No joke - A billion-dollar capacitor problem.

[Rick Law]

If you think you have bad capacitor problem, here's one that will blow your mind - in more ways than one.

Since many would not want to read a long article, here is my summary of the main points in the article (except those in quotes, the rest are my words):


"Defects found in a $5 electrical component will delay the Navy and Air Force nuclear warhead refurbishment program by 18 months and cost more than $1 billion to fix, a National Nuclear Security Administration official said during a congressional hearing Wednesday."

During the cold war, 70% of the missile parts were made in house.  Now 70% are off-the-shelf parts due to cost reduction initiatives.

While in early tests, those $5 commercially available capacitors were meeting specification meet requirement today, however during later stress tests, a few of those failed.

"The problem is, these parts used in the warhead upgrades must survive for decades, up to 30 years after production, Verdon said. However, the quality of each capacitor production lot varied, which led to the stress testing failure. Instead of using the capacitors and risking readiness in the future, Verdon said his agency decided to delay the upgrade work, initially scheduled to begin in December.
[Charles Verdon is the deputy administrator for defense programs at the National Nuclear Security Administration]

Replacement capacitors are being produced but will cost about $75 per unit, compared with the $5 per unit cost of the off-the-shelf capacitors that failed stress testing."

Here is the USNI September 25, 2019 article:
https://news.usni.org/2019/09/25/faulty-5-parts-cause-18-month-1-billion-delay-to-navy-air-force-nuclear-upgrades

EDIT:
Changed the word I used in the summary "While in early tests...meeting specification" to the wording in the original article "While in early tests...meet requirement today".  Requirement and Specification can be very different.

[Circlotron]

Moral of the story is - beware of mil-spec eBay parts.

[ataradov]

They need to switch the operating model to a more consumer one. Why have a very expensive warhead that has to last 30 years? Why not have a cheap one that has to be replaced every 5 years. Yes, overall quality will be lower, but on the up side you can make a ton of them fast and cheap.

Having more warheads about to expire is also good for training and design refinements.

[bd139]

They tend to work on the basis they will need to survive perhaps two decades into a total supply chain collapse.

[Rick Law]

Please note: I made a change in wording in the OP to make my summary reflect the original meaning better.

EDIT:
Changed the word I used in the summary "While in early tests...meeting specification" to the wording in the original article "While in early tests...meet requirement today".  Requirement and Specification can be very different.

[DimitriP]

Industry best practices were used to stress the components beyond their design planned usage as a way to establish confidence that they will continue to work over the necessary lifetime of the warhead,” Verdon said. “During stress testing, a few of these commercially available capacitors did not meet the reliability requirements.”

Ahh....yes...."best practices" but which industry? Gotta love those! 

[Dundarave]

Those old RIFA brand caps are showing up everywhere...   

[ataradov]

Those old RIFA brand caps are showing up everywhere...   

Warheads are supposed to explode. All you have to do is time it right.

[CatalinaWOW]

If there is any military product that I would want to work 20 years without maintenance it would be a nuke.  Every time you take one out of a vault or wherever they are stored you increase the risk of something bad happening.  A theft.  A train wreck.  A lightning strike.  Even a meteor.   Or a long list of other bad things.  None of these is likely, but in a really high stakes game like this you don't want to roll snake eyes.

Best answer is to get rid of them, but that doesn't seem to be the world we live in, so I am happy to see them handled as little as possible.

It does seem that it there would be cheaper ways to get the caps they want.  Just buy up the rest of the lots that pass the stress test.  I am sure that the smallest batch size of commercial parts is huge relative to the number of nuclear warheads.

[Ampera]

I mean if that billion is being used to fire US-run factories, that's not /as/ bad, as it goes back into our economy.

[peter-h]

Best thing is to not use electrolytic capacitors, in anything which is to last decades.

It can be very difficult but is possible.

There is a big variation in quality, and it isn't possible to predict who will make a low quality batch one day.

Look at the motherboard "popping capacitor" problem of a few years ago. Those were chinese caps; with chinese stuff there is absolutely no way of knowing.

[Kjelt]

Best thing is not to build any warheads 

[Ampera]

Best thing is not to build any warheads 

And let the military industrial complex collapse? As a hot blooded American dedicated to the development of new and incredible ways to kill surprisingly few soldiers while decimating civilian populations for the sake of financial gain and economic enhancement, I find the concept of a peace time economy against the freedom to manifest our destiny in any way possible!

[scatterandfocus]

Best thing is not to build any warheads 

And let the military industrial complex collapse? As a hot blooded American dedicated to the development of new and incredible ways to kill surprisingly few soldiers while decimating civilian populations for the sake of financial gain and economic enhancement, I find the concept of a peace time economy against the freedom to manifest our destiny in any way possible!

And not spend 50% of our tax dollars (that we know of) on building their toys of power and destruction.  Imagine how much better science and  engineering education could be, and how many more people could be educated, with even half of that money.  But that is business I guess.

[Rick Law]

Best thing is not to build any warheads 

And let the military industrial complex collapse? As a hot blooded American dedicated to the development of new and incredible ways to kill surprisingly few soldiers while decimating civilian populations for the sake of financial gain and economic enhancement, I find the concept of a peace time economy against the freedom to manifest our destiny in any way possible!

Actually, while it is called "military industrial complex", useful civilian technology do come out of it.  It was US Defense's DARPA (Defense Advanced Research Projects Agency) who invented the internet.

Without defense spending on research, I suspect if we were solely on civilian-commercial spending for R&D, satellite and rocket technology would be far more expensive - far too expensive for something like weather satellite or civilian communication satellites.

I'd rather see our federal government focus on defense than other stuff that should be done by individuals or NGO's.  If useful stuff come out of it, great, that is extra side-benefits.

[Kjelt]

Not building warheads does not mean not investing in defense.
Just imagine what devastation would happen if a nuclear warhead goes off in its silo within 100 miles of a large city in the US due to a failing component.

[bd139]

Unless you’re in Russia. Nyonoksa radiation accident...

[zl2wrw]

You might be interested in reading about "strong link / weak link" systems with respect to nuclear weapons. The basic concept is that the weak link in the system is what makes it go prompt critical, and the strong link is what stops it from doing so. In the event of a mishap, eg fire, rocket malfunction on launch, etc the weak link is designed to fail before the safe link, rendering the bomb a dud...

I understand that convincing ones opponents of the reliability of ones nuclear arsenal is a requirement for an effective deterrent "2nd strike capability", however given the size of the US (and Russian) nuclear arsenals relative to the worlds other nuclear weapons states, maybe Trump & Putin should negotiate another round of "megatons to megawatts"? (converting nuclear material from bombs into civilian nuclear reactor fuel - a good way to get rid of them bombs)

[langwadt]

If there is any military product that I would want to work 20 years without maintenance it would be a nuke.  Every time you take one out of a vault or wherever they are stored you increase the risk of something bad happening.  A theft.  A train wreck.  A lightning strike.  Even a meteor.   Or a long list of other bad things.  None of these is likely, but in a really high stakes game like this you don't want to roll snake eyes.

Best answer is to get rid of them, but that doesn't seem to be the world we live in, so I am happy to see them handled as little as possible.

It does seem that it there would be cheaper ways to get the caps they want.  Just buy up the rest of the lots that pass the stress test.  I am sure that the smallest batch size of commercial parts is huge relative to the number of nuclear warheads.

add to that the with nukes there's a bunch of treaties to follow too

[MT]

Unless you’re in Russia. Nyonoksa radiation accident...

The one in Spain 1966 Palomares accident was quite bad as well, 4 hydrogen bombs went loose and one or two exploded (not the plutonium) after the B52 and KC135 tanker exploded in mid air during refill! The place are still radioactive today!

[Kjelt]

(converting nuclear material from bombs into civilian nuclear reactor fuel - a good way to get rid of them bombs)

I thought the high radioactive material for nukes was a by-product from nuclear powerplants not the other way around?

[LeonR]

Not building warheads does not mean not investing in defense.
Just imagine what devastation would happen if a nuclear warhead goes off in its silo within 100 miles of a large city in the US due to a failing component.

Aren't silos supposed to contain the fallout mass of such cases?

[Seekonk]

Like the nuclear missile silo that blew up near Atlanta (Damascas Arkansas) hardly anyone knows about? Watch Command and Control, PBS 2017.

[langwadt]

(converting nuclear material from bombs into civilian nuclear reactor fuel - a good way to get rid of them bombs)

I thought the high radioactive material for nukes was a by-product from nuclear powerplants not the other way around?

https://www.nytimes.com/2009/11/10/business/energy-environment/10nukes.html

[Bud]

They need to switch the operating model to a more consumer one. Why have a very expensive warhead that has to last 30 years? Why not have a cheap one that has to be replaced every 5 years. Yes, overall quality will be lower, but on the up side you can make a ton of them fast and cheap.

Having more warheads about to expire is also good for training and design refinements.

In a big organization upgrades are a massive undertaking. I can tell seemingly simple upgrade of the servers fleet 5 years is a bare minimum anyone would agree to and then the project may last 1 or 2 years. Do not ask me why, this is the way it is.