Some new information.
One of my Inverters died today.
The inverter directly above the fire.
Naturally I cleaned the inverters inside after the fire, mostly due the sand/clay dust that came with putting out the fire with sand.
Pressurised air and a dry, unused paintbrush, inside and out, disassembled all components and put humpty Dumpty back together.
How many weeks since the fire? About a month?
Picture from internal:
Rust.
No, not from water, the other inverters have NOT a rust problem.
I can share more pictures of Rusty screws, bolts and nuts, this is clear enough.
It also (for me) absolutely confirms the reason of failure, self discharge of the cell, probably failure from cascade of micro errors.
This inverter was above the fire, and the smoke detector is directly above there.
Smoke will have passed via the inverter, probably (almost) all, as it's push-pull 2*2 120mm fans...
After the smoke detector alarm went off, I stoped all electricity with one disconnect (contactor based for solar and battery
. )
The smoke was acid.
Electrolyte apparently is acid chemical stuff that when it does vent/ go in fire, that smoke is dangerous for all electronics.
I would not advise to breath it either.
Putting out the fire was not fun, mainly due the chemical smoke.
Not smoke from melted wires or wood or paint or.... Never smelled anything like that before, but felt really toxic.
Sure, I have COPD, so smoke quickly makes breathing a problem for me.
This was different.
Chemical, bad, bad, stuff.
@HighTechLab your cell fire looked like mine, the leaking flames.
I don't know if you smelled the smoke, but probably can confirm, it is nasty.
I really tried not to react on
@mrzed001 as I feel he just likes to make a quick score. It keeps bugging me so.... :
With your comment that no sheets between is an "error" or fault, you also make a joke of
@Will Prowse in your eyes. What video do you see Will using separator sheets??
Perhaps knowledge you obtain recently??
All people who have no separator sheets are in direct danger?
And yes, apparently you didn't read the thread.
Some people say it's crazy dangerous to have LiFePO4 cells inside the house.
The safe type lithium.
Yet, they have the Unsafe lithium by their bed side.......
What explains for you the video of explosion of phone battery.
I doubt it's explanation was needed for others who have read the thread and don't do cherry picking response.
Why agressieve towards your post?
As it feels lik you just try to score and point to blame.
Not being productive or supportive at all in that post.
I doubt you even know how to top balance, in your idea cells reach 3.4v are fully charged...
Try to charge a few cells yourself and you know that's not realistic.
Discharge test will reveal 5-10% imbalance between the cells.
Only really fully charged cell is a cell a few hours @ 3.65v
I doubt that is worth any dispute.
But welcome to do this via PM.
We will make together a new thread based on the outcome to explain the forum and world when a LiFePO4 cell is 100% SOC.
Perhaps several days on 3.4 might get it higher charged and more close together.
Yes, a laserwelder would be nice also..... Unobtainable.
Off grid = no electricity except what you make. About 10 hours solar.
Not 24 hours grid....
I can't bent space and time.
For those who can..
Have the environment that makes almost anything possible, it's easy
.
Most of us need to work with limited access to tools, time, electricity, financial funds, etc.
Now that is of my chest...
Back to the inverter.
Yes, failed cell.
Why, many small errors, starting with too many sheets in a housing, failed BMS (active balancer BMS, worked 3 days), not compress the cells, damaged threads due a few remounts with the screws of the seller.
(One time build would be not a problem, if also the BMS would not have failed, etc etc etc)
No torque meter to tighten the nuts equally over the whole pack.
And relocation, not as one unit but cells.
What wasn't an error
- no sheets between.
- wires.
That it is a lot doesn't make it unsafe
- my choice of clamping BMS wires between nuts instead of on the bus-bar. With reason as BMS do fail and lugs need to be replaced, at the moment threads in the weak aluminium will fail...
Not optimal, sure, sadly only laserweld is optimal.
Perhaps an error:
- thread repair. Not the way of making, it should not be done at all in the first place, drilling a hole., Tapping threads.
It does place stress on the thermal, as cutting threads uses force.
Even with oil, it is a stressing factor.
The seller or garage workshop that cut the threads did not use oil, (when you make close up you can see the difference between with and without oil thread cutting)
That's already "damaged" before it's installed.
Cascade of micro errors, not all made by me.
- using the screws provided...
Failure?? Will is using them. Without different information, it's not an error or fault.
It just turns out not to work as expected.
- buying China BMS (low quality like Daly, Jikong, where low quality is debatable, but... They do fail.
Fail lot more then electrodacus, Batrium, Orion or other high quality BMS.)
With failure, nuts untighten or screw removed = protential thread loss.
For all accounts:
If I would have known better, I would have done better.
Doing something that turns out not to work is not making an error.
Edison did not make 1000 errors before the light kept burning for a few seconds.
Error is negative, like making a fault.
Mistakes where made, and shared
For mine and your entertainment, and sometimes educational.
Pointing a blaim or trying to say something is stupid....
That doesn't help anyone.
Just an easy score.
Failure is part of life, that how most of us learn.
Sometimes funny, like a kitten first steps, falling, trying and at the end jumping and running around.
DIY LiFePO4 is new.
Especially in the high ampere regions.
What was true for 20Ah cells, they came standard with thread, isn't true for 200Ah cells.
20A cells won't standard face 40A, C0.2 charge,
or 0.5, 100A
Thinking that the challenge for 20A cells is the same as 200A...
Is unwise...
Like a 10A fuse placed in an 100A setup and expect that it will work fine with 100A loads.
February 2020 was my start of LiFePO4 adventure.
Anyone who wants to point, should do that with the available knowledge on that time, in this forum.
Some begin installations of Will might now be seen as Unsafe.
His knowledge also grows.
The ones who like easy scores should comment on those early videos.
Always easy score looking in the past with knowledge of today.
Just remember, the knowledge you have today is there because of the past.