question about nuclear fast reactors

[mr friendly guy]

I have been reading about some of the so called Generation 4 nuclear reactors and one of the designs is the fast neutron reactor. On the surface it sounds neat since it can utilise more of the uranium vs older designs. So I would like to know why have the countries who do have it, Japan, US, Russia etc haven't pursued further research in it.

I understand its uneconomical (and I would like more details about it) but also if there are any other downsides to it.

Any help would be appreciated.

[Rabid]

Basically, as I understand it, the technology was judged to be too unsafe at the time, what with the use of a (rendered highly radioactive) coolant that could explode if it where to come into contact with water (liquid sodium). And the systemic risk of the reactor blowing up if mismanaged, because most of the designs, as far as I know, aren't fail-safe.

After this, there was the whole "stop nuclear power" shtick that made it difficult to fund research into innovative fission reactor designs, and so the world's nuclear agencies were forced to make a choice with their budgets, and they chose to continue with a known, proven technology (light water reactor) and apply incremental upgrade to it and the existing reactor in exploitation ; rather than take the risk to lose billions-of-dollar-equivalent in the development of a brand new type of reactor whose chances of commercial success were deemed remote at best.



TL;DR : The exigence for safety was the biggest blow, and the shrinking/limited budgets in nuclear research was the coup-de-grace.


However, nothing preclude that if fusion doesn't prove viable in the 30-50 years to come the technology could make a come-back, made safer and more economical by roughly half a century of research and innovation.



Note : I made it all up just by reading the wikipedia article and with some trivia running in the back of my memory, so I could be talking bullshit. Best to not trust my words on the subject.

[bz249]

The thing about nuclear energy is so that fuel is dirt cheap. The cost of the fuel represent less than 10% of the actual production cost and the majority of the fuel cost comes from the investment spent into the enrichment infrastructure. So unlike a fossil fuel plant (where thermal efficiency is the king) a better burning ration is not really important (there is more than enough uranium all over the world, and factor of two increase in the ore price makes the reserves roughly ten times greater or so). The crucial stats are the cost of the investment and operation and maintenance. And that is the problem with Gen IV., the standard Light Water Reactor is more or less a mature technology (or at least the core of it, there is always room to research this and that) unlike the Gen IV, which requires a massive investment in research. And even if Gen IV. reactor types will be available they generally requires a much more exotic machinery... which may decrease the O&M cost, but an increase is more probable.

So unless LWR reactors will be banned any revolutionary design is in a serious economic disadvantage simply by requiring extra investment in the research.

[Magis]

Technologically, it is primarily a problem of kinetics. The average neutron generation time (which is the average amount of time between the birth neutron A, and the birth of neutron B which was produced by fission caused by neutron A), for a fast reactor is very small. This means reactivity changes will produce a much faster power growth rate in a fast reactor, because the multiplication period is on a smaller timescale. And by much faster, I mean power increases that are potentially thousands of times faster than thermal reactors. The power growth/decay rates are so large for fast reactors that mechanical control systems are unable to respond fast enough to counter normal operational perturbations in the chain reaction.

Another problem is with the delayed neutron fraction (referred to as beta). When a fission reaction takes place and released neutrons, some of those neutrons are released promptly (on the order of a nanosecond), and some are released later on, anywhere between seconds later to months later. Having a large fraction of delayed neutrons is good for controllability, because the control system can see changes in the prompt neutron population and make control adjustments preemptively before the delayed neutrons belonging to the same generation even appear. The delayed neutron fraction of fast reactors is much smaller than thermal reactors. Also, with a small beta, the margin between normal operating conditions (criticality sustained by contributions from both prompt and delayed neutrons) to a condition called prompt criticality (criticality sustained by prompt neutrons alone) is much smaller. As a result, reactivity perturbations, such as those arising from a failure of a component or some other accident, that would be tolerable for a thermal reactor can push a fast reactor into the prompt supercritical regime, which basically means that the reactor is boned.

There are also nuclear data issues with fast reactors. Nuclear reaction probabilities, fission neutron yields, etc., are not well known in the fast part of the energy spectrum. There has been a comparative lack of experimentation at those energies, and so the industry has been unable to establish the values of physical variables with any large degree of confidence. This means that simulations and other calculated predictions of fast reactor behavior are not necessarily reliable.

I should also mention that a lot of bread-and-butter nuclear engineering theory has been developed by implicitly assuming application to a thermal reactor. For example, a common assumption is to assume that neutron scattering is entirely elastic, which completely breaks down in system dominated by fast effects where inelastic scattering is important.

I guess that's a decent summary, even if it's a bit heavy on technicals.

[Hamstray]

Basically, as I understand it, the technology was judged to be too unsafe at the time, what with the use of a (rendered highly radioactive) coolant that could explode if it where to come into contact with water (liquid sodium).

There are much safer alternatives to the sodium coolant: Pb or PbBi.
Lead also has several advantages to water cooling:
1. it does not contaminate (although Bi can transmutate to Polonium)
2. the reactor operates under normal pressure (also with sodium)
3. lead shields off additional radiation
the downside with liquid metal reactors is that they are hard to salvage once they freeze over

in the recent past there has also been increased interest in small modular reactors based on submarine technology (such as PWRs and a few soviet LFRs), though this is in conflict with big energy corporations since it goes down the decentralization route.
amongst those:
hyperion 25MWe (i think i stumbled across it on this board somewhere)
gidropress 10MWe (the company that did the BM40A reactor, website seems to be down)
these claim to be on the ready for shipment in the very near future though i'm a bit sceptical since european efforts like the MYRRHA experimental reactor are decades away. i've read in one of their papers though that they claim europe is experienced in fast reactor technology, which is almost certainly a lie.
as for bigger reactors i principally deem them unsafe, though not as unsafe as BWRs.

...
This means reactivity changes will produce a much faster power growth rate in a fast reactor, because the multiplication period is on a smaller timescale. And by much faster, I mean power increases that are potentially thousands of times faster than thermal reactors. ...

What i've read on fast reactors like the IFR it is possible for the kinetic energy to be strong enough to have the core expand in a rate that slows down the reaction.

[Hamstray]

There are also nuclear data issues with fast reactors. Nuclear reaction probabilities, fission neutron yields, etc., are not well known in the fast part of the energy spectrum. There has been a comparative lack of experimentation at those energies, and so the industry has been unable to establish the values of physical variables with any large degree of confidence. This means that simulations and other calculated predictions of fast reactor behavior are not necessarily reliable.

I'd say experimentation on fast fission behavior has been excessively done and is more than sufficient. The problem is the industry barely had access to the data due to confidentiality issues.

[Sea Skimmer]

Reactors not cooled by water are a limited idea; you get a bad leak from say… an earthquake and suddenly pouring in river water or seawater isn't such an easy emergency option. Anything powered by liquid metal or liquid sodium has far too many operating problems to be worthwhile; if the plant ever goes cold its likely ruined and coolant leaks (these simply are going to happen, it is reality) become VERY big problems. The Soviets built liquid metal reactors for the Alfa class submarines... thought that external heaters would be sufficient to keep the coolant warm but it didn't work and they ended up running the reactors full time, and loosing one reactor to a failure. The USN had one liquid metal reactor, which it ended up ripping out of the sub.

Gas cooling is a better as many gas cooled designs can cool by natural circulation with no power making loose of coolant less critical... but that can still run into problems and some of the designs can actually have the reactor core catch on fire as happened at Windscale in 1957. Generally it is much easier and cheaper to go with known technology, and if fuel supply becomes an issue... reprocess the damn fuel. Lots of people have been studying Gen IV for a long time, some prototypes exist but the cost of validating such designs is staggering.

[Hamstray]

Reactors not cooled by water are a limited idea; you get a bad leak from say… an earthquake and suddenly pouring in river water or seawater isn't such an easy emergency option. Anything powered by liquid metal or liquid sodium has far too many operating problems to be worthwhile; if the plant ever goes cold its likely ruined and coolant leaks (these simply are going to happen, it is reality) become VERY big problems...

How so? if the coolant (i.e.: lead) goes cold it isn't going anywhere and is unlikely to leak into the groundwater or anything. You probably want to keep stuff like water out of the reactor, because water is a neutron moderator and would have devastating effects when it gets into a fast reactor core.

The USN had one liquid metal reactor, which it ended up ripping out of the sub.

Yes they loaded a sodium cooled reactor into a submarine. What on earth were they thinking?

[Sea Skimmer]

How so? if the coolant (i.e.: lead) goes cold it isn't going anywhere and is unlikely to leak into the groundwater or anything. You probably want to keep stuff like water out of the reactor, because water is a neutron moderator and would have devastating effects when it gets into a fast reactor core.

I'm not saying you should put water in a fast reactor, I'm saying that's simply a reason to never build them in the first place.

Yes they loaded a sodium cooled reactor into a submarine. What on earth were they thinking?

She was only the second nuclear submarine ever; USS Nautilus was not very fast and they hoped that the sodium reactor would allow for 30+ knot submarines on relatively small displacements. USS Sea Wolf wasn't any faster; but the main point was her total machinery plant was much smaller for about the same speed. Sodium was not just supposed to be more thermally efficient, it also allowed for much higher temperature steam allowing for a small higher powered turbine. As it turned out PWR designs got much better very quickly and building a 35 knot nuclear submarine with one soon proved totally feasible; though the USN was never able to come up with a small and yet still combat effective attack submarine. The shear size of the required sonars and demand for ever more volume and weight for silencing pretty much precluded making them small anyway. The Soviet Alfa class was had a very twisted evolution and was basically an attempt to make a fighter plane out of a submarine. Boy was that a mistake.

[bz249]

How so? if the coolant (i.e.: lead) goes cold it isn't going anywhere and is unlikely to leak into the groundwater or anything. You probably want to keep stuff like water out of the reactor, because water is a neutron moderator and would have devastating effects when it gets into a fast reactor core.

The really nice thing about water that it is an inert thing, sure there are some metal-water vapor reactions, but compared to a solid metal-liquid metal it is a really a gentle material. The second bad thing about liquid metals that they are solid at room temperature, so if the pipes/machinery gets too cold you end up with a zillion dollar worth of slag because there is no way to clean the thing up. Water is liquid so it will flow out, or at most can be blown out by pressed air. So unless a meltdown happens the investment is safe (and even in case of a meltdown part of the machinery is still salvageable). So the only Gen. IV reactor type which has a slim chance to actually function as a commercial reactor is the Supercritical Water Reactor (though the design team badly needs a marketing expert, who tells them that Supercritical and Nuclear Reactor in the same sentence is really bad advertisement). But the development there also costs money and it is likely that this design would be also more expensive than a standard PWR/BWR thus the increased thermal efficiency/burnout does not help much. Even reprocessing is barely viable with the actual ore prices.

[Magis]

The really nice thing about water that it is an inert thing, sure there are some metal-water vapor reactions,

Yes, like an exothermic Zirc/steam reaction that destroys fuel cladding integrity at high temperatures and produced hydrogen, naturally leading to hydrogen explosion risks, i.e. exactly what happened at Fukushima.

So the only Gen. IV reactor type which has a slim chance to actually function as a commercial reactor is the Supercritical Water Reactor (though the design team badly needs a marketing expert, who tells them that Supercritical and Nuclear Reactor in the same sentence is really bad advertisement).

There is currently no known material suitable for SCWR fuel cladding and pressure tubes. Economic feasibility aside, there are enormous technological challenges to overcome in the materials department before an SCWR can even be designed let alone constructed.

[Magis]

I'd say experimentation on fast fission behavior has been excessively done and is more than sufficient. The problem is the industry barely had access to the data due to confidentiality issues.

I assume you are referring to weapons-related research. Unfortunately, the data needed for fast reactors and the needed for nuclear weapons do not entirely overlap. Specifically regarding U238 processes (which are almost entirely absent in a weapon), and higher actinides that are produced during the life cycle of a plant which are again absent in a weapon system. Similarly, delayed neutron characteristics are extremely important for fast reactor control considerations, but delayed neutrons generally don't even appear during the power pulse of a nuclear weapon. So, the data really isn't there for fast reactors like it is for thermal reactors. Numbers exist, but the confidence associated with them is quite low by comparison.

[bz249]

The really nice thing about water that it is an inert thing, sure there are some metal-water vapor reactions,

Yes, like an exothermic Zirc/steam reaction that destroys fuel cladding integrity at high temperatures and produced hydrogen, naturally leading to hydrogen explosion risks, i.e. exactly what happened at Fukushima.

So the only Gen. IV reactor type which has a slim chance to actually function as a commercial reactor is the Supercritical Water Reactor (though the design team badly needs a marketing expert, who tells them that Supercritical and Nuclear Reactor in the same sentence is really bad advertisement).

There is currently no known material suitable for SCWR fuel cladding and pressure tubes. Economic feasibility aside, there are enormous technological challenges to overcome in the materials department before an SCWR can even be designed let alone constructed.

Relatively speaking. Compared to a liquid metal water is really friendly. And the said steam+Zr reaction have a rather high activation barrier and wont work below ~1000°C. At this temperature sodium is already boiling and the Pb-Bi eutectic eaten up the whole reactor core (Zr is soluble enough in liquid Pb anyway at any temperature).

[Hamstray]

Relatively speaking. Compared to a liquid metal water is really friendly. And the said steam+Zr reaction have a rather high activation barrier and wont work below ~1000°C.

Even without reaching that temperature there can be complications. The pressure inside of the primary loop inside of an PWR is about 158 Bar and inside a BWR about 57 Bar, the pressure vessel becomes brittle from neutron irradiation and may explode if poorly maintained.

At this temperature sodium is already boiling and the Pb-Bi eutectic eaten up the whole reactor core (Zr is soluble enough in liquid Pb anyway at any temperature).

If dissolving of the reactor core in molten lead prevents critical reassembly from taking place then I'd say that's a positive thing.