Shunt snapped and caused a fire🔥
- Thread starter nezjnr
- Start date
robbob2112
Doing more research, mosty harmless
From what we see in the pictures we can infer the basic design of 1 battery that is repeated 4 times. 3 shelves, bottom is 16 cells in compression, second is 16 cells in compression (8*2 rows), 3rd shelf had lengthwise busbar across all 5 batteries. The BCM is visible in red heatsink at one end down low. The breakers are mounted on the ends up high. The shelves appear to be 16 or 18 inches wide, but could be 24
This gives us 48v in parallel 10 times. To put 2 cells side by side I assume 105ah cells. . Someone else can calculate the total capacity.
On the busbars, one negative one positive you can see the breakers connected per shelf. So 2 breakers on the ends, one for upper and one for lower. Bottom side connects to battery, top to bussbar.
On bussbar it appears there are leads from the PV array several places coming in from the opposite side.
Banalance leads from the end using aligator clips which has already been covered.
This gives us 48v in parallel 10 times. To put 2 cells side by side I assume 105ah cells. . Someone else can calculate the total capacity.
On the busbars, one negative one positive you can see the breakers connected per shelf. So 2 breakers on the ends, one for upper and one for lower. Bottom side connects to battery, top to bussbar.
On bussbar it appears there are leads from the PV array several places coming in from the opposite side.
Banalance leads from the end using aligator clips which has already been covered.
OffGridForGood
Catch, make or grow everything you can.
these look bigger than 105's to me. not sure, but look at post # 17 pic 131500 .
93 x 0.550kW = 51.1kW PV wow!
(i wish i had that much!)
robbob2112
Doing more research, mosty harmless
So, asked my soon to be brother-in-law a few questions and showed him the pictures. He is a fireman down in texas with a medium sized town. He agrees with the shelf collapse theory from what he can see.
And as a side note he says the extinguishing agent was likely class D powder to smother the active flames. Seems Like that could be a way to build a self extenguisher based on some retaining wire or thin rope with a bottom opening bin above the battery bank.
Sodium cloride based - Heat from the fire causes it to cake and form a crust excluding air and dissipating heat from burning metal.
Which means there was no thermal shock to break the shunt, I assume simple mechanical/heat failure
And as a side note he says the extinguishing agent was likely class D powder to smother the active flames. Seems Like that could be a way to build a self extenguisher based on some retaining wire or thin rope with a bottom opening bin above the battery bank.
Sodium cloride based - Heat from the fire causes it to cake and form a crust excluding air and dissipating heat from burning metal.
Which means there was no thermal shock to break the shunt, I assume simple mechanical/heat failure
Last edited:
robbob2112
Doing more research, mosty harmless
Assuming you are right the insulation issue was between the battery and the breaker
robbob2112
Doing more research, mosty harmless
Someone better at guessing size from looking will have to take a stab it the size then
k490
Solar Enthusiast
The 2nd post shows he had 16 cells per shelf 2 rows of 8 if they are 280Ah batteries that about 12lbs each I'm rounding up that would be 192 lbs on each shelf. I have put well over 600 lbs. on 3/4 inch particle board shelves are built like these without any issue of collapse. I know the ones that are sold at Home Depot, and use to be sold at Costco both had 1200 lbs ratings. I don't believe these ratings are bogus they would be liable they are testing these shelves. I have even had oil soaked into mine had a 350 small block Chevy engine sitting on it for 5 years.
I think something got hot his circuit protection didn't work started a small fire that caught the wood above it once that glue in the wood starts to fuel the fire it gets hot quick.
I was thinking of using one of these shelves now I'm re-thinking this batteries need steel around them treat them like a lipo battery. Even with fuses and circuit breakers can still start a fire and draw less current then the circuit protection is rated for. My grandparents home burned down from a lamp plug got smashed behind some furniture. The 15 amp breaker never pulled enough to trip it.
Last edited:
OffGridForGood
Catch, make or grow everything you can.
OP is in Australia, his suppliers are not the same as your suppliers. the shelf material may be thinner - we don't know.
From other posts, OP has a long conversation about humidity forming in the garage affecting equipment - that doesn't mean water or humidity affecting these shelves but sure is concerning.
The dates of the OP's posts about purchasing the 93 550W panels are in 2022. Not sure about the cells. From the pics and postings he used Daly BMS - or at least one is Daly. I have had some issues with Daly and their factory settings are not ideal. others have posted issues too.
His Inverter appears to be an MPP 11kW 230 volt - and additional SCC MPP PM60's which are only 150 volt max. some discussions on 2s 4p strings
Can you imagine how many strings? - only 8 PV at a time to wire up all 93 panels?! and if this is how it ended up in the actual system the amperage would be something! We are not seeing the whole picture but starting to wonder about high amperage PV strings, and loss of control over the charging of a pack.
This could easily lead to an over-voltage situation, cell venting and ignition. if a shelf failed before or after the fire started is not clear, but seems unlikely to "just happen" out of the blue.
51kW of PV at 51.2 volts is 1,000 Amps!
even with ten packs, 100 Amp each.
now what happens if one or two of those 3P breakers trip - we see some tripped - that 1,000 Amps is going to fewer packs, and much higher amperage.
Also being southern hemisphere, they are in late summer conditions right now, not winter like those of us in the northern hemisphere. The garage may be hot to begin with, and then high amperage charging could add to that heat. OP notes a window A/C unit in one post.
I highly doubt the shunt is the cause.
Last edited:
This is what I used on all 6 wood sides and structure builds inside my solar trailer .. 35 buck a quart bottle …they say two coats are better than one….I used 7 coats….let it soak in and dry , then repeat as much as you want …
I tested it on some soft wood fir strips with two coats and could not get it to burn with a lighter , it would char but not support flame and smolderedright out as soon as the flame was removed..
I even cut some slices along the sides to give the fire a chance to grab hold ..no fire ..it’s on Amazon …….I used two bottles inside the trailer…
They have a quite impressive video of what effect it has on 2 dog houses soaked in gasoline then lit on fire …..the untreated one was destroyed ……The treated one went out as soon as the gas burned off…
Not sure it would help as much if the wire and plastic inside got ignited early on …it’s meant for wood and fabrics ( movie theater curtains) and stuff to retard active burning in places people congregate…
…no smell…non toxic….BUT DO NOT GET IT ON COPPER OR METAL GEAR AT ALL..TURNS IT GREEN IN A SHORT TIME…
I got some mist on some 4/0 copper lugs and they had serious heavy green spots the next day…had to 100 grit sand them off.
And NO …I HAVE NO VESTED INTREST IN THIS PRODUCT OR COMPANY…OR GET ANYTHING FOR SAYING I LIKE IT…….SO THERE…!
EDIT….. BTW ….IT IS PERFECTLY CLEAR…YA CANT SEE IT…
J.
Attachments
Last edited:
robbob2112
Doing more research, mosty harmless
Ok, I did the math for 5 batteries, 16 cells, 304ah = 24302ah
robbob2112
Doing more research, mosty harmless
I
I was never good at higher math like this, which is why I wanted someone else to do it. I can feel this being my next reading topic.
k490
Solar Enthusiast
OP said 5 rows of shelves each shelf had 2 battery banks that is 10 battery banks lets assume 304ah batteries wired in series to get 51.2 V each shelf grand total of 3040 amps. Around 145-150k watts of power. That's a lot of power in a garage .
Brucey
Solar Wizard
Testament to the relative safety of LFP cells, even with that much capacity, if that was NCM they'd prob still be trying to put the fire out.
Sverige
A Brit in Sweden
Was this wire passing through a hole in the metal frame at a point where the wood shelf would have been above, pressing down on the wire on the far side? If the wiring was routed in this way with pressure on the cabling from the shelves it would compress the insulation and over time with some thermal expansion too and fro could slice thru and cause a short from the cable to the metal frame.
Edit: looking again, maybe it’s not a hole in the frame? Kinda looks that way but possibly an optical illusion + me imagining things!
Edit: looking again, maybe it’s not a hole in the frame? Kinda looks that way but possibly an optical illusion + me imagining things!
OffGridForGood
Catch, make or grow everything you can.
I better do some checking on my own system today...just to be sure.
The more I look at the pictures the more I am thinking that the failure comes down to 2 areas. First off is construction of battery banks and shelving was not done to best practices. The second is simply environmental and neglect. Over time conditions that the batteries were stored in allowed corrosion. Components allowed to corrode increase resistance which creates heat. At some point this heat set off combustion of materials.
I do thank the OP for being willing to post this. It gives me impetus today to do a complete power down of my home to check my panels and connections. It has been more than a year since I did my last major upgrade and it is easy to not want to turn things off and simply let time pass.
I do thank the OP for being willing to post this. It gives me impetus today to do a complete power down of my home to check my panels and connections. It has been more than a year since I did my last major upgrade and it is easy to not want to turn things off and simply let time pass.
OffGridForGood
Catch, make or grow everything you can.
It is a little unclear but some of their other posts lead to thinking this is pretty recently set up, perhaps Oct-Nov 2023. Meaning it isn't old, and may not have been running (or not at full scale) for all that long. There are posts about buying 93 550W panels, but It is not clear how many were installed or if this was done in stages.
What is clear: this is a large system, and the OP says 'running 7kW load steady' - BCM system - that is 168kWh per 24hours.
What I have learned from the event:
Have a smoke/heat detection system in the ESS area, to alert you in the case of an event.
Use quality parts, equipment, crimps. Be sure.
Check on things on a schedule, use a laser-temperature gun to confirm temps of connections under high load conditions (charge & discharge)
Like a chain, a system is only as strong as it's weakest link: one failure can bring the whole thing down.
