Can plutonium be used in small reactors for ships and other applications?

[vemvare]

Imagine if we had a plutonium economy, where nuclear reactors breeding 239-Pu from 238-U and burning it were the standard in large-scale grid power production.

Could this readily available reactor grade plutonium be used in compact reactors for ships, et cetera? Can small reactors of perhaps 10-100MWe be built utilizing plutonium?

I'm finding references to MOX fuel being used in today's PWR/BWR reactors and to higher-enriched uranium (8-100%) being used to power small nuclear reactors, but I haven't found any reference to "high Pu" MOX-burning small reactors, or how small breeding reactors can theoretically be.

I'd be thankful for any help!

EDIT: I'm evidently not good at googling: http://www.iaea.org/inis/collection/NCLCollectionStore/_Public/35/062/35062698.pdf a Japanese paper on a 300MWt PWR,
http://www.diva-portal.org/smash/get/diva2:727866/FULLTEXT01.pdf European effort (research reactor ELECTRA).

Apologies to all for this, I've got some reading to do.

 

[QuantumPion]

Imagine if we had a plutonium economy, where nuclear reactors breeding 239-Pu from 238-U and burning it were the standard in large-scale grid power production.

That is already the case now. Standard light water reactors produce as much as 50% of their energy from fissioning of plutonium created from U-238.

Could this readily available reactor grade plutonium be used in compact reactors for ships, et cetera? Can small reactors of perhaps 10-100MWe be built utilizing plutonium?

You can make small compact reactors using U-233, U-235 or Pu-239, they are equivalent for the purposes of power reactors.

I'm finding references to MOX fuel being used in today's PWR/BWR reactors and to higher-enriched uranium (8-100%) being used to power small nuclear reactors, but I haven't found any reference to "high Pu" MOX-burning small reactors, or how small breeding reactors can theoretically be.

I'd be thankful for any help!

Some commercial reactors do use MOX, which comes from decommissioned nuclear weapons. They use MOX because the government gives it to them for free to dispose of instead of having to buy uranium normally. In general you can make a reactor any size with any fuel size you want, just depending on the application. Smaller reactors for things like submarines or aircraft carriers use high enrichments because they are not designed to be refueled regularly - with high enrichments they can produce power for decades before having to be replaced. Thermal breeder reactors tend to be large because they use natural uranium, which has a low fuel content. So to get the same total energy you need more fuel volume. But fast breeder reactors are small because they use high enrichments and are cooled by liquid metal or salt.

 

[Astronuc]

But fast breeder reactors are small because they use high enrichments and are cooled by liquid metal or salt.

Fast reactors use high enrichments because they are small, and in the case of breeder reactors, they use U-238 as a blanket. The liquid metal is used since it doesn't moderate (thermalize) the neutrons, it has very high thermal conductivity compared to water, and it doesn't corrode metal alloys at high temperature as water or steam does. Fast reactors can also run at higher power densities than even the highest density LWR, since departure from nucleate boiling (DNB in a PWR) or critical heat flux (CHF) to dryout, is not a concern. Voiding in sodium however is a concern.

The size and enrichment of core depends on it's mission and the technology being employed. As was pointed out, propulsion reactors are normally designed for a whole-core single life time. Alternatively, power reactors are designed to shutdown on a schedule (annual/12-mo, 18-mo, or 24-mo) for refueling in which some fraction of the core is replaced with fresh fuel. Commercial fuel has regulatory limits on burnup (energy produced per unit mass of initial heavy metal (U, Pu)), and in some cases, residence time, and on enrichment.

Small research reactors have used highly enriched uranium, but most research reactor fuel has been replaced with lower enrichments (of up to about 20% U-235). U-233 and Pu-239/240/241 could be used, but there are restrictions based on proliferation concerns.

 

[mheslep]

Voiding in sodium however is a concern.

Why would that be, as it has no moderation effect thus no reactivity effect? A sodium void would effect heat transfer but heat buildup must inevitably be managed reactivity controls.

 

[Astronuc]

Why would that be, as it has no moderation effect thus no reactivity effect? A sodium void would effect heat transfer but heat buildup must inevitably be managed reactivity controls.

Sodium is a neutron absorber and thus a sodium void means more neutrons are absorbed by the fuel, and that means a positive reactivity in a fast reactor.

 

[Zackary Miller]

Sodium is a neutron absorber and thus a sodium void means more neutrons are absorbed by the fuel, and that means a positive reactivity in a fast reactor.

From my limited understanding voiding in sodium coolant in a fast reactor is worse than voiding when you have water moderator/coolant as the sodium only absorbs neutrons as far as the nuclear reaction itself goes (yes, it cools the core, preventing some expansion from heat, but not much else).

Thus removing sodium should, as you said, cause less absorption by non fuel and more absorption by fuel. How bad this is depends upon reactor design. Now If we are talking about water cooled and moderated reactors, from my limited understanding voiding results in a loss of moderation, thus increasing the neutron absorbtion by non-fuel and leakage, resulting in negative reactivity.

 

[snorkack]

That is already the case now. Standard light water reactors produce as much as 50% of their energy from fissioning of plutonium created from U-238.

But they are not designed to be built with separated Pu.

You can make small compact reactors using U-233, U-235 or Pu-239, they are equivalent for the purposes of power reactors.

Not quite.
Highly enriched U-235 has something like 3 times the critical mass of Pu-239 or U-233.
Between Pu-239 and U-233, Pu-239 is hotter itself and more liable to be contaminated with other, even hotter Pu isotopes.

In general you can make a reactor any size with any fuel size you want, just depending on the application.

You cannot make a reactor smaller than critical mass.

 

[vemvare]

Thanks for the great replies! One thing I wonder about is how the Integral Fast Reactor were supposed to work .If I've understood it correctly, one of the points of re-processing the metallic fuel used in the EBR II is to mix in "fresh" Pu from the blanket into the reprocessed (fission products removed) "core metal", so that the balance between 239-Pu and 240-Pu doesn't shift enough to shut the whole thing down, the blanket metal containing proportionally more 239. Is perhaps 240-Pu bred into fissile actinides and burned, but only if enough 239 is present?

 

[mheslep]

Sodium is a neutron absorber and thus a sodium void means more neutrons are absorbed by the fuel, and that means a positive reactivity in a fast reactor.

Thanks.

I see that such a void would gain positive reactivity for a fast reactor, not that the total reactivity is driven positive, especially given fast reactor designs have strong heat driven negative reactivity. I imagine localised increases to reactivity from a nearby void have their own localised problems like causing unwanted spatial power gradients leading to fuel damage.

There must still be a command driven method of stopping a fast reactor that I can't quickly locate in the literature. The only apparent method would be changing the geometry to deprive the reactor of criticality, though I can't visualise how that's practical.

 

[Astronuc]

But they are not designed to be built with separated Pu.

Some actually are. Separated Pu is typically blended with depleted U to match the equivalent fissile content of UOx fuel. Commercial power reactors would not be fueled with PuO₂, since the fissile content wuold be to high without being blended into non-fissile material. Separated Pu has a high fissile content, since it is mainly Pu-239 and -240. The Pu could be blended into ThO₂, or some inert matrix.

Not quite.

Highly enriched U-235 has something like 3 times the critical mass of Pu-239 or U-233.

Between Pu-239 and U-233, Pu-239 is hotter itself and more liable to be contaminated with other, even hotter Pu isotopes.

The comment about highly enriched fissile isotopes applies to the pure form, and the pure form is definitely not used in a commercial power reactor, and it is used in certain special reactor applications.

For a commercial power reactor with fuel with an enrichment of 5% U-235, the fissile content is limited and the mass of fuel would be the same, but enrichments different based the fissile content of the Pu or U. For reactor grade MOX, the Pu content would be about 6 to 8% Pu to match the equivalent to U-235.

You cannot make a reactor smaller than critical mass.

Well, as the core needs to be able to achieve criticality with some excess reactivity to accommodate the depletion of fissile isotopes. Regardless of the fissile isotope, fuel cladding is subject to limits on heat flux and temperature.

 

[Astronuc]

Thanks.

I see that such a void would gain positive reactivity for a fast reactor, not that the total reactivity is driven positive, especially given fast reactor designs have strong heat driven negative reactivity. I imagine localised increases to reactivity from a nearby void have their own localised problems like causing unwanted spatial power gradients leading to fuel damage.

There must still be a command driven method of stopping a fast reactor that I can't quickly locate in the literature. The only apparent method would be changing the geometry to deprive the reactor of criticality, though I can't visualise how that's practical.

With a proper core design, power peaking can be flattened, and the operator maintains a margin to the power level that would cause local voiding. I work with some folks who were involved with FFTF and some with EBR-II, so I'll try to find out if there were any events, or what the mitigating procedures were in the event of a positive reactivity or voiding event. It's been more than 30 years since I've had to think about fast reactor transients. I'd have to look at the accident/transient scenarios to see what precursor conditions would lead to Na voiding, because that is what the protection system should respond to.

 

[snorkack]

Some actually are. Separated Pu is typically blended with depleted U to match the equivalent fissile content of UOx fuel. Commercial power reactors would not be fueled with PuO₂, since the fissile content wuold be to high without being blended into non-fissile material. Separated Pu has a high fissile content, since it is mainly Pu-239 and -240. The Pu could be blended into ThO₂, or some inert matrix.

The comment about highly enriched fissile isotopes applies to the pure form, and the pure form is definitely not used in a commercial power reactor, and it is used in certain special reactor applications.
For a commercial power reactor with fuel with an enrichment of 5% U-235, the fissile content is limited and the mass of fuel would be the same, but enrichments different based the fissile content of the Pu or U. For reactor grade MOX, the Pu content would be about 6 to 8% Pu to match the equivalent to U-235.

The original question was about small reactors with goal of portability. They would then need to have high enrichment.

Well, as the core needs to be able to achieve criticality with some excess reactivity to accommodate the depletion of fissile isotopes. Regardless of the fissile isotope, fuel cladding is subject to limits on heat flux and temperature.

And these temperature limits depend on the chemical composition of fuel.
But yes, a small reactor with highly enriched fuel necessarily has smaller power at the same temperature.

 

[Astronuc]

And these temperature limits depend on the chemical composition of fuel.

Somewhat. The temperature limits on the cladding depend on the cladding properties, and are usually related to the materials corrosion behavior in the coolant, and strength/creep. The general design requirement is that the fuel is coolable, and the core maintains a controllable geometry. Metal fuel has a lower melting point than ceramic fuel, so the temperature limit on metal fuel is lower than that of ceramic fuel.

But yes, a small reactor with highly enriched fuel necessarily has smaller power at the same temperature.

The temperature of the fuel depends on the coolant and flow rate, and size of the core, and power density. Some small core, e.g., HIFR, can develop a very high power density, which is much greater than the power density in a much large commercial core. The power level is 85 MW, but it has a unique geometry and does not generate electricity, and it is about the size of a large washing machine. It's primary function is a high neutron flux.

The Al cladding means the cladding temperature must be relatively low - much less than a commercial power reactor. "Under these conditions the inlet coolant temperature is 120°F (49°C), the corresponding exit temperature is 156°F (69°C), and the pressure drop through the core is about 110 psi (7.58 × 10⁵ Pa)." That's a high pressure drop for a small reactor, but the coolant exit temperature is not sufficient for an appreciable electrical generation.

https://neutrons.ornl.gov/hfir
https://neutrons.ornl.gov/hfir/core-assembly

Most small research reactors have been converted from highly enriched U-based fuel to relatively low enriched fuel (~20%), and they are primarily used as neutron sources for experiments and isotope production.

One could construct a small 85 MWt reactor to generate 25 to 40 MWe, depending on the thermodynamic cycle design, but it would be substantially larger than HIFR. In addition, rather than a short (~ monthly) cycle, the plant would need to operate with some reasonable capacity factor for a year or longer, or maybe semi-annually.

In reactor, the purpose of the fuel is provide heat (thermal energy), but it must also retain fission products. In addition, the fuel must maintain geometric stability during it's lifetime in the core, so that the core can be shutdown in an emergency in order to prevent an accident, such that the staff and the public are not exposed to radiation, and the plant is not damaged. The fuel temperature is limited to prevent geometric distortion and maintain structural integrity.

The general requirements for fuel and reactors are enumerated in the General Design Criteria in 10 CFR 50 Appendix A. These are mandatory and legal requirements.
http://www.nrc.gov/reading-rm/doc-collections/cfr/part050/part050-appa.html

There are numerous sections of 10 CFR that apply to regulation of the nuclear systems and radioactive material. Other nations have similar regulations. There are also numerous acts, e.g., Atomic Energy Acts, and related Acts, in the US Code, as well as NRC Regulatory Guides that pertain to nuclear power reactors, other types of reactors, and the supporting infrastructure.
http://www.nrc.gov/reading-rm/doc-collections/reg-guides/

 

[Astronuc]

From my limited understanding voiding in sodium coolant in a fast reactor is worse than voiding when you have water moderator/coolant as the sodium only absorbs neutrons as far as the nuclear reaction itself goes (yes, it cools the core, preventing some expansion from heat, but not much else).

Thus removing sodium should, as you said, cause less absorption by non fuel and more absorption by fuel. How bad this is depends upon reactor design. Now If we are talking about water cooled and moderated reactors, from my limited understanding voiding results in a loss of moderation, thus increasing the neutron absorbtion by non-fuel and leakage, resulting in negative reactivity.

In water moderated reactors, water voiding reduces the moderation. In a pressurized water reactor, the voiding would generally add negative reactivity. Some high power density (high duty) PWRs may have some limited nuclear boiling in the core, but most PWR cores are designed to limit the extent of nuclear boiling since it can lead to high corrosion of the cladding, and can also lead to the precipitation of crud in the boiling areas.

Boiling water reactors are designed to have boiling in the upper part of the core. The boiling starts in the lower half of the core, usually within the first 1 meter. The boiling progresses from nucleate boiling to bulk boiling. Exit qualities may be as high as 75 - ~80% in hot BWR channels.
http://ocw.mit.edu/courses/nuclear-...0/lectures-and-readings/MIT22_06F10_lec13.pdf

The high void in BWRs hardens the neutron spectrum and allows for Pu to be produced thus saving on enrichment, since the Pu can be used as fuel in the latter part of the cycle and lifetime of the fuel.

 

[snorkack]

Somewhat. The temperature limits on the cladding depend on the cladding properties, and are usually related to the materials corrosion behavior in the coolant, and strength/creep. The general design requirement is that the fuel is coolable, and the core maintains a controllable geometry. Metal fuel has a lower melting point than ceramic fuel, so the temperature limit on metal fuel is lower than that of ceramic fuel.

And Pu has, for one, much lower melting point than U, and for another several allotropic transformations.

One could construct a small 85 MWt reactor to generate 25 to 40 MWe, depending on the thermodynamic cycle design, but it would be substantially larger than HIFR. In addition, rather than a short (~ monthly) cycle, the plant would need to operate with some reasonable capacity factor for a year or longer, or maybe semi-annually.

What was the thermal power of Kosmos 954 reactor?

 

[Astronuc]

And Pu has, for one, much lower melting point than U, and for another several allotropic transformations.

And that is why no one in their right mind would propose unalloyed metal fuel for a commercial power reactor. As it is, metal fuel (even alloyed) is problematic.

What was the thermal power of Kosmos 954 reactor?

I'm not sure why one would bring a up a military reactor designed for satellites. It would be inappropriate for a commercial power reactor, any more than the auxiliary power unit on a Boeing 747 would be appropriate.

Apparently the BES-5 reactor thermal power was 100 kWt and the system produced a few (?) kWe, for its mission. I've seen numbers from 3 to 10 kWe.
http://www.svengrahn.pp.se/trackind/RORSAT/RORSAT.html

 

[snorkack]

While low enriched reactors have been operated on ships, most reactors, whether on submarines, aircraft carriers or icebreakers use high enriched fuel for considerations like space constraints.
So: why would high enriched fuel reactor using metal alloy rather than ceramics and fast neutrons not make sense on a civilian ship?
How about a compact and fast civilian reactor designed to use bred Pu-239 rather than enriched U-235?

 

[Astronuc]

While low enriched reactors have been operated on ships, most reactors, whether on submarines, aircraft carriers or icebreakers use high enriched fuel for considerations like space constraints.

Submarines must stay underwater for long periods, and fossil power systems require air to function. Certainly nuclear systems eliminate the need for fuel storage and resupply. In most cases, naval reactors are designed to operate without periodic refueling.

So: why would high enriched fuel reactor using metal alloy rather than ceramics and fast neutrons not make sense on a civilian ship?

Economics. Highly enriched material is expensive.

How about a compact and fast civilian reactor designed to use bred Pu-239 rather than enriched U-235?

Compact fast reactors and the fuel are much more expensive the small LWRs and their fuel.

Some background on marine propulsion.
http://www.world-nuclear.org/info/Non-Power-Nuclear-Applications/Transport/Nuclear-Powered-Ships/
http://www.world-nuclear-news.org/NN_Full_steam_ahead_for_nuclear_shipping_1811101.html

The head of the large Chinese shipping company Cosco suggested in December 2009 that container ships should be powered by nuclear reactors in order to reduce greenhouse gas emissions from shipping. He said that Cosco was in talks with China's nuclear authority to develop nuclear powered freight vessels. However, in 2011 Cosco aborted the study after three years, following the Fukushima accident.

In 2010 Babcock International's marine division completed a study on developing a nuclear-powered LNG tanker (which requires considerable auxiliary power as well as propulsion). The study indicated that particular routes and cargoes lent themselves well to the nuclear propulsion option, and that technological advances in reactor design and manufacture had made the option more appealing.

http://www.world-nuclear.org/info/Nuclear-Fuel-Cycle/Power-Reactors/Small-Nuclear-Power-Reactors/

The articles do not address specific design information. Of course, naval propulsion system details are not available in the public domain for obvious reasons.

There are some critical design considerations that must be addressed. The choice of power, power density, and thermodynamic cycle necessitate the 'correct' choice of fuel and structural materials that can withstand the temperatures and in-core irradiation environment. Part of the fuel choice is the refueling strategy.

 

[vemvare]

Having read up a little, in Na-cooled fast reactors voiding leads to an increase in power, a.k.a a positive void coefficient. Lead,or LBE on the other hand boils at a much higher temperature but has the disdvantage of attacking/dissolving structural materials. Molten salts seem interesting, but can such a coolant loop + reprocessing units possibly be made small?

What kind of fast or thermal reactor would be easier to "handle" (construct, maintain and refuel) if the fuel is to be reactor grade plutonium, with another parameter being that it and any primary coolant loop must be small in order for it to be used on ships.

--
I'm thinking a fast molten-salt fast reactor would be self-limiting the same way an aquaeous homogenous reactor is, if it boils the total volume of the liquid expands, and the core geometry is suddenly wrong and no longer critical.

 

[Ahmed Shaker]

Having read up a little, in Na-cooled fast reactors voiding leads to an increase in power, a.k.a a positive void coefficient. Lead,or LBE on the other hand boils at a much higher temperature but has the disdvantage of attacking/dissolving structural materials. Molten salts seem interesting, but can such a coolant loop + reprocessing units possibly be made small?

What kind of fast or thermal reactor would be easier to "handle" (construct, maintain and refuel) if the fuel is to be reactor grade plutonium, with another parameter being that it and any primary coolant loop must be small in order for it to be used on ships.

--
I'm thinking a fast molten-salt fast reactor would be self-limiting the same way an aquaeous homogenous reactor is, if it boils the total volume of the liquid expands, and the core geometry is suddenly wrong and no longer critical.

Gas, gas is non-corrosive and has a no void coefficient. A gas cooled fast reactor can have a power density of 200 MW/M³ , however due to the low thermal inertia of gas, meltdown could occur rapidly if a primary loop is lost.

 

[Astronuc]

In-core structural materials in any reactor will be challenged. The challenge for the scientist/engineer is to determine the optimal material for a given life-cycle. Temperature and irradiation have very significant effects on materials in terms of dimensional stability, strength, transmutation, and erosion/corrosion (loss of material).

Fuel systems must contend with the accumulation of fission products, especially noble gases which are released to the void volumes in the fuel, or create void volume due to precipitation and migration of fission gas bubbles, or fuel relocation, as well as fuel-cladding chemical attack. For every fission, two atoms are produced, which leads to an increase in volume of the fuel, which can be partially accommodated with porosity. At some point the available (initial) porosity gets filled and the fuel material begins to expand.

BTW - https://www.iaea.org/INPRO/cooperation/5th_GIF_Meeting/GFR_Stainsby.pdf
https://www.oecd-nea.org/science/meetings/ARWIF2004/2.01.pdf

Core coolability and reactivity control must be maintained - always.

 

[vemvare]

They changed the design parameters from 600 to 2400MWth, the graph on page 21 on the second document is particularly interesting. Miniaturizing a gas-cooled fast design might be impractical, or at least it'd be dangerous since the volumetric power density increases with smaller sizes. Also interesting that natural convection (be it provided that the gas pressure doesn't drop), is calculated as being capable of cooling a system with such a high volumetric power density. I'm guessing that the possible need for a secondary coolant loop with heat-exchangers (big due to lower thermal conductivity of gas) might make such a system even less likely to be miniaturized.

Would "shine" from fission products in the primary coolant loop be much of a problem?

 

[Astronuc]

They changed the design parameters from 600 to 2400MWth, the graph on page 21 on the second document is particularly interesting.

Usually, there are economies of scale related to containment and balance of plant. It would be less expensive to build one 2400 MWt system than four 600 MWt units.

Miniaturizing a gas-cooled fast design might be impractical, or at least it'd be dangerous since the volumetric power density increases with smaller sizes.

The power density is set by operational restrictions and fuel design. SMRs based on LWR technology tend to have lower power density than the larger PWRs.

Also interesting that natural convection (be it provided that the gas pressure doesn't drop), is calculated as being capable of cooling a system with such a high volumetric power density.

The natural convection is an operating regime in the decay heat removal (DHR) system, and decay heat is a fraction of the full power conditions.

I'm guessing that the possible need for a secondary coolant loop with heat-exchangers (big due to lower thermal conductivity of gas) might make such a system even less likely to be miniaturized.

Other than BWRs, nuclear power plants tend to have a primary cooling loop since it is possible that fuel failures will occur and release fission products into the cooling system, as well as coolant other than He would become activated. It is desirable to have little or no activity in the power generation cycle.

The 'shine' of fission products, which consists of beta particles and gamma rays, is inherent, and most of that is deposited in the fuel, core structural materials or coolant. Cores are enshrouded by structures and coolant, and are contained in pressure vessels. The pressure vessels are within a containment building which also provides for shielding.