diy solar

diy solar

Need help Figuring this FIRE out!

So since these cells are probably Second life, Which do you think was the most likely source of ignition, the electrolyte or a design issue, (connections)?
Without pictures of the cell packs bolted together without the cardboard covering them, we cannot see the connections.
However, the burned views show there were large washers on the bolts, that is unusual.

The terminal tabs need close tight bonding to get the correct amps through. Perhaps your washers dug into the pouches?

Any assembling pics before the cardboard stage?
 
Kinda seems like the OP has BMS avoidance syndrome .... all statements and questions about BMS ignored.
Kinda wondering if other packs have been built in the same way.

Agreed.

The idea that a set of LFP cells placed in series will maintain any sort of balance is something of a fantasy. If you get lucky with premium cells with perfectly matched everything, it's almost realistic to expect.

I too followed this "I don't need a BMS" path once. I needed a quick 12V battery made from whatever I had. Simple 4S LFP with CALB cells originally obtained from a legit North American distributor ("NA" prefix to SN). Top balanced and simply floated at 13.8V via an RV converter (no charging phases, single voltage output). I don't think these cells were ever cycled even once.

The 4 cells had the same capacity within 5% and the same IR within 10%. Again, they were top balanced.

Outcome: After about 2 months, one cell went way over voltage, the other 3 were within the operating range, though semi-healthy. Once the battery was removed, the over-voltage cell settled to 0V pretty quickly. The other 3 seem to be unscathed but settled at very low SoC.

While it's impossible to nail down the exact cause of failure, it is most likely that the failed cell was the BEST cell of the bunch, at least in terms of self-discharge. The other three cells likely had a very slight XmA self-discharge at the float voltage. This forced all 4 cells to draw a very very small current. The "good" cell didn't lose charge as fast as the others, so this very low sustained current over a long period of time drove its voltage higher to the point of failure.

No bloat or fire.
 
Last edited:
The statement about that pack being built for somebody else was a little scary.

OP, please don't take offense to this, but the lack of proper safety features and best practices show that you have no business building batteries for people yet. It's very fortunate for the owner of the cart (and you) that nobody was hurt, and nothing was burnt down. If I purchased those batteries from you, and that happened, I would insist that you pay for the damages.

Please spend some time on this forum doing research about battery management systems, proper charging techniques, fuses, and general battery safety. Everybody is capable of learning about this stuff, and it would be terribly unfortunate if you hurt somebody next time.
 
Thanks all for your reply.
Like I said before, it has been a humbling experience.
We have learned allot since then. (Friends and I)
Now we add a BMS to every built.


I will try to address all questions:
I dont have more photo detail of the original battery before construction. But here are some examples of following packs, also a few months old.
I would like a detailed explanation of exactly how you top balanced them…
Top Balancing, charging to 3.55v (charger setting for Lifepo4). checked all cells at the same voltage just after finished balancing.
1654628407085.png


this pack is a 8S2P, Just to show @Supervstech the connection on the tabs. simple, mechanical, tab to tab, used fender washer on the original, more contact area. Loctite red. the cardboard on the original was to protect from shorting the screws on the tabs. (this one used battery building adhesive foam, more distance betwen tabs)
1654628677425.png


Regarding BMS, I did not installed a BMS on this pack. Subsequent to the fire, we went back and installed BMS on all packs.
@sunshine_eggo , thanks for sharing your personal experience with No BMS. thank G_d you had no fire.
Since BMS now is a given, I wanted more feedback in the design and construction, any improvements and any possible risks.
Again, thanks all for your input.
 
Last edited:
pack burns because no BMS,
The OP insist to talk about carboard and what not.
 
The middle battery received the least amount of damage most likely because it saw fewer charge amps than the outer batteries.
Thats a great observation!
I agree! thats pushing me over to faulty charger
 
Oh, and put the crocodile clamps on diagonally opposed corners, not on the same side of the battery.
 
Thats a great observation!
I agree! thats pushing me over to faulty charger

I don't agree with that conclusion. Parallel batteries wired the way you had them will by nature end up with more charge on the outer batteries. It's not the responsibility of the charger to deal with this.
 
I meant that if his charger was faulty, in that parallel configuration, It made sense that the outer batteries recieved more damage from more amps. More stress from overcharging, etc..
But then again, could have been a bad connection..
 
I meant that if his charger was faulty, in that parallel configuration, It made sense that the outer batteries recieved more damage from more amps. More stress from overcharging, etc..
But then again, could have been a bad connection..

My suggestion mainly applies to batteries wired in parallel. Since his batteries were in series (36v for a golf cart is about right) the same may not apply.

This failure may have been a perfect storm. No BMS and a charger that is designed for lead acid batteries. The middle battery may simply have lucked out.
 
Hi

Actually they were 48v batteries connected in parallel

OK, then I'll go back to my original suggestion on why the middle battery wasn't damaged as much as the outer batteries.

There is a thread in the General Discussion subforum that explains why the middle battery will see lower voltage when connected using the diagonal method. Here is a link to that thread:

 
Thanks all for your reply.
Like I said before, it has been a humbling experience.
We have learned allot since then. (Friends and I)
Now we add a BMS to every built.


I will try to address all questions:
I dont have more photo detail of the original battery before construction. But here are some examples of following packs, also a few months old.

Top Balancing, charging to 3.55v (charger setting for Lifepo4). checked all cells at the same voltage just after finished balancing.
View attachment 97703


this pack is a 8S2P, Just to show @Supervstech the connection on the tabs. simple, mechanical, tab to tab, used fender washer on the original, more contact area. Loctite red. the cardboard on the original was to protect from shorting the screws on the tabs. (this one used battery building adhesive foam, more distance betwen tabs)
View attachment 97705


Regarding BMS, I did not installed a BMS on this pack. Subsequent to the fire, we went back and installed BMS on all packs.
@sunshine_eggo , thanks for sharing your personal experience with No BMS. thank G_d you had no fire.
Since BMS now is a given, I wanted more feedback in the design and construction, any improvements and any possible risks.
Again, thanks all for your input.
Ok, so you charged the packs in parallel…

Please list details. Exactly what voltage did you charge at, what charger did you use, what voltage did you stop at, what was the voltage on each pack after separated?
Top balancing is a slow and tricky practice… those wires wrapped in the holes add resistance to the pack. You really need perfect contact, in parallel to effectively top balance.
The packs should sit a day after charging to see if any drift when finished.
 
Thanks all for your reply.
Like I said before, it has been a humbling experience.
We have learned allot since then. (Friends and I)
Now we add a BMS to every built.


I will try to address all questions:
I dont have more photo detail of the original battery before construction. But here are some examples of following packs, also a few months old.

Top Balancing, charging to 3.55v (charger setting for Lifepo4). checked all cells at the same voltage just after finished balancing.
View attachment 97703


this pack is a 8S2P, Just to show @Supervstech the connection on the tabs. simple, mechanical, tab to tab, used fender washer on the original, more contact area. Loctite red. the cardboard on the original was to protect from shorting the screws on the tabs. (this one used battery building adhesive foam, more distance betwen tabs)
View attachment 97705


Regarding BMS, I did not installed a BMS on this pack. Subsequent to the fire, we went back and installed BMS on all packs.
@sunshine_eggo , thanks for sharing your personal experience with No BMS. thank G_d you had no fire.
Since BMS now is a given, I wanted more feedback in the design and construction, any improvements and any possible risks.
Again, thanks all for your input.
Let's see your new packs with BMS. What BMS are you running now.
 
Another thing with golf carts, they can consume a lot of current at startup. If I was going to put lithium batteries in a golf cart, I would design a motor controller that controls the ramp up from a dead stop so you don't overstress the batteries. Even with a BMS on each battery, you may hit the current limit on startup.

I have driven plenty of golf carts that come with lithium batteries. You can tell the acceleration is somewhat muted.
 
Last edited:
Back
Top