Hacker News Reader

Comment by instagraham | Hacker News Reader

"Most designs will require fuel with a higher concentration of uranium-235 that’s not currently used in today’s reactors"

The problem with making such tech mainstream is you'd need impossible levels of security to prevent even one store of such material from being stolen.

View on HN

Some of the designs are not made to be refueled. In those the fuel gets loaded in at the factory and sealed up. Though in that case they are loaded up with quite a large amount of fuel (like a couple tons of plutonium for a ~30y lifespan)

In such designs you really would only need high security at the factory as stealing fuel from a reactor that is not meant to be opened (you would have to cut into it) and that is on without someone noticing is really really hard.

edit: Would also like to add that such designs are not actively being pursued anymore. The tradeoffs just are not worth it. We know how to securely transport fuel around so that is not that big of an issue really.

That doesn't really reduce security requirements for the reactor itself. It just means you don't have to secure additional fuel in transit.

I'd imagine that a big concrete bunker in the ground effectively burying this thing would probably be pretty good security. Taking a look at how nuclear missiles are in silos and having a variant of that approach. Not saying it's the best approach, just that it's should be fairly solvable.

Then again, it is nuclear that we're talking about, so.....

How exactly are you going to steal the fuel from inside the reactor if it is running?

Only option is to cut into the reactor (and thus destroy it). And now you have just cut a hole into a nuclear reactor that is ON and killed yourself (or triggered a whole bunch of safety features stopping the nuclear reaction and maybe not kill the attacker but for sure notified the authorities)

Basically such a reactor is started once installed and never turned off until end of its lifetime.

> And now you have just cut a hole into a nuclear reactor that is ON and killed yourself (or triggered a whole bunch of safety features stopping the nuclear reaction and maybe not kill the attacker but for sure notified the authorities)

So somebody can strap a bunch of explosives to it and cause a good bit of quite radioactive material to spill (I don't mean a chain reaction, just the normal radioactive waste). You don't need to steal it if your goal is to have a dirty bomb.

That's certainly a concern - but if we're stipulating people with huge amounts of explosives and ill-will, they can probably also do a lot of damage without the nuclear reactor anyway.

Why huge amounts? These reactors are supposed to fit on a semi truck (see e.g. the articles "infographic").

I assume that containment is going to be serious business on these things even without factoring in attacks.

The spill is contained within the containment facility. I guess they could use enough explosives to also breach the containment facility. But if they're capable of that, they probably have the means to do even more damage by targeting a skyscraper. Multiple orders of magnitude more people died in 9/11 than in Fukushima.

We're talking microreactors here? There's no huge containment facility. They're not all that large.

Right there's no huge containment facility, but there's still a concrete dome over the reactor isn't there?

I'm still not seeing how our hypothetical terrorist A-team inflicts more damage by attacking a microreactor versus attacking a populated area directly.

> Right there's no huge containment facility, but there's still a concrete dome over the reactor isn't there?

I'm pretty sure these aren't intended to be used with a concrete dome or anything as substantial as that. Look at the pictures in the article, and some of the referenced benefits:

Can be used for emergency response to help restore power to areas hit by natural disasters

- "Can be quickly removed from sites and exchanged for new ones"

- "Can be used for emergency response to help restore power to areas hit by natural disasters"

That doesn't really mesh with needing to build a concrete dome on site.

Or it can be installed underground, as the same website has proposed in other articles on small reactors: https://www.energy.gov/ne/articles/5-key-resilient-features-... It's a lot easier to dig a pit than build a giant 5 meter thick dome.

Regardless, the point remains: the question that needs to be asked isn't "are these sites vulnerable to attack" it's "can an attacker inflict more damage here than through conventional means?"

You can have tamper sensors and a great deal of tamper resistance. Add canary pinging of a remote service and alerting and you can have an armed response team on site within minutes of any attempt to hack the thing.

>a couple tons of plutonium

Is that a typo? tons?

Yes if it is supposed to run for 30 years.

See the "Reactors with Lifetime Cores" section of

https://www.nrc.gov/reading-rm/doc-collections/commission/sl...

Thanks, I had no idea they consumed that much.

The ref. is for 200MWe so I'd guess that for 1-5MW it would be less but no idea if it scales linearly.

Also, could you load all the fuel at once and leave it running? Or would you have to refuel every n years?

> Also, could you load all the fuel at once and leave it running?

Some designs call for exactly this. After all it's heavy, so the volume isn't much.

Keep in mind that one cubic feet of plutonium weighs ~1200 pounds.

A basketball made of solid plutonium would weigh 300 pounds or so.

Two (metric) tonnes of plutonium takes up 100 liters of space, or about 3.5 cubic feet.

Probably not really. It's much less than normal reactors - e.g. the infamous Chernobyl 4th block had 217 tons of uranium fuel in it, so a couple tons = two orders of magnitude less.

I once heard that stealing nuclear material from a reactor design like this was akin to stealing a lit bar-b-bq grill by hand. Not impossible but not very easy or practical.

If you mean it would be a suicide mission, those happen with some regularity.

Except that you need to survive in order to successfully steal the fuel. And you won't survive. QED

Not necessarily if you're part of a team.

That is dark.

It is, and it's standard practice. Suicide bombers are almost always part of a team that provides them with instructions and know-how, sometimes equipment and money, and first and foremost with a goal to die for. Also, lots of these attacks are part of a greater political scheme, and I would assume pretty much all attackers are aware of that. If your sacrifice means your side gets access to actual freaking atomic bombs, I mean, I could see that be pretty enticing.

And if not there's always the option to make people do lethal things against their will. People have been sent to their deaths by the millions over the course of history, that's totally doable.

IMO, dotting the landscape with such reactors is just asking for some state or other to take them and use them to leapfrog a lot of the effort required to build nuclear bombs.

Sure, but maybe your dirt poor family will get a million dollars and/or rich country passports.

Or less if you don't have long to live anyway.

Those are "carrot" versions. There are also "stick" ones.

You do realize the lengths states like Iran are going to enrich enough Uranium for a few simple bombs? If this works as a shortcut, it doesn't have to be easy or practical. And besides, it might be a lot easier if you don't care whether the people cracking the thing open will survive the ordeal, or whether radioactivity gets released to the environment.

More important is that you have to do it before a response can be deployed.

If you treat these things like nuclear weapons they won't be very useful. Securing those bombs is enormously costly. You can't do the same for thousands or more of such reactors deployed for decades, all over the world, under civilian control.

Given that even the US Military doesn't have infinite resources, how can you guarantee there will be a fast response, or any response at all, or even that anyone will notice 20 years after deployment if someone with the resources of a nation state tries to get at that Pu? There might be wars where that reactor is deployed, there might be natural disasters, people might switch to 100% renewables and lose interest in the reactor, ...

>> You can't do the same for thousands or more of such reactors deployed for decades, all over the world, under civilian control.

So what? Put them under military control, then.

Hooraa!

NWO, there you go!

My thought is, 30 years isn't enough. In 30 years you need to open the reactor and swap out the fuel, which is too tricky an operation (securing the fuel going in & out) to do cheaply and at scale. 30 years is within the lifetime of most people, so people are hesitant to cause a problem that they'll have to deal with. If these things can be buried in lead & concrete for 60+ years then maybe it'll happen.

If it pays for itself and then some then in 30 years you recycle the used reactors and replace them with new ones.

Recycling would mean sending the spent reactor back to the manufacturer who would have the tooling and skilled labor needed to safely open the reactor and extract the spent fuel (and then dispose of it safely).

Or with thorium breeding, fuel input is not a big deal, since thorium is pretty much not dangerous.

The micro-reactor designs I looked up generally fell into one of three categories:

1. Very high enrichment fuel because they need fast load-following, e.g. for naval use.

2. Very high enrichment fuel because they're for use in space, and lower mass is better.

3. Reactors for ordinary power use, with fuel that's enriched more than most, but still nowhere near enough to be practically useful for weapons.

An example of #3 is the Westinghouse eVinci micro-reactor, which uses fuel enriched to a little under 20% U-235. Weapons-grade uranium is closer to 90%.

People often confuse the difference in enrichment needed for weapons and reactors. Typically a reactor is 3% where a weapon is 97%. It becomes significantly harder to create that much enriched uranium which is why experts call Iran's production a bluff. Past their agreed levels, but nowhere near enough for weapons. The whole Iran deal relies on this fact actually.

Could make a dirty bomb still, but the risk from those is more psychological than real, which is probably why no one has ever tried to do it.

s/no one/very few/

https://time.com/4728293/uranium-underworld-dark-secrets-dir...

What is the threat model for fuel theft? Weaponizing uranium requires enrichment well above even what maritime reactors use. Note that the difference between 90% enrichment and 99.9% enrichment isn't a factor of 1.1 difference, it's a two-order of magnitude difference - the remaining impurities need to be reduced by a factor of 100, each of those extra nines is another order of magnitude. And even if fuel does get enriched, it requires extensive nuclear physics research and engineering effort to produce a fission device. The only groups proven to be capable of this are state actors, which can set up their own enrichment facilities and have no real reason to steal uranium.

An alternative threat model is not producing a bomb, but dumping radioactive waste in a populated area. While damaging, radiation is a relatively slow killer save for the most intense doses. The amount of damage that it would be able to inflict is minimal relative to the amount of effort requires to break into a nuclear facility and exfiltrate fuel. The question that should be asked is, does this present a larger potential for harm as compared to the same amount of effort put into conventional means of destruction (bombing buildings, shooting up crowded areas)?

Enriching HEU to weapons grade is trivial by comparison to enriching natural uranium to even just 20%. Really, the difference in effort needed is enormous. Yeah, you still need centrifuges, but you need many fewer centrifuges and many fewer passes.

Also, you can make a bomb with uranium enriched to much less than weapons grade -- you'll just need a lot more of it compared to weapons grade uranium.

> Enriching 90% uranium to weapons grade is trivial by comparison to enriching natural uranium to even just 20%

Naturally occurring uranium is 99% U238, and 1% U235 (and some U234, but that's a trivial amount). 20% enrichment means that 80% of the fuel is U238. You only need to take out 1/5th of the U238. In order to weaponize uranium, U238 concentration needs to be reduced to 10% or below. You need to take out about 90% of the U238. Sure, if you're staring with 20% enriched uranium you'll need a factor of 20 less input material. But that's not very significant, uranium isn't particularly expensive. The factor that's limited by centrifuge capability is the final enrichment purity. That's why one of the main things nuclear weapons inspectors look at are centrifuges.

> Also, you can make a bomb with uranium enriched to much less than weapons grade -- you'll just need a lot more of it compared to 99% enriched uranium.

No, you can't. The presence of too much U238 prohibits a runaway nuclear reaction regardless of the amount of uranium. You can get fission from lower enrichment, but not runaway fission. It's like flare versus a firecracker. And that's a desirable trait in electric power generation. The main principle of weapons inspection is to ensure that a country's enrichment facilities can only enrich uranium to a purity sufficient for power generation, but below weapons applications. The height and diameter of the centrifuges are the main factors here.

Reply on news.ycombinator.com