Accelerator-Driven Subcritical Reactors

[sanman]

There seems to be a lot of talk going on about the the new dielectric wall accelerator technology:

http://www.physorg.com/news103909668.html

http://www.aip.org/pnu/2007/split/833-2.html

http://www.technologyreview.com/Biotech/19084/

http://www.economist.com/displaystory.cfm?story_id=9537553

I'm wondering to what extent this technology could make accelerator-driven subcritical reactors more feasible?

 

[Morbius]

I'm wondering to what extent this technology could make accelerator-driven subcritical reactors more feasible?

sanman,

There will have to be HUGE gains - the currents reasonably achievable in accelerators
are orders of magnitude below what one can achieve in a critical system.

The other question is "...to what purpose?"

Why would one want an accelerator-driven system over a critical system?

There is the oft claimed advantage that accelerator systems are immune to meltdown;
but as discussed earier in this thread - that is just NOT TRUE!

Yes - one can stop the fission reaction immediately by stopping the accelerator; just
as one can stop the fission reactions immediately by use of control rods, or any of
a number of prompt feedback characteristics like Doppler broadening.

However, that's NOT the problem! Meltdowns are NOT caused by fission power; they
are caused by the decay heat of the fission products; and an accelerator-driven system
is EVERY BIT as susceptible to melting due to decay heat as is a critical system.

Dr. Gregory Greenman
Physicist

 

[ZapperZ]

As someone who is currently working on dielectric-loaded accelerating structure for particle accelerators, I find all of this to be rather puzzling.

First of all, I've already mentioned several times in this sub-forum why using particle accelerators as fusion reactor is close to a fallacy. The wall-plug efficiency is NEGATIVE!

Secondly, we are already facing significant issues in terms of dielectric structures which we are currently trying to solve. This includes breakdown issues, secondary emission, etc.. I hate to think the types of issues that would be faced if this were to be used as a fusion device.

Thirdly, there is a HUGE difference between dielectric as accelerating structures (which those links are referring to), versus using them for fusion devices. They are not even in the same specie! One simply can't blindly connect one to the other and hope for something rational to come out.

Zz.

 

[mheslep]

As someone who is currently working on dielectric-loaded accelerating structure for particle accelerators, I find all of this to be rather puzzling.

First of all, I've already mentioned several times in this sub-forum why using particle accelerators as fusion reactor is close to a fallacy. The wall-plug efficiency is NEGATIVE!

I believe the OP was referring to some kind fission application?

With regards to accelerator based fusion power(Q>1), there's still ample room for skepticism, but its increasingly unwarranted to dismiss it out-of-hand in light of recent work:

Park, Nebel, Stange and Murali, Phys Rev Lett 95, 015003 (2005)
"Experimental Observation of a Periodically Oscillating Plasma Sphere (POPS) in a Gridded Inertial Electrostatic Confinement Device"

POPS is in answer to the well recognized problems created by coloumb collisions in beam-beam plasmas.

"...A new electrostatic plasma equilibrium that should mitigate this problem has been proposed by Barnes and Nebel [12,13]. This concept uses electron injection into a spherical device to produce a virtual cathode with a harmonic oscillator potential (constant electron density). An ion cloud immersed in the virtual cathode [referred to as the periodically oscillating plasma sphere (POPS)] will then undergo a harmonic oscillation where the oscillation frequency is independent of the amplitude. By tuning the external radio-frequency (rf) electric fields to this naturally occurring mode, it is then possible to phase lock the ion motions. This simultaneously produces very high densities and temperatures during the collapse phase of the oscillation, when all ions converge to the center with their maximum kinetic energies. It has been shown that an analytic solution for the POPS oscillation exists and has the remarkable property that it maintains the ions in local thermodynamic equilibrium at all times [12]. In particular, the equilibrium state survives even though the plasma density and the temperature may vary by several orders of magnitude during the POPS oscillation..."

This LA website describes the idea in general
http://www.lanl.gov/p/rh_pp_park.shtml

There have also been recent accelerator fusion (IEC) based dissertations published. McGuire did work to improve ion focus problems for IEC.
http://ssl.mit.edu/publications/theses/PhD-2007-McGuireThomas.pdf, "Improved Lifetimes and Synchronization Behavior in Multi-grid Inertial Electrostatic Confinement Fusion Devices", MIT PhD dissertation, 2007

...show that the greatly increased confinement
allows for the development of significant collective behavior in the recirculating ions. The plasma self-organizes from an initially uniform state into a synchronized, pulsing collection of ion bunches. In simulations, these bunches are observed to be long-lived with lifetimes on the order of at least a tenth of a second, exceeding 20,000 passes. This represents a 3 order of magnitude improvement in confinement time and device efficiency...

Finally, Rostoker retorts that the Rider loss criticisms are overly generallized
Rostoker et al, "Colliding Beam Fusion Reactors", J. Fusion Energy, 22, 2 (June 2003)

Abstract The recirculating power for virtually all types of fusion reactors has previously been calculated [1] with the Fokker–Planck equation. The reactors involve non-Maxwellian plasmas. The calculations are generic in that they do not relate to specific confinement devices. In all cases except for a Tokamak with D–T fuel the recirculating power was found to exceed the fusion power by a large factor. In this paper we criticize the generality claimed for this calculation. The ratio of circulating power to fusion power is calculated for the Colliding Beam Reactor with fuels D–T, D–He3 and p–B11. The results are respectively, 0.070, 0.141 and 0.493.

 

[ZapperZ]

Yes, I'm aware that the OP is talking about using accelerators for fusion generation. And this issue has been talked about in another thread in this sub-forum, so I will not repeat my original objections.

The references you gave have more to do with plasma confinement devices than "accelerators". None of these actually would qualify as particle accelerators. A tokomak certainly has never been classified as an accelerating structure. In fact, a scour of the abstracts and publications from the last two Particle Accelerator Conferences did not show anything resembling the kinds of studies that you are highlighting here. This tells me that the accelerator physics community either aren't aware, are not participating, or do not consider these things as "particle accelerators".

This particular thread deals with a misuse of the dielectric-enabled acceleration mechanism. The OP mistakenly thought that just because one can use a dielectric structure as a possible source of acceleration, then it might be useful as a "fusion device" (which is yet to be explained). This misunderstanding makes the whole issue rather moot.

Zz.

 

[sanman]

Hi Zapper, yes, as mheslep said, I was thinking of accelerator-driven fission by proton beam.

mheslep, those those links you gave on the POPS (periodically oscillating plasma sphere) were very interesting too. Perhaps I can take them up in a separate thread.

 

[Morbius]

Hi Zapper, yes, as mheslep said, I was thinking of accelerator-driven fission by proton beam.

sanman,

Why proton-induced fission?

Neutron-induced is easier; and there are no advantages to proton-induced vis-a-vis
neutron induced.

Dr. Gregory Greenman
Physicist

 

[sanman]

http://www.uic.com.au/nip47.htm

http://www.europhysicsnews.com/full/06/article8/article8.html

http://www.eoearth.org/article/Accelerator-driven_nuclear_energy

 

[Morbius]

http://www.uic.com.au/nip47.htm

http://www.europhysicsnews.com/full/06/article8/article8.html

http://www.eoearth.org/article/Accelerator-driven_nuclear_energy

sanman,

That doesn't answer my question.

Any of the things you can do with an accelerator-driven system, like
actinide burning; you can do better with a critical system:

http://www.pbs.org/wgbh/pages/frontline/shows/reaction/interviews/till.html

Those that say that accelerator-driven systems are safer are just plain
WRONG! They ascribe this added safety to the fact that the accelerator
can be turned off.

Big deal! Critical systems can be turned off easily too.

However, BOTH suffer EQUALLY from the problem that you can't turn
off the decay heat. If you make energy, you make fission products;
and you have the IDENTICAL problem with decay heat.

Dr. Gregory Greenman
Physicist

 

[sanman]

Well, what about ease of throttling, for a propulsion application?

Even if there's a certain amount of "thermal inertia", whereby you can't stop your power output instantaneously, it still seems easier to scram the reactor by shutting off the beam than by having to mechanically remove control rods.

 

[ZapperZ]

Again, this has produced no new concepts or even addressed the original set of problems that I laid out in another thread on accelerator-driven fusion reactor. Those of us who work in accelerator systems, and as someone who is also familiar with the beam dynamics of a synchrotron center, we simply shake our heads when something like this is being proposed. It appears that these things are thought up by people who are not working in this field.

1. Luminosity problems. To get any significant "fusion", you need a large number of collisions. This is the "luminosity" issue that even places like the Tevatron faced. It is very hard to do this with neutral particles such as neutrons because you can't direct them to where you want to go. If you do this with protons, god help you because...

2. You will need a LOT of protons, and thus, you encounter SPACE CHARGE problems. Since you want to tightly confine the protons to a very thin beam (to increase the luminosity), you encounter A LOT of space charge problems that simply want to blow up the beam, thus reducing your luminosity.

3. Because of #1 also, you encounter a large emittance problem, simply because your beam will have way too much transverse momentum. So good luck with the confinement.

4. If you want to fuse protons (which is one of the silliest thing to do for a fusion reactor), you will need collide them at such a high energy that you will require accelerating structures powered by klystrons, etc.. This will simply KILL your wall-plug efficiency. In other words, the energy you get out of the fusion reaction is LESS than the energy you put in in the first place. Transferring RF power from a klystron to a structure can be highly inefficient, and large accelerator complex requires several of these things to power a number of LINACs.

Again, I've listed these problems already and so far, there hasn't been ANY responses to tackle these problem. Yet, we keep getting these "accelerator fusion devices" periodically. I would suggest that until some paper written by someone who is working in the accelerator physics field comes out in favor of such a technique, that this issue is laid to rest.

I will assume that since we are not back to discussing some generic accelerator-driven fusion device, that the issue of using dielectric-loaded structure is now DEAD.

Zz.

 

[Morbius]

Well, what about ease of throttling, for a propulsion application?

Even if there's a certain amount of "thermal inertia", whereby you can't stop your power output instantaneously, it still seems easier to scram the reactor by shutting off the beam than by having to mechanically remove control rods.

sanman,

Reactors have inherent shutdown mechanisms such as Doppler broadening - you don't
have to move the control rods.

Besides, it's just as easy to scram with control rods - you turn off a switch - that
de-energizes the magnets that hold up the control rods and they fall by gravity.

That's not going work in your space application, of course. In that case, the rods
would probably be inserted hydraulically by a system operating off a source of stored
pressure such as an accumulator.

Dr. Gregory Greenman
Physicist

 

[Morbius]

Again, this has produced no new concepts or even addressed the original set of problems that I laid out in another thread on accelerator-driven fusion reactor. Those of us who work in accelerator systems, and as someone who is also familiar with the beam dynamics of a synchrotron center, we simply shake our heads when something like this is being proposed. It appears that these things are thought up by people who are not working in this field./QUOTE]
Zapper,

Thank You for your well-informed post. Yes too often people just put together ideas to
sound "high tech" without any real understanding or consideration of the physics involved.

Dr. Gregory Greenman
Physicist

 

[mheslep]

Again, this has produced no new concepts or even addressed the original set of problems that I laid out in another thread on accelerator-driven fusion reactor. Those of us who work in accelerator systems, and as someone who is also familiar with the beam dynamics of a synchrotron center, we simply shake our heads when something like this is being proposed. It appears that these things are thought up by people who are not working in this field.

I reviewed some of the older threads to which you refer and it appears that part of the issue is nomenclature: the word accelerator in your context means nothing less than a GeV or TeV particle smasher, with large (m or km) mean free path. Fair enough, 'accelerator' is so defined. Certainly that type of apparatus has no practical application to fusion.In the links on confinement I posted above, 'accelerator' refers merely to acceleration, as in F=ma where the F is typically provided by a kV sized electrostatic field applied to charged particles, and appropriate for nuclear coulomb barriers.

A Tokamak certainly has never been classified as an accelerating structure.

Agreed, a Tokamak is a completely thermal device in concept and has almost nothing in common with electrostatic acceleration inertial confinement (IEC) fusion approaches to which I refer. Now, IEC does have much in common with 'accelerators', or particle smashers as many of the physics issues you reference below here still obviously apply to these kV confinement cases:

1. Luminosity problems. To get any significant "fusion", you need a large number of collisions. This is the "luminosity" issue that even places like the Tevatron faced. It is very hard to do this with neutral particles such as neutrons because you can't direct them to where you want to go. If you do this with protons, god help you because...

2. You will need a LOT of protons, and thus, you encounter SPACE CHARGE problems. Since you want to tightly confine the protons to a very thin beam (to increase the luminosity), you encounter A LOT of space charge problems that simply want to blow up the beam, thus reducing your luminosity

Electrostatic confinement concepts use deuteron's, tritium, He3, or even p-11B because of the much greater fusion reactivity cross section (just as in Tokamaks of course).
Luminosity and space charge are issues here as well (density is the common term in the referenced literature, but perhaps luminosity is more apt as its still beam physics that apply). Hence, some degree of neutrality is usually proposed by inclusion of electrons. If the electrons are allowed to thermalize w/ the ~50kV ions they radiate all the power way, so clever schemes have been proposed to prevent this (unrealized so far) such as keeping the electron life times short, etc.

4. If you want to fuse protons (which is one of the silliest thing to do for a fusion reactor), you will need collide them at such a high energy that you will require accelerating structures powered by klystrons, etc.. This will simply KILL your wall-plug efficiency. In other words, the energy you get out of the fusion reaction is LESS than the energy you put in in the first place. Transferring RF power from a klystron to a structure can be highly inefficient, and large accelerator complex requires several of these things to power a number of LINACs.

Again nobody is trying to fuse protons in my references, and drive power is provided by the electrostatic field for my cases, not much mention of RF though there's some resonance inducing attempts via RF.

 

[ZapperZ]

Actually, even MeV range of energy acceleration can be considered as an accelerator. I work at one in which the highest energy of the electrons we accelerate is around 15 MeV.

What you described was not what I took issue with. If you look at the OP, this is certainly not what you are describing. If you read the links given in the OP, you'll notice that there is an attempt to use dielectric accelerating structure and turn it (how, I don't know) into a fusion device. I can only imagine that the parent nuclei are somehow accelerated and then collided into each other to cause fusion. My argument is that using a standard particle accelerator, as done in those references and referred to in another thread on here, is one of the, if not the, most inefficient way to do this. This is on top of all the other physical problems that I had outlined.

Such pie-in-the-sky idea pops up on PF every so often, and it is puzzling why. That is the reason why I had try to clearly outline the problems surrounding it, so that the next person who tries to do this can first of all address these issues.

Zz.

 

[mheslep]

Eh? All four OP links in this thread are on application of new high voltage insulators to cheaper more compact medical proton therapies, the term fusion is not used therein. I became interested when your post raised fusion in conjunction w/ particle acceleration, or its non-feasibility, and still am interested to see what I learn from the particle accelerator community as to how the applicable accelerator physics (or engineering) might be applied to confinement (not, the GeV/MeV devices themselves).

 

[mheslep]

...h the highest energy of the electrons we accelerate is around 15 MeV.

15 MeV electrons? Seems like that would be just a big X-Ray machine since
P_{Br} [\textrm{Watt/m}^3] = \left[{n_e \over 7.69 \times 10^{18} \textrm{m}^{-3} }\right]^2 T_e[\textrm{eV}]^{1/2}

 

[ZapperZ]

Eh? All four OP links in this thread are on application of new high voltage insulators to cheaper more compact medical proton therapies, the term fusion is not used therein. I became interested when your post raised fusion in conjunction w/ particle acceleration, or its non-feasibility, and still am interested to see what I learn from the particle accelerator community as to how the applicable accelerator physics (or engineering) might be applied to confinement (not, the GeV/MeV devices themselves).

But that's what I meant! I didn't make the connection of those links to fusion. The OP did!

15 MeV electrons? Seems like that would be just a big X-Ray machine since
P_{Br} [\textrm{Watt/m}^3] = \left[{n_e \over 7.69 \times 10^{18} \textrm{m}^{-3} }\right]^2 T_e[\textrm{eV}]^{1/2}

It could be. Our aim isn't getting to some high energy, but rather the GAIN in energy. About 2 months ago, we managed to get a significant milestone in which we managed to show a 100 MV/m gradient in the wakefield generated in such a dielectric accelerating structure. Note that a conventional copper accelerating structure can only manage up to 40 MV/m. This could lead to a more efficient and compact accelerator, just the type mentioned in the OP.

It is significant enough that our funding agency immediately agreed to pay for the "accessories" needed for a second Klystron that we planned for.

Zz.

 

[mheslep]

Congratulations