So this happened today

[Sanwizard]

Everyone is talking about all the things someone needs to pay attention to when setting up a pack, except - as @Supervstech pointed out - this person had a fundamental flaw in understanding parallel connections. He/she had 6 5p "cells" and 2 1p "cells" with the 1p being on the most positive and most negative. Charging and discharging would have beat the crap out of those to single cells. On charging, they would have gotten cranked way, way above the 3.65V max. And on discharging I'm not even sure how far down they would go. Maybe to nearly 0. Whether or not that would cause a fire is interesting to speculate. The more appropriate point is that this person should not have been attempting to build a LiFePO4 battery. Period.

I don't mean to be judgmental. I just think that DIY for this is like DIY on lots of things: Many (most?) people maybe just should not try. I don't really believe we can cover everything that someone needs to know in a FAQ or DIY guide or book or anything.

Amen.

 

[FilterGuy]

My 2 cents:
After reading the whole thread I conclude there just is not enough info to make any kind of assertion about what happened.

Several people pointed out the lack of BMS.... but was the system operated that way?

@Supervstech noted a bus-bar oddity that would be worth looking closely at, but we don't know if the system in the before pic was they way it was run.

All we can say for sure from the data we have is that there was a fire in a structure that had a lot of cells. For all we know the wood stove caused the fire and the electrical system had nothing to do with it.

The OP says it is not his set-up, and no one on the forum has claimed it was theirs so we can't even ask basic questions. All we can do is look at the ugly pictures and speculate......And that is very unfortunate, if the system was involved in starting the fire, it would be an excellent opportunity for us all to learn and avoid similar mistakes. If the OP could invite the owner to the forum to help us understand the event it would be very helpful. The owner can even set up an anonymous log-in to protect his identity if he/she is embarrassed by the event.

 

[SlowBill]

it looks like there are also some 5P cells that are connected + to + and - to - with the next series 5P cells.
not sure which is + and which is - on these cells, but the string is not wired correctly in addition to the already mentioned first and last cell problem.

 

[plaidcabin]

My 2 cents:
After reading the whole thread I conclude there just is not enough info to make any kind of assertion about what happened.

Several people pointed out the lack of BMS.... but was the system operated that way?

@Supervstech noted a bus-bar oddity that would be worth looking closely at, but we don't know if the system in the before pic was they way it was run.

All we can say for sure from the data we have is that there was a fire in a structure that had a lot of cells. For all we know the wood stove caused the fire and the electrical system had nothing to do with it.

The OP says it is not his set-up, and no one on the forum has claimed it was theirs so we can't even ask basic questions. All we can do is look at the ugly pictures and speculate......And that is very unfortunate, if the system was involved in starting the fire, it would be an excellent opportunity for us all to learn and avoid similar mistakes. If the OP could invite the owner to the forum to help us understand the event it would be very helpful. The owner can even set up an anonymous log-in to protect his identity if he/she is embarrassed by the event.

This seem as good a comment to address as any. I have invited the owner of the burnt trailer to the forum, and indeed he has browsed here before and was inspired by Will Prowse's youtube channel to get into Solar Power and LiFePo4 batteries. He is not an electrician by any means, as the fine members here have deduced. The main thing preventing him from posting here is that his laptop was among the items destroyed by fire.

The wood stove was not in use. It was during a bright sunny day in August, and his Solar Panel array was likely generating a boatload of power when this happened. He has about 10 335w mono panels on the adjacent structures roof

I do know for sure that he was not using BMS. This had previously been a point of contention between us, as I bought new cells and Daly BMS units for my system, and he bought several crates of new old stock batteries for some cancelled power project and has been disassembling them and using the cells for various projects. They are stored at nearly 0V in the crates, and apparently have never been charged before. He still has many more of these. They have been working for him for some time in this manner, but as stated in OP - this was a new battery assembly and there were likely mistakes made.

As for the legal concerns: The trailer was not insured. The contents and structure are a total loss. There is no investigation, as there was no crime. Just a stupid, preventable accident, and we are glad that no one was hurt. The Fire Department released the scene and are taking no further actions.

All things considered, he's taking it rather well - and will be much more careful in the future. No time to dwell on the mistakes though, he has a lot of work to do now before winter returns!

 

[FilterGuy]

it looks like there are also some 5P cells that are connected + to + and - to - with the next series 5P cells.
not sure which is + and which is - on these cells, but the string is not wired correctly in addition to the already mentioned first and last cell problem.
View attachment 60285

Are you basing that on the color of the studs (Black vs silver)? If so, is that a reliable indicator of polarity?

 

[mrzed001]

it looks like there are also some 5P cells that are connected + to + and - to - with the next series 5P cells.
not sure which is + and which is - on these cells, but the string is not wired correctly in addition to the already mentioned first and last cell problem.
View attachment 60285

Another hawk eye ... good catch

Are you basing that on the color of the studs (Black vs silver)? If so, is that a reliable indicator of polarity?

The terminals have silver and black color under the busbars.

 

[DerpsyDoodler]

LFP is safe. The haphazard way some people use the technology is what's not safe. A 5p16s battery is like playing Russian Roulette. A 5p16s battery without a BMS is like playing Russian Roulette with a bullet in all six cylinders! Murphy never sleeps.

eh, a bullet in all 6 cylinders? That's suicide, not RR....

 

[DerpsyDoodler]

Everyone is talking about all the things someone needs to pay attention to when setting up a pack, except - as @Supervstech pointed out - this person had a fundamental flaw in understanding parallel connections. He/she had 6 5p "cells" and 2 1p "cells" with the 1p being on the most positive and most negative. Charging and discharging would have beat the crap out of those to single cells. On charging, they would have gotten cranked way, way above the 3.65V max. And on discharging I'm not even sure how far down they would go. Maybe to nearly 0. Whether or not that would cause a fire is interesting to speculate. The more appropriate point is that this person should not have been attempting to build a LiFePO4 battery. Period.

I don't mean to be judgmental. I just think that DIY for this is like DIY on lots of things: Many (most?) people maybe just should not try. I don't really believe we can cover everything that someone needs to know in a FAQ or DIY guide or book or anything.

If that's true, then sites like this shouldn't exist. Personally, I'm glad it does and I don't think ignorance should be allowed to squash that. Perhaps a huge disclaimer on the landing page:

"WARNING: DIY Electrical can be hazardous to your health. Use this site at your own risk."

 

[DerpsyDoodler]

This seem as good a comment to address as any. I have invited the owner of the burnt trailer to the forum, and indeed he has browsed here before and was inspired by Will Prowse's youtube channel to get into Solar Power and LiFePo4 batteries. He is not an electrician by any means, as the fine members here have deduced. The main thing preventing him from posting here is that his laptop was among the items destroyed by fire.

The wood stove was not in use. It was during a bright sunny day in August, and his Solar Panel array was likely generating a boatload of power when this happened. He has about 10 335w mono panels on the adjacent structures roof

I do know for sure that he was not using BMS. This had previously been a point of contention between us, as I bought new cells and Daly BMS units for my system, and he bought several crates of new old stock batteries for some cancelled power project and has been disassembling them and using the cells for various projects. They are stored at nearly 0V in the crates, and apparently have never been charged before. He still has many more of these. They have been working for him for some time in this manner, but as stated in OP - this was a new battery assembly and there were likely mistakes made.

As for the legal concerns: The trailer was not insured. The contents and structure are a total loss. There is no investigation, as there was no crime. Just a stupid, preventable accident, and we are glad that no one was hurt. The Fire Department released the scene and are taking no further actions.

All things considered, he's taking it rather well - and will be much more careful in the future. No time to dwell on the mistakes though, he has a lot of work to do now before winter returns!

Glad to see this update and to know that the person whom this incident most affected is gathering their wits and moving on/forward. Even if it's later down the road, it would be awesome to have him come here and detail the build/answer questions.

 

[Supervstech]

it looks like there are also some 5P cells that are connected + to + and - to - with the next series 5P cells.
not sure which is + and which is - on these cells, but the string is not wired correctly in addition to the already mentioned first and last cell problem.
View attachment 60285

I have studied the picture, and I agree… this also was done… there MUST have been sparks galore when this was wired… cells self discharging into each other, and the first last set 1P… ugh… those poor cells.

 

[A.Justice]

Look at the before picture… notice where he has the cables connected to the cells… there is a space between the terminal with the wire and the rest of the paralleled cells…
So his 5P set is 5P on all the cells in the pack, except the first set in series and the last set in series… meaning the main pack is 400Ah, and the first and last cells in the pack are 80Ah… they didn’t stand a chance of surviving without a BMS…

What, you've never seen a 7s1p5p10p8s pack before? ?

Sorry, fires aren't funny, and I feel for the owner, but that is a terrible setup.

 

[DerpsyDoodler]

I have studied the picture, and I agree… this also was done… there MUST have been sparks galore when this was wired… cells self discharging into each other, and the first last set 1P… ugh… those poor cells.

If that were the case, no way he could've made it so far as to have it charging. Looking at those pictures i can't confidently say anything about anything. Nor would I even try to figure it out without using photo software to analyze the pictures and markup to make notes while doing it.

I hope op makes it here, at some point.

 

[upnorthandpersonal]

So every step every best practice, all over the place.
It needs all to be put together like a book (and that needs a table of contents). To organize all. To have a model of a perfect build.

That was the idea behind the wiki...

 

[plaidcabin]

I have studied the picture, and I agree… this also was done… there MUST have been sparks galore when this was wired… cells self discharging into each other, and the first last set 1P… ugh… those poor cells.

Stored at 0v, never been charged before.

 

[mrzed001]

That was the idea behind the wiki...

I checked it one time and it seemed to be empty (with only 2 items in the navigation).
It took me now 2 minutes to find the little Pages hypertext link on the top left corner to see all (OK I am slow today ).
It is a good idea, but not a very user friendly interface.

Maybe a treeview hierarchical structure ? That you can open down (like a file directory).
Where the tree leaves order also suggest the order the user should walk trough

Battery
- DiY from Cells
- - Before buying Cells
- - Cell buying options
- - Before building a pack
- - - Cell identify
- - - Cell testing
- - - Cell top balance (not optional)
- - Building a pack
- - - Cell placement/cabinet
- - - Cell separation
- - - Cell compression
- - - Cell order series and parallel
- - - Busbar options and todo when connecting the cells
- - - - Seller given nickel plated copper
- - - - User made flexible cable
- - - - Aluminium busbars
- Manufactured like 12V battery
- - Battery buying options
- - Battery pack build
- - - Battery placement
- - - Battery order series and parallel
- - - Battery connectors, cables

BMS (not optional)
- Types
- - FET based
- - Relay based
- Balance types
- - passive balance
- - active balance
- Before buying a BMS
- Buying options
- BMS to battery build
- - BMS placement
- - BMS cable order
- - BMS cable management
- How to start the BMS
- How to monitor the BMS
- BMS connections like RS485

etc ...

 

[upnorthandpersonal]

We're limited by the plug-in for the forum software and really don't have a lot of available resources to customize and maintain changes.
I'll see what's possible...

 

[NwCali]

eh, a bullet in all 6 cylinders? That's suicide, not RR....

Or like playing RR with a magazine fully loaded...or...well..if a magazine, with one bullet in it.

 

[mrzed001]

We're limited by the plug-in for the forum software and really don't have a lot of available resources to customize and maintain changes.
I'll see what's possible...

OK. If a html page can be called or inserted as an iframe that works too.
A hierarchy treeview is simple: DIVs inside of DIVs.
On DIV Onclick on every leaf set first child (that holds al the DIVs for the sub level) display to block from none (default style="display:none;" set to style="display:block;").
Only silly static html.

 

[circus]

it looks like there are also some 5P cells that are connected + to + and - to -

He probably would have caught one mistake at the end of the string but two mistakes made the ends come out right. My gut feeling says it just wouldn't work, not start a fire. We should work out the math on how much juice could flow through the tiniest bottleneck. If even possible, how much would it take and what would happen if you forced flow backward through cells. What reverse voltage (charging) did those cells see?

 

[upnorthandpersonal]

OK. If a html page can be called or inserted as an iframe that works too.
A hierarchy treeview is simple: DIVs inside of DIVs.
On DIV Onclick on every leaf set first child (that holds al the DIVs for the sub level) display to block from none (default style="display:none;" set to style="display:block;").
Only silly static html.

Yes, that could work. The Wiki has a HTML editor and that one should support inline CSS.