Exploded energy discharge capacitor

[MRMILSTAR]

Someone working at Laurence Livermore National Laboratory posted this picture a few months ago on FB. This is what the guts of a large energy discharge capacitor looks like. I believe it was 300 uF rated at 24 KV. The source states that it was used in laser fusion experiments and was pushed a little too hard.  This capacitor is made by ICAR. From the outside it looks similar to one that I have except mine is an Aerovox.

I must say that the inside of this looks nothing like I expected. It appears to be comprised of a bank of smaller capacitors immersed in oil. I was expecting a stack of alternating parallel sheets of foil and dielectric all immersed in oil.

Does anyone know if all of these large energy discharge capacitors are constructed in this manner?

[davekni]

Wow, thank you for the picture Steve!  Wonder if my ancient paper/oil quarter-shrinker cap is similar/

[klugesmith]

Never would have guessed that, but the picture doesn't look fake.  Maybe the capacitor is from a recent program, e.g. NIF, and capacitor contractor is trying a new technology.  There must be important variables other than voltage and energy in the "pulse discharge" market, even if we consider only Big Science.  Perhaps no one size fits all.  Or maybe they're all like that, and always have been, but I doubt it.

Consider the internal construction options for 9 volt batteries and car-starting batteries.

[MRMILSTAR]

The picture is definitely real. The lab employee couldn't give out all of the details because the details were proprietary. He also stated that handling these exploded capacitors was very risky because each one of the individual capacitors comprising the unit had to be discharged before it could be safely handled.

I also forgot to state that several hundred of these were used in combination in some sort of parallel-series arrangement. Maybe this one developed an internal short and the other ones discharged their full energy load into it causing a very big BOOM! The risks of parallel energy discharge capacitors.

[Mads Barnkob]

Does anyone know if all of these large energy discharge capacitors are constructed in this manner?

It is perfectly normal, it is the only way to control current distribution. Large plates would have "cold" corners and be prone to hot spot damages.

Many smaller capacitors follow the same rule as a MMC, we distribute the voltage/current sharing, have better heat dissipation and manufacturing is easier as you can use off the shelf products/manufacturing methods.

[klugesmith]

Well said, Mads.  You've made a believer out of me.

As Steve said, the ruptured case probably represents a capacitor that failed internally and got the energy of its bank mates.

The stored energy of a big box cap by itself can make a lot of noise, or heat the whole mass of capacitor by a tiny fraction of one degree C.

[klugesmith]

Physical damage to a capacitor from its own stored energy might be worse than my guess in previous post.  Depends on how efficiently electrical energy can be converted to mechanical work, instead of local heat, in a fault.

Steve's coin shrinking capacitor, at full charge, holds 0.06 watt hours per kg.
In theory that's enough to lift the whole capacitor by 24 meters.  So an absolute upper bound on physical damage, is that from dropping capacitor 24 meters.
Contest time: actually elevate HV capacitor banks using their own stored energy. 

To put that in perspective, consider damage to a car starting battery from electrical overloading or failure.  With typically 40 Wh/kg, a car starting battery could elevate itself by 15 kilometers.  A supercapacitor (5 Wh/kg) might elevate itself by 1800 meters, but doesn't need a thick steel case.  Of course they are designed for discharging no faster than 10 minutes, or 10 seconds, instead of less than a millisecond.

[MRMILSTAR]

I'll have to put that capacitor-raising project at the bottom of my list of projects. 

[Mads Barnkob]

Here is a paper on metallized polypropylene capacitors and their thermal properties.

Look at a fig. 3 to get an idea about why large capacitors are worse than many smaller.