Long-term robustness of Soviet and Russian nuclear armaments

[cheater]

Hi all,
first of all, I do not wish to make a political topic here. This is purely a technical question that I hope people will answer based on facts they can present.

I was wondering recently how probable it is that Russian nuclear ICBM warheads are in fact functional. Assumedly like in any arsenal some percentage will be duds, but the question is what percentage. I soon figured that I don't know anything at all about those systems:

- What sort of technology is used for the warhead and detonation device? What sort of assembly? I don't even know details like is it vacuum tube, magnetic, or mechanic; and are they using wires or boards; are they soldering, crimping, wire wrapping, brazing, using screw terminals, springs, ... Are there any documents on what such a device looks like in Russia's armament? What were the general manufacturing techniques used by the Soviet union in those days? I honestly don't even know during what years those things were all produced.

- What sort of maintenance is necessary for the items to continue to function - and what is the manpower needed to do that? Are there any reports of the maintenance being especially good or bad?

- I understand that nuclear tests of any sort of size can be sensed from anywhere in the world. And of course you don't know if a weapon works until you fire it. When's the last time such a test has been carried out by the Soviet or Russian Government?

- What sort of knowledge is necessary to continue maintaining the items? Is it conceivable that very few people really know how to confidently work on those things in the first place?

- Other than the explosives themselves, what other systems are necessary for a successful delivery of such a payload? Does the ground site have to be special in some way, does it have to have special equipment? How difficult to operate or maintain could this equipment be?

I would ask that everyone keep politics out of this thread. Of course we know what's going on and everyone is grown up and knows who's right or wrong. However if stuff like that happens this thread will be derailed and there will be no knowledge to be gained from it. We all know we can go to Twitter or Facebook to talk about politics.

Please avoid reductionist arguments, or bad faith arguments. Stuff like "Oh it's Russia with their kleptocracy, of course things don't work" just doesn't cut it. Saying "I saw Russian trucks weren't turned so the tires on them broke" doesn't cut it as well. Keep it to what you specifically know either of the devices themselves or of Russian manufacturing industry from the time the devices were built.

Please avoid arguments such as "even one working is bad enough". We know it's bad. We're trying to learn here, not make geopolitical decisions. I am not the head of NATO, I'm just a guy who doesn't know how a type of device works, and just like a toaster or a VCR, I'd like to understand what is involved in keeping it in function.

[xrunner]

Good question I was wondering the same thing the other day. I guess because of the state of affairs. But it must be something akin to say having a FM radio and making sure it would actually make sound from the speaker without turning the audio stage on. You can make visual tests and applying power to certain parts of the system. But in the end if all the tests you can do (without actually firing it off) pass, the you assume it has to work, because it can't do anything else but work, if it passes all the tests up to a certain point. According to the laws of physics it would have no other outcome other than than to blow up ... 

Of course all that presumes it is delivered properly by some other system that might fail ... Yea so a certain % of the systems are not even going to make it to the target anyway. Who knows how many?

[cheater]

Who knows how many?

Right, that's the point! I'm sure there are nuclear engineers or physicists who are able to answer this question better than you or I.

[exmadscientist]

So I know just enough about this stuff to be dangerous... may as well have a go. (Disclaimer, DoE has in the past paid my salary, but not NNSA... I actively preferred not to work for them.)

- What sort of technology is used for the warhead and detonation
device? What sort of assembly? I don't even know details like is it vacuum tube, magnetic, or mechanic; and are they using wires or boards; are they soldering, crimping, wire wrapping, brazing, using screw terminals, springs, ... Are there any documents on what such a device looks like in Russia's armament? What were the general manufacturing techniques used by the Soviet union in those days? I honestly don't even know during what years those things were all produced.

Initiator is probably plastic explosives. I don't know the mechanical construction or how they age over time. This is the most sensitive bit, at least on a fission weapon. Fusion adds complications of its own. The fusion physics packages are remarkable objects. They definitely go bad over time; I think at least one component of US designs is volatile. But the initiators are even worse, so they're still a major bottleneck.

- I understand that nuclear tests of any sort of size can be sensed from anywhere in the world. And of course you don't know if a weapon works until you fire it. When's the last time such a test has been carried out by the Soviet or Russian Government?

There are test ban treaties in place. The major powers have actually respected these, so it's been a while. Those treaties were a major driver of supercomputer development by the US DoE.

- What sort of knowledge is necessary to continue maintaining the items? Is it conceivable that very few people really know how to confidently work on those things in the first place?

The physics packages are VERY specialized. (DoE has in the past lost information on how to work on one fusion design. That was a problem.) The rest of it, not so much.

- Other than the explosives themselves, what other systems are necessary for a successful delivery of such a payload? Does the ground site have to be special in some way, does it have to have special equipment? How difficult to operate or maintain could this equipment be?

The rest of it is not special. Obviously an ICBM has its own skills needed, but the physics package doesn't make it any easier or harder to build one of those.

In the past, Soviet design tended to favor making bigger versions of crude, inefficient, reliable designs. So that suggests most warheads and delivery systems are probably quite functional. But the fanciest warheads will be the most complex and most subject to needing maintenance or periodic refresh. I imagine they have done these things, or why else would Mayak be so busy? It isn't every day you get to botch an order for a neutrino experiment source, and they have to stay busy....

[Cerebus]

Take a look at Carey Sublette's Nuclear Weapon Archive probably the largest single repository of non-secret information on nuclear weapons design and deployment.

Probably best to start at the FAQs (http://nuclearweaponarchive.org/Nwfaq/Nfaq0.html) and work outwards.

[cheater]

@exmadscientist thanks, what do you make of this info? (Found on the page kindly provided by Cerebus)

http://nuclearweaponarchive.org/Russia/Sovwarhead.html

There isn't much to go on...

But let's talk about the "usual" stuff. Even a perfectly working radio, if you just leave it sitting for 50-70 years, it'll stop working due to progressing tarnish/corrosion, chemical incompatibilities, etc. Probably way worse if it's next to a radiation source for those 70 years... I heard even metal changes its properties. In what ways could such old electronics be affected?

[bdunham7]

I doubt anyone could give you a good estimate of the percentage of decades-old Soviet nuclear warheads that are going to function, but I'd guess from what public analysis there is that any problems in delivering a nuclear detonation are more likely to be due to some other cause.  You actually can test a nuke without violating any treaties--you just remove the plutonium pit and install a bunch of sensors and cameras and see how the timers and explosives perform.  A uranium gun-fired design would not be as easily tested, but those are considered very reliable.  A multi-stage device also could only be tested for the operation of the primary compression explosives, but there probably isn't much need.  As far as I know (which probably isn't much) everything beyond the initial stage is pretty stable except of course for the tritium, which can be measured and monitored separately.

[exmadscientist]

But let's talk about the "usual" stuff. Even a perfectly working radio, if you just leave it sitting for 50-70 years, it'll stop working due to progressing tarnish/corrosion, chemical incompatibilities, etc. Probably way worse if it's next to a radiation source for those 70 years... I heard even metal changes its properties. In what ways could such old electronics be affected?

These things aren't that radioactive. Not pleasant, sure, but not so bad. And they're not 70 years old. Perhaps the Russians still have those in service (I hear they don't throw things away?) but the antiques are not the ones to worry about. Also, there shouldn't be any electronics in the physics package itself. Nearby, to time things, but not inside it. Warheads don't need much smarts to operate. They are not fly-by-wire devices!

I doubt anyone could give you a good estimate of the percentage of decades-old Soviet nuclear warheads that are going to function, but I'd guess from what public analysis there is that any problems in delivering a nuclear detonation are more likely to be due to some other cause.

Agreed.

You actually can test a nuke without violating any treaties--you just remove the plutonium pit and install a bunch of sensors and cameras and see how the timers and explosives perform.

Fission weapons, yes. Fusion weapons, no. Those things have secondaries initiated by radiation pressure. This is the sort of thing you'd like to test out, rather than trust your simulator on, but that's not allowed anymore. The RIPPLE design is an example of this. It supposedly yields a lot more efficiently than a Teller-Ulam device, but could not be made reliable before tests were banned. Is it in the field now? Who knows. Certainly not me.

As far as I know (which probably isn't much) everything beyond the initial stage is pretty stable except of course for the tritium, which can be measured and monitored separately.

So the US design I mentioned being "lost" involved a very special material codenamed FOGBANK. The methods of production were lost, partly due to classification mistakes, and so they had to figure out what that stuff was and how to make more, or the warheads can't go splodey anymore. It is now generally believed that FOGBANK was a lithium deuteride aerogel that relied on acetonitrile impurities to actually be manufacturable. I submit that lithium deuteride aerogel is pretty bloody weird stuff, and if you've got to maintain the likes of that, you had best be alert for anything!

[AndrewBCN]

Just have joesmith test them...

[EEVblog]

I was wondering recently how probable it is that Russian nuclear ICBM warheads are in fact functional. Assumedly like in any arsenal some percentage will be duds, but the question is what percentage.

100% chance enough are functional to be a real threat.
There are stories coming out about how they didn't have the budget and/or systems to maintain the trucks which have failed in the field, but whether that's the same for the nukes no one knows.
Still, 100% chance enough are functional.

[AndrewBCN]

...
I am not the head of NATO, I'm just a guy who doesn't know how a type of device works...

How admirably modest of you! 

[Berni]

They are likely built to pretty high quality standards so they do have a good chance of being pretty reliable.

Failures tend to be more statistical rather than just black and white. If they make enough of these, it is inevitable that eventually one would end up faulty even when brand new. Then there might be some that do work but die early due to a manufacturing fault, then way on there might be some that die of old age once the slower deteriorating factors kick in. But even those factors are not uniform as they might have been stored in different conditions. You might also not get full failures, it is possible for it to be good enough to go off, but produce a much worse yield. Even one that completely fails to start the reaction is not a nice thing to deal with, it still gets blown to bits by the initiating explosive, spewing radioactive stuff around the area.

But that being said even if 3/4 of them have become duds by now, this still leaves a lot of working ones due to the sheer number of them.

[EEVblog]

They are likely built to pretty high quality standards so they do have a good chance of being pretty reliable.

It also depends on the type.
If we are talking short/medium/long range ICBM missles, either land or sub based, then you have the added layers of reliability of the missle itself and the missle launch systems etc.
But there are simpler nukes like the 2S7M based battlefield nuke artilery, or the old school gravity bombs which you could be pretty confident are still almost 100% reliable and could be carried and dropped by any capable warplane.
Like I said, 100% guaranteed capability in some form.

[T3sl4co1l]

Like I said, 100% guaranteed capability in some form.

Not only that, but the western arsenal is likely significantly above 0% capability.  Even if the first strike is all duds, our retaliation devastates not just untold miles of noncombatants and wilderness, but carries fallout over much of the entire hemisphere.  You can't win a nuclear battle; at best, you survive it.

Tim

[cheater]

But that being said even if 3/4 of them have become duds by now, this still leaves a lot of working ones due to the sheer number of them.

What is the sheer number?

[Berni]

But that being said even if 3/4 of them have become duds by now, this still leaves a lot of working ones due to the sheer number of them.

What is the sheer number?

Not many people are privileged to know the actual numbers of that, but they are guessing it at being around 6000 for Russia and same for the US
https://www.armscontrol.org/factsheets/Nuclearweaponswhohaswhat

Safe to say that only a small percentage of those is already enough to blow up entire nations. Hence why it is still bad even if a lot of those don't actually work.

[daqq]

If you want to see how military electronics of their time were built, take a look at these channels:

https://www.youtube.com/user/msylvain59/videos
https://www.youtube.com/channel/UC4Kz0sEXSqNkfi7W1PBKyMg/videos

Mind you, this tells you nothing about the way that the current big stuff is manufactured and on the level of build quality.

As to how much of the arsenal is fully functional and would work as intended, provided it's even 1% it's still a problem.

[Haenk]

To my experience, most russian electronics is very simple and robust, but outdated by at least a decade.
So I'd assume most of the electronics will just work, even after decades, unless they have completely abandoned maintenance.

[SilverSolder]

During WWII,  a hundred year old fortification (manned by raw recruits and pensioners), armed with a handful of 40- to 50-year-old torpedoes designed in 1866 of Austro-Hungarian manufacture, destroyed a heavy cruiser so new, its crew was still finishing training...   the last known operational use of a Whitehead torpedo.

https://en.wikipedia.org/wiki/Battle_of_Dr%C3%B8bak_Sound

https://en.wikipedia.org/wiki/Whitehead_torpedo


The moral of the story is that discounting old weaponry can be an expensive mistake...

[cheater]

Like I said, 100% guaranteed capability in some form.

First off I'm guessing you can completely exclude airplane based gravity devices completely. Here's what I understood reading up on nuclear defense initiatives and SDI (Strategic Defense Initiative) a good 10-15 years ago: once nuclear strike becomes a threat, anything flying in your direction gets shot down immediately from land, no questions asked. In this scenario, NATO will have moved their troops onto Ukraine and every country bordering with Russia or Belarus to minimize response time. So that really only leaves ICBMs and (for strikes on Europe or Asia) very fast cruise missiles. However a military expert will have actual competence on the topic which I don't have.

Regarding the remaining items. How do we know they wouldn't all suffer a mechanical failure before then? Say, using screw terminals, and they didn't use thread locker? What sort of construction methods were being used in missiles and warheads during those times? They certainly did it in /some/ way.

Here's the thing though. It's by now well know that the Russian organized crime, allegedly led by the government itself, has been extorting the Russian military, and that includes the RVSN, the Strategic Rocket Forces of the Russian Federation, the separate branch of Russian military that manages the ICBMs. Look at the second headline here:

https://mobile.twitter.com/kamilkazani/status/1502682176848027658

You can see they're talking about the "РВСН", which is cyrillic for RVSN. The full headline says that personnel of the RVSN is being extorted by the Russian mafia and they have to pay them tribute, which is a recurrent payment.

https://en.wikipedia.org/wiki/Strategic_Rocket_Forces

Extortion money comes from the budget... the same budget necessary to maintain the missiles and warheads. So let's think about it, you're a Russian official, you are being extorted, where will you embezzle the money from? Probably not pay some of the conscripts, and also probably some general maintenance across the board. Some things need to work - conventional missiles that you'll be using in Chechenya or Georgia need to work because they're going to be fired. However you've got those nuclear payloads and they haven't been used in decades. There's no nuclear war on the horizon. That looks like a pretty good target for embezzlement. No one's checking, no one will find out, and if they do they try, that's an end of the world scenario and you're probably dead anyways, so you don't care.

So let's think about it together, how much money is required to keep a nuke alive and working? And would they be using their own delivery methods not interchangeable with normal ICBMs and cruise missiles? My guess is yes, but I can't know. If yes, then what's necessary to keep those delivery methods alive?

Agreed, no one can give us a definite percentage (not even the owners of those nukes know), but we can try and figure some things out to better understand the situation.

[cheater]

During WWII,  a hundred year old fortification (manned by raw recruits and pensioners), armed with a handful of 40- to 50-year-old torpedoes designed in 1866 of Austro-Hungarian manufacture, destroyed a heavy cruiser so new, its crew was still finishing training...   the last known operational use of a Whitehead torpedo.

https://en.wikipedia.org/wiki/Battle_of_Dr%C3%B8bak_Sound

https://en.wikipedia.org/wiki/Whitehead_torpedo


The moral of the story is that discounting old weaponry can be an expensive mistake...

Right, but that's in Norway, where budget embezzlement isn't a daily thing that's proven and reported on widely by press, like it is in Russia. So that gives at least reason to suspect and investigate how involved the upkeep is and whether it's probable that someone's been signing the inspection sheet and not doing it.

[SilverSolder]

During WWII,  a hundred year old fortification (manned by raw recruits and pensioners), armed with a handful of 40- to 50-year-old torpedoes designed in 1866 of Austro-Hungarian manufacture, destroyed a heavy cruiser so new, its crew was still finishing training...   the last known operational use of a Whitehead torpedo.

https://en.wikipedia.org/wiki/Battle_of_Dr%C3%B8bak_Sound

https://en.wikipedia.org/wiki/Whitehead_torpedo


The moral of the story is that discounting old weaponry can be an expensive mistake...

Right, but that's in Norway, where budget embezzlement isn't a daily thing that's proven and reported on widely by press, like it is in Russia. So that gives at least reason to suspect and investigate how involved the upkeep is and whether it's probable that someone's been signing the inspection sheet and not doing it.

A wise man once said, "Stuff like "Oh it's Russia with their kleptocracy, of course things don't work" just doesn't cut it." 

Enough of their stuff will probably work to be a real problem, especially if used in a surprising / unexpected way.

[daqq]

The moral of the story is that discounting old weaponry can be an expensive mistake...

Of course. Even if the thing was controlled by clockwork and ran on steam it's still a formidable weapon.

[Psi]

yeah, I too was wondering the other day, about those ~100 Russian 1 kiloton suitcase nukes that were lost/misplaced.
If any of them are still around if they would actually still function.

[dietert1]

It's wishful thinking to assume Russian nuclear weapons won't work. Nobody wants a proof.
Radioactive materials used in those warheads decay at a certain rate. Even thermonuclear warheads include fission bombs as starters. To my understanding that means you can not have a stock of warheads for let's say 10 years and expect to unpack and use it. Let's assume each one of those weapons is under tight monitoring and maintainance.
And those lost bombs probably won't work very well anymore.

Regards, Dieter