What would be the technical considerations of a practical dense plasma pinch based neutron source for running a nuclear fission process?

I can think of a few:
1. Energy Efficiency - is this actually a problem? Would a neutron source produce enough neutrons per energy input to not harm the power plant total efficiency?
Depends on the fission reaction you are planning. Thorium could be mixed with ‘high level’ actinide waste and end up with a lump of a fuel type that is almost as fissile as the operating point of a regular uranium reactor. Not sure if anyone has tried that.

The ‘almost’ is then a clue as to how to drive and run such a reactor - ‘all’ that is then needed to drive such a reactor is an input of neutrons that will tip the neutron budget of the fuel over towards ‘energy producing’, because thorium won’t run a chain reaction of neutron production like uranium so the extra neutrons are needed to ‘make up for’ the extra neutrons that would otherwise come out of the uranium process.
Thorium itself takes quite a bit of neutron irradiation to get it reacting enough but as the reactor runs the required input of neutrons will come down as the actinides build up. This scheme has been proposed as the ‘Rubbia Energy Amplifier’, and is driven by a spallation source of neutrons (in which protons are accelerated into lead, which could then also act as the coolant once in melts.)
Whereas if a thorium reactor had a blend of actinides then the required neutron input would be low initially but go up as the actinides decay away. You could then either ramp up the neutron rate, or re-blend the fuel.
(Not sure if many folks have yet considered making thorium fuel more fissile by blending with actinide waste, which also has the added benefit that the waste gets eaten up?)
So the answer to your question rather hinges on whether you are seeking to drive a fissile process essentially exclusively by neutrons (thorium energy amplifier), partially (my scheme with blended actinides) or not at all with neutrons (regular uranium slow reactor).

2. The rate at which neutrons are produced being at a sufficient rate/power to vary the thermal output of a fission process… would be very nice for load following in a power system. Slower neutron production can gradually reach full power through the build up of fissile byproducts.

This, depends on the mix, but in theory there would be a blend where the thorium and the actinide blend would be relatively stable in its demand to be driven by a neutron source. It would also be relatively safe, because you switch off the neutrons and you switch off the reaction. Obviously, it would still have a heat half-life to consider, but at least the fuel would not go critical by itself.

3. Can something be scaled up properly and still be feasibly built? Are there practical limits?

In order to have high output and good control (and maintain a safe K_eff), the driver neutron source needs to supply what is generally considered to be a huge amount of neutrons: ~1E+15 n/s per MWth or so. When you talk about a 20-MWth ADS power plant, generally you might expect to need 2E+16 n/s. Continuously, around the clock, with high availability!

Obviously, amateur Farnsworth fusors are out of the running. IEC devices have been proposed for the application (by George Miley and others), but the technological readiness or even plausibility remains undemonstrated. If one considers only neutron source technology that is presently well-developed and is reasonably reliable, you come down to the following more-or-less distinct possibilities (with wall-plug efficiency estimates):

-Spallation (3E+09 neutrons / joule)
-Low-energy reactions (p, Li), (p, Be), etc. (1E+07 neutrons / joule)
-Photonuclear reactions (I don’t know enough about present facilities)
-Fusion (typical commercial DT generator is ~3E+07 neutrons / joule)

Spallation wins on efficiency. It also will win on reliability, since the big spallation research sources (SNS, LANSCE, etc.) have been tooling along at 80-90% availability for years. You need a half-mile-long linac to get proton energies in the GeV range, so that’s a possible downside of spallation. But at least this is a technology that’s ready to go. For your 20-MWth plant, you would want to have an accelerator like LANSCE (800 kW beam).

A spallation spectrum consists of mostly fast neutrons, typically peaking in the MeV range, but having some particles with 10- and 100-MeV range energies. “Thermal” neutrons have a Maxwellian energy spectrum at whatever ambient temperature is applicable, e.g. a mean energy of 0.025 eV near room temperature.

In an ADS, the spallation source provides a small contribution to the total neutron population. Most neutrons are the result of fission in the fuel, and will have the characteristic Watt spectrum of fission. There will be a few (n,xn) neutrons in there too, because the high-energy tail of the spallation spectrum will cause these reactions. The source may or may not need special moderation in addition to whatever moderator surrounds the fuel. It’s not obvious to me—but rather a research question—what sort of flux tailoring scheme will best serve various types of fuel and various kinds of spallation targets.

The lifetime of the target in a spallation source is typically determined by gas production and consequent embrittlement of the target if it is solid and has to retain some kind of mechanical or thermal integrity. This is because spallation produces all kinds of light nuclear byproducts such as hydrogen and helium in addition to neutrons. If the target is liquid, the gas production issue applies to the container it’s in.

The short answer is the PF is unlikely to make an impact on sub-critical fission systems.  Current plasma focus devices operate at a peak of ~1E12 neutrons per shot (DD) at ~2MA.  If one assumes the 100X gain by going to DT and it is a big if, you need more than 3X the current.  A 7 MA driver running at repetition rate has not been demonstrated.  I don’t know of a plasma focus that has operated above 3 MA.  The electrode erosion, vacuum chamber design, confirmation that the fast ion beam does not substantially exceed the peak in the DT fusion cross section, next generation pulse power technology beyond the requirements of FoFu-1 and years of engineering failure analysis make this approach unlikely.  The accelerator systems are much closer to reality with countries like India pouring money into ADS. An accelerator like SNS could be viable for an ADS.  The operational time needs to improve but many of the problems are known and could be addressed.

A spallation source uses protons to knock neutrons out of heavy metal atoms. Liquid mercury is used as it doesnt suffer damage to a solid structure, although I am curious as to what the mercury is gradually converted into.

Consider that the DPF produces alpha particles of a high energy… Does the concept of alpha particle spallation make sense? I know we use charged particles because we can guide them (as opposed to the willy nilly go everywhere neutral neutrons) into a spallation target, resulting in neutrons traveling in roughly the same direction.

Spallation sources rely on very high energy protons (~1 GeV) to work.  A PF alpha particle or proton source is unlikely to exceed 10 MeV with most of the particles of interest far below this value.  The PF would serve as an intense ion source but you need the particle accelerator in front of it to accelerate to the desired speed. 

If you want to go to an (alpha,n) reaction with something like Be it is possible but you have to deal with a target accepting a high instantaneous power, low average power situation which makes the target difficult to thermally manage where the beam impacts the target.  I’ve damage steel with a plasma focus ion beam.

andrewmdodson - 30 January 2013 08:38 PM

A spallation source uses protons to knock neutrons out of heavy metal atoms. Liquid mercury is used as it doesnt suffer damage to a solid structure, although I am curious as to what the mercury is gradually converted into.

If you knock a neutron out of a mercury nucleus, you get mercury.

When a mercury nuclei is hit with a 1GeV proton it more or less explodes. A neutron is not knocked out. But some of the fragments are neutrons and they travel a long way while the charged fragments don’t.

but if you do knock neutrons out of mercury, then sometimes you will get gold

err no. Knocking out neutrons does not change the atomic number.

This reaction is as follows:

198Hg (9.7% abundance)—> 197Hg + n; 197Hg (electron capture h.l. 64 h)—> 197Au

Oh yea. Right. But again that is not how these spallation source work. They shatter the nucleus.

Somehow this feels like we got to talking about alchemy.  I mean, I know nuclear reactions fit the definition of alchemy but it still feels weird when someone is talking about turning quicksilver into gold.

Once energy is plentiful it will probably be more cost effective to extract gold from sea water that transmute it from other elements.
Transmutation might be used for anti-matter production.