Nuclear reactor design with refuel period of more than 20 years

  • Thread starter Thread starter corealist
  • Start date Start date
Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
9 replies · 803 views
corealist
Messages
5
Reaction score
1
Nuclear fast reactors are capable of breeding more fuel than they consume. I was wondering if it would be possible to design a nuclear reactor that (initially) breeds just a little more fuel than they consume. The breeding ratio will reduce a little over time, but with a felxible design it should be able to keep the output power roughly the same for a very long period of time without having to replace the fuel. If only 15% of the u238 in that breeder reactor is converted into fuel, it should be able to run about 10 times as long (about 20 years) without refueling as current designs, which would dramatically lower maintenance, processing and storage costs.
Since nobody is designing a reactor like that, I assume that this is physically not possible. Does anybody on this forum know why this is not possible.
 
Physics news on Phys.org
I think breeding ratio is controlled by blanket fuel rod allocation design. Keeping criticality for a long time is another issue. Fuel is extracted by reprocessing process. Consuming them during reactor operation seems impossible to achieve in reactor design.
 
Last edited:
I read that even in a light water reactor the breeding ratio is about 0.6 ( no breeding blanket), and breeding reactors with breeding blanket can achieve a ratio of about 1.4 so to me it doesn’t seem to be farfetched that a appropriately designed fast reactor ( I was thinking lead based fast reactor) could reach a breeding ratio of around 1.05 without a blanket, which should be sufficient to compensate for a increase in neutron absorption by nuclear fission products.
 
If the goal is to operate the reactor for an extended period with little or no additional fuel supplied from outside, the issue is not simply one of the breeding ratio; rather, it requires an integrated design that takes into account core burnup, neutron economy, and reactivity changes over the fuel cycle. It is not denied in principle and is an interesting but a challenging issue.
 
Last edited:
I understand that breeding is just one factor.
I assume It would require some way to periodically ( yearly ?) rearrange fuelrods in the core to maintain an even burn, but if that is physically possible it would result in an order of magnitude less maintenance, radioactive waste and processing and fuel costs
 
Reply
  • Like
Likes   Reactions: anuttarasammyak
corealist said:
Nuclear fast reactors are capable of breeding more fuel than they consume. I was wondering if it would be possible to design a nuclear reactor that (initially) breeds just a little more fuel than they consume. The breeding ratio will reduce a little over time, but with a flexible design it should be able to keep the output power roughly the same for a very long period of time without having to replace the fuel. If only 15% of the u238 in that breeder reactor is converted into fuel, it should be able to run about 10 times as long (about 20 years) without refueling as current designs, which would dramatically lower maintenance, processing and storage costs.
It's not so simple and straightforward. One has to maintain criticality and an appropriate power distribution, which may be accomplished by burnable absorbers and/or control rods, as well as margins to technical limits. One cannot avoid the accumulation of fission products from the fission process, and these will accumulate. Fission gases, Xe and Kr isotopes must be accommodated by a void volume (upper and optionally lower plenum volumes); some of the Xe and Kr will decay to Cs and Rb, respectively, which will eventually decay to Ba and Sr, respectively. Precursors to Xe and Kr are Te -> I -> Xe, and Se -> Br -> Kr. Depending on the fuel operating temperature, partial pressures of I and Br, and possibly Cs and Rb may be considered. Each fission produces two atoms for one atom of U or Pu that fissions, so one must consider solid swelling of the fuel, and with high enough temperature, fuel restructuring, in addition to the fuel (rod) internal pressure (and gaseous swelling in the fuel from fission gases that do not escape into the void volume).

Most power reactors shuffle fuel during refueling outages to even the burnup. Assemblies close to the periphery of the core experience strong flux gradients (and consequently power and burnup gradients), so such assemblies may be moved to opposite side of the core. Some gradient assemblies may experience dimensional distortion, which must be controlled to prevent anomalies in reactivity distribution and control rod operation.

One has to consider burnup limits related to some of the above considerations, which affects power densities, and cumulative radiation damage to the structural materials, e.g., cladding, and in the case of fast reactors, fuel assembly shroud/duct. Embrittlement, corrosion and dimensional stability are factors to consider.

corealist said:
I read that even in a light water reactor the breeding ratio is about 0.6 ( no breeding blanket), and breeding reactors with breeding blanket can achieve a ratio of about 1.4 so to me it doesn’t seem to be farfetched that a appropriately designed fast reactor ( I was thinking lead based fast reactor) could reach a breeding ratio of around 1.05 without a blanket, which should be sufficient to compensate for a increase in neutron absorption by nuclear fission products.
LWRs and liquid metal-cooled reactors are two different technologies.

Traditionally, LWRs have been licensed to use fuel up to 5% enrichment (235U) or equivalent with MOX. Recently, and effort has been underway to increase enrichments up to 6% and potentially 8% 235U, in order to accommodate new high burnup fuel designs. Some plants (most BWRs and PWRs of moderate power density) already operate on 24-month fuel cycles. Some utilities prefer to maintain 18-month cycles, while other prefer 24 month cycles. Fuel typically operates for 2 or 3 cycles, or 3 years to 4.5 years for 18-month cycles, or 4 to 6 years on 24-month cycles, although I have seen core designs that allowed fuel for a 4th 18-month or 24-month cycles, so 6 years or 8 years, respectively.

In addtion to refueling and fuel shuffling, outages are used to perform maintenance on the reactor system and balance of plant.

In designing a plant and core/fuel, one must consider the operating temperature of the coolant and fuel, and consider the effects of material degradation. In a neutron and gamma field, one must consider the cumulative displacements per atom (dpa), which relates to the cumulative radiation damage to structural material, as well as effects of transmuation, which of course changes the structual alloy. Alloy creep is another factor to be considered.

On top of normal operation, one must consider potential transients from anticipated operational occurences as well as various hypothetical accidents. The fuel design must mitigate release of fission products to the system (and especially to the environment), as well as allow for shutdown (cessation of the fission process) upon demand.

In LWR-space, thorium breeder reactors have been considered for the production of 233U. Initial fuel uses fuel enriched with 235U; either highly enriched 235U must be produced and blended into Th, usually in oxide form, UO2+ThO2, or ThO2 bearing rods would have be removed from carrier assemblies and placed into power assemblies. The latter process requires removing fuel rods from one assembly to another assembly, which is essentially a reconstituion process, which is carried out ex-core.

Some reactors, e.g., CANDU are designed for online refueling. In CANDU, refueling (or fuel handling) machience are coupled to each end of a coolant channel, or pressure tube, while the reactor is running. one machine pushes 'new' or 'fresh' fuel into the core, while the machine at the other/opposite end of the pressure tube takes used fuel, which can then be moved to another pressure tube (coolant channel) for continued operation.

Some reactors for marine propulsion have been designed for single core operation for 15 years, and more modern designs may achieve 20 years and possibly 25 years without refueling. Upon reaching service life, the entire reactor and pressure vessel is removed.
 
Reply
  • Informative
Likes   Reactions: Alex A and berkeman
Thanks for the detailed response. The last paragraph about marine propulsion is what I was looking for. A reactor core that lasts long enough that there is no need for complicated, dangerous, expensive recycling of uranium/plutonium.

Anyway based on you response it seems that fission gases and radiation damage to structural part of the reactor ( e.g cladding) are the biggest problem. Are these problems so complicated to solve that it is not even worthwhile considering it ?
 
The idea in the OP reminds me of the traveling wave design most recently (?) promoted by Bill gates.
 
Yes it is a similar idea, but the traveling wave seems to breed the new fuel on the outside of the active part of the reactor, and move slowly from the inside to the outer rings. To me this seems very complicated to design correctly, as the ring gets wider and wider but the output energy must stay more or less the same.
If fission gasses are the main problem of the design I was thinking of, it should not be too difficult to design something to occasionally vent those gasses to somewhere else in the reactor. That would probably be easier than the traveling wave design.