New Nuclear Reactor Design: Desired Features & Goals

[edpell]

What features do you want in a new nuclear reactor design?

I want:
1) no pressure above 1 atm in reactor (OK in secondary heat loop and turbine)
2) passively cooled on shutdown (planned and unplanned)
3) waste lifetime < 1000 years
4) small amount of waste (relative to current day LWR)
5) known reserves of fuel for 200 years of planet at 10 billion people each at 1KW continuous

As far as I can tell LFTR reactors meet all of these. What are you looking for and do you see a reactor design that meets all your goals?

 

[Astronuc]

One will not get less than 1 atm of pressure in the core, because all material have mass/density and pressure goes with mgh or ρgh (head) and fuel has to have some volume = lateral area x height. One may ask for vapor pressures less than some value. Even liquid metal reactors operate at up to 5 atm of pressure - but they are liquid metal.

Then there is the secondary/tertiary power loop which might be the Rankine steam cycle, and many steam cycles operate with as high as pressure as possible. PWR secondary side and BWR primary side has high pressure of ~1000 psia.

The waste (fission products) will be there. Given about 200 MeV/fission, there is not a lot of way to reduce the fission products per unit energy. One can use thorium which produces less transuranics, but is does produce some because thorium based systems produce and use U233, which can become U234, U235, U236, . . . , but much less TU than with U238 in the system. And the structural materials become activated.

 

[gmax137]

I want the passive cooling on shutdown. In my mind, the decay heat is the biggest difference between a reactor and any other energy source. The lack of a true 'off switch' is a fundamental property of the nuclear unit. A completely robust passive solution would go a long way to improve safety.

 

[QuantumPion]

A smallish, turn-key plant with fixed design, cost, and pre-licensed for operation, $500 million for 300 MWe, with passive safety capability. Like a cheaper, miniature ESBWR.

 

[ulriksvensson]

If we're talking future designs, I'd like to see something that doesn't need to boil water. Or at least better than 35 % efficiency.

 

[Drakkith]

Given that the reactors in the Japan incident were 40 years old, what do the newer plants already have for additional safety features?

 

[Astronuc]

Given that the reactors in the Japan incident were 40 years old, what do the newer plants already have for additional safety features?

Passive cooling - for using natural convection and condensation to transfer heat in containment.

I expect they also have their emergency diesel generators in a safe place, but ostensibly they don't count on the need for electrical power to cool a reactor.

 

[Drakkith]

Passive cooling - for using natural convection and condensation to transfer heat in containment.

I expect they also have their emergency diesel generators in a safe place, but ostensibly they don't count on the need for electrical power to cool a reactor.

Excellent.

 

[mheslep]

If we're talking future designs, I'd like to see something that doesn't need to boil water.

The existing http://en.wikipedia.org/wiki/Pressurized_water_reactor" for all types of heat cycles, and thus molten salt presents a solution for a high energy-conversion-efficiency reactor.

 

[mheslep]

A smallish, turn-key plant with fixed design, cost, and pre-licensed for operation, $500 million for 300 MWe, with passive safety capability. Like a cheaper, miniature ESBWR.

Yes, that seems to the most promising business model, except for the security aspect. That is, one has to be able to scale down the small fortune spent on security (upfront and recurring) at a 1GW plant down 3X to your 300 MW turn-key, and I don't think the NRC or the industry will allow that to happen, and thus threaten the status quo.

 

[mheslep]

One will not get less than 1 atm of pressure in the core, because all material have mass/density and pressure goes with mgh or ρgh (head) and fuel has to have some volume = lateral area x height. One may ask for vapor pressures less than some value. Even liquid metal reactors operate at up to 5 atm of pressure - but they are liquid metal.

A molten salt reactor could theoretically run at ambient vapor pressure, and it is the vapor pressure that drives the design of the pressure vessel (what, 20cm steel in a 150 atm PWR?).

Then there is the secondary/tertiary power loop which might be the Rankine steam cycle, and many steam cycles operate with as high as pressure as possible. PWR secondary side and BWR primary side has high pressure of ~1000 psia.

With molten salt the idea would be to go with http://en.wikipedia.org/wiki/Brayton_cycle" operate(d) at 650C and up.

The waste (fission products) will be there. Given about 200 MeV/fission, there is not a lot of way to reduce the fission products per unit energy.

True, though in a molten salt reactor the fission products merge only as ionic salts (e.g. fluoride salts), not as elements that might readily enter biological pathways if exposed to the environment.

One can use thorium which produces less transuranics, but is does produce some because thorium based systems produce and use U233, which can become U234, U235, U236, . . . , but much less TU than with U238 in the system.

Yes, far less transuranics is a rather large advantage of Thorium as it is the TUs that extend the waste hazard window out to thousands or years versus tens or hundreds for a system with no TU.

And the structural materials become activated.

Yes, but is that not true for any neutron producing nuclear reaction?

 

[QuantumPion]

Yes, that seems to the most promising business model, except for the security aspect. That is, one has to be able to scale down the small fortune spent on security (upfront and recurring) at a 1GW plant down 3X to your 300 MW turn-key, and I don't think the NRC or the industry will allow that to happen, and thus threaten the status quo.

The plants would most likely be built as add-on units to existing sites so I don't think the marginal security cost would not that significant.

 

[mheslep]

The plants would most likely be built as add-on units to existing sites so I don't think the marginal security cost would not that significant.

Good idea, though it restricts one of the large advantages of small-modular: distribution close to the demand/load, esp. in developing countries where there are no existing big nukes.

 

[QuantumPion]

Good idea, though it restricts one of the large advantages of small-modular: distribution close to the demand/load, esp. in developing countries where there are no existing big nukes.

Well I am more focused on desired features for America/Europe, where the big problem is the risk involved for private companies building multi-billion dollar reactors that take years to construct. China/India etc have different priorities in that they are more likely to have government financed construction of new large multi-reactor power stations for their growing economies.

 

[mheslep]

Well I am more focused on desired features for America/Europe, where the big problem is the risk involved for private companies building multi-billion dollar reactors that take years to construct. China/India etc have different priorities in that they are more likely to have government financed construction of new large multi-reactor power stations for their growing economies.

I'm not sure can remove developing from developed (nuclear) in this case, as one of the enablers for small-modular is large quantities and thus greater economy of scale which gets one to the cost of $1-2 per Watt you referenced.

 

[ulriksvensson]

The existing http://en.wikipedia.org/wiki/Pressurized_water_reactor" for all types of heat cycles, and thus molten salt presents a solution for a high energy-conversion-efficiency reactor.

Yes, I'm a PWR-physicist myself. Every power plant (fossile or nuclear) boils water. Only difference is how. I was more thinking of something that can convert radiation to electricity directly. Kind of like a solar cell. It's done all the time in neutron detectors but I guess it would be impractical for power generation and probably give less efficiency than boiling water.

 

[mheslep]

Yes, I'm a PWR-physicist myself. Every power plant (fossile or nuclear) boils water.

If it uses water.

Only difference is how. I was more thinking of something that can convert radiation to electricity directly. Kind of like a solar cell. It's done all the time in neutron detectors but I guess it would be impractical for power generation and probably give less efficiency than boiling water.

You may be interested in this 2009 thread in this forum, https://www.physicsforums.com/showthread.php?t=307262&highlight=photon", which discusses some of the two step nuclear radiation conversion work (nuclear radiation -> UV fluorescer -> UV photons -> UV photovoltaics) done by George Miley and Prelas in the 1990s, among other related topics.

http://prelas.nuclear.missouri.edu/Publications/PIDEC Fission Conversion.pdf"

 

[QuantumPion]

Yes, I'm a PWR-physicist myself. Every power plant (fossile or nuclear) boils water. Only difference is how. I was more thinking of something that can convert radiation to electricity directly. Kind of like a solar cell. It's done all the time in neutron detectors but I guess it would be impractical for power generation and probably give less efficiency than boiling water.

The neutrons themselves only account for ~5% of the energy released from fission. A majority of the energy is the kinetic energy of the fission fragments (i.e. heat).

You can turn the heat of fission directly into electricity using a Stirling engine, however it is less efficient then the currently used steam cycles.

There are also materials which can directly turn radiation into electricity called beta voltaics. However the problem is that the radiation tends to quickly degrade the material.

 

[ulriksvensson]

Yes, I forgot this. But agree that much of power generation is done by converting heat to electricity, a technology that is about 150 years old. "There's got to be a better way", David Olive.

 

[Drakkith]

Yes, I forgot this. But agree that much of power generation is done by converting heat to electricity, a technology that is about 150 years old. "There's got to be a better way", David Olive.

The wheel is thousands of years old too, but we still use it. Would you say we haven't had any improvements on the wheel in that whole time?

Steam turbines and other ways to capture heat to produce electricity has improved drastically in 150 years.

 

[mheslep]

Production of electricity using heat engines is also 'drastically' limited by thermodynamics. If I had to guess I'd say heat engines won't see the 22nd century, giving way to some kind of electromagnetic process without those efficiency limitations.

 

[Drakkith]

Production of electricity using heat engines is also 'drastically' limited by thermodynamics. If I had to guess I'd say heat engines won't see the 22nd century, giving way to some kind of electromagnetic process without those efficiency limitations.

I could see this only if:
A. We find some way to convet heat to something else instead of using it in a turbine or other heat engine.
B. If our primary means of producing power goes to something like aneutronic fusion where the primary source of power is charged particles.

 

[mheslep]

I could see this only if:
A. We find some way to convet heat to something else instead of using it in a turbine or other heat engine.
B. If our primary means of producing power goes to something like aneutronic fusion where the primary source of power is charged particles.

https://www.physicsforums.com/showpost.php?p=3313487&postcount=17

 

[Drakkith]

https://www.physicsforums.com/showpost.php?p=3313487&postcount=17

Interesting!

 

[minerva]

If we're talking future designs, I'd like to see something that doesn't need to boil water. Or at least better than 35 % efficiency.

But why is using water as the working fluid intrinsically bad?

Sure, it limits the system temperature a bit, and thus limits thermodynamic efficiency. But it's cheap, abundant, has a high specific heat capacity, and its chemical, thermodynamic, radiochemical and hydraulic characteristics are extremely well characterized.

We can get a bit higher than 35% out of the heat engine, but getting roughly ~50% is about the best you will get in the real world. As a famous nuclear power plant operator once said... In this house, we obey the laws of thermodynamics!

What features do you want in a new nuclear reactor design?

I want:
1) no pressure above 1 atm in reactor (OK in secondary heat loop and turbine)
2) passively cooled on shutdown (planned and unplanned)
3) waste lifetime < 1000 years
4) small amount of waste (relative to current day LWR)
5) known reserves of fuel for 200 years of planet at 10 billion people each at 1KW continuous

As far as I can tell LFTR reactors meet all of these. What are you looking for and do you see a reactor design that meets all your goals?

Either IFR or LFTR both seem to tick all the boxes here.

 

[zapperzero]

I want
- dual-loop, passive circulation air cooling (to be used for electricity generation and as process heat),
- 100% fuel burnup
- no leftover daughter nuclides and if that means lower energy efficiency, so be it
- passive safety features that control reactivity with no automation or user interaction required
- total independence from the power grid (to include an ability to store/use power locally instead of scramming when there is low/no demand)
- unattended operation - startup to safe shutdown with no user interaction required, automation designed to cover multiple simultaneous failure modes.
- direct conversion from hard radiation, alphas and betas to electricity

Oh, and a core catcher that works. And zero emissions (yes, Kr and Xe do count).

And a pony?

 

[nikkkom]

What features do you want in a new nuclear reactor design?

I want:
1) no pressure above 1 atm in reactor (OK in secondary heat loop and turbine)
2) passively cooled on shutdown (planned and unplanned)
3) waste lifetime < 1000 years
4) small amount of waste (relative to current day LWR)
5) known reserves of fuel for 200 years of planet at 10 billion people each at 1KW continuous

Add
(6) not-reactive coolant. (In particular, sodium cooling is bad. Imagine how much worse Fukushima would be if its reactors were leaking not just radioactive water, but radioactive, corrosive, burning SODIUM!)

As far as I can tell LFTR reactors meet all of these. What are you looking for and do you see a reactor design that meets all your goals?

I also like lead-bismuth fast reactor. It seems to fulfil all of the above, sans "passively cooled on shutdown". But safer post-shudown-cooling is not an insurmountable problem for any reactor - just have a large enough pond nearby, and piping to feed water from it by gravity alone...

 

[edpell]

But why is using water as the working fluid intrinsically bad?

Either IFR or LFTR both seem to tick all the boxes here.

I agree on the second point and wonder why IFR research was stopped.

On the first point, water is bad in the primary cooling of the core because it is high temperature, high pressure water if there is a leak it flashes to steam so a large high pressure containment dome is needed and primary cooling is lost. A leak with say liquid salts just means there is a lump of solid salt on the floor and loss can be made up by pumping more liquid salt into the loop. The pumping is easy because it is at room pressure not 400 psi.

Water in the secondary loop is fine. If it fails that just means the turbine stops. It fails safe.

 

[nikkkom]

I agree on the second point and wonder why IFR research was stopped.

On the first point, water is bad in the primary cooling of the core because it is high temperature, high pressure water if there is a leak it flashes to steam so a large high pressure containment dome is needed and primary cooling is lost.

The bad thing about water is actually only high pressure. High temp is intrinsically needed with any coolant, in order to obtain good conversion efficiency. The leak also can cause loss of cooling with any coolant, not limited to high pressure one. However, high pressure coolant, such as water, will boil out and thus escape even through a leak at the top. Molten salt or molten lead wouldn't.

 

[Astronuc]

But why is using water as the working fluid intrinsically bad?

Basically - it is the corrosion metals in water that is the issue, as well as the high pressure, which increases with temperature. Metals prefer to revert to oxides if given the chance.

 

[mheslep]

My chemistry is a rusty but I believe corrosion is mainly due to the water's pH.

 

[Astronuc]

My chemistry is a rusty but I believe corrosion is mainly due to the water's pH.

It's actually much more complicated, and involves the influence of each cation/anion species and their relative amounts.

pH affects passivity of metal surfaces, and is more important to dissolution of metal from certain surfaces (materials), e.g., stainless steel and Ni-bearing alloys such as Inconels, and precipitation of metal oxides on the fuel or elsewhere in the system.

Soluble oxygen in water is important.

Also, in the core, another key factor is radiolysis, which compounds the chemistry, and then electrochemical potential. (Corrosion after all is electrochemical or galvanic).

Finally, the driver between corrosion of the fuel materials, primarily the cladding is the heat flux and temperature.

Of course, the above discussion refers to water reactor systems.


Liquid metal systems have their perculiar issues with respect to element dissolution which changes the surface characteristics of alloys, and then redepostion elsewhere.

Noble gas coolants are chemically inert, but CO2 has it's perculiarities depeding on the fuel system.

 

[swl]

What features do you want in a new nuclear reactor design?

I would like a plant that includes a financial insurance feature to compensate all costs related to it's worst possible failure mode. If the plant design is safe enough, then such insurance should be feasible.

Our current practice of running without insurance places a tremendous and inequitable burden on the public relative to their proximity to the plant.

 

[gmax137]

[rant]

I would like a plant that includes a financial insurance feature to compensate all costs related to it's worst possible failure mode. If the plant design is safe enough, then such insurance should be feasible.

I'm pretty sure every insurance policy on anything has a maximum payout. In other words, nothing is insured for it's 'worst possible failure mode.'

Our current practice of running without insurance places a tremendous and inequitable burden on the public relative to their proximity to the plant.

If you want to see a real burden on the public, try living downwind from a big dirt-burner. And, they don't compensate the thousands who die each year from the respiratory consequences of the normal plant operation, let alone its failure modes.

[/rant]