I have seen broken strands of wire caused by over crimping cause fires. I have seen loose connections cause it also.
Fire!! Never cover LiFePO4 with wood!!!
- Thread starter Frank in Thailand
- Start date
It could also mean the busbar conducted heat away and thus leaves the terminal surface relatively intact. I still think the weight pressed down and caused a problem internal, right underneath the terminal.
That’s my guess as well.
From some tear-down videos, the connection between the aluminum terminal and the pouch is through a large number of very fine (and fragile) wires.
Any damage to any of those wires no doubt results in increased internal resistance and increased heat when current is being run through that cell.
His current levels overnight were modest, but someone can do the math to determine how much IR is needed to heat a terminal up to wood-charring temps over 10 hours at those current levels.
Once the terminal and stud get hot enough, they can melt plastic and char wood. The busbar shielded the terminal surface from smoke/char residue, but so what? It’s not the aluminum terminal that was burning, just all the non-metallic material it was in contact with.
ArthurEld
Solar Wizard
So you are saying the terminal was so hot it could heat the busbar and screw hot enough to start a fire (which is pretty much red hot)
So all of those items and more were red hot but the terminal and busbar didn't melt or even show a sign of heat.
edit: The terminal wouldn't have to be red hot if there was an arc or short above the busbar
Last edited:
Hedges
I See Electromagnetic Fields!
- Joined
- Mar 28, 2020
- Messages
- 20,497
Doesn't need to be red hot. Even a soldering iron will char wood.
Ignition Temperature of Wood - Fire Engineering - Firefighting
What is the ignition temperature of wood?
www.fireengineering.com
Fahrenheit 459, although judging by the Dr. Suess experience, that may not be in school libraries anymore.Doesn't need to be red hot. Even a soldering iron will char wood.
Ignition Temperature of Wood - Fire Engineering - Firefighting
What is the ignition temperature of wood?
ArthurEld
Solar Wizard
I just looked up MDF and it is similar. It is going to be tough making my battery without wood or other combustible materials.Doesn't need to be red hot. Even a soldering iron will char wood.
Ignition Temperature of Wood - Fire Engineering - Firefighting
What is the ignition temperature of wood?
Fhorst's battery looks like all of the heat is on the top.
You guys could be right. I just like to look at all of the angles.
I'll even say you guys are probably right. But I am not 100% convinced.
Other minor things could have started the fire on top. Even a shorted balance lead.
But the intensity is definitely around that terminal. It wouldn't be that way from a shorted balance lead.
After I shorted my battery I decided I had to make my batteries as safe as possible.
If I don't do everything right I could have a fire.
Last edited:
Charring starts even lower (320F): https://startwoodworkingnow.com/how-hot-does-wood-burn/
“The initial process of wood burning is at 160-260 degrees Celsius (320 degrees Fahrenheit). Irreversible changes begin to appear in the wood, ending with fire. The ignition temperature of the wood varies between 200-250 degrees Celsius (392-482 degrees Fahrenheit).”
Hedges
I See Electromagnetic Fields!
- Joined
- Mar 28, 2020
- Messages
- 20,497
1/2" of sheetrock makes a 15 minute fire barrier, 5/8" is 30 minutes. At least for house fires.
If nothing else, having that between you and the battery should give you time to escape.
Double layer was used on walls between condo units. Made studfinder useless.
Is concrete backer board any good as a fire barrier?
Yes. However, it's a PITA to drill into and can be quite dusty.
ArthurEld
Solar Wizard
That's the problem with getting rid of all the wood. Then you don't have anything easy to screw in to.
How about fire retardant plywood - https://www.homedepot.com/p/3-4-in-...-Fir-Pressure-Treated-Plywood-97874/203482483
Hedges
I See Electromagnetic Fields!
- Joined
- Mar 28, 2020
- Messages
- 20,497
I think stick-frame houses (with sheetrock) are more fire resistant than metal-stud construction.
If you make a plywood box and line it with sheetrock that should hold off fire for quite a while.
For course, the paper surface will char if exposed to heat.
Fire resistant safes have plaster poured into a gap.
If you make a plywood box and line it with sheetrock that should hold off fire for quite a while.
For course, the paper surface will char if exposed to heat.
Fire resistant safes have plaster poured into a gap.
ArthurEld
Solar Wizard
Drywall might be good for lining wire shelves like this - https://www.homedepot.com/p/Husky-S...in-W-x-78-in-H-x-24-in-D-MR482478W5/203828250
Or for making a cover to protect the terminals.
It's not very good to screw into but it could work.
Maybe use fire retardant plywood for the fixture.
Or for making a cover to protect the terminals.
It's not very good to screw into but it could work.
Maybe use fire retardant plywood for the fixture.
Last edited:
That's actually what I was thinking off, with a concrete backer board lining. Mount the BMS on the outside, since so far I've not had any cell get remotely warm yet.
ArthurEld
Solar Wizard
The concrete backer board would be very good to line my shelves. It would protect each shelf from the battery below.
ArthurEld
Solar Wizard
I'm going to continue beating this dead battery.
So, I agree that the fire most likely happened from heat at the terminal.
The reason I think that is because it is obvious from the picture that the main intensity of the fire is centered around that terminal.
Just for reference, my short was on the top side of the busbar and it still melted the terminal and busbar.
But lets just say fhorst's short was at the terminal or below the terminal.
fhorst said his terminal was lowered a small amount.
He thought it was from kneeling on the wood resting on the stud. That might be the cause or it might not.
I assume it would take quite a bit of weight to actually push the terminal down. Especially with all those busbars.
His studs are so long it probably would tip easier than push the terminal down.
It is possible the terminal just lowered because of the heat. Heat tends to distort most things.
People really want to come up with a quick determination and blame fhorst so they don't have to worry that this might happen to them.
That is the easiest.
I'm not easily convinced. There still are multiple possibilities.
So, I agree that the fire most likely happened from heat at the terminal.
The reason I think that is because it is obvious from the picture that the main intensity of the fire is centered around that terminal.
Just for reference, my short was on the top side of the busbar and it still melted the terminal and busbar.
But lets just say fhorst's short was at the terminal or below the terminal.
fhorst said his terminal was lowered a small amount.
He thought it was from kneeling on the wood resting on the stud. That might be the cause or it might not.
I assume it would take quite a bit of weight to actually push the terminal down. Especially with all those busbars.
His studs are so long it probably would tip easier than push the terminal down.
It is possible the terminal just lowered because of the heat. Heat tends to distort most things.
People really want to come up with a quick determination and blame fhorst so they don't have to worry that this might happen to them.
That is the easiest.
I'm not easily convinced. There still are multiple possibilities.
It's best to be the owner of the battery. We can only look at the data/evidence through the info here.
But anyways, I'll take a turn at bat with beating this.
I believe it would be most likely that a small wire was pinched or compressed against a stud or some buss bar. The insulation is pretty thin on the small wires. It would have started warming at that pinch point, quickly heating the wire/insulation to have other shorting locations. The temperature sensor wire is a good possibility.
Protecting the small wires is important. Sometimes DIY'ers overkill one design criteria, and underkill others. You can either protect all the small balancing/thermal wires, or have them the shortest length. Doing both is difficult.
I'm sure there's debates on the resistance/foot of the balancing leads and how it will affect the operation. I'm not concerned much. I'd lean towards protecting the small wires.
At work we use clear tubing to avoid chaffing or shorting. I believe it's PTFE. It's rugged and high temp. Pushing the small wires through this tubing would protect them, but this would lengthen the wires a little.
Usually, the very last thing that's been modified has a connection to the fault. An amp meter was added and immediately a short occurred when connecting 12V to meter. Did the short cause wire insulation damage?
ArthurEld
Solar Wizard
I am not very experienced with this kind of thing but it looks to me like the most heat is obviously around one terminal.
A small wire could cause a fire but it wouldn't cause the the heat pattern around the terminal.
@fhorst should be able to test that cell to find out if there is an internal short.
There also appears to be some extra heat around the main negative terminal on the left. Maybe that is a clue.
fhorst thought there could have been a short though the wood between the circled terminals.
Wood seems unlikely but could something else have caused a short between those two terminals?
Would the negative terminal get hot if there was an internal short in the other cell?
A small wire could cause a fire but it wouldn't cause the the heat pattern around the terminal.
@fhorst should be able to test that cell to find out if there is an internal short.
There also appears to be some extra heat around the main negative terminal on the left. Maybe that is a clue.
fhorst thought there could have been a short though the wood between the circled terminals.
Wood seems unlikely but could something else have caused a short between those two terminals?
Would the negative terminal get hot if there was an internal short in the other cell?
Similar threads
