Shunt snapped and caused a fire🔥

[nezjnr]

Got a call Sunday from the wife about smoke coming from the garage and fire trucks were already on the way.

After they had there fun spraying many cans all over the garage I got to have a look inside.

I have 5 rows of shelves with 2 shelves for in each with a battery and then busbar along the top for interconnecting everything.

I had each battery on a 300a shunt, looks like my failure was the shunt actually snapped right in half, then probably made contact with the wood underneath/metal frame and looks like shelf collapsed onto the bottom battery and probably 10 cells let off the smoke/burned .

Definitely should have had a fire alarm in the garage hooked in with the house.
And should have mounted the back of the shunts on isolators.

I tried braking the other shunts but couldn't get any to give think it was just a defective one.

 

[time2roll]

Was the shunt mounted to something solid? A before picture would be interesting.
Sorry for the troubles.

 

[OffGridForGood]

Wow what a mess,
I don't think I have ever read of a fire caused by a shunt, what make and model ?

 

[42OhmsPA]

Wow that looks miserable, fortunately it wasn't worse.

Are you sure the shunt was the root cause? Is it possible one of the particle board shelves collapsed causing a dead short and the shunt to fail?
The shunt was on the negative, correct?

 

[SupraSPL]

That's was I was wondering, shelf collapse break the shunt?

 

[nezjnr]

It's really hard to determine because the fire fighters sprayed and hosed it deforming a lot of the plastics and wood.

But the shunt was on the negative, going to a breaker, it had no strain on it so the battery dropping wouldn't made the shunt snapped like that.

I think it was poor design on my part really should have had an isolated under the non busbar side for support but also the metal mustn't have been great because it's sheared it's self inflicted half.

 

[nezjnr]

I just got power restored hoping my Home assistant caught the cell voltages and current flow at the time, wil add if I can get.

 

[nezjnr]

Shunt and Monitor was one of these.

https://a.aliexpress.com/_m0YM7ro

I got the 300a shunts, highest current was only 60a

 

[seneysolar]

Holy crap!!!!!
Do you have "before" pictures???

 

[seneysolar]

I have a feeling with that style shelving, with the typical partical board inserts they come with, one of the shelves finally gave out and dropped some batteries. They do not hold heavy objects very long...

 

[rhino]

Were those shelves the ones you get at menards/Home Depot that are particle board and the cells were on them?

 

[OffGridForGood]

Shunt and Monitor was one of these.

https://a.aliexpress.com/_m0YM7ro

I got the 300a shunts, highest current was only 60a

looks like the shunt is 50A max under constant load.

 

[rhino]

looks like the shunt is 50A max under constant load.

I got the impression they must have a 300A version since that 50A one doesn't look like what is shown in the burnt pictures.

In any case I would not trust a critical piece of your power system that all power in your system goes through to some not well known Chinese brand.

 

[rhino]

Should also say we are sorry for your loss and thankful it just appears to be that area of the home/garage that was damaged. Thank you for sharing so others may be able to learn.

 

[nezjnr]

looks like the shunt is 50A max under constant load.

I bought a different model, but it was this model/design

 

[nezjnr]

Holy crap!!!!!
Do you have "before" pictures???

Not really before pictures but from the other end.

I'll have a look later for before pictures.

 

[nezjnr]

I have a feeling with that style shelving, with the typical partical board inserts they come with, one of the shelves finally gave out and dropped some batteries. They do not hold heavy objects very long...

These sit on a lip for support and were rated for the weight of the batteries...

 

[nezjnr]

Should also say we are sorry for your loss and thankful it just appears to be that area of the home/garage that was damaged. Thank you for sharing so others may be able to learn.

Yeah, thankfully it's just some money and time and a lot of stress.

But wife is fine, cats are safe garage plater just needs a little bit of touching up and paint.

Hopefully it helps someone.

 

[Brucey]

Not really before pictures but from the other end.

I'll have a look later for before pictures.

That shunt doesn't look like it has an isolation base. A Victron for example:

 

[rhino]

Thanks for the extra pictures.. couple questions, why are there alligator clips on all those cells? also is there some insulation between the cells?

In any event even if those shelves are rated for the weight you had on them I would never trust particle board for something like that. It is not a stable "wood" and can easily warp especially when exposed to moisture. If it warped at all it would cause the collapse of the shelf.

 

[Checkthisout]

Can see why the shunt snapped.

They aren't fastened to anything and supporting the weight of the cables.

I would say there was a bad crimp/lose bolt at the shunt that failed

OR

There was some other kind of failure that caused high current draw across the shunt.

OR the shunt was one of the bad ones of the batch.

Either way, fast those shunts to a solid surface so they aren't supporting the cables.

Edit: Also let me make it clear that I'm also not attacking the OP. That's not the point of the "up in smoke" section of this forum. Plus I'm in no position to judge.

 

[OffGridForGood]

Not attacking the OP, just want to point out some things to be warry of,
For, Guests, others learning - review the pics-

1. we see all-thread rods running along side the cells without any physical isolation - the exterior of the cells is just a thin layer of shrink wrap and the cases are electrically connected to NEG terminals - if you use all-thread/cables/steel straps (any conductive materials) always put some type of covers over them. {correction, I am told by more knowledgable people the cases are not neg but pos, indirectly connected, point remains - don't let anything metal touch the cells, isolation is safest}

2. we see aligator clamps - these appear to have been used for BMS sense wires. The BMS is the safety device to shut down things if the cells get too high voltage, get too hot, get too cold, get too low voltage - if the aligator slips off a bus bar, the BMS can't do it's job. If it touches another bus, trouble can result.

3. we suspect the cells were supported on particle board - cells are heavy use a support worthy of respect. Particle board is not a very strong material under shock load, or if it gets damp in high humidity. The cells are heavy and their constant load on the shelf over a long period of time may not be ideal solution. If one set of batteries fell, the sudden load on the next shelf down is going to lead to a cascading failure.

4. we are not seeing separation sheets between the cells. Maybe just unclear from the pics with all the dust and debris in the shot. Cells need separator sheets, FR-4 is recommended product.

5. we see all-thread rods, with thick plywood ends. we don't see any cushion material anywhere to allow the cells to expand, we don't see any springs at either end of the all-thread rods to allow the cells to expand. Depending on which cells these are, we need to allow up to 1mm per cell for a full charge expansion. Without any room, the cells can be over-compressed and vent. The cells could have been set in the rods at full 100% charge, not become over-compressed. but seems unlikely. Better to have no compression rather than over-compression. Maybe the fire burnt away some springs and we don't see them now.

 

[Checkthisout]

Not attacking the OP, just want to point out some things to be warry of,
For, Guests, others learning - review the pics-

1. we see all-thread rods running along side the cells without any physical isolation - the exterior of the cells is just a thin layer of shrink wrap and the cases are electrically connected to NEG terminals - if you use all-thread/cables/steel straps (any conductive materials) always put some type of covers over them.

2. we see aligator clamps - these appear to have been used for BMS sense wires. The BMS is the safety device to shut down things if the cells get too high voltage, get too hot, get too cold, get too low voltage - if the aligator slips off a bus bar, the BMS can't do it's job. If it touches another bus, trouble can result.

3. we suspect the cells were supported on particle board - cells are heavy use a support worthy of respect. Particle board is not a very strong material under shock load, or if it gets damp in high humidity. The cells are heavy and their constant load on the shelf over a long period of time may not be ideal solution. If one set of batteries fell, the sudden load on the next shelf down is going to lead to a cascading failure.

4. we are not seeing separation sheets between the cells. Maybe just unclear from the pics with all the dust and debris in the shot. Cells need separator sheets, FR-4 is recommended product.

5. we see all-thread rods, with thick plywood ends. we don't see any cushion material anywhere to allow the cells to expand, we don't see any springs at either end of the all-thread rods to allow the cells to expand. Depending on which cells these are, we need to allow up to 1mm per cell for a full charge expansion. Without any room, the cells can be over-compressed and vent. The cells could have been set in the rods at full 100% charge, not become over-compressed. but seems unlikely. Better to have no compression rather than over-compression. Maybe the fire burnt away some springs and we don't see them now.

So what do the holding fixtures for these cells look like when they are installed in EV's?

They are in a constant tension fixture??

 

[SparkyJJO]

They do not use springs on EVs.
But that's an age old argument here, not something to (yet again) rehash in this guy's thread.
Sorry for the OP's loss, that's expensive. But glad it didn't go even worse!

 

[Checkthisout]

Gandering at the pictures more.

That bus bar is just floating there not fastened to anything.

And also! Holy crap! Just realized how many batteries the OP had. That setup is massive!

That must have been an interesting sight for the fireman. Those now green cables would have been glowing intense orange in color! Hot spaghetti!