SLOWFAST, reactor being both slow and fast

[artis]

While reading about neutron interactions in a book I noticed that there is some amount of fast fission going on even in a thermal reactor with a water moderator/coolant. So essentially after each neutron that survives to the thermal range and manages to produce a fission event , on average 2 to 3 neutrons are born which at the first moment are born as fast neutrons of course.
Some of these fast neutrons before they leak out of the fuel manage to cause some fast fission events in the very fuel , after all they haven't been slowed down by scattering in the moderator yet.
Ok so theory aside here is my question/scenario.

Leaving out the practical aspects of engineering and more on a theoretical note.

Imagine a box filled with U fuel plates like the plates of a parallel plate capacitor, between each plate is a separation distance.
The plates and the distance between them are arranged such that each fuel plate is 2x as thick as each separation distance.
So let's say 1cm for fuel plate and 0.5cm for separation.
The separations are filled with water that acts as moderator/coolant. Now the idea is to have a 2:1 fuel to moderator ratio or thereabout , or in other words so that each fast neutron born in one fuel plate wouldn't get slowed down as fast before it enters at least 2 or 3 more adjacent fuel plates while still being "fast".
For this to happen the water layers between fuel plates need to be as thin as possible.

At the same time each fast neutron born in a fuel plate would either produce a fast fission in adjacent fuel plates or get slowed down whgile passing through more plates and water separations and become a slow neutron producing thermal fission. The way in which one could cool a reactor like this with such low water content is to say increase the flowrate of water through the box of plates , and allow the water to turn to steam at the end where it could be direct to a turbine , maybe even directly to a turbine very close by.
The only thing I'm not sure is whether the decrease in steam output temp would be compensated by the much larger amoutn of steam produced from the high flowrate and vaporization of the steam.
Just interested to hear has something like this been ever attempted in any form and is it even possible to produce a reactor which could utilize both thermal fission and fast fission at any considerable ration?

 

[dRic2]

There are some issues that should be addressed in you post, but a part from that I don't get the point of you question. U-235 has a very low fission cross section at high energies (that is, for fast neutrons) so you actually want to slow down neutrons as much as possible in order to max the probability of fission. So if you work with Uranium you don't want fast neutrons flying around. If you want fast fission to occur you can try to add other fissile material that has an higher cross section for higher neutron energies (an hybrid reactor maybe? I don't know if they exist). The geometry of the reactor you have in mind wouldn't make any significant difference since fast fission is very unlike to happen for U-235. By the way, why did you think of a slab configuration in particular ? Why not the standard rod configuration used for all water-moderated reactors?

 

[Rive]

Just interested to hear has something like this been ever attempted in any form and is it even possible to produce a reactor which could utilize both thermal fission and fast fission at any considerable ration?

Well, as I understand fast fission happens in every core, but due the low enrichment and the presence of moderator in 'conventional' designs it is thermal fission which dominates.
On the other hand, in fast reactors due the lack of moderator and higher enrichment the fast fission dominates over the occasional thermal neutrons.

Since both ends of the scale exists I see no reason why anything between could not exist too: however such hybrid systems would just throw away the benefits of both (requires high enrichment and moderator too) while I do not think they could gain anything substantial.

Maybe some special cases might exists.

 

[artis]

@dRic2 Well guess what, the U235 in any conventional thermal power reactor is no more than 5% of the fuel , aka 5% enriched, the rest is U238 or natural Uranium as it is found in nature.
U238 is fertile but can undergo fast fission which produces other isotopes that then are fertile like Pu239, that's all the rage about the fast reactors that they can breed fissile fuel from their fertile fuel.
Same thing with Thorium 232.

Now maybe I'm all wrong about my idea I was just thinking , you see there are 2 ,main problems , the first one is that fast reactors require no moderator only coolant but all practical coolants are liquid metals or salts as they having a higher mass number don't slow down the neutrons as much , in slow reactors neutrons are also lost and some absorbed in the water moderator.
But water at least is easy to work with and is not toxic and doesn't react violently with other substances like Sodium does in the fast reactors.

A good nuclear reactor is one where neutron losses are as minimal as possible and the spectrum can be used to produce as much fission as possible , including fast one.
So with this I was just thinking could one minimize the thermal absorbtion while also increase the fast fission rate. That is why the moderator/coolant is there as little as possible and the fuel is spread out in sheets so that each neutron as it is born has the capability to strike a fast fission or travel a longer distance get scattered and cause thermal fission.

One problem ofcourse which I suspect could be the necessary enrichment as the decreasing of moderator to fuel ration would probably increase the minimum enrichment needed for a given mass of fuel to achieve criticality. The question is how much , which is probably a maths problem to solve.

 

[artis]

@Rive, yes fast fission happens to some extent in every core , it's just that if one could increase this fast fission rate and decrease neutron losses one would get a longer burnup for fuel and bit less long lived waste within the fuel. Fast neutrons tend to "eat up" those isotopes better.

 

[Rive]

Fast neutrons tend to "eat up" those isotopes better.

Fast neutrons are the source of thermal neutrons, they are always present.

So I think at the end what you seek is actually about keeping the core in operation with (naturally happening or artificially created) higher burnup, so fission products could be further 'burned'.

 

[dRic2]

it's just that if one could increase this fast fission rate and decrease neutron losses one would get a longer burnup for fuel and bit less long lived waste within the fuel

Instead of increasing fast fission rate, you could reduce neutron losses by just reducing leakages (ie. with reflectors). There is no point is trying to increase fast fission contribution in a water moderated reactor because thermal fission is simply better under those circumstances. If you have fast neutrons it's better to try to slow them down instead of producing fast fission (if you want to work with water as a coolant/moderator, otherwise you could go for a fast reactor)

PS: I don't think U-238 fast fission is much relevant in water-reactors.

 

[artis]

Well maybe I'm wrong physics wise but I feel you still haven't got the idea I was trying to explain.

The idea is to have local fast fission with overall thermal fission. If you have like I mentioned fuel in the form of thin plates and water between them you can think of this as a parallel plate capacitor or pancake reactor. Now the water layer between each fuel plate is very thin, the neutrons that are born from fission are always fast before they get slowed down, being born fast in one specific fuel plate they have some chance of causing a fast fission in the same fuel plate before they leave but this is rare , now even if they leave they would then travel through some neighboring fuel plates and either cause a fast fission there or eventually travel far enough to be slowed down, it's just that they don't get slowed down immediately because the water layers between the fuel plates are thinner than the mean free path for the fission neutrons so such thin water structures won't slow them down considerably, only after a neutron travel through multiple such water layers will it be slowed down.In ordinary thermal water reactors the amount of water is much larger and distance between adjacent fuel rods also larger.

In my idea there could be more fast neutrons fission as adjacent fuel is very close and neutrons born in one plate won't slow down before going through adjacent plates. Overall there is still thermal fission because neutrons that do go further and cross multiple such fuel plate and water layers get slowed down enough to get thermalized.

Something like this.
Why water? Well as I said water is the easiest to work with. Fast reactors would be a reality by now if it wasn't for the nasty coolant choices which complicate everything and even make the situation unstable.

I was just thinking would it be possible to use water and increase the fast fission spectrum.
Well for a single closed loop system maybe even heavy water would be better, it absorbs even less neutrons.

 

[dRic2]

it's just that they don't get slowed down immediately because the water layers between the fuel plates are thinner than the mean free path for the fission neutrons so such thin water structures won't slow them down considerably,

if the layers are much thinner than the mean free path it's pretty much the same as to consider just a block of Uranium (I remember I had to do an exercise like that...) because the moderating effect of water would be too small. I think it's known from experimental evidence that a bare block of U-238 can not sustain a chain reaction.

 

[trurle]

.. like I mentioned fuel in the form of thin plates and water between them you can think of this as a parallel plate capacitor or pancake reactor. Now the water layer between each fuel plate is very thin, the neutrons that are born from fission are always fast before they get slowed down, being born fast in one specific fuel plate they have some chance of causing a fast fission in the same fuel plate before they leave but this is rare , now even if they leave they would then travel through some neighboring fuel plates and either cause a fast fission there or eventually travel far enough to be slowed down, it's just that they don't get slowed down immediately because the water layers between the fuel plates are thinner than the mean free path for the fission neutrons so such thin water structures won't slow them down considerably, only after a neutron travel through multiple such water layers will it be slowed down.In ordinary thermal water reactors the amount of water is much larger and distance between adjacent fuel rods also larger.

In my idea there could be more fast neutrons fission as adjacent fuel is very close and neutrons born in one plate won't slow down before going through adjacent plates. Overall there is still thermal fission because neutrons that do go further and cross multiple such fuel plate and water layers get slowed down enough to get thermalized.

Due to large fraction of fuel in your reactor, you will need to use weakly enriched uranium to avoid over-criticality. This is similar to fast breeder reactor.

The pancake design will have severe cooling, control and maintenance problems though. Especially if you think of what happens after irradiated fuel plates will start to swell. Finally, the tendency to switch from bubble boiling to film boiling will be difficult to counter - too little space to disperse bubbles.

In brief, pancake reactor will require a tremendous fuel load, but will have small specific power.

 

[artis]

@dRic2, well a fast reactor apart from it's coolant is basically just a spread out " lump of Uranium" , yet it works, sure it has to be enriched to about 20% for there to be enough neutrons to sustain criticality in the absence of thermal fission.

My ideas was basically to keep as much of the fast spectrum as possible while also having the slow one and having a coolant that isn't super toxic or dangerous.
Not saying that such a plate like reactor is any good , it was just an example.

 

[Astronuc]

So essentially after each neutron that survives to the thermal range and manages to produce a fission event , on average 2 to 3 neutrons are born which at the first moment are born as fast neutrons of course.

Either 2 or 3 neutrons are produced in fission, with average depending on the energy of the incident neutron. Some neutrons are release with delay (delayed neutrons) from certain fission products, which allows one to control the reactor, but control the absorption of a small fraction of neutrons.

One neutron is required to induce fission for the subsequent generation (the definition of criticality), so the other neutrons are absorbed (in fuel, but without fission; in fission products; in structural alloys or coolant; in burnable poisons, e.g., boron (B-10), gadolinium or erbium; or they leak out of core and are absorbed by the surrounding structure).

As I recall, about 7-8% of fissions in a thermal (e.g., light water reactor (LWR)) reactor are from fast fission in ²³⁸U, otherwise the fissions are in ²³⁵U. Some neutrons are absorbed by ²³⁸U without fission, which produces

239

U, which decays by beta emission producing ²³⁹Np, which decays by beta emission to ²³⁹Pu. ₂₃₉PU has a greater fission cross-section than 235U in the thermal energy range, but also in the fast range. As the fuel is used, more of the ²³⁸U is converted to ₂₃₉Pu, and with depletion of 235U, more fissions actually occur in ²³⁹Pu, such that a burnup of about 5% FIMA (or 50 GWd/tU), about half the fissions occur in Pu. ²³⁹Pu can also absorb neutrons without fission and produce ²⁴⁰Pu and ²⁴¹Pu, and successive neutron capture can also lead to production of Am and Cm isotopes.

One can affect the conversion of U to Pu by adjusting the relative amount of fuel to moderator. One can have more fuel (undermoderated) with a harder spectrum, which leads to more Pu production, as opposed to an over-moderated system. In the case of boiling water reactors (BWRs), one can reduce flow and have more boiling in the core, which reduces moderation (and hardens the spectrum, i.e., more fast neutrons) in order to produce Pu in the fuel. This is so-called spectral shift, implementing a particular fuel assembly design allows a process called bundle-enhanced spectral shift. After Pu builds up in the fuel, the flow is increased and the boiling reduced resulting in more moderation and fission in the Pu.

To reduce neutron loss from a core, enrichment is varied within a fuel rod with so-called blanket zones, or sections of fuel that have lower enrichment than the bulk of the fuel. For example, one might have 3.3 m of fuel with an enrichment of 4 to 5%, and the upper and lower 15 cm (0.15 m) have an enrichment of 3%. Similarly, some assemblies have lower enrichment, and they are placed at the edge of the core, or high-burnup fuel, which operates at low power, is placed at the edge of the core.

In the case of fast reactors, there is usually a blanket region, or reflector, around the core. Assemblies of lower enrichment surround a core of fuel with higher enrichment. Because the fission cross-section at fast neutron energies, E > 0.1 MeV, is much lower than the fission cross-section at thermal energies, the enrichments of fast reactor fuel considerably greater (18-24%) than those of thermal (LWR) reactors (up to 5%). Some special fuel can have higher enrichment but that is handled with special conditions, and usually not in power reactors.

Power reactors use fuel that is fairly consistent and uniform, since the principal objective is power generation.

Most LWRs use a square lattice of fuel rods in an assembly, although a triangular (hexagonal) lattice is possible. Fast reactors typically use a hexgonal lattice.

Furthermore, BWR and fast reactor assemblies are surrounded by a shroud to direct flow along the fuel and prevent cross flow, while PWR fuel assemblies are open, and allow some cross-flow, which is a concern with respect to grid-to-rod fretting.

The fuel rods allow for heat transfer from the fuel, through the cladding and into the coolant. The cladding serves to retain or prevent loss of fission products to the coolant and also to maintain geometry of the fuel and core. In PWR, some lattice positions in the assembly are occupied by guide tubes which allow for insertion of control rods. Fast reactors have similar arrangements. In BWRs, the control rods sit between fuel assemblies, on the outside of the channel (shroud).

During reactor operation, PWR control rods are withdrawn from the core, with tips sitting just above the core in the upper portion of the guide tubes. Burnable absorbers in the fuel and soluble boron in the coolant are used to control reactivity. In BWRs, groups of control rods are inserted into the core, with most out of the core, in order to control reactivity and power distribution. BWR control rods are periodically moved (pattern adjustment) or exchanged, i.e., those in the core come out while others are inserted (sequence exchange). Some PWRs use grey rods, for axial and radial power distribution control, or for quick power maneuvering as in frequency control or load-follow.