Last fire.. :-(

Supervstech

Administrator
Staff member
Moderator
Here you are absolutely wrong.
Not ZERO reason for the wires to be placed like that.
...

Some people like an easy score.
Those people frustrate me, and are the reason many people opt out sharing their mistakes, what is the most educational!

I didn't think/expect @Supervstech is one of them!!


If you don't understand what you see, ASK!
please don’t misunderstand me. I did ask for high resolution pictures. Electricity is a wild and dangerous thing.
I would really like to look everything over and see where all the wiring goes.
you may not think the mass of wires could possibly be the culprit, but it could. 1.5A on balance wires, Cris’s crossed around, and you have heat concentrated on overlapping conductor insulation. Heat could eventually allow insulation issues, and failures can happen. Imagine the BMS thinking one cell was at cutout, and trying to balance when the cell actually wasn’t.
as you pointed out in the first post… professional built and factory BMS makes all the difference. Welded contacts, and perfect banding make for a very reliable pack. Something we mortals just cannot handle at home.

Don’t think I am stating the spaghetti IS the issue… but it very well could have been a part that instigated the failure.
obviously, if the high res pictures don’t show anything, looking at it post Mortum isn’t going to reveal much…

I’m not trying to blame you for the fire… I’m trying to impress upon every member reading this thread how little it takes to go wrong and cause serious issues.
 

fhorst

making mistakes so you don't have to...
The Line " there is zero reason" isn't hard to misunderstand.

No BMS yet.
And, even if there would be at that time, as you could read, it was empty, 48v stopped by the inverter, and just a short time started with 49v total voltage, early in the morning, clouded rainy day +/- 09.00.
It received a few hundred watts charge.

On a PCB, sure.
We humans can't solder as nice as a robot can.
2 wires together isn't done by a robot.
And with shrink-wrap, almost as safe as original wire coating.

You probably didn't read wel, I worked extremely careful.
Made absolute sure no wire had bare contact anywhere.
No lose contacts, nothing twisted or folded.

That 160 wires look like a mess, sure.

On the picture you see the 24 X 152Ah cells, 24* 4 almost 100 wires.

IMG_20210706_113928_907.jpg
This is one set of 4 wires I'm holding up.

Other picture:
IMG_20210706_113854_096.jpg
Just because it's more then 17 wires doesn't mean it's not safely done!

BMS plastic white plug and my plastic white plugs are same safe.

Just not 17 but 64 wires per 16 cells.
And not one big plug but 32x for each 2 wire.

My soldering qualities:
Oven baked + soldering iron.
2021-7-26 11-0-38.jpg


Daly soldering
2021-7-26 11-1-56.jpg

3.2v, max 1.5A. 5 watts.
AWG 22 tinned silicone wire, that self fuses at 10A, if there is a short.

Sure 1 or more wires can always make a short.
They can get damaged by animal, accidents do happen.

99.8% sure that the wires I made where safe, totally safe during installation.

The BMS Cell units where not placed yet, not installed.

Monitoring via Jikong active balancer.

As I wrote half hour, one hour earlier I checked.
Voltage between 2.9 and 3.0v
Nothing special.

What did start the fire?

Most likely the cascade of issues during the 18 months, one cell decided to give up, self discharge and take the rest with it.

Why?
It was self discharge!!
Just fusing wires with silicone coating could never create enough heat to let other cells self discharge also.

They where not a pile of wires as shown in the picture anymore, they where "stretched", sets of 4 wires

Maximal 8 close together, where 4 are 2* + and 2* -

I can rest with one cell, to see how it would react.

I did test one wire on my bloated 280Ah cell.
That fused almost in the center, and contact stoped.
The silicone coating in tact.

The 0.2% from the 99.8, is a possible animal.
Making larger portions bare wire.
And perhaps building a nest, food for a fire.

Even then...
Before that is hot enough to make the cell self discharge...
 
Last edited:

400bird

Solar Enthusiast
I think you shouldn't insult others reading comprehension. Your posts are quite long, tent towards rambling, and do jump around quite a bit. (I also shouldn't point fingers as my posts can do the same)

Yes, you need lots of wires with that many temperature sensors and cells. But, it gives everyone a better impression of your build of the wiring is tidy. Clean wiring also wouldn't bring doubt on the spaghetti.

And, the BMS not being connected doesn't matter in my mind. The BMS wiring was. It is connected to the cells and is a possible cause.

There is also the possibility that it was an manufacturing defect from EVE. There are so many possibilities, I think everyone here is trying to help both you and anyone else working through a build.
 

Supervstech

Administrator
Staff member
Moderator
The Line " there is zero reason" isn't hard to misunderstand.

No BMS yet.
And, even if there would be at that time, as you could read, it was empty, 48v stopped by the inverter, and just a short time started with 49v total voltage, early in the morning, clouded rainy day +/- 09.00.
It received a few hundred watts charge.

On a PCB, sure.
We humans can't solder as nice as a robot can.
2 wires together isn't done by a robot.
And with shrink-wrap, almost as safe as original wire coating.

You probably didn't read wel, I worked extremely careful.
Made absolute sure no wire had bare contact anywhere.
No lose contacts, nothing twisted or folded.

That 160 wires look like a mess, sure.

On the picture you see the 24 X 152Ah cells, 24* 4 almost 100 wires.

View attachment 57659
View attachment 57659
One thing I think I see wrong with the build is difficult to see in the blurry pictures, but it appears like your balance wiring rings are sitting up on nuts, then locked down with additional nuts… is this the case?

Stainless steel is a terrible conductor. If the balance leads are not in DIRECT contact with the bussbars, they are sensing voltage through a resistor. Do you have any pictures showing the wiring in focus?

Please do not feel insulted or anything. We ALL want to know how to avoid a fire, and you have had two this year. That is scary, and if examining your build will help members build more safely I REALLY want to promote that.
please post images we can see in detail. It is clear you think everything is safe and I hope you are right, and we can blame the cells… but THOUSANDS of members are using this type of cell. Are they all getting ready to lose everything?
 

Will Prowse

Admin
Staff member
Moderator
I have to agree with Supervstech. I understand that you are mad that your pack caught on fire, and you are taking it personally when people question your wiring job.. but they have a point. Your wiring job does look dangerous.

Having that many wires touching each other and raw battery terminals is just asking for issues. They should be neatly organized, no matter how many cables you are dealing with.

Especially with an active balancer pushing high current for hours on end. Those conductors need to dissipate a certain amount of heat, and if they cannot do that, they will develop hotspots. These hotspots can melt insulation or cell insulation causing smoke.

I remember in the past that you had another fire occur. And most people here use the same cells that you use. I do not think the issue is the cells. I have piles and piles of these cells, and never had these issues. I think the problem is somewhere in that mess of wires.

Very good point about using stainless steel. Your lugs need to be copper and flush with the terminal.

I just cannot for the life of me understand people's logic with these active balancers. I don't get the point. Let them be out of balance and top balance the best you can. The liability of using an active balancer just does not seem justified.
 

fhorst

making mistakes so you don't have to...
Will, I'm not trying to pick a fight here, but you hardly have a voice / knowledge in this.

You charge your cells on daily basis to full.
Nothing wrong with that.

On the other hand, you advertise how people can get almost double the cycle count by not charge them 100% bit use 10-90% as charge space.

This effectively stops top balance.
You have no experience in how dis affects the imbalance, or even tried how a Daly 0.08A (total, not per cell) can balance the pack when you decide to charge full again.

We talked few times PM about this.

Easy to write that you don't understand the need for active balancer, if you don't want to use it.

For most of the post of supervstech, I agree with him.
His word choice that there is ZERO reason to make wires like this is dumb., Feels like easy score without knowing what he talks about.

And that part, I try to make him see the error of that.

Unless he knows an wireless way of making 4 contact per cell, 75cm or better 3 meter per cell, will look messy.
Not unsafe.

I did not order 30 meter coating to bundle the 4 wires nice together.
Local stores don't have it.

Visual it would have made it look more nice, for safety.. no difference.

Yes!
Nut on a nut!
And with absolute good reason.

You can try to bash me how that is a bad contact and bla bla bla.
Yet, measure Voltage on the studs is normal..
High resistance?
Nope. Higher then copper.
True.
Not high resistance.
I'm not trying to push-pull 100A.
Maximal 1.5A per cell.
5 watt.

Really??
Have you tried it?
Place a stainless steel stud between 2 wires and "blast" 5 whole watts via it?

Where you able to measure a thenth it a less temperature increase??

Please. Stay realistic.

Why I didn't not share this information sooner.

For this type of discussion.
Get realistic.

Look at your 12v (not 3.2v) 300w (not 5!) and see stainless steel bolts and studs as main contact to your sigaret lighter....

Apparently I'm the only one who has troubles with th strength of the grubscrew in the cells.
Making it a mandatory one time mount.
Having long enough grubscrew, to be able to add the BMS wires AFTER you installed the bus-bar is a really good way of working.

We "all" know BMS mostly don't live the lifetime of the LiFePO4 cells.
Especially the China ones.

I know at least 10 forum members who have one or more BMS mounts.

We are talking about sensing the voltage. That's what we do.
With my high sensitive battery meter I could not measure Voltage difference between the top of the stainless stud and the aluminium terminal.
We talk about 0.000x voltage
That sensitive.

Sure if you have more sensitive, there eventually will be difference.
I still need to see the BMS that uses 0.0001v as a significant number.

So please.

Stay realistic people.
In Dutch we call this " op alle slakken zout leggen" or
"spijkers op laag water zoeken" or
"Mieren neuken"

I don't know the English equivalent.

If this is the direction it needs to go.
Please close this thread for further discussion.

Exactly this is why many people don't share their failure.

@Will..
Once again.
Please, if you want to have any right of talking about active balancer and charging 10-90% ...
Do it!!!
And see how the nice Daly does not work getting the pack back in balance after 3 months.

2 sets.
One with and without active balancer, both Daly BMS 0.08A balance per setup

Then talk about how it does or doesn't make a difference.

Without it, it's just your personal preference, based on a failure of not connect a wiree that you Blaim to be the fault of the Balancer as no BMS would have a probelm with a lose wire.
That's right. It's not a BMS...

You last video as far as I remember about active balancer who you first thought worked fine..
Thousands of YouTube video show they do, unless you mess up..
Disconnect one wire :)

You know a lot, not this (active balancer) field, please educate and work with it before you make a statement.
 

Will Prowse

Admin
Staff member
Moderator
Will, I'm not trying to pick a fight here, but you hardly have a voice / knowledge in this.

You charge your cells on daily basis to full.
Nothing wrong with that.

On the other hand, you advertise how people can get almost double the cycle count by not charge them 100% bit use 10-90% as charge space.

This effectively stops top balance.
You have no experience in how dis affects the imbalance, or even tried how a Daly 0.08A (total, not per cell) can balance the pack when you decide to charge full again.

We talked few times PM about this.

Easy to write that you don't understand the need for active balancer, if you don't want to use it.

For most of the post of supervstech, I agree with him.
His word choice that there is ZERO reason to make wires like this is dumb., Feels like easy score without knowing what he talks about.

And that part, I try to make him see the error of that.

Unless he knows an wireless way of making 4 contact per cell, 75cm or better 3 meter per cell, will look messy.
Not unsafe.

I did not order 30 meter coating to bundle the 4 wires nice together.
Local stores don't have it.

Visual it would have made it look more nice, for safety.. no difference.

Yes!
Nut on a nut!
And with absolute good reason.

You can try to bash me how that is a bad contact and bla bla bla.
Yet, measure Voltage on the studs is normal..
High resistance?
Nope. Higher then copper.
True.
Not high resistance.
I'm not trying to push-pull 100A.
Maximal 1.5A per cell.
5 watt.

Really??
Have you tried it?
Place a stainless steel stud between 2 wires and "blast" 5 whole watts via it?

Where you able to measure a thenth it a less temperature increase??

Please. Stay realistic.

Why I didn't not share this information sooner.

For this type of discussion.
Get realistic.

Look at your 12v (not 3.2v) 300w (not 5!) and see stainless steel bolts and studs as main contact to your sigaret lighter....

Apparently I'm the only one who has troubles with th strength of the grubscrew in the cells.
Making it a mandatory one time mount.
Having long enough grubscrew, to be able to add the BMS wires AFTER you installed the bus-bar is a really good way of working.

We "all" know BMS mostly don't live the lifetime of the LiFePO4 cells.
Especially the China ones.

I know at least 10 forum members who have one or more BMS mounts.

We are talking about sensing the voltage. That's what we do.
With my high sensitive battery meter I could not measure Voltage difference between the top of the stainless stud and the aluminium terminal.
We talk about 0.000x voltage
That sensitive.

Sure if you have more sensitive, there eventually will be difference.
I still need to see the BMS that uses 0.0001v as a significant number.

So please.

Stay realistic people.
In Dutch we call this " op alle slakken zout leggen" or
"spijkers op laag water zoeken" or
"Mieren neuken"

I don't know the English equivalent.

If this is the direction it needs to go.
Please close this thread for further discussion.

Exactly this is why many people don't share their failure.

@Will..
Once again.
Please, if you want to have any right of talking about active balancer and charging 10-90% ...
Do it!!!
And see how the nice Daly does not work getting the pack back in balance after 3 months.

2 sets.
One with and without active balancer, both Daly BMS 0.08A balance per setup

Then talk about how it does or doesn't make a difference.

Without it, it's just your personal preference, based on a failure of not connect a wiree that you Blaim to be the fault of the Balancer as no BMS would have a probelm with a lose wire.
That's right. It's not a BMS...

You last video as far as I remember about active balancer who you first thought worked fine..
Thousands of YouTube video show they do, unless you mess up..
Disconnect one wire :)

You know a lot, not this (active balancer) field, please educate and work with it before you make a statement.

I do understand active balancer circuits and why people use them. And I will say it again, it is not worth the trouble.

So your argument for which you wish to push here is that you wish to cycle between 10-90%. The typical reasoning is for increasing cycle life of your pack, no? If that is the case, and you are scared that excessive cell drift will occur and lower the life of your pack, and that an active balancer will prevent this, so that you can theoretically (from most literature available), double or triple your cycle life to 80%? In this instance, I still do not think it is worth it. The heat stability of LiFePO4 is quite high, especially compared to NMC and other cobalt based chemistries. Most cycling bandwidth studies pertain to NMC, from what I have found. This chemistry is very sensitive to high and excessively low SOC. The degradation when cycling 0-100% is quite high. Some cells are rated to 500 cycles for this reason. If you wish to modify your cycle bandwidth for this reason, so be it. But I do not agree that you need an active balancer. Balancing is not an issue here. What you should probably do instead is modify your balancing trigger voltage. Because if your absorption is lower than the balancing trigger voltage, the cells won't balance. This is why most people on this forum run into this issue. If you change these settings and do a solid top balance, they will not deviate at all.

If cell drift is excessive, but you are not pushing the C rate limit of a specified cell, then you probably have mismatched cells. In this instance, a solid top balance will fix this issue.

If you are pushing the C rate of your cell for your chemistry and cell design, and excessive drift is occurring because there is a deviation in internal resistance values across your pack, and one cell is dissipating heat faster than the others, then an active balancer may be warranted. Such as a lipoly pack, or LTO. But LiFePO4? Absolutely not. If you are running these for solar, the highest C rate is .2 for most systems.

Ever notice that hybrid cars quickly cycle with modified micro cycles every few minutes, and they do not go out of balance for years, and they are not using an active balancer. I could argue, and through my personal experience pushing 100C through lipoly packs, that again, this is unnecessary. Balance charging would be most appropriate for this specific application. Which is obviously what people do with these packs in multiple industries.

The next possible reason for using an active balancer is because the cells are heavily degraded, and the capacity does not match at all. In this instance, you can hide your problem with an active balancer. It will not solve the issue. And the degradation rate will not change. And you will not get any more capacity during a cycle. I have done this test a few times. I did .1C and used 60/100Ah cells in series to see what would happen. A high quality active balancer would add only a single amp hour. Not 40. So in this instance, the active balancer is a waste of money.

Professionally designed, UL listed packs do not have active balancers. They use passive dissipative balancers and they work great. Would I trust these engineers that are conforming to strict regulations, or someone online who has caused multiple electrical fires? Oh, but you argue that you are different, and that you are trying to increase the cycle life by modifying the cycling limits. And in this instance, I still disagree.

I understand that everyone today seems to be milking these LiFePO4 cells to last forever, but why? They are cheap and they can handle the abuse. Push them to their thermal limits, and you will still get more cycles than you know what to do with.


You have other battery builders here that have never had a fire, and have systems that 100X better than yours, and you do not want to hear what they have to say. Your wires are a mess, and in a quest to diagnosis specifically what happened, you need to clean it up. How could you lose by learning what others have to say?
 
Last edited:

fhorst

making mistakes so you don't have to...
Thousands of members are using the cells.
Just a few use the 152Ah cells
The ones who do, I guess I'm the only one with 48 of then.

They are no longer build.
The 120 and 135 are.
Guess why.

At that time it seemed like a good buy.
Sadly pushing 152 in 120Ah housing does has a limit.

Like pushing 350ah in the 280Ah, by simply adding more sheets.

You might be able to build it
And in laser welded casing and bus-bar, it probably will work.

In the "out side of specification" installation we do, it doesn't.

Making parallel is increased difficulty, a lot.

I can only advice people to buy more BMS and build single strand.

2 sells parallel is doable.
3 is really pushing the edge.

And that the cells are already "edge"..

You have the recipe for my troubles.

18 months ago no one talked about not to use the screws, use grubscrew.

Compression was also something a few talked about, nothing serious yet.

And tape a few rolls, isn't compression :)

Sure, I'm sad that I lost my cells.
Also relieved.

After a few mishaps your walking on quicksand, it's not strong anymore and failures cascade.
39 of 160 terminals needed new threads...
That's my quicksand.

Many have seen my failures, and learned from it
Helping them not to make the same "mistakes".

Is it a mistake to use the screws provided by the seller?
Or thinking that you can use the thread 10 times?

Especially... New, uncharted territory.

Now we know.

I wish everyone loads of success.
My build is finished.
And I truly wish it would have been available 18 months ago.

Open your eyes and see what you are playing with isn't build or intended for this.

Yes, it can work.
One installation Only.
No playing around.
Not change location that need to redo the bus-bar and no BMS failure!!
 

fhorst

making mistakes so you don't have to...
I might be wrong, reading about the exceptional increase in lifecycle only on the forum, but think to remember getting this information from one of your videos.

If I could afford it and send you a set of 16* 152Ah Eve cells, you will see how "high quality brand" does have its misfits.

And sadly, I don't have to money to just push them aside as they behave absolutely not the same as my 280Ah cells did.

Day and night difference.
2 different sellers the 152Ah.
A year after purchase I talked about them about it, where the informed me that the 152Ah are "a bad batch", I got unlucky.

End of the line, I do toss them away, and have a really good high quality replacement with a build that us DIY can only dream of.

Will, you have had your share of issues, and no longer financial challenged.

If you need to provide your self and family with electricity, during your time in RV, and had the leaking cell trouble as you had in one of your last video's....
You probably would have used all, except the leaking one.

Or other video, the cells slightly bloated.
If you don't have money to replace.
And alterative is no electricity.
(Start of your RV time)
Be honest, you would really have tried to make it work, even knowing that it's not the strong base you hoped for, but it's all you got.

I doubt you have those cells installed now. Not the rest of the leaking pack, not the slightly bloated ones.

For the simple reason, you have alterative and can afford to get different/ new / less challenging cells.

With me, many (not USA) people around the globe who simply have months/ years financial issues to replace. (Probably also a lot in the USA)

Why I write not USA?
I've seen the average income numbers, and while several costs can be high also.
$300 is "nothing" where more then half of the globe has less for 2 weeks, 6 days a week, 10 hours a day work.

No offense, and I'm happy for you all.
I'm also not complaining with my incapacitated payments.
But doesn't come close to the "additional funds" the average USA person has to spend.

Trow away is easy to say of you can afford.
If you can't...
Well.
Then your building like me on quicksand.

I don't know how many of the thousands have this same situation.
And I hope for them that they don't run into the troubles I did.

Probably going with 120Ah instead of 152, would have made this story a lot different.

Also buying the latest version of BMS... Happy puppy for 3 days.
Now 18 months later, I have the same Jikong.. grown up, no longer pre-mature..

Bad luck magnet :)
 
Last edited:

Mcgivor

Solar Addict
Simplicity has a quality all of it own, keeping the amount of contact points down to a minimum reduces the potential of something going wrong, particularly with active wiring carrying current.

Food for thought, in the last 5 years my pickup has lost an injector, turn signal, wiper motor fuel filler door. My tractor, fuel gauge, starter motor and headlights. Solar system, array feeder cable short circuit, inverter failure twice,charge controller failure and BMS shutdown. The cause of every fault was due to rodents, wit the exception of inverter once by a gecko, then by lightning, lightning for charge controller.

Thailand is tough on equipment especially in rural settings, my equipment has been rodent/gecko proofed to minimize problems, which although simple to repair in most cases, are an annoyance.

Having excessive, and in my personal opinion, unnecessary wiring which ballancers introduce, move away from the simplicity theroy. Somehow my cells have manage to stay ballanced without intervention for 3 years, but to each their own.

The likelihood of rodents being involved directly or indirectly is probably pretty high, based on my experience in Thailand, it's part of life here cover any opening where they would have a tendency to go, with stainless mosquito mesh, also keeps out the geckos.
 

coolbz

New Member
Simplicity has a quality all of it own, keeping the amount of contact points down to a minimum reduces the potential of something going wrong, particularly with active wiring carrying current.
Having excessive, and in my personal opinion, unnecessary wiring which ballancers introduce, move away from the simplicity theroy. Somehow my cells have manage to stay ballanced without intervention for 3 years, but to each their own.

I agree with you. Simplicity could avoid many complicated issues. I'm thinking the possibility of liquid (splashed, condensed, etc) or insects drop onto the exposed battery packs and wires, or compression rods rub/cut wires, causing shorts.
 

fhorst

making mistakes so you don't have to...
:)

Yes.

That is why I went from 1 BMS setup towards 2 BMS and finally one BMS per string.

Failure of Jikong, 250A Dumb Daly and 150A smart Daly (sadly... Again first delivery, full of bugs, 3 major hardware changes I know of since then)

Have left me for a long long time with me as BMS and BT active balancer to get information about the cell voktage.
Doable, easy, just keep it between 15-85 and the risk of over (dis) charge is minimal.

In the meantime build DIYBMS.
WiFi, and contactor based.
I'm still going to use it, as secondary BMS combined with the Jikong.

Set balance to 3.70 (the Jikong does all the work) and only in emergency, the DIYBMS will burn off excess.

Contractors control the solar charge and battery connection.

One of the reasons to be able to quickly stop the charge / discharge, push my emergency button, all contactors release.

DIYBMS will do this automated in case of fire.

Simplicity.
I am probably again the only one with the problem, having a lug under a nut, tighting the nut will twist my lug!
Especially BMS wires.
Simple solution: place it on top of the nut, and tighten it enough (but not 4Nm) for reading the voltage.

Perhaps I'm too simple for this.

Adding a washer helps, but..
A lug is a washer, with a wire.
Everywhere you read, having 2 or more washers is BAD news.

Geckos are lovely creatures.
And yes, can't have them in electronics :)
They love the warmth..

We mostly use aluminium mosquito mesh, but those buggers are persistent.
555 (for non Thai, that's LOL) number 5 is pronounced "Ha"

Animals, heat and moisture, clay dust...(who leaves deposit on damp surfaces, and is hydroscopic, wil become wet clay during rain season)

We lost one airconditioning blower, totally out of balance till leaves snapped, clay dust finding a nice home.

Ants seem to like inkjet printer with flatbed scanner, perfect for hatching eggs...

It did continue to work for a few weeks, then without visual damage..
No longer.

Thailand is great.
Living off grid is amazing, many challenges.
Even simpler things as drinking water.

Combine this with my "lucky magnet"
This even has been my job for about a year.
New software, ready for launch, last test... Me 🤣🤣
And so many times back to development.
 

Short_Shot

Photon Sorcerer
Not here to comment on any cause of the failure, but just the wiring in general.

Looking at the photo, my biggest concern is not the rats nest of wiring so much as the fact that many of the wires lay across bus bars and terminals all over the place.

If one decides to "self fuse" as you put it or get chewed on where is sits on top of a bus bar or terminal then it'll just immediately be exposed to this bus bar or terminal and cause a secondary short circuit condition.

If this happens to occur between two cells and it just so happens the voltage difference low enough that the current stays under 10 amps.... would it be possible then to overcharge and rupture one of the cells as a result? Maybe. There's too many variables to diagnose that.
 

fhorst

making mistakes so you don't have to...
As you might notice In the later pictures, this is what it was during time of fire :

IMG_20210725_101520.jpg

No wires laying over Bus-bars, going "straight up" to the next level.
 

Short_Shot

Photon Sorcerer
I see several wires touching studs in multiple locations.

I'm not saying that was or was not the cause in this case. I'm simply pointing out that they should be organized more neatly as best practice.

However it seems odd that it was in service at the time, as you claim, and this shows an incomplete battery pack without any connections between cells.
 

fhorst

making mistakes so you don't have to...
:)

2 layers.

First layer 16x 280 Ah and 16x 152Ah share the same main bus-bar, provide power for the system for long time.

In that time, build wires and add them, not by release bus-bar, add on top, as it's 99.8% of the time just sensor for voltage.

These wires have plugs.
This type:
2021-7-26 16-12-34.jpg
When they didn't went up, looks like rat nest.
IMG_20210706_113854_096.jpg

Placing "shelf" for second layer:
IMG_20210725_101500_edit_85423254815089.jpg

And end result on the 280's
(I did not make an overall picture, I'm documenting some, not all I'm doing!)
IMG_20210725_101520.jpg

The top of the 152's looked the same as this.

And, the top did not have a problem.

Location of the red dot is where I think the fire started.

That Cell looks like this now:
2021-7-16 14-18-44_copy_270x522.jpg

The rest looks 'normal'

Also, this is the location where the slow/calm flames came from.

I expect burning Ox-guard, that self egnite at 400c or so.
But can be a flame from the cell as I have seen in some you tube videos.

To kill it all:
IMG_20210715_111429.jpg
Sand :)
Many many buckets.
36 hours later, several cells on the bottom layer too hot to touch by hand...
 

fhorst

making mistakes so you don't have to...
IMG_20210726_164705.jpgIMG_20210726_164710.jpgIMG_20210726_164720.jpg

I wonder if lithium is worth something for recycling.

One way to easily know if a cell is defective/ dangerous to use:
Shake it.
If you hear any liquid sounds..
Like water bottle..

Just don't use them.
Even if they still have charge.
The ones with bursted vent are 0.0v,
Othera between 0.2, many 2.0 a few 2.5 and some even 3.3v
Those are probably from the top layer and bloated due the heat.

12 of the 280 aren't bloated and don't sound like they have water inside.

The 152 where already bloated "at the start" 10 don't make "water sounds"

22 of the 80 possible reusable.
But..
In steel casing!
 

Leon

Solar Enthusiast
@fhorst , Thanks for continuing to share your experiences. I don't have any knowledge to contribute, but appreciate the discussion about causes and prevention of such fires. I have just installed a (much smaller) battery of this type in my shed, and will now be giving this extra consideration.
 

Supervstech

Administrator
Staff member
Moderator
The issue with steel washers between sense leads isn’t heat from the leads or balancers… it is misreadings of voltage during charge discharge sessions.

A misreading can cause over or under charge reporting that can allow the cells to be over or under charged causing heat…

But if you are unwilling to examine the possibilities, I don’t know what else to say.

I’m glad you have a new setup. I wish the new setup was available here in the USA…
 
Top