diy solar

diy solar

Fire!! Never cover LiFePO4 with wood!!!

Thank you very much for sharing. I'm sorry this difficulty, but very glad the results were not worse.

Even though we don't know the cause for certain, your experience impresses on me the potential hazard represented by the massive kWh stored in my battery bank. It speaks to the need for all reasonable precautions;
Proper fusing
Care in mechanical assembly
Assuring that stray debris does not accumulate on cells
Care with respect to combustible materials in the vicinity
Consideration of mechanical stresses imparted through rigid bus bars
A dedicated smoke/temperature detector/alarm (which I have not had, but will install immediately)
General good workmanship-like practices.

I will look at my setup even more critically now. The un-fused BMS leads have always nagged at me as a potential hazard. Normally, they carry no current, but are there hazardous modes of BMS internal failure that happen? But there are so many of these leads that I haven't thought of a workable solution. I don't want to unreasonably over-react, but maybe a fuse in the at least the BMS positive and/or negative terminal leads would be a worthwhile precaution? They are all such small gauge wire they might serve will enough as their own fuse?
Installing a fuse in the balance leads will lead to increased resistance in the circuit leading to erroneous V readings for the cells by the BMS. It's the same reason balance leads should be kept the exact same length as differing lengths of wire have different resistance. Some will argue such a short distance would not affect the V at the end, however resistance is resistance and when dealing with 3.65v per cell, even a small resistance has a huge effect.
 
No, just for Johncfii's concern about balancing wires being something that could short. I think they would blow out and splatter, probably nothing more after that.

As for fhorst's battery pack, I saw how there was thermal damage all around the terminal, and it extended for a ways onto busbar.
I didn't think that mild heating of stud by MDF would do so much to the battery case. Stud is poor thermal conductor and busbar is good thermal conductor with good contact to terminal.

So I figured terminal is what got hot. If not external poor contact, then internal poor contact (either requires current flow). I suppose internal short could do it too. Shoving terminal against layers of other electrode? Not sure of exact construction.
If you get insulating layer(thread locking compound for example..) between the terminal and the busbar the stud conducts all the current and gets hotter than anything else. Stack of busbars alleviates the problem.

OP also mentioned that battery bank was on low load when it chatch on fire. But It is possible that ot has been slowly cooking for several hours or even longer.
 
Metal, concrete (perhaps tile backer board), sheetrock.
Sheetrock/plaster has the particular characteristic of absorbing heat as it breaks down and releases water. It is used in fire resistant safes.

Weight is important for my truck camper. Also robustness from physical damage.

I think I'll go with polyisocyanate insulation board with a non-conducting surface (not aluminum skin).

The electrical insulation should be pretty simple. It doesn't need to be too thick for low voltages, Maybe a sheet of silicone rubber.

Here's a torch test for different insulations. Mineral Rock Wool is probably the best, but I think it will crumble if it's in a vehicle due to vibration.

I've used ceiling tiles for generator boxes. They can handle the heat, but crumble if they get wet or are being moved.

 
Last one could be.

From time to time there was a few kg (not my total weight of 80kg, but quite some) resting on 24 /36 threaded rods, and 2 are a little higher...

And no....
I didn't reconnect the smoke detector yet (what is powered by the BMS setup)
Cry...

So stupid, as I have battery operated alarms in the bedroom, and if the door would be open..
They sure would have been triggered.

Other lesson learned....

Anyways...
Time to rebuild and take out 16 cells.
They go to grandfather back up power, outside his house.
I probably make brick housing for them.

As long as they work, it's better then the alternative, nothing.

But I don't thrust them inside the wooden shed he calls his house.
And he is with right proud of his build.
In the west we call it a shed, at best, that's not intended negative here.

Some parts straw roof, metal sheets and loads of dry, termite eaten wood...

That will catch fire to quickly.
32* 152 and 32x 280 is still a nice large LiFePO4 battery
"And no....
I didn't reconnect the smoke detector yet (what is powered by the BMS setup)
Cry..."

Every time I have had a problem it is during some type of transition. Moving pack, new inverter, etc. I set up these temporary measures and it comes back to bite me. Never anything to this level but it has come close.

I now have covered my plywood with welding felt. More money, probably overkill but given enough time stuff does go wrong.
 
Weight is important for my truck camper. Also robustness from physical damage.

I think I'll go with polyisocyanate insulation board with a non-conducting surface (not aluminum skin).

The electrical insulation should be pretty simple. It doesn't need to be too thick for low voltages, Maybe a sheet of silicone rubber.

Here's a torch test for different insulations. Mineral Rock Wool is probably the best, but I think it will crumble if it's in a vehicle due to vibration.

I've used ceiling tiles for generator boxes. They can handle the heat, but crumble if they get wet or are being moved.

I used this stuff in my underbelly around the RV tanks. I haven't reopened it but having handled the stuff I would be extremely surprised if it crumbled. https://www.rockwool.com/uk/product...osm1kWKrVhiw66f1L8RoC-PwQAvD_BwE&gclsrc=aw.ds
 
Weight is important for my truck camper. Also robustness from physical damage.

I think I'll go with polyisocyanate insulation board with a non-conducting surface (not aluminum skin).

The electrical insulation should be pretty simple. It doesn't need to be too thick for low voltages, Maybe a sheet of silicone rubber.

Here's a torch test for different insulations. Mineral Rock Wool is probably the best, but I think it will crumble if it's in a vehicle due to vibration.

I've used ceiling tiles for generator boxes. They can handle the heat, but crumble if they get wet or are being moved.

Those are some good ideas. Ceiling tiles are simple and easy- https://www.lowes.com/pd/Armstrong-...-in-Drop-Acoustic-Panel-Ceiling-Tiles/3045521
 
With many people making compression fixtures, if the plywood ends extended sufficiently above terminal studs to be clear of any bowing, a sheet of plywood on top would be supported and you could walk on it.

(I don't believe in using MDF "Beaver Barf")
 
Have we determined the cause of this fire?

Could an internal short heat up the screw to start the fire?
Could Loctite block the terminal connection and start a fire?
Could a loose terminal connection cause enough heat to start a fire without melting the terminal?

It's a mystery to me
 
Smoke detectors are pretty cheap insurance. They are normally not suggested for a garage since car exhaust can set them off, but I spent $19 on one for the garage where I am building/testing my batteries. I wouldn't be comfortable charging or discharging unattended or while sleeping otherwise.

On an established stable system I might not be as concerned, but with anything new or when making updates, change typically increases the chance for something to go wrong.
 
Have we determined the cause of this fire?

Could an internal short heat up the screw to start the fire?
Could Loctite block the terminal connection and start a fire?
Could a loose terminal connection cause enough heat to start a fire without melting the terminal?

It's a mystery to me

Flash Point - Not Applicable
Auto Ignition Temperature - 425 - 475 degrees F
Flammable Limits – Wood Panel, Piloted flame ~500 degrees F.
Fire Extinguishing Media - Water Spray, Carbon Dioxide

So somewhere above 425F it will ignite. But if that much temp bellow I would expect more serious damage to the terminal itself.

This may just need to remain a mystery as much of what happens in Thailand..... The lyric below might give you some hints....

Bangkok, Oriental setting
And the city don't know that the city is getting
The creme de la creme of the chess world
In a show with everything but Yul Brynner
Time flies doesn't seem a minute
Since the Tirolean spa had the chess boys in it
All change don't you know that when you
Play at this level there's no ordinary venue
It's Iceland or the Philippines or Hastings or
Or this place!
One night in Bangkok and the world's your oyster
The bars are temples but the pearls ain't free
You'll find a god in every golden cloister
And if you're lucky then the god's a she
I can feel an angel sliding up to me
One town's very like another
When your head's down over your pieces, brother
It's a drag, it's a bore, it's really such a pity
To be looking at the board, not looking at the city
Whaddya mean?
Ya seen one crowded, polluted, stinking town
Tea, girls, warm, sweet
Some are set up in the Somerset Maugham suite
Get thai'd, you're talking to a tourist
Whose every move's among the purest
I get my kicks above the waistline, sunshine
One night in Bangkok makes a hard man humble
Not much between despair and ecstasy
One night in Bangkok and the tough guys tumble
Can't be too careful with your company
I can feel the devil walking next to me
Siam's gonna be the witness
To the ultimate test of cerebral fitness
This grips me more than would a
Muddy old river or reclining Buddha
And thank God I'm only watching the game controlling it
I don't see you guys rating
The kind of mate I'm contemplating
I'd let you watch, I would invite you
But the queens we use would not excite you
So you better go back to your bars, your temples
Your massage parlours
One night in Bangkok and the world's your oyster
The bars are temples but the pearls ain't free
You'll find a god in every golden cloister
A little flesh, a little history
I can feel an angel sliding up to me
One night in Bangkok makes a hard man humble
Not much between despair and ecstasy
One night in Bangkok and the tough guys tumble
Can't be too careful with your company
I can feel the devil walking next to me
 
Smoke detectors are pretty cheap insurance. They are normally not suggested for a garage since car exhaust can set them off, but I spent $19 on one for the garage where I am building/testing my batteries. I wouldn't be comfortable charging or discharging unattended or while sleeping otherwise.

On an established stable system I might not be as concerned, but with anything new or when making updates, change typically increases the chance for something to go wrong.

fhorst actually did more, wiring the smoke detector to a shutdown relay. That would have worked for a series resistance but stopping current flow, but not an internal short. A very good idea, considering that connectors can become loose or oxidized and heat up.
 
Could Loctite block the terminal connection and start a fire?
I don't know, but I had some loctite overflow and harden on a cell terminal I was doing a capacity test on last week. Within 2-3 minutes after starting the test (I think it was around 80a discharge), the bus bar was almost too hot to touch and the voltage on that cell was showing way off on the BMS. I don't know what would have happened if I didn't catch it, but it was going bad in a hurry. Something is going to melt or catch fire at some point if a terminal or bus bar gets hot enough.
 
I don't know, but I had some loctite overflow and harden on a cell terminal I was doing a capacity test on last week. Within 2-3 minutes after starting the test (I think it was around 80a discharge), the bus bar was almost too hot to touch and the voltage on that cell was showing way off on the BMS. I don't know what would have happened if I didn't catch it, but it was going bad in a hurry. Something is going to melt or catch fire at some point if a terminal or bus bar gets hot enough.
I used a 48V battery without cleaning up the loctite well enough and it sucked the loctite into the terminal connections.
I was trying to do a 12A top balance and after days of no change in voltage I realized something was wrong. Then I cleaned all the terminals but had problems again later so I cleaned up those terminals better.
 
I used a 48V battery without cleaning up the loctite well enough and it sucked the loctite into the terminal connections.
I was trying to do a 12A top balance and after days of no change in voltage I realized something was wrong. Then I cleaned all the terminals but had problems again later so I cleaned up those terminals better.
I am just reinstalling pack and purchased loctite. I think I am going to just rely upon serrated flange nuts on my pack. Loctite other connections..
 
I am just reinstalling pack and purchased loctite. I think I am going to just rely upon serrated flange nuts on my pack. Loctite other connections..
I was using loctite to permanently install my studs, not to attach the nuts. It's really about avoiding the stripping problem and making the stud install solid. Red loctite is used for this and you would never want to use it on the nuts because it's basically permanent. Blue loctite is used for assemblies which might need to be taken apart again.

I did a write up of the red loctite stud install process (link below) that explains how to do it without making a mess of your terminals. After making a mess of my first pack, I figured out the process and the rest of the cells were fine.

 
It looks to me like the terminal & bus bar connection got hot probably due to a bad connection. The MDF, even under humidity, wouldn't have shorted anything out. Something went wrong with the battery. Without high rez pictures it's hard to tell (you'd need to upload elsewhere and link them here).
Correct,

I have been down this avenue with investigations on some other electrical systems. Wood does not conduct electricity, unless soaking wet, at these voltages. Soaking wet wood does not ignite though.

It is almost certain that the terminal got to ~450F, and ignited the wood. At that temperature the terminal itself would not show any obvious damage.

Those bus bars are a bit of a mess, with a triple stack on some of them.

With that piece of wood blocking almost all airflow, it would only take 20-50W of dissipated power to cause that damage.


RULE NUMBER ONE, do not block air flow to high current conductors!
RULE NUMBER TWO, do not put wood in direct contact with conductors!

There is a very good reason that all electrical codes required connections to be made inside a non flammable enclosure. If you are interested, the AYBC codes from older years can often be found for free. They are a good reference.

A perfect example is how many home owners DIY electrical stuff, and don't use junction boxes. A wire nut gets loose, and is sitting in contact with some wood or insulation backing paper. Load the circuit up on a hot day, and it hits 450F, and a smoldering fire starts. Bye-bye house.
 
Last edited:
I suspect a short. You were talking about electro magnetic field before.
I think these batteries have certain spots that are just dying to short.
I suspect condensing relative humidity when he speaks of rain, or a leak in the roof.
 
Back
Top