They will be HUMONGOUS because they need a large surface to radiate away the heat for the passive safety, so they can't be easily put into a containment building.
A core of a PWR plant is _tiny_ for the amount of power it produces (around 3GWt!), just around 5 meters in diameter and 15 meters in height.
The pebble bed reactor in the article (HTR-PM) is around the same size, but it produces a mere 0.25 GWt.
Pebbles themselves are also problematic, they tend to swell, crack, and they can't be reprocessed using the current technologies. They MASSIVELY increase the amount of waste.
credit_guy 7 days ago [-]
This reactor indeed has lower power density than a PWR, but not by a factor of 12. I compared it with NuScale's reactor, which is a PWR SMR. Details about both can be found in [1]. The HTR-PM reactor pressure vessel has a volume of about 640 m3, and yields 105 MWe, while NuScale has a volume of 101 m3 and yields 77 MWe. The power densities come to be 6.1 m3/MWe vs 1.3 m3/Mwe, and the ratio is 4.7x.
Still, this is a good price to pay for getting a meltdown-proof reactor.
> Pebbles themselves are also problematic, they tend to swell, crack, and they can't be reprocessed using the current technologies.
It is simply not true that pebbles tend to swell and crack. Quite the opposite happens: fuel elements in the current generation PWRs tend to swell, crack and burst. This happens because some fission products and decay products are gasses, such as xenon, kripton, radon. They build up in time and create internal pressure. The same happens inside the fuel kernels in the pebbles used in this reactor, but those kernels are specially built to withstand much higher internal pressures.
Here's a relevant quote from [2]:
> As the pressure vessel size is reduced, the more efficient spherical geometry can be adopted, and the required wall thickness drops dramatically to the point that a 35 μm SiC layer can indefinitely contain gas pressures in excess of 100 MPa. This compares to the main reactor steel pressure vessel which may go as high as 20MPa or the Zircalloy cladding which can have pressures up to 10s of MPa in limited conditions – temperatures far below what the ceramic pressure vessel tolerate
As for the reprocessing part, I think you are jumping the gun. There is no reprocessing done in the US, at all, for any type of fuel. Even where reprocessing happens, as in France, the benefit is quite reduced. One can extract some plutonium and unburned uranium, but in the end that will allow you to extract maybe 10-20% more energy from the original amount of natural uranium. It will not make you extract one hundred, or 10 times, or even just twice as much energy. Reprocessing is simply not a game changer. It is not clear at all if it makes economic sense to build the highly complex facilities that do reprocessing, for the limited benefit.
> As for the reprocessing part, I think you are jumping the gun.
The problem is that if the pebbles aren't reprocessed, you now have to store their very large volume. The moderator, graphite, is integrated into them. This is unlike a LWR, where the moderator is water that the spent fuel can be simply lifted out of (after cooling). LWR fuel can be stored after a few years into dry casks; the equivalent for pebbles would be vastly larger and more expensive, and involves storing the spent fuel with its moderator, increasing concerns of criticality (although I imagine they'd be doused with borate or something to prevent that).
yeknoda 7 days ago [-]
PWR pressure is on order 2-3x higher than an HTGR, so that’s a 2-3x thicker vessel. So good savings there
pfdietz 6 days ago [-]
Cost of a pressure vessel is roughly proportional to pressure x volume. HTGR doesn't come out on top by that metric.
Thickness of a pressure vessel wall is proportional to pressure x linear dimensions of the pressure vessel, increasing at a given pressure as the vessel is scaled up.
huijzer 6 days ago [-]
Is NuScale’s design a good one to compare against? Isn’t it a theoretical design only?
I held their stock for a while until I realized they don’t exist to make a reactor. They exist to get funding.
Sounds crazy, but look through their actions. All press releases are just talk about what they will research and with whom they talked or made a “memorandum of understanding”. The CEO CV is also interesting since it lists a whole lot of board positions and titles but it’s not clear what he has actually done.
credit_guy 6 days ago [-]
This is a very cynical take.
I picked NuScale because it's a PWR SMR, and is the only SMR design that was approved by the Nuclear Regulatory Commission. HTR-PM is an SMR, and comparing it with a full scale PWR reactor is not entirely fair, because reactors benefit from the square-cube law: the larger a reactor is the more efficient its neutron economy is, so you can extract more power per unit of volume.
As for NuScale being a scam, it would be probably the most elaborate scam in the history of financial scams. They were founded in 2007 and went public in 2022. People don't spend 15 years to run a scam, especially if this involves not one by two government agencies (the NRC and the SEC). There's a huge probability that such a scam would not work in the end, with an additional likelihood that you get relocated to a correctional facility (see Elizabeth Holmes).
pfdietz 6 days ago [-]
The power density of the reactor vessel of a conventional large PWR is close to 20 MW/m^3 (inverse of 0.05 m^3/MW); a BWR is around half that. The peak power density in a PWR's core is around 100 MW/m^3.
NuScale integrates the steam generator in the "reactor" so the volume is larger, but this means using it is comparing apples and oranges. NuScale's design is also intended to use natural convection instead of forced circulation in accident conditions. This further reduces the power density allowed.
NuScale isn't a scam, but it appears to be founded on faulty principles (that the thing holding back nuclear power was safety concerns, rather than cost) so its business case doesn't appear to be working.
huijzer 6 days ago [-]
> As for NuScale being a scam, it would be probably the most elaborate scam in the history of financial scams. They were founded in 2007 and went public in 2022. People don't spend 15 years to run a scam, especially if this involves not one by two government agencies (the NRC and the SEC).
I wouldn’t call it a scam, but something that runs for 15 years with the involvement of 2 government agencies (read: large bureaucracies) is not unlikely to produce very little.
For example, see NASA’s space program, the European space program, and Boeing.
instagraham 6 days ago [-]
interesting comment - made me look up the stock and it just jumped. Propelled by this news perhaps?
huijzer 6 days ago [-]
price != value
Otherwise Buffett wouldn't be so rich.
cyberax 6 days ago [-]
> This reactor indeed has lower power density than a PWR, but not by a factor of 12.
It's more.
> while NuScale has a volume of 101 m3 and yields 77 MWe
VVER1200 has the inner vessel _diameter_ of 4.2m, height of 11m for the internal volume of 153m^3, and 1200MWe capacity (so around 3GWt).
_THIS_ is what you're comparing it with.
> Quite the opposite happens: fuel elements in the current generation PWRs tend to swell, crack and burst
Nope. A swollen or a ruptured fuel rod in a regular reactor is a reason for SCRAM. The water inside the reactor vessel is constantly monitored for fission products. The individual fuel tablets swell, but they are contained inside zirconium rods.
freeone3000 7 days ago [-]
So for a 3GW pebble bed reactor, we’re looking at a core the size of small house instead of a master bedroom? I don’t see a huge difference here; it’s the same amount of everything else (cooling, pumps, turbines, security) since it produces the same amount of heat/power.
cyberax 7 days ago [-]
> So for a 3GW pebble bed reactor, we’re looking at a core the size of small house instead of a master bedroom?
No, we're looking at a core the size of a small residential tower. Probably around 30 meters in height.
lambdasquirrel 7 days ago [-]
Sounds terrible for a relatively dense place like Western Europe or Japan, but I think this would be fine in the US and China.
LargoLasskhyfv 5 days ago [-]
Why is that? Are current powerplants built in prime real estate of downtown?
JojoFatsani 7 days ago [-]
That’s not too big man
pfdietz 7 days ago [-]
It is, when you consider the problem of dealing with spent pebbles. These are not like the fuel rods of a LWR that occupy a fraction of a much smaller volume. The dry casks to store them would be immense.
Etheryte 7 days ago [-]
One room worth of radioactive waste vs a residential tower worth of radioactive waste is a big difference.
JumpCrisscross 6 days ago [-]
> One room worth of radioactive waste vs a residential tower worth of radioactive waste is a big difference
Most nuclear waste is stuff near the reactor, not the fuel per se. And nuclear waste isn't the Armageddon stuff it's portrayed as in mass media. If this works, waste volume won't be an issue.
akira2501 6 days ago [-]
> Most nuclear waste is stuff near the reactor,
And now your containment has a much larger surface area.
> And nuclear waste isn't the Armageddon stuff it's portrayed as in mass media.
When it's contained. Uncontained waste has been the source of multiple tragedies.
> waste volume won't be an issue.
Process frequency will be.
JumpCrisscross 6 days ago [-]
> Uncontained waste has been the source of multiple tragedies
In the way rolled steel has. We still roll steel, though, because despite the tragedies having narrative heft they're infrequent and small relative to the benefits of the product.
akira2501 6 days ago [-]
> they're infrequent and small
So far. Now you're introducing new technology. Are you comfortable prognosticating that it always will? Are there no lessons to be learned from previous tragedies, in particular, how _small_ radioactive sources are _way_ more dangerous due to peoples inability to identify them and tendency to pick them up?
You introduce medical radiation sources into a country without radiation controls and just a few years later you have an outbreak of tragedies. I'd rather not learn the lesson again the hard way.
JumpCrisscross 6 days ago [-]
> Are you comfortable prognosticating that it always will?
Much more than accepting the deadly status quo!
> Are there no lessons to be learned from previous tragedies
Sure there are. Never take any risks ever isn't one of them.
> how _small_ radioactive sources are _way_ more dangerous due to peoples inability to identify them and tendency to pick them up?
You're describing MRIs more than nuclear power plants, which also produce lots of small nuclear waste.
Maybe he's confusing it with PET scans or perhaps gamma ray imaging with technetium (used for cancer imaging).
Interestingly, the gadolinium in MRI contrast agents, while composed of stable isotopes, is a ferociously good thermal neutron absorber. The element is used in burnable absorbers in some reactors to keep reactivity level as the fuel is burned down.
pfdietz 7 days ago [-]
The other problem is mechanical abrasion of the pebbles, creating radioactive dust.
One of Germany's PBRs had to filled with concrete after it was defueled, they couldn't decontaminate it enough to dismantle it.
shepherdjerred 7 days ago [-]
Is GWt a common abbreviation for gigawatt? I first read that as gigawatt-tons which is a… confusing unit
howenterprisey 7 days ago [-]
Gigawatt thermal, as opposed to gigawatt electric. Gigawatt thermal is the heat your power plant makes, whereas gigawatt electric is the electricity that the heat is used to generate. They're not the same because not all the heat can be converted into electricity, and the percent of heat that gets converted varies from power plant to power plant.
shepherdjerred 7 days ago [-]
Oh that makes total sense! Thanks for explaining
retzkek 6 days ago [-]
There is a similar unit used for tracking nuclear fuel “burnup,” which is how much energy it produced: GWd/t (or GWd/MTU).
Grosvenor 7 days ago [-]
Seems like that would be fine for places where you have < 0.25GWt energy needs, and need a safe power source. Like remote installations/towns. Antarctic research stations, etc.
cyberax 7 days ago [-]
Pebble bed reactors are indeed researched as a source for process heat (e.g. for steel or concrete production). But I really dislike that.
If you just need 250MW of power, then just use electricity sourced from a regular PWR for heating. It'll be cheaper.
Retric 7 days ago [-]
The maximum temperature from nuclear reactors is far below what you need for steel production because otherwise you’d melt the steel used to construct them.
Waste heat can be useful for district heating systems because houses don’t need to reach high temperatures, but few designs give you access to even 300C and nothing currently hits 1,000+C.
pfdietz 6 days ago [-]
Anything at 300 C or below can be provided from renewable sources using industrial heat pumps. Heat is very storable, compared to electrical energy, so if the end user needs heat they could make the heat when renewable energy is cheap and store it, potentially for very long periods. Some district heating systems involve seasonal storage, essentially artificially charged geothermal. Nuclear thermal would be competing with otherwise curtailed renewable energy, which can be very cheap on the margin.
cyberax 6 days ago [-]
Sorry, you're right. Pebble bed reactors can provide helium at around 600C (reasonably up to 700C, but that's at the limits of material science).
klysm 6 days ago [-]
I’m not convinced the size is actually a problem - is that the most constrained resource here? What do we gain from decreasing power density?
I remember that article. An older friend had worked on fluidised-bed reactors at Tsinghua in the 1990s, and I sent it to him.
Using helium indicates a problem with kinematic viscosity of cooling gases? Nitrogen would be non-reactive enough, I assume?
7 days ago [-]
semi-extrinsic 7 days ago [-]
Helium has excellent heat transfer properties (low Prandtl number) and does not undergo nuclear reactions when subjected to a neutron flux.
Nitrogen will undergo an (n-p) reaction to produce carbon-14 which has a half-life of 5700 years.
cyberax 7 days ago [-]
Yup. That's why regular PWRs take care not to nitrogen to pressurize or flush the primary coolant loop after maintenance.
pfdietz 7 days ago [-]
And why reactors that use nitride fuels use fuel made with the isotope nitrogen-15.
The front runner steel for use in fusion reactors, EUROFER-97, contains a necessary small amount of nitrogen. This is enough under some nations' rules to render it into intermediate level radioactive waste after use, due to the carbon-14 content.
cyberax 7 days ago [-]
Oh yeah. I have a friend who was working on researching fusion-safe steels. Solving it fully is going to be a real engineering challenge.
Apparently, even a small natural niobium contamination would make it a low-grade waste.
pfdietz 7 days ago [-]
Yes, niobium is the other problem. I think steel makers are used to adding alloying elements, but not set up for excluding them to ppm levels. I saw someone bemoaning that the steel alone for DEMO would cost 3 billion euros (or dollars?), and I've wondered if this is the reason.
"[...] in the first
layer, nearest the plasma, the rate of production of 94Nb – via neutron capture (n,γ)
reactions on the stable 93Nb of niobium – is so high that Eurofer in this region is
predicted to exceed the France-LLW limit within the first year of operation, and
consequently would not be disposable as LLW under French regulations for more than
1000 years."
It's not just the steel. Beryllium typically contains about 100 ppm U, and an estimate of the cost of purifying it enough to avoid excessive fission products was another billion.
chasil 6 days ago [-]
It will likely be necessary to start with steel made before atmospheric nuclear tests began.
I don't think that would make any difference. The problem isn't traces of radioisotopes in the initial steel, the problem is vastly larger amounts of radioisotopes created in the steel under intense neutron bombardment due to absorption in certain stable impurity elements.
pfdietz 6 days ago [-]
And when I say "intense", I mean it: the integrated neutron flux the first wall of a DT fusion reactor would be exposed to over its operational lifespan is comparable to, or even greater than, that experienced by components of a nuclear weapon in the brief interval of a nuclear explosion.
HPsquared 7 days ago [-]
Nitrogen is a big neutron absorber in gas-cooled reactors. It's actually used as a secondary shutdown mechanism in the UK's AGR reactors. So if it was to leak out (lose pressure) you'd see an increase in power at the same time as a loss of cooling... Not a great idea!
yk 7 days ago [-]
> Other pebble beds: The pebble bed technology and design has previously been used in prototype reactors in China and Germany, but not a larger-scale plant like Shidaowan.
That's wrong, Hamm-Uentrop was a full scale commercial reactor. It did run in total for a week or so between 1985 and 1989 and was then shut down. The fundamental problem is, that the pebbles grind against each other, and being of the same material as pebbles they can grind each other down. (Now if you wonder why this wasn't discovered at the experimental reactor in Juellich, those guys just never mentioned that they lost fuel.)
This video explores an incident with a reactor of a similar design, and very rudimentarily explains the way pebbles and the helium gas is used.
dongobongo 7 days ago [-]
This actually happened in August and September of 2023 and it’s great validation for High Temperature Gas-cooled Reactor (HTGR) at larger scales. I hope they have the guts to also do a full loss of coolant test. I’ve also heard the these two reactors have been turned off for quite a while due to issues with the primary heat exchangers, temperature fluctuations, and uneven cooling - characteristically disadvantages of pebble beds.
There’s of course 2 flavors of HTGR (prismatic and pebble bed), and people choose the pebble version for continuous refueling despite all the drawbacks [1]. But there’s a lot of reasons to do prismatic. Can’t wait to see China’s prismatic HTGR.
Hmmm. The Titanic had sixteen watertight compartments.
Did the pebble-bed reactor's "commercial-scale inherent safety" also pass a test with an compromised container, which would admit air that would cause the graphite to burn?
JumpCrisscross 6 days ago [-]
> which would admit air that would cause the graphite to burn?
Rats, then it would cause about as many deaths per GWh as the coal it's replacing.
natmaka 6 days ago [-]
The real underlying subject isn't "coal or nuclear" but "nuclear or renewables", and a potential nuclear plant not inducing any risk of major nuclear accident (nor any annoying waste? nor any dependency towards any fuel?...) would induce quite new way to consider it.
JumpCrisscross 6 days ago [-]
> real underlying subject isn't "coal or nuclear" but "nuclear or renewables"
Not in the real world. We're deploying solar and wind as quickly as we can because they're the cheapest sources of power. The bottlenecks are production and permitting, and there is no indication either of those are seeing a step change in the next decade.
There is a reason America and Europe, where anti-nuclear sentiment runs ripe, are building gas power plants and infrastructure at the fastest clip in history. In the West, the choice is gas or nukes. In the China and India, it's coal or nukes.
> Nuclear is plateauing, and renewables are booming
Nobody said otherwise. I literally said "we're deploying solar and wind as quickly as we can because they're the cheapest sources of power." You're arguing against a straw man.
Renewables are booming. So is gas [1]. Your analysis fails on two counts. One, it ignores the substitution effect [2]. If we're talking about grid stability, et cetera, this is fine. If we're looking at emissions, it's not.
Two, you're lumping together fossil fuels. That masks the fact that we've added about as much natural gas capacity as solar. The growth rates are different. But so are the base levels.
I'll say it again: we're building renewables as fast as we can. We can't build them substantially faster. That means there is never a choice between renewables and something else; it's always renewables by default. Where there is a choice, therefore, it must be between the other options.
The convenient lie the gas industry has sold the nuclear nervous is that it's a competition between solar and wind and fossil fuels. It's not. We're investing trillions of dollars in gas infrastructure with 20 to 40-year investment theses despite renewables booming because power demand is booming too, and the difference has to be made up somehow. That gas infrastructure's thesis only works if we exclude nuclear energy. (It also precludes us reaching our 2030 and 2050 emissions commitments, but nobody seems to care about those anyway.)
We're building gas instead of nuclear in the West for the same reason China and India are building coal plants: it's cheaper than nuclear. Nobody is acting on emissions.
And also for the same reason China installed 180x more PV than nuclear (on a rated power basis) in 2023: PV is cheaper than nuclear.
China also installed more than 4x PV than coal on a rated power basis in 2023. I don't know the capacity factor of their coal -- I hear it's operated at fairly low capacity factor -- so PV installs could well exceed coal installs (especially net coal installs) on a levelized basis there.
natmaka 5 days ago [-]
Carbon intensity of electricity generation in China:
2011: 752g eqCO2 (production: 4712 TWh)
2023: 582g eqCO2 (production: 9459 TWh)
You wrote "bottlenecks", I showed that we deploy more and more. The reader can decide.
You first criticize me for merely providing data, then for my "analysis". Go figure.
Your points (substitution effect and methane) are pertinent, but my answer wasn't about this but about the fact that there is no sign of abatement of renewables' growth.
> We can't build them substantially faster
One more: nothing, especially historical data, sustains this.
We burn less and less fossil fuel thanks to renewables, which more and more replace it, and there is no clear indicator of this trend to change. This is too slow, granted, however there is no magic wand and demanding an immediate perfect solution is playing upon the nirvana fallacy ( https://en.wikipedia.org/wiki/Nirvana_fallacy ).
> We're building gas instead of nuclear
... because since the 2000's:
- industrial renewables appeared and are (very quickly) more and more adequate. I other words to replace fossil fuels nuclear isn't the only contender anymore. Thanks to renewables: no risk of any major accident propagating very dangerous long-term stuff difficult to recover, no hot waste, no dependency towards any provider of fuel (such as uranium), no weapons proliferation risk, quick and easy deployment, no decommission-related nightmare (see the ongoing case in the U.K.)... No wonder renewables are booming.
- all nuclear-building projects are more-or-less resounding failures
GeoAtreides 6 days ago [-]
the compartments weren't watertight 100%, they weren't sealed at the top
pineaux 7 days ago [-]
Can someone explain to me how this reactor is meltdown proof?
Summary: As the temperature rises, neutron absorption increases, reducing fission and thus temperature.
chickenbig 7 days ago [-]
> As the temperature rises, neutron absorption increases, reducing fission and thus temperature.
Negative fuel temperature coefficient is not an unusual feature.
The real question is whether the heat removal system of the reactor as a whole is sufficient to remove the decay heat to keep the fuel within the limits.
Gibbon1 7 days ago [-]
I remember talking to an engineer at the old GE nuclear research facility in San Jose. He said you can design reactors to be cooled by natural convection.
chickenbig 6 days ago [-]
> He said you can design reactors to be cooled by natural convection.
That's the point of the Westinghouse AP1000; the containment (steel liner) and protection from the outside world (concrete wall) are separated, allowing the liner to cool by convection and water dripping from above. Admittedly you need to top up the water tank at the top, but that is less of a task than trying to push water into the containment.
pfdietz 7 days ago [-]
It just makes them larger. And it makes the building containing them larger. And this makes them more expensive.
NuScale's reactor was originally motivated by the desire to make it safer by using natural convection. But it ends up requiring 1/3rd more labor hours to build a NPP using their reactors than it does to build a conventional large reactor power plant.
cyberax 7 days ago [-]
The reactor vessel is humongous, so the natural convective cooling can carry away the decay heat. The pebbles themselves can tolerate extremely high temperatures (literally glowing white-hot) without burning.
DennisL123 7 days ago [-]
It’s meltdown proof in principle since the process goes into an equilibrium rather than into a runaway process.
euroderf 7 days ago [-]
IIRC there is a question about graphite fires.
AtlasBarfed 7 days ago [-]
Can pebble beds have a cooldown pan similar to a LFTR, where a plug melts and the "pebbles" fall and spread into a pan where they won't stay critical because they are too separated / unconcentrated?
Because the real problem with solid rods is that they ... are solid rods, and if they start "overreacting" you can't split up the rods, unlike a pile of pebbles/spheres.
cyberax 7 days ago [-]
The unique "melt plug" safety story of LFTRs is mostly a fairy tale.
Modern PWRs also have this safety feature, if a core melts down, the molten mass will be contained in a core catcher. Where it'll be mixed with inert material that can provide enough surface area and thermal mass to prevent further fuel mass migration.
The biggest problem in the core catcher design was to make sure that the molten fuel lava spreads out enough for the passive cooling to stop it from melting through concrete.
Pebble bed reactors will have a similar problem. You can "drain" pebble beds somewhere, but then you need to make sure that this "somewhere" can conduct away the decay heat without melting.
AtlasBarfed 6 days ago [-]
... why?
You're saying LFTRs can't have a pool of cooling liquid that the overheating fuel liquid can't mix with and cool down even faster than solid rods?
So how could LFTR safety plug and cooling pool be a fairy tale?
cyberax 6 days ago [-]
A failure that results in the plug melting will mean that the reactor is beyond salvaging. It's essentially the same situation as with PWRs equipped with core catchers.
> can't have a pool of cooling liquid
What "liquid"? Water?
Do you realize what's going to happen if molten salt drops into water? First, there's going to be a steam explosion that will atomize the fuel and spread it through the whole containment building (because the water vapor can't be contained in a reasonable volume).
Then the water will boil away from decay heat, and the fuel lava will continue chewing through your reactor building.
That's why LFTR reactors with "melt plugs" will essentially use the same approach as PWRs: spread the molten fuel across sacrificial concrete cladding.
AtlasBarfed 5 days ago [-]
Gee, I dunno, maybe any of a thousand different materials/chemical reactions to absorb heat? Yeah, I didn't say water. Maybe you just have the dump pool be a bunch of molten thorium or even solidfied thorium salt and that also plummets the neutron economy as the neutrons get absorbed by thorium as the hot uranium salt melts the thorium salts. Maybe you keep the thorium salts liquid.
But fundamentally since the nuclear fuel is a liquid, the cooling pool is very wide and shallow, separating the fuel well past sustained nuclear fission. This is the problem with solid rods, they go runaway, the fuel is a solid rod, you can't separate apart the fuel without shoving some moderator into it. Yeah I don't know the viscosity of Uranium tetrafluoride, maybe that's a problem, but I doubt it.
Why does the plug only melt past some point of no return? The plug can melt at whatever temperature point is desired. A "meltdown" can just be part of the usual fuel flow and recirculation.
The difference is that the liquid spread because, you know it is a liquid. In liquid form? Rather than a solid form or some semi-solid uranium lava. See the difference? What am I missing here?
In the dump pool, it will cool into solid salts. Then you just need to reheat it to recover the fuel, and pass it back through the salt reprocessing systems.
Plus I thought LFTRs have some mechanism for self-moderation by expansion of the fluid when it gets hot, separating the fuel apart and reducing neutron economy.
You seem stuck in the limitations of solid fuel rods, solid fuel reactors, and their inherent inconvenience. Yeah, the liquid is 650 degrees or something like that, but it's still liquid and you can do things with liquids that you can't do with solid rods
cyberax 5 days ago [-]
> Gee, I dunno, maybe any of a thousand different materials/chemical reactions to absorb heat?
Well, I definitely don't know which reaction can absorb on the order of 2GWh of residual decay heat within the first 2 days.
I'm assuming a 3GWt reactor, something that at least can be competitive with PWRs.
To give you some perspective, this amount of energy is enough to vaporize more than 3000 tons of water. More than an Olympic swimming pool.
> Maybe you just have the dump pool be a bunch of molten thorium or even solidfied thorium salt and that also plummets the neutron economy as the neutrons get absorbed by thorium as the hot uranium salt melts the thorium salts. Maybe you keep the thorium salts liquid.
Sigh. It's not the fission that is a problem. Fission will be quenched by all the neutron poisons. Even in Chernobyl or Fukushima the fission stopped immediately after the accident.
It's the decay heat that has to be conducted away.
> Why does the plug only melt past some point of no return? The plug can melt at whatever temperature point is desired. A "meltdown" can just be part of the usual fuel flow and recirculation.
If normal recirculation works, then there's no problem with supplying cooling water. The touted advantage of molten salt reactors is their passive safety, they are supposed to fail safe even if EVERYTHING fails.
> You seem stuck in the limitations of solid fuel rods, solid fuel reactors, and their inherent inconvenience. Yeah, the liquid is 650 degrees or something like that, but it's still liquid and you can do things with liquids that you can't do with solid rods
Yeah, because I actually worked in the nuclear power industry.
pfdietz 5 days ago [-]
> Well, I definitely don't know which reaction can absorb on the order of 2GWh of residual decay heat within the first 2 days.
The sales pitch for salt-cooled reactors is the lack of any coolant that would become pressurized hot gas in an accident. Heat can stay in salt or in other low vapor pressure materials.
The problem with LWRs is the water goes to steam in accidents, and this steam must be contained. This drives the size of the containment building, and the containment building is costly.
An alternative for LWRs would be to filter and vent the steam instead of trying to contain it. This would allow small quantities of radioactivity to escape (including all the noble gas fission products), but the filtering can actually be quite good, reducing emissions by many orders of magnitude. Second generation filtered containment venting systems can filter iodine as well as cesium and strontium. If Fukushima had had such systems the impact would have been far lower.
> The sales pitch for salt-cooled reactors is the lack of any coolant that would become pressurized hot gas in an accident. Heat can stay in salt or in other low vapor pressure materials.
Sodium-cooled reactors and the upcoming lead-cooled reactor also have this property. It turns out to not be such a huge advantage, we have plenty of experience working with pressurized water.
> The problem with LWRs is the water goes to steam in accidents, and this steam must be contained. This drives the size of the containment building, and the containment building is costly.
No, it's really not a problem. The loop doesn't suddenly loose compression if something bad happens. If there's electric power, there's more than enough time to slowly cool down the reactor.
And a containment building (that also protects against external threats like an airplane ramming into the reactor) has more than enough volume if the primary loop is de-pressurized and the water flashes into steam.
> An alternative for LWRs would be to filter and vent the steam instead of trying to contain it.
The water in the primary loop is clean. It's constantly purified by filtration through ion exchange resins. Once the activated oxygen decays (in ~1 hour) you can swim in it (although I wouldn't drink it).
PWRs (actually, all thermal power plants) have areas where steam can be dumped. If you watched "Chernobyl" series, the ridiculous scene with divers was supposed to happen inside such an area ("barboter pool").
Modern PWRs are also designed to do that safely. There's plenty of capacity to condense all the water from the primary loop after the loss-of-cooling. Of course, after that the fuel will melt down, and chew through the reactor vessel.
The filtering system you linked is not strictly necessary for modern PWR designs. They will still be safe in case of an accident with total loss of cooling, but the containment building will be hopelessly contaminated internally. This filtering system can allow the steam to be vented into the atmosphere, perhaps giving more time to fix the emergency cooling systems.
pfdietz 4 days ago [-]
> No, it's really not a problem. The loop doesn't suddenly loose compression if something bad happens. If there's electric power, there's more than enough time to slowly cool down the reactor.
It can in design basis accidents, for example a complete break of a main circulation pipe leading the loss of coolant (LOCA) into the containment. The emergency cooling system would then operate by spraying water into the core that would evaporate into steam that would go right out of the reactor vessel. The containment has to be sized for such an accident.
As an example of such an accident, consider what would have happened at Davis-Besse had the erosion of the lid of the reactor vessel progressed to an actual perforation. As it was, the steel was removed in an area down to the inner stainless steel liner, a liner that was never intended to be load bearing against the internal pressure.
> And a containment building (that also protects against external threats like an airplane ramming into the reactor) has more than enough volume if the primary loop is de-pressurized and the water flashes into steam.
Right, it does. That's why it's so big and expensive, with so much internal volume. If it didn't have to, it could be made much smaller. The airplane requirement doesn't change this; it's easier to make a smaller containment building resistant to aircraft impact than a larger one.
> The water in the primary loop is clean. It's constantly purified by filtration through ion exchange resins. Once the activated oxygen decays (in ~1 hour) you can swim in it (although I wouldn't drink it).
That's true in normal operation, where you might have some small number of fuel rods with cracks or perforations (but even that is getting pretty uncommon these days). It would not be true in a design basis accident, where some or all of the fuel may have partially or completely melted, and where cladding will have been compromised by high temperature reaction with steam. The design must assume essentially all the volatile fission products have gone into the water. At TMI, fission products carried in the water (and also noble gases) raised radiation levels in the containment building to 800 rem/h during the accident.
> The filtering system you linked is not strictly necessary for modern PWR designs.
I offered up the possibility that such systems could replace the large volume containment of modern systems (or at least reduce its size and cost). Sure, they're not obviously necessary if you have a large volume containment already (although some countries ended up requiring them anyway since some accident scenarios do involve venting, as happened at Fukushima, which admittedly had pre-modern designs.)
kevin_thibedeau 7 days ago [-]
It's not, just statistically unlikely assuming no fuel pebbles crack and coalesce their fragments.
m3kw9 6 days ago [-]
The automatic mechanism could fail right?
paranoidrobot 6 days ago [-]
The safety of the system comes from it's design. It doesn't rely on anyone, or any mechanical process cooling it.
> The pebble-bed reactor is designed so that this effect is relatively strong, inherent to the design, and does not depend on moving parts. This negative feedback creates passive control of the reaction process.
From the Wikipedia article on the PBR that was linked in a neighbouring comment.
It should, at least if built correctly, be impossible for it to get into a run-away state. Even if it lost all power, and everyone walked away, it should not melt down.
> New testing done at China’s Shidaowan nuclear power plant has confirmed its ability to be naturally cooled down, an industry-first milestone for achieving commercial-scale inherent safety, according to researchers.
Amazing. Well done! How far this country has come in the last few decades is nothing short of breathtaking.
cynicalsecurity 7 days ago [-]
The obsession with "green" energy wouldn't do us any good, would it.
Rendered at 14:05:25 GMT+0000 (Coordinated Universal Time) with Vercel.
They will be HUMONGOUS because they need a large surface to radiate away the heat for the passive safety, so they can't be easily put into a containment building.
A core of a PWR plant is _tiny_ for the amount of power it produces (around 3GWt!), just around 5 meters in diameter and 15 meters in height.
The pebble bed reactor in the article (HTR-PM) is around the same size, but it produces a mere 0.25 GWt.
Pebbles themselves are also problematic, they tend to swell, crack, and they can't be reprocessed using the current technologies. They MASSIVELY increase the amount of waste.
Still, this is a good price to pay for getting a meltdown-proof reactor.
> Pebbles themselves are also problematic, they tend to swell, crack, and they can't be reprocessed using the current technologies.
It is simply not true that pebbles tend to swell and crack. Quite the opposite happens: fuel elements in the current generation PWRs tend to swell, crack and burst. This happens because some fission products and decay products are gasses, such as xenon, kripton, radon. They build up in time and create internal pressure. The same happens inside the fuel kernels in the pebbles used in this reactor, but those kernels are specially built to withstand much higher internal pressures.
Here's a relevant quote from [2]:
As for the reprocessing part, I think you are jumping the gun. There is no reprocessing done in the US, at all, for any type of fuel. Even where reprocessing happens, as in France, the benefit is quite reduced. One can extract some plutonium and unburned uranium, but in the end that will allow you to extract maybe 10-20% more energy from the original amount of natural uranium. It will not make you extract one hundred, or 10 times, or even just twice as much energy. Reprocessing is simply not a game changer. It is not clear at all if it makes economic sense to build the highly complex facilities that do reprocessing, for the limited benefit.[1] https://aris.iaea.org/publications/SMR_catalogue_2024.pdf
[2] https://www.usnc.com/triso/
The problem is that if the pebbles aren't reprocessed, you now have to store their very large volume. The moderator, graphite, is integrated into them. This is unlike a LWR, where the moderator is water that the spent fuel can be simply lifted out of (after cooling). LWR fuel can be stored after a few years into dry casks; the equivalent for pebbles would be vastly larger and more expensive, and involves storing the spent fuel with its moderator, increasing concerns of criticality (although I imagine they'd be doused with borate or something to prevent that).
Thickness of a pressure vessel wall is proportional to pressure x linear dimensions of the pressure vessel, increasing at a given pressure as the vessel is scaled up.
I held their stock for a while until I realized they don’t exist to make a reactor. They exist to get funding.
Sounds crazy, but look through their actions. All press releases are just talk about what they will research and with whom they talked or made a “memorandum of understanding”. The CEO CV is also interesting since it lists a whole lot of board positions and titles but it’s not clear what he has actually done.
I picked NuScale because it's a PWR SMR, and is the only SMR design that was approved by the Nuclear Regulatory Commission. HTR-PM is an SMR, and comparing it with a full scale PWR reactor is not entirely fair, because reactors benefit from the square-cube law: the larger a reactor is the more efficient its neutron economy is, so you can extract more power per unit of volume.
As for NuScale being a scam, it would be probably the most elaborate scam in the history of financial scams. They were founded in 2007 and went public in 2022. People don't spend 15 years to run a scam, especially if this involves not one by two government agencies (the NRC and the SEC). There's a huge probability that such a scam would not work in the end, with an additional likelihood that you get relocated to a correctional facility (see Elizabeth Holmes).
NuScale integrates the steam generator in the "reactor" so the volume is larger, but this means using it is comparing apples and oranges. NuScale's design is also intended to use natural convection instead of forced circulation in accident conditions. This further reduces the power density allowed.
NuScale isn't a scam, but it appears to be founded on faulty principles (that the thing holding back nuclear power was safety concerns, rather than cost) so its business case doesn't appear to be working.
I wouldn’t call it a scam, but something that runs for 15 years with the involvement of 2 government agencies (read: large bureaucracies) is not unlikely to produce very little.
For example, see NASA’s space program, the European space program, and Boeing.
Otherwise Buffett wouldn't be so rich.
It's more.
> while NuScale has a volume of 101 m3 and yields 77 MWe
VVER1200 has the inner vessel _diameter_ of 4.2m, height of 11m for the internal volume of 153m^3, and 1200MWe capacity (so around 3GWt).
_THIS_ is what you're comparing it with.
> Quite the opposite happens: fuel elements in the current generation PWRs tend to swell, crack and burst
Nope. A swollen or a ruptured fuel rod in a regular reactor is a reason for SCRAM. The water inside the reactor vessel is constantly monitored for fission products. The individual fuel tablets swell, but they are contained inside zirconium rods.
No, we're looking at a core the size of a small residential tower. Probably around 30 meters in height.
Most nuclear waste is stuff near the reactor, not the fuel per se. And nuclear waste isn't the Armageddon stuff it's portrayed as in mass media. If this works, waste volume won't be an issue.
And now your containment has a much larger surface area.
> And nuclear waste isn't the Armageddon stuff it's portrayed as in mass media.
When it's contained. Uncontained waste has been the source of multiple tragedies.
> waste volume won't be an issue.
Process frequency will be.
In the way rolled steel has. We still roll steel, though, because despite the tragedies having narrative heft they're infrequent and small relative to the benefits of the product.
So far. Now you're introducing new technology. Are you comfortable prognosticating that it always will? Are there no lessons to be learned from previous tragedies, in particular, how _small_ radioactive sources are _way_ more dangerous due to peoples inability to identify them and tendency to pick them up?
You introduce medical radiation sources into a country without radiation controls and just a few years later you have an outbreak of tragedies. I'd rather not learn the lesson again the hard way.
Much more than accepting the deadly status quo!
> Are there no lessons to be learned from previous tragedies
Sure there are. Never take any risks ever isn't one of them.
> how _small_ radioactive sources are _way_ more dangerous due to peoples inability to identify them and tendency to pick them up?
You're describing MRIs more than nuclear power plants, which also produce lots of small nuclear waste.
This https://en.wikipedia.org/wiki/MRI_contrast_agent ?
Interestingly, the gadolinium in MRI contrast agents, while composed of stable isotopes, is a ferociously good thermal neutron absorber. The element is used in burnable absorbers in some reactors to keep reactivity level as the fuel is burned down.
One of Germany's PBRs had to filled with concrete after it was defueled, they couldn't decontaminate it enough to dismantle it.
If you just need 250MW of power, then just use electricity sourced from a regular PWR for heating. It'll be cheaper.
Waste heat can be useful for district heating systems because houses don’t need to reach high temperatures, but few designs give you access to even 300C and nothing currently hits 1,000+C.
Using helium indicates a problem with kinematic viscosity of cooling gases? Nitrogen would be non-reactive enough, I assume?
Nitrogen will undergo an (n-p) reaction to produce carbon-14 which has a half-life of 5700 years.
The front runner steel for use in fusion reactors, EUROFER-97, contains a necessary small amount of nitrogen. This is enough under some nations' rules to render it into intermediate level radioactive waste after use, due to the carbon-14 content.
Apparently, even a small natural niobium contamination would make it a low-grade waste.
https://scipub.euro-fusion.org/wp-content/uploads/eurofusion...
"[...] in the first layer, nearest the plasma, the rate of production of 94Nb – via neutron capture (n,γ) reactions on the stable 93Nb of niobium – is so high that Eurofer in this region is predicted to exceed the France-LLW limit within the first year of operation, and consequently would not be disposable as LLW under French regulations for more than 1000 years."
It's not just the steel. Beryllium typically contains about 100 ppm U, and an estimate of the cost of purifying it enough to avoid excessive fission products was another billion.
There are many uses for such steel already.
https://en.m.wikipedia.org/wiki/Low-background_steel
That's wrong, Hamm-Uentrop was a full scale commercial reactor. It did run in total for a week or so between 1985 and 1989 and was then shut down. The fundamental problem is, that the pebbles grind against each other, and being of the same material as pebbles they can grind each other down. (Now if you wonder why this wasn't discovered at the experimental reactor in Juellich, those guys just never mentioned that they lost fuel.)
[1] https://en.wikipedia.org/wiki/AVR_reactor
[2] https://en.wikipedia.org/wiki/THTR-300
This video explores an incident with a reactor of a similar design, and very rudimentarily explains the way pebbles and the helium gas is used.
There’s of course 2 flavors of HTGR (prismatic and pebble bed), and people choose the pebble version for continuous refueling despite all the drawbacks [1]. But there’s a lot of reasons to do prismatic. Can’t wait to see China’s prismatic HTGR.
[1] https://lvenneri.com/blog/pebble-bed-nukegumball
Did the pebble-bed reactor's "commercial-scale inherent safety" also pass a test with an compromised container, which would admit air that would cause the graphite to burn?
Rats, then it would cause about as many deaths per GWh as the coal it's replacing.
Not in the real world. We're deploying solar and wind as quickly as we can because they're the cheapest sources of power. The bottlenecks are production and permitting, and there is no indication either of those are seeing a step change in the next decade.
There is a reason America and Europe, where anti-nuclear sentiment runs ripe, are building gas power plants and infrastructure at the fastest clip in history. In the West, the choice is gas or nukes. In the China and India, it's coal or nukes.
World: Energy: https://ourworldindata.org/grapher/primary-energy-share-nucl...
Electricity: https://ourworldindata.org/grapher/electricity-fossil-renewa...
China, while being the most efficient nation at deploying nuclear, illustrates it perfectly: https://ourworldindata.org/grapher/primary-energy-share-nucl...
https://ourworldindata.org/grapher/electricity-fossil-renewa...
India: https://ourworldindata.org/grapher/primary-energy-share-nucl...
https://ourworldindata.org/grapher/electricity-fossil-renewa...
Europe: https://ourworldindata.org/grapher/primary-energy-share-nucl...
https://ourworldindata.org/grapher/electricity-fossil-renewa...
> Nuclear is plateauing, and renewables are booming
Nobody said otherwise. I literally said "we're deploying solar and wind as quickly as we can because they're the cheapest sources of power." You're arguing against a straw man.
Renewables are booming. So is gas [1]. Your analysis fails on two counts. One, it ignores the substitution effect [2]. If we're talking about grid stability, et cetera, this is fine. If we're looking at emissions, it's not.
Two, you're lumping together fossil fuels. That masks the fact that we've added about as much natural gas capacity as solar. The growth rates are different. But so are the base levels.
I'll say it again: we're building renewables as fast as we can. We can't build them substantially faster. That means there is never a choice between renewables and something else; it's always renewables by default. Where there is a choice, therefore, it must be between the other options.
The convenient lie the gas industry has sold the nuclear nervous is that it's a competition between solar and wind and fossil fuels. It's not. We're investing trillions of dollars in gas infrastructure with 20 to 40-year investment theses despite renewables booming because power demand is booming too, and the difference has to be made up somehow. That gas infrastructure's thesis only works if we exclude nuclear energy. (It also precludes us reaching our 2030 and 2050 emissions commitments, but nobody seems to care about those anyway.)
We're building gas instead of nuclear in the West for the same reason China and India are building coal plants: it's cheaper than nuclear. Nobody is acting on emissions.
[1] https://en.wikipedia.org/wiki/Primary_energy#/media/File:Glo...
[2] https://ourworldindata.org/energy-substitution-method
China also installed more than 4x PV than coal on a rated power basis in 2023. I don't know the capacity factor of their coal -- I hear it's operated at fairly low capacity factor -- so PV installs could well exceed coal installs (especially net coal installs) on a levelized basis there.
https://ourworldindata.org/grapher/carbon-intensity-electric...
https://ourworldindata.org/grapher/electricity-generation?ta...
You first criticize me for merely providing data, then for my "analysis". Go figure. Your points (substitution effect and methane) are pertinent, but my answer wasn't about this but about the fact that there is no sign of abatement of renewables' growth.
> We can't build them substantially faster
One more: nothing, especially historical data, sustains this.
We burn less and less fossil fuel thanks to renewables, which more and more replace it, and there is no clear indicator of this trend to change. This is too slow, granted, however there is no magic wand and demanding an immediate perfect solution is playing upon the nirvana fallacy ( https://en.wikipedia.org/wiki/Nirvana_fallacy ).
> We're building gas instead of nuclear
... because since the 2000's:
- industrial renewables appeared and are (very quickly) more and more adequate. I other words to replace fossil fuels nuclear isn't the only contender anymore. Thanks to renewables: no risk of any major accident propagating very dangerous long-term stuff difficult to recover, no hot waste, no dependency towards any provider of fuel (such as uranium), no weapons proliferation risk, quick and easy deployment, no decommission-related nightmare (see the ongoing case in the U.K.)... No wonder renewables are booming.
- all nuclear-building projects are more-or-less resounding failures
Summary: As the temperature rises, neutron absorption increases, reducing fission and thus temperature.
Negative fuel temperature coefficient is not an unusual feature.
The real question is whether the heat removal system of the reactor as a whole is sufficient to remove the decay heat to keep the fuel within the limits.
That's the point of the Westinghouse AP1000; the containment (steel liner) and protection from the outside world (concrete wall) are separated, allowing the liner to cool by convection and water dripping from above. Admittedly you need to top up the water tank at the top, but that is less of a task than trying to push water into the containment.
NuScale's reactor was originally motivated by the desire to make it safer by using natural convection. But it ends up requiring 1/3rd more labor hours to build a NPP using their reactors than it does to build a conventional large reactor power plant.
Because the real problem with solid rods is that they ... are solid rods, and if they start "overreacting" you can't split up the rods, unlike a pile of pebbles/spheres.
Modern PWRs also have this safety feature, if a core melts down, the molten mass will be contained in a core catcher. Where it'll be mixed with inert material that can provide enough surface area and thermal mass to prevent further fuel mass migration.
The biggest problem in the core catcher design was to make sure that the molten fuel lava spreads out enough for the passive cooling to stop it from melting through concrete.
Pebble bed reactors will have a similar problem. You can "drain" pebble beds somewhere, but then you need to make sure that this "somewhere" can conduct away the decay heat without melting.
You're saying LFTRs can't have a pool of cooling liquid that the overheating fuel liquid can't mix with and cool down even faster than solid rods?
So how could LFTR safety plug and cooling pool be a fairy tale?
> can't have a pool of cooling liquid
What "liquid"? Water?
Do you realize what's going to happen if molten salt drops into water? First, there's going to be a steam explosion that will atomize the fuel and spread it through the whole containment building (because the water vapor can't be contained in a reasonable volume).
Then the water will boil away from decay heat, and the fuel lava will continue chewing through your reactor building.
That's why LFTR reactors with "melt plugs" will essentially use the same approach as PWRs: spread the molten fuel across sacrificial concrete cladding.
But fundamentally since the nuclear fuel is a liquid, the cooling pool is very wide and shallow, separating the fuel well past sustained nuclear fission. This is the problem with solid rods, they go runaway, the fuel is a solid rod, you can't separate apart the fuel without shoving some moderator into it. Yeah I don't know the viscosity of Uranium tetrafluoride, maybe that's a problem, but I doubt it.
Why does the plug only melt past some point of no return? The plug can melt at whatever temperature point is desired. A "meltdown" can just be part of the usual fuel flow and recirculation.
The difference is that the liquid spread because, you know it is a liquid. In liquid form? Rather than a solid form or some semi-solid uranium lava. See the difference? What am I missing here?
In the dump pool, it will cool into solid salts. Then you just need to reheat it to recover the fuel, and pass it back through the salt reprocessing systems.
Plus I thought LFTRs have some mechanism for self-moderation by expansion of the fluid when it gets hot, separating the fuel apart and reducing neutron economy.
You seem stuck in the limitations of solid fuel rods, solid fuel reactors, and their inherent inconvenience. Yeah, the liquid is 650 degrees or something like that, but it's still liquid and you can do things with liquids that you can't do with solid rods
Well, I definitely don't know which reaction can absorb on the order of 2GWh of residual decay heat within the first 2 days.
I'm assuming a 3GWt reactor, something that at least can be competitive with PWRs.
To give you some perspective, this amount of energy is enough to vaporize more than 3000 tons of water. More than an Olympic swimming pool.
> Maybe you just have the dump pool be a bunch of molten thorium or even solidfied thorium salt and that also plummets the neutron economy as the neutrons get absorbed by thorium as the hot uranium salt melts the thorium salts. Maybe you keep the thorium salts liquid.
Sigh. It's not the fission that is a problem. Fission will be quenched by all the neutron poisons. Even in Chernobyl or Fukushima the fission stopped immediately after the accident.
It's the decay heat that has to be conducted away.
> Why does the plug only melt past some point of no return? The plug can melt at whatever temperature point is desired. A "meltdown" can just be part of the usual fuel flow and recirculation.
If normal recirculation works, then there's no problem with supplying cooling water. The touted advantage of molten salt reactors is their passive safety, they are supposed to fail safe even if EVERYTHING fails.
> You seem stuck in the limitations of solid fuel rods, solid fuel reactors, and their inherent inconvenience. Yeah, the liquid is 650 degrees or something like that, but it's still liquid and you can do things with liquids that you can't do with solid rods
Yeah, because I actually worked in the nuclear power industry.
The sales pitch for salt-cooled reactors is the lack of any coolant that would become pressurized hot gas in an accident. Heat can stay in salt or in other low vapor pressure materials.
The problem with LWRs is the water goes to steam in accidents, and this steam must be contained. This drives the size of the containment building, and the containment building is costly.
An alternative for LWRs would be to filter and vent the steam instead of trying to contain it. This would allow small quantities of radioactivity to escape (including all the noble gas fission products), but the filtering can actually be quite good, reducing emissions by many orders of magnitude. Second generation filtered containment venting systems can filter iodine as well as cesium and strontium. If Fukushima had had such systems the impact would have been far lower.
https://www.nrc.gov/docs/ML1224/ML12248A021.pdf
Sodium-cooled reactors and the upcoming lead-cooled reactor also have this property. It turns out to not be such a huge advantage, we have plenty of experience working with pressurized water.
> The problem with LWRs is the water goes to steam in accidents, and this steam must be contained. This drives the size of the containment building, and the containment building is costly.
No, it's really not a problem. The loop doesn't suddenly loose compression if something bad happens. If there's electric power, there's more than enough time to slowly cool down the reactor.
And a containment building (that also protects against external threats like an airplane ramming into the reactor) has more than enough volume if the primary loop is de-pressurized and the water flashes into steam.
> An alternative for LWRs would be to filter and vent the steam instead of trying to contain it.
The water in the primary loop is clean. It's constantly purified by filtration through ion exchange resins. Once the activated oxygen decays (in ~1 hour) you can swim in it (although I wouldn't drink it).
PWRs (actually, all thermal power plants) have areas where steam can be dumped. If you watched "Chernobyl" series, the ridiculous scene with divers was supposed to happen inside such an area ("barboter pool").
Modern PWRs are also designed to do that safely. There's plenty of capacity to condense all the water from the primary loop after the loss-of-cooling. Of course, after that the fuel will melt down, and chew through the reactor vessel.
The filtering system you linked is not strictly necessary for modern PWR designs. They will still be safe in case of an accident with total loss of cooling, but the containment building will be hopelessly contaminated internally. This filtering system can allow the steam to be vented into the atmosphere, perhaps giving more time to fix the emergency cooling systems.
It can in design basis accidents, for example a complete break of a main circulation pipe leading the loss of coolant (LOCA) into the containment. The emergency cooling system would then operate by spraying water into the core that would evaporate into steam that would go right out of the reactor vessel. The containment has to be sized for such an accident.
As an example of such an accident, consider what would have happened at Davis-Besse had the erosion of the lid of the reactor vessel progressed to an actual perforation. As it was, the steel was removed in an area down to the inner stainless steel liner, a liner that was never intended to be load bearing against the internal pressure.
> And a containment building (that also protects against external threats like an airplane ramming into the reactor) has more than enough volume if the primary loop is de-pressurized and the water flashes into steam.
Right, it does. That's why it's so big and expensive, with so much internal volume. If it didn't have to, it could be made much smaller. The airplane requirement doesn't change this; it's easier to make a smaller containment building resistant to aircraft impact than a larger one.
> The water in the primary loop is clean. It's constantly purified by filtration through ion exchange resins. Once the activated oxygen decays (in ~1 hour) you can swim in it (although I wouldn't drink it).
That's true in normal operation, where you might have some small number of fuel rods with cracks or perforations (but even that is getting pretty uncommon these days). It would not be true in a design basis accident, where some or all of the fuel may have partially or completely melted, and where cladding will have been compromised by high temperature reaction with steam. The design must assume essentially all the volatile fission products have gone into the water. At TMI, fission products carried in the water (and also noble gases) raised radiation levels in the containment building to 800 rem/h during the accident.
> The filtering system you linked is not strictly necessary for modern PWR designs.
I offered up the possibility that such systems could replace the large volume containment of modern systems (or at least reduce its size and cost). Sure, they're not obviously necessary if you have a large volume containment already (although some countries ended up requiring them anyway since some accident scenarios do involve venting, as happened at Fukushima, which admittedly had pre-modern designs.)
> The pebble-bed reactor is designed so that this effect is relatively strong, inherent to the design, and does not depend on moving parts. This negative feedback creates passive control of the reaction process.
From the Wikipedia article on the PBR that was linked in a neighbouring comment.
It should, at least if built correctly, be impossible for it to get into a run-away state. Even if it lost all power, and everyone walked away, it should not melt down.
[1] https://en.wikipedia.org/wiki/Pebble-bed_reactor#Safety
Amazing. Well done! How far this country has come in the last few decades is nothing short of breathtaking.