Short Circuited my 48V 120Ah Build (I'm fine)

[Hedges]

Did you say the short occurred through the fuse? Presumably something like class T?
The 17 milliohm cell internal resistance suggests to me about 22kA, so with class T rated 20kA I would expect it to interrupt successfully.

How long was current flowing? You didn't mention having to use a crowbar to remove the offending metal.
If current only flowed momentarily I have a hard time believing any internal damage occurred, just scorched outside due to arc, splatter, etc.
As Luthj said, takes time to cause enough heat (spread over internal mass)

Other people have had warm busbars at moderate current due to poor contact. I think you just did that with higher I^R
It should have simply blown the fuse, no harm, no foul other than your wrench. But you had a weaker link.

The melted cell terminal looks a bit more aggressive than the process of intentionally welding bars on terminals, could have overheated it enough to damage battery internally near the terminal.

I'm betting that any cell without such terminal damage will just shrug it off.

 

[ArthurEld]

My short was from a socket on a flanged nut and the ratchet handle hitting another busbar. The stud is gone and that busbar looks like yours.
The busbar the handle hit just has a small melted spot.
The terminal that the socket was on melted away. But there is a glob of metal left that I can test voltage on.
Here's a pic. There's a few pieces of metal I picked up and the nut. The top of the stud is in the nut but the bottom is gone.

 

[Luthj]

Yeah, those bus bars are no where near heavy enough for 400A. The contact isn't fantastic either. If you had better busbars, or a lower rated fuse, it should have successfully blown, preventing this damage.

Its also possible you just had a bad contact there, which became the weak area, limiting peak current to below the fuses trip rating.

Its not often talked about, but bus bars and their contact at the terminals should be considered when fusing the pack. In the same way any cabling would be. This is especially important at higher voltages where arcs like this can be more energetic.

Here is a class T trip curve. For less than 100ms, you need 4X+ the rated current. So 1,600A. I wouldn't expect that bus bar to vaporize in that short of time, which makes me think that nut may have been loose. That would have initiated the arc.

 

[chops728]

I can't help but wonder-- what is the power rating for the Golf Cart ---- might be a thought to have some kind of fusible links in multiple places throughout the pack -- you got me thinking about my own packs

 

[Luthj]

For comparison, my vans 12V 580AH pack is designed for 350A continuous (I typically never see more than 280A). I am fused with a 400A class T. The bus bars are 2/0 equivalent, and the main cabling is 4/0. If I had used the light bus bars supplied with the cells (similar to the ones you are using), I don't think I would fuse any greater than 200A, even if I was doubling them up.

 

[ArthurEld]

It appears that AussieInSeattle had a completely different issue than me.
I assume fuses wouldn't help in my case.

What you guys are describing is frightening. I only have 100A fuses on my batteries.

 

[Hedges]

Especially with the low bolt torque, I would expect battery cable motion to cause the connection to unscrew, which would trigger a similar failure.
The short ones AussieInSeattle has which go to a bulkhead look like they would avoid that problem because they are secured.
But anyone with a long cable going to inverter or other hardware has a lever arm bigger than the torque wrench they used. Just heating and cooling of the cable should wiggle the ring terminal back and forth. Ought to be secured to the clamp frame around the batteries, I think.

 

[AussieInSeattle]

Did you say the short occurred through the fuse? Presumably something like class T?
The 17 milliohm cell internal resistance suggests to me about 22kA, so with class T rated 20kA I would expect it to interrupt successfully.

How long was current flowing? You didn't mention having to use a crowbar to remove the offending metal.
If current only flowed momentarily I have a hard time believing any internal damage occurred, just scorched outside due to arc, splatter, etc.
As Luthj said, takes time to cause enough heat (spread over internal mass)

Other people have had warm busbars at moderate current due to poor contact. I think you just did that with higher I^R
It should have simply blown the fuse, no harm, no foul other than your wrench. But you had a weaker link.

The melted cell terminal looks a bit more aggressive than the process of intentionally welding bars on terminals, could have overheated it enough to damage battery internally near the terminal.

I'm betting that any cell without such terminal damage will just shrug it off.

Yes there was a fuse in place that I had on my old FLA pack a ANN-400 fuse which did not blow - have since learnt I should have had a much lower fuse like a ANN-250.

The cell with the melted terminal still has a voltage the same as the other cells but its trash now.

So assuming the other 15 cells are fine I should just buy another 120Ah Higee cell from amy/xuba?

 

[AussieInSeattle]

I can't help but wonder-- what is the power rating for the Golf Cart ---- might be a thought to have some kind of fusible links in multiple places throughout the pack -- you got me thinking about my own packs

500amp controller (you can set peak amps) but I didn't get that far - I was going to set it at 200. Carts can pull 200 amps under heavy load going up a hill.

 

[Hedges]

400A fuse vs. 250A wasn't the problem. It takes some amount of time for them to blow anyway.
There was enough resistance at that one connection that it went into runway and all the energy was deposited there.

Have you dealt with aluminum terminal native oxide and inhibiting corrosion? I'm guessing not, unless this was just an un-torqued connection.

 

[AussieInSeattle]

Especially with the low bolt torque, I would expect battery cable motion to cause the connection to unscrew, which would trigger a similar failure.
The short ones AussieInSeattle has which go to a bulkhead look like they would avoid that problem because they are secured.
But anyone with a long cable going to inverter or other hardware has a lever arm bigger than the torque wrench they used. Just heating and cooling of the cable should wiggle the ring terminal back and forth. Ought to be secured to the clamp frame around the batteries, I think.

I did not check the terminal nuts after putting the pack in the cart - I built the pack in my basement and they could have come loose while I was moving the pack upstairs and into the garage where the cart was. Moving a 100lb pack was an exercise in itself!

I'm leaning towards just selling the remaining 15 cells if I can get anything for them. Like other Lithium stuff there are drop-in solutions for golf carts but I was trying to save money.

 

[Luthj]

So assuming the other 15 cells are fine I should just buy another 120Ah Higee cell from amy/xuba?

Sure, though maybe buy a spare just in case.

 

[Luthj]

I'm leaning towards just selling the remaining 15 cells if I can get anything for them. Like other Lithium stuff there are drop-in solutions for golf carts but I was trying to save money.

You have already gotten this far. It seems much easier and cheaper to just replace the one damaged cell. Once you load test the pack you can see if any connections are loose. You just were unlucky.

 

[AussieInSeattle]

400A fuse vs. 250A wasn't the problem. It takes some amount of time for them to blow anyway.
There was enough resistance at that one connection that it went into runway and all the energy was deposited there.

Have you dealt with aluminum terminal native oxide and inhibiting corrosion? I'm guessing not, unless this was just an un-torqued connection.

Its looking like that connection was dodgy - I have other pics from earlier in the night that elude to there not being as much stud thread sticking out the top of the nut which might have meant it was not tight on that terminal.

 

[AussieInSeattle]

You have already gotten this far. It seems much easier and cheaper to just replace the one damaged cell.

And just monitor the other 15 cells for heat issues via the bms to make sure they are not damaged? Or do some sort of testing on the remaining 15 cells?

 

[Hedges]

Its looking like that connection was dodgy - I have other pics from earlier in the night that elude to there not being as much stud thread sticking out the top of the nut which might have meant it was not tight on that terminal.

But how did you prepare the terminals? Are you familiar with native oxide on aluminum making poor contact, and the success some members have had fixing previously hot busbars?

 

[Luthj]

And just monitor the other 15 cells for heat issues via the bms to make sure they are not damaged? Or do some sort of testing on the remaining 15 cells?

You don't need to test the cells. We know that a poor connection caused the arcing. The length of short circuit was brief enough.

As I said early I think that connection was loose. You may have missed it when torqueing things down.

 

[Hedges]

And just monitor the other 15 cells for heat issues via the bms to make sure they are not damaged? Or do some sort of testing on the remaining 15 cells?

Especially if you didn't blow the fuse, I wouldn't worry at all about current flow damaging the cells. Just this one that got a terminal heated enough to melt.

 

[AussieInSeattle]

But how did you prepare the terminals? Are you familiar with native oxide on aluminum making poor contact, and the success some members have had fixing previously hot busbars?

I am not familiar with that - I just wrapped the busbars in some heat shrink to keep the double ups together and put them on there with 20mm stainless studs and serated flange nuts

 

[Luthj]

These terminals aren't anodized, so you don't need to do much but wipe with solvent. Non flat busbars are the main issue most folks encounter.

Those double bus bars should be good for 100-150A. Brief pushes up to 200A might be okay. With the minimal voltage sag of the LFP packs, you won't need as much current as you did with the lead pack.