2 pack faling behind with charging and get unbalanced (update 9/07/2020)

[Redpacket]

The OP's issue is clearly cell a critical failure in the middle of the cell with nothing to do with the ends of the cell.
As not2bme pointed out above, the most likely cause is physical impact.
A dent, etc might have been concealed if the cell's shrink wrap sprung back to the "normal" shape after impact.
Replacement of that cell will likely fix the issue for the OP.
The damaged cell should be put somewhere fireproof asap & discharged to zero (safe then) with eg a 100 ohm resistor.
As mentioned already, while is has any significant charge it is a high fire risk.
Given the shape & orientation of the heat/damage, my opinion is nearby cells would be unharmed & at most got a few degC warmer (harmless).
If the burn had faced other cells, then agree adjacent cells might have suffered.

 

[gauss163]

The OP's issue is clearly cell a critical failure in the middle of the cell with nothing to do with the ends of the cell [...]

As I explained briefly in post #15, the reason for a post-mortem examination of the top is to check for any evidence that the cell attempted to vent through the top vent holes (or see if they are blocked). Knowing that helps to narrow down the possible modes of failure. In particular, since this cell did not appear to vent on top, that increases the probability that the event was caused by an internal short nearer to the can sidewall (away from the central vent tube (mandrel) - see images 1 & 2 below), where gases are more likely to get trapped - unable to reach the central vent - resulting in sidewall ruptures - as in image 3 (in 1 second thermal runaway can generate gas having 50 times thecell volume).









excerpted from Computed Tomography Analysis of Li-Ion Battery Case Ruptures (2020), by L. Kong et al.

Given the shape & orientation of the heat/damage, my opinion is nearby cells would be unharmed & at most got a few degC warmer (harmless).If the burn had faced other cells, then agree adjacent cells might have suffered.

Since 18650 cells have been reported to reach up to 700C in thermal runaway (cf. 2020 paper above), there is certainly a chance that it caused thermal damage to neighboring cells (no matter the location of the rupture). Thermal damage can wreak havoc on cells.

[...] As not2bme pointed out above, the most likely cause is physical impact. A dent, etc might have been concealed if the cell's shrink wrap sprung back to the "normal" shape after impact.

Yes, as I remarked in post #9, mechanical damage is one possibility, but usually it takes a significant dent to the sidewall to cause a short since the jelly role has opportunity to deform a bit without damageunder such radial forces (unlike axial compression on the top). Below is an image excerpted from a 2012 paperdiscussing shorts caused by sidewall dents.

But there are also many other possible causes of internal shorts, e.g. copper/lithium dendrites from over (dis)charges, metal contaminants (shrapnel) introduced during manufacture, thermal damage, etc.

 

[Redpacket]

gauss163 thanks again for another informative post.
I'm just looking at the simple side of it as shown in the OP's picture.
We love all the detailed info you bring but please allow the practical & pragmatic side some space here too.

The OPs pic shows some localized spot heating which when combined with the other info is not likely to have been deep in the cell.
Compared to other typical failure result pics this is very mild - OP was lucky!
Even the cell's own sleeve isn't melted much apart from the hole.
The OPs cell cap was "normal" after this.
The cell is even (mostly) holding voltage & taking a charge! This suggests much of the internals are not destroyed.
It's as though this event was a"slow blow resistive" not a "fast blow full short"
Thermal runaway obviously did not take hold.
So temps are not likely to have gone to 700degC & pic shows only a wisp of smoke on other cells.
All this points to something in the outer layers not the cell middle.
In this case, pressure venting or winding design doesn't seem to be that relevant (although still interesting).

I agree it could have been caused by a dendrite too and since the OP said they can't see a dent maybe this is a likely cause.

 

[gauss163]

[...] please allow the practical & pragmatic side some space here too.

I have no idea what you mean by "please allow". As I stressed in the past, my view on safety education is that we should provide users with as much pertinent information as possible. Only then do they have the best chance possible to make wise decisions. Too many folks have suffered serious injuries from Li-ion batteries only because no one properly warned them of the inherent risks.

The OPs pic shows some localized spot heating which when combined with the other info is not likely to have been deep in the cell. Compared to other typical failure result pics this is very mild - OP was lucky!

We'll have to disagree on that. That gaping hole likely requireda serious thermal event.

Even the cell's own sleeve isn't melted much apart from the hole

That implies very little without knowing the wrapper material and its melting point. Even PVC has formulations with very high melting points - up to 500F (260C) - which is far more than needed to thermally damage a (neighboring) cell.

And there is good reason for top-tier manufacturers to use a wrapper with a high melting point because it helps to prevent runaway chain reactions (if a thermal event melts the wrappers between two side-by-side series cells then their cans may short together, which shorts the neighbor, then that goes into runaway, which shorts another neighbor ... (domino effect).

In the OP, we don't have anywhere near enough evidence to rule out thermal damage toneighboring cells.In such cases the safe way to proceed is to assume the worst and replace the neighbors too.

In this case, pressure venting or winding design doesn't seem to be that relevant (although still interesting).

Since you seemed to think that my asking for photos of the vent was not relevant, I replied explained why this is a standard step in such post-mortem analyses. Further, understanding such "under the hood" matters is relevant since it may help one to use cells more safely. An educated Li-ion user is a safer user.

 

[daromer]

That cell cant have been that hot. The cells that go full thermal have melted the sleeve totally and the casing Will be discoloured all the way around.

Remove the sleve and check.

Edit; it almost look like a melted hole due to some Contact on the outside?

 

[gauss163]

^^^ Obviously it did not go full thermal, but even localized thermal events like this can generate enough heat to damage neighboring cells. The hole needn't be from external contact since sidewall ruptures like that are characteristic of internal shorts - see post #22 and the paper linked there.

 

[barry]

*update*

I changed the cell andneighboring cells with the same capacity all the packs are in balance now and works how it is designed.
and just in time last night we had now power

 

[Redpacket]

[...] please allow the practical & pragmatic side some space here too.

I have no idea what you mean by "please allow". As I stressed in the past, my view on safety education is that we should provide users with as much pertinent information as possible. Only then do they have the best chance possible to make wise decisions. Too many folks have suffered serious injuries from Li-ion batteries only because no one properly warned them of the inherent risks.

Please allow = There's a difference between helping this forum's readers understand Li-Ion issues & risks vs hammering the safety issue every time with a sledge hammer.
We as a group want to build packs & power systems. Dumping a ton of bricks on every possible risk isn't helping, it's loosing people's interest.

I note you have not provided links to injuries/damage from DIY Li-Ion. Yes people have been hurt, etc but not from DIY so far.
Before you jump on that - yes we want to understand & learn & be as safe as reasonable.
This very forum helps people learn & follow reasonable practices.

I stand by what I've said, it's plain obvious, as I explained based on the details the OP provided.
It was (fortunately) a mild event & cell burst/explosion clearly did not occur so in THIS case internal venting design is not of primary relevance it's secondary.

A wrapper stopping thermal run away? I don't think so. Even in this mild case (yes quite bad enough but there's much worse) the wrapper was burnt through.
Cell wrappers are only heatshrink plastic & that doesn't stand a chance. Kapton tape would do better, but I can still burn that with a soldering iron.
As many videos & results show, full thermal runaway = most bits crispy.

An educated Li-ion user is a safer user.

Agree.

 

[gauss163]

A wrapper stopping thermal run away? I don't think so.

It seems your above comment refers (too briefly) to my remark below:

And there is good reason for top-tier manufacturers to use a wrapper with a high melting point because it helps to prevent runaway chain reactions (if a thermal event melts the wrappers between two side-by-side series cells then their cans may short together, which shorts the neighbor, then that goes into runaway, which shorts another neighbor ... (domino effect).

In that case, yes, wrappers - as electrical insulators - can prevent shorting of neighboring cells/packs. As a consequence they may help prevent those neighbors from also reaching runaway. Such electricalinsulation is the primary purpose of the wrapper.

Recall that the entire cell can is negative, and neighboring (touching) cans in (laptop) packs can have different potential (as newbie pack harvesters often learn the hard way by accidentally shorting them during disassembly). If the wrapper had a low melting point then it would not work well as an electrical insulator (between neighbors) duringthermal events. So wrappers with higher melting points are preferred. That's whyit is not unusual to see the wrapper survive seemingly intact after localized thermal events such as in the OP (though all bets are off with lower-tier cells whose wrappers may melt at very low temps- Ultrafire'sare well-named!)

If you wish to further discuss the meta topic of how to best discuss general safety matters then I think that would be better placed in a thread devoted to that purpose.

 

[Korishan]

An educated Li-ion user is a safer user.

Agree.

Double Agreed

If you wish to further discuss the meta topic of how to best discuss general safety matters then I think that would be better placed in a thread devoted to that purpose.

Triple agree

 

[Korishan]

Recall that the entire cell can is negative, and neighboring (touching) cans in (laptop) packs can have different potential (as newbie pack harvesters often learn the hard way by accidentally shorting them during disassembly). If the wrapper had a low melting point then it would not work well as an electrical insulator (between neighbors) duringthermal events. So wrappers with higher melting points are preferred. That's whyit is not unusual to see the wrapper survive seemingly intact after localized thermal events such as in the OP (though all bets are off with lower-tier cells whose wrappers may melt at very low temps- Ultrafire'sare well-named!)

Yes, we know that the whole can is Negative. However, in a pack of parallel cells, it doesn't matter if the can touches another as they are all at the same potential. Another thing to note is that the OP had the packs built inside of a constructed housing that kept space between the cells of one pack from those of another.

However, in this instance, you can see that that particular pack was up against something "other" than another pack. So we know that it wasn't a voltage difference between packs. Buuuuut, what was that pack up against? Only the OP can inform us.

This is just a theory: Perhaps the unit that these packs were stored in was metal. Perhaps also the Negative side of the string was grounded to the storage rack somehow, perhaps the inverter/charger was bare metal to bare metal. This would make the rack essentially 0V potential. This particular pack, being in the middle some where, would have a different potential relative to the metal rack, perhaps even only 4V at full charge if this was the 2nd pack in the string. Then the OP pushed the pack up against the back of the metal rack. At the point of contact, by pure coincidence, there was a metal bur on the metal rack that slightly pierced the shrink wrap.
This metal bur could possibly even be on the end of a bolt that held the rack together. Then over time, that spot slowly heated up until it finally burned the wrapping and created a small hole. Basically it was a slow spot welding. Until the metal can burned enough metal away to disconnect from the bur.

I ask barry to check the back of the rack he installed the packs on and check for any damage to the rack itself.

Of course, this theory is all moot if he built a wooden rack to install these on. In which case, it would appear something else pierced the wrapping.

 

[gauss163]

Yes, we know that the whole can is Negative. However, in a pack of parallel cells, it doesn't matter if the can touches another as they are all at the same potential.

It appears you missed the word "series" in "two side-by-side series cells". These (and other mixed potential neighbors) occur frequently in commercial packs (e.g. laptops, powerbanks, etc), esp. those having odd shape. Of course the manufacturers must spec the wrappers to handle these cases, not only parallel neighboring cases as in DIY powerwalls.

As for "we know....", my experience is that almost all beginners have no idea that the entire can is negative, since that is contrary to most cells at consumer level. One of the primary purposes of safety warnings is to help protect beginners - where most (but not all) accidents occur (e.g. vaperpants-on-fire accidents due to carrying naked cells in pockets with keys, coins, etc). Andof course every active site has manybeginners at any given time (often hidden due to lurking). As such, it is helpful to emphasize (in passing) these common oversights, since osmosis often works wonders for learning.

 

[Korishan]

Yes, we know that the whole can is Negative. However, in a pack of parallel cells, it doesn't matter if the can touches another as they are all at the same potential.

It appears you missed the word "series" in "two side-by-side series cells". These (and other mixed potential neighbors) occur frequently in commercial packs (e.g. laptops, powerbanks, etc), esp. those having odd shape. Of course the manufacturers must spec the wrappers to handle these cases, not only parallel neighboring cases as in DIY powerwalls.

I didn't miss the word "series".

As for "we know....", my experience is that almost all beginners have no idea that the entire can is negative, since that is contrary to most cells at consumer level. One of the primary purposes of safety warnings is to help protect beginners - where most (but not all) accidents occur (e.g. vaperpants-on-fire accidents due to carrying naked cells in pockets with keys, coins, etc). Andof course every active site has manybeginners at any given time (often hidden due to lurking). As such, it is helpful to emphasize (in passing) these common oversights, since osmosis often works wonders for learning.

As far as I can tell, barry isn't a newbie and is quite aware of the issues of shorting cells that are in series. Also, we're talking about a pack that all cells are connected in parallel. Discussions about a topic should stay on topic based on what the OP has stated and asked. Not fill the thread full of information that doesn't directly pertain to the OP's thread. If you wish to add extra information, please create a new thread and link to it ("If you'd like more information about such-n-such, please read this thread.....")

 

[gauss163]

I didn't miss the word "series".

Then it's not at all clear why you replied to my remarks on wrappers for general commercial packs with much more specific remarks about a particular pack topology (such as OP's). Obviously manufacturers must design cells for the general case, not some specific case.

As far as I can tell, barry isn't a newbie [...]

Again you appear to be (mis)reading general remarks (on safety) much more specifically than intended. No one claimed that the OP was a newbie. Please try to read more carefully. It is such misreadings that are primarily responsible for the thread veering onto tangential meta matters.

 

[Korishan]

I didn't miss the word "series".

Then it's not at all clear why you replied to my remarks on wrappers for general commercial packs with much more specific remarks about a particular pack topology (such as OP's). Obviously manufacturers must design cells for the general case, not some specific case.

As far as I can tell, barry isn't a newbie [...]

Again you appear to be (mis)reading general remarks (on safety) much more specifically than intended. No one claimed that the OP was a newbie. Please try to read more carefully. It is such misreadings that are primarily responsible for the thread veering off-topic onto meta matters.

I choose to not go into details as it has nothing to do with the OP or his situation. Please stay on target and keep answers within the realm of knowledge of the OP and participants. We don't need super basic replies, or ones that are so advanced that most are lost in them.

 

[gauss163]

[...] We don't need super basic replies, or ones that are so advanced that most are lost in them.

Could you please have the courtesy to stop writing "we" when you mean "I".

Sorry that my posts are either too basic or too advanced for you. But most of them were not targeted specifically at you (or anyone for that matter, since I oftenstriveto consider general matters in order to make my postsas useful as possible).

 

[Korishan]

It is not "I", it is "we". because several have posted about it. So if it is more than 1, it is "we".

Your content isn't too basic or advanced for "me". What you are doing is not sticking within the boundaries of the thread. Keep your comments in reference and boundaries of the thread. If you have more detailed information you wish to add that goes beyond the thread, post a new thread and link to it.

This conversation is done, over with. The thread has been derailed long enough. It will return back to on topic of the OP and his situation.

As I mentioned earlier, it would be interesting to know if that pack was pushed up against the back of a metal of rack that possibly punctured the cell wrap. We won't know that answer until Barry responds.

 

[barry]

just what Korishansaid i am not a nooby .
yes all my pack are well isolated and rest on nylon L brakets so nothing touches one and other and the skrews are well hidden.

thanks guys for all the info and help!!

 

[gauss163]

^^^ Not to worry, no one called you a noob. Good job on the quick fix.

 

[Redpacket]

Glad to hear you got it fixed Barry
Nylon pack slides sound like a great idea!