, or nothing.
From a physicists perspective, one can wave their hands and omit the power requirements for external coils because they could be made superconducting. That is 'just a problem for the engineers'. Doing so is not a lie, per say, but it is important to know the context if you are not familiar with their reasoning.

Another 'trick' you'll run into is extrapolating experimental D-D results to equivalent D-T results. This is because tritium is nasty stuff to deal with in real life and drags in extremely expensive handling systems and extra regulations. The behavior of the two reactions is well known and waving your hands at is is a legitimate approach in initial experimental stages.

No one is saying these issues are minor or will magically go away, just that in principle they shouldn't fundamentally matter and proving so is beyond the limitations of the existing experimental device. When you have exhausted what your experiment can teach you, you design and build a more sophisticated experiment which can address the shortcomings of the previous iterations.
Chris Bradley wrote:tokamaks have not yet been driven by a bootstrap current, which they would require
Neither of these points are correct. A quick search turns up "Quasisteady High-Confinement Reversed Shear Plasma with Large Bootstrap Current Fraction under Full Noninductive Current Drive Condition in JT-60U" or "Experimental study of neoclassical plasma flow and bootstrap current in the tokamak textor" among other papers. Also, bootstrap currents are a nice bonus that could boost efficiency under certain conditions, but not a fundamental requirement for a tokamak scheme.
Chris Bradley wrote:have not yet been driven by a superconducting magnet
Not true. See HT-7, EAST, KSTAR, Tore Supra, or upgrades to JT-60. In other words, most of the newest generation of large reactors.
Chris Bradley wrote:why don't the experiments recover that energy now?
Again, those with superconducting coils must do this any time they power down their large magnets. Some may dump it into a resistive dummy load or some may pass it back onto the grid but it has to go somewhere. Super conducting magnetic energy storage is a well developed and implemented idea in its own right.

Most older reactors use resistive copper coils which you can't recover energy from like you could feasibly do with superconducting coils. In the ideal end game scenario, you would never have to shut your SC magnets off. Simple copper is much cheaper and more practical on smaller scales. From a physics standpoint, an amp through a copper turn gives identical results as an amp through one of expensive superconductor. Scientists tend to look at the more interesting and unknown aspects first. Any experiment has limits on scope, current technology level and a finite budget. It dosen't always make sense to add the extra cost and time to integrate every possible feature.

Today, we are in the transition between emphasis on pure experimental physics of the components and the practical engineering of combining them into a full working system.
A physicist doesn't really care how you make a 5 Tesla field happen, an engineer does.

The newer existing reactors listed above are essentially partnered under the ITER umbrella and have been testing superconducting coils and non-inductive heating schemes for many years. The idea is ITER will take things demonstrated in JET, EAST, JT-60 etc and improve on them. Only the newest devices will have cutting-edge technology. Why dosen't my rotary phone have bluetooth?
Chris Bradley wrote:absolutely no guarantee that they are tractable problems
That's why its called research. Tokamak Fusion Test Reactor, Experimental Advanced Superconducting Tokamak, International Thermonuclear Experimental Reactor and so on.

Damn those Wright brothers for not landing on the moon.
Chris Bradley wrote:Under what circumstances would 'tokamak' ever be deemed a failed experiment?
If someone proves a fundamental physics limitation either theoretically or experimentally, or new generations of devices stop progressing and improving beyond their predecessors, or some other fusion scheme (MTF,ICF, RFP, DPF to name a few) is proven to work on a faster or cheaper scale, I suppose commercial tokamaks will have 'failed'.

So far, none of these things has happened yet.
User avatar
Dennis P Brown
Posts: 3160
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Do we need breakeven to consider fusion power sources?

Post by Dennis P Brown »

The cost for ITER is getting more and more expensive every year (still haven't even built a single real component, as far as I've seen) and I have strong doubts even if it were to cost just 10-12 billion it'd not be shorted by governments who are facing very tight budgets resulting in failure. In that sense, ITER and any future tokamacs in general are failures - simply by the fact that the cost to research dollar isn't exactly workable; that is, ITER, at present cost growth will never get finished. Any design that can't even be cost estimated within a factor of ten (!) isn't, in my opinion, a viable design - yet, as you say, it is just an experimental machine - I'd hate to see the final cost of even just a proto-type reactor.

On the other hand, as Chris pointed out, and if ITER had somehow stayed on an even reasonable budget, a tokamac fusion-breeder is capable (with currently projected performance - a big if) of producing net energy.

There is currently only one big machine that was brought in under budget and has some hope to achieve "steady state" fusion for any real length of time - the Wendelstein-7 stellarator. Whether it does ever become operational is still an open question but the basic design has been proven to work, and budget to real world cost for a finished test reactor was well understood and came in under cost estimates - tell me the last time that ever happened in a major fusion project?

While this was delayed it was primarily due to the first company unable to build the complex magnetic coils and remain solvent - that issue has been addressed, is rather well understood, and was solved.

Also, while this is a toy compared to even a proto-type power reactor, at least if it works, stellarators would appear to be a far better bet for scale up than any tokamac.
User avatar
Chris Bradley
Posts: 2930
Joined: Fri May 02, 2008 7:05 am
Real name:

Re: Do we need breakeven to consider fusion power sources?

Post by Chris Bradley »

prestonbarrows wrote:
Chris Bradley wrote:tokamaks have not yet been driven by a bootstrap current, which they would require
Neither of these points are correct. A quick search turns up "Quasisteady High-Confinement Reversed Shear Plasma with Large Bootstrap Current Fraction under Full Noninductive Current Drive Condition in JT-60U" or "Experimental study of neoclassical plasma flow and bootstrap current in the tokamak textor" among other papers. Also, bootstrap currents are a nice bonus that could boost efficiency under certain conditions, but not a fundamental requirement for a tokamak scheme.
Oh really? Neither point correct? So, running a tokamak on a fraction of a bootstrap current, or running one in a computer is a practical demonstration of actually driving a tokamak with a bootstrap current? I think not. And if not a self-sustaining current, how do you see the current being sustained? What practical example, here today, shows the route forward without bootstrap currents?
prestonbarrows wrote:
Chris Bradley wrote:have not yet been driven by a superconducting magnet
Not true. See HT-7, EAST, KSTAR, Tore Supra, or upgrades to JT-60. In other words, most of the newest generation of large reactors.
So, errr... what were the best fusion rates for these devices, or have you noticed if they are just hydrogen-only plasma experiments.

The issue of the line of conversation above was directed towards real and practicable implementation of a fusion reactor, not a fancy research plasma. Give me some details of the neutron rates out of these super-conducting tokamaks and I'll take a greater interest. Maybe they have by now, the Koreans are dead keen to get going (good for them), but I have not heard much lately and have not gone looking for it.
prestonbarrows wrote:
Chris Bradley wrote:why don't the experiments recover that energy now?
Again, those with superconducting coils must do this any time they power down their large magnets. Some may dump it into a resistive dummy load or some may pass it back onto the grid but it has to go somewhere. Super conducting magnetic energy storage is a well developed and implemented idea in its own right.
So what? Again to say, this thread we are talking here about actual stuff, not prospective and hoped-for outcomes. If it's that easy, why haven't they demonstrated it by now?
prestonbarrows wrote:
Chris Bradley wrote:absolutely no guarantee that they are tractable problems
That's why its called research. Tokamak Fusion Test Reactor, Experimental Advanced Superconducting Tokamak, International Thermonuclear Experimental Reactor and so on.
You are taking all of my responses out of context, if you think your retorts are justified, because I was explicitly talking about here-and-now actual outcomes in fusion research per the original question that '[the OP] would say they are almost there...', not wanna-be research projects.

Your aim is mis-judged. There is little point in taking pot-shots at me here because I am quite supportive of tokamak (and few here cheer such projects). But they have shown very little in the way of being 'almost there' and that was the origin of this particular discussion.
prestonbarrows wrote:
Chris Bradley wrote:Under what circumstances would 'tokamak' ever be deemed a failed experiment?
If someone proves a fundamental physics limitation either theoretically or experimentally
But that will never happen. You can't disprove a limitation in a field you have never mastered before. This is a recipe to go on forever because someone will always have another 'what-if'. As we approach the 4th generation of scientists now getting into tokamak research, it is clear that if one lifetime of research is not enough to convince you to give up then the next generation will always think they have some insight the old-fogies never did, and they'll carry on regardless.

My opinion on takamak is that it does show promise, but they are going about it in the wrong way. They are spending all their money on hydrogen based experiments and materials testing, and they still haven't shown the damned thing can get a fusion reaction going to pay back even 1/10th of the input energy. My prioritisation would be very different and focussed on getting fusion out of the thing.

In regards cost, bear in mind that JET came in on budget at around GBP 200 million 3 decades ago and is still the only reactor capable of running tritium experiments. JET's main costs have been its running costs, not its manufacture. Years ago when ITER started and it was still being pitched as a euro 5 bn project, I was alerting folks to my opinion that it would end up as euro 30 bn and I was ignored as being way off. Then it escalated to 8 bn. Then 10 bn. Now it is 15 bn looking to head up to 18~20 bn. I will stick with my prediction of 30 bn because it is an experiment with a poorly defined objective, and not really having a good plan always escalates costs.
User avatar
Richard Hull
Moderator
Posts: 14992
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Do we need breakeven to consider fusion power sources?

Post by Richard Hull »

Nice followup Chris! I cannot and dare not add anything to it. You kinda' told it like it was. Power producing fusion talks and b-ll-hit walks.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Post Reply

Return to “Fusor and/or General Fusion Theory (& FAQs)”