# Can a Li-Ion cell failing in a battery pack damage the other cells?

If you have a Lithium Ion battery, made from multiple 18650 cells in parallel, can any failure of one cell damage the other cells when only in electrical contact with the other cells?

More specifically, if each cell in the pack is physically isolated in all ways except for electrical contact, and in the event that a cell vents it's electrolyte, assume it's impossible for the venting gasses and liquids to contact the other cells.

Can any failure of a single cell induce electrical conditions in the other cells that will damage them?

If I didn't explain well enough, please allow me to rephrase things.

• It depends on what electrical safety mechanisms are in place and what kind of pack it is. For example if it is a parallel pack, one cell could fail and act like a short circuit for the other cells, causing them to fail, too. If series, one cell failing short could cause other to be over-charged. Etc. Commented Nov 2, 2016 at 2:45
• Ah, so it is possible for a cell to fail into a short. Would individual protection circuits for each cell be overkill? Commented Nov 2, 2016 at 2:52
• Given the premise of your question, where it seems you are designing full containment for each cell, I think maybe just an individual fuse for each cell. The "weak" cell, if it fails short/low impedance, will hopefully shunt enough current from the other cells for its fuse to blow, taking it out of the pack, effectively. Otherwise I would consider using a protection circuit for each cell. But it depends on many things, including how fast you need to discharge the cell, etc. Commented Nov 2, 2016 at 3:20
• is the failing cell on fire? Commented Nov 2, 2016 at 17:08

If you have a Lithium Ion battery, made from multiple 18650 cells in parallel, can any failure of one cell damage the other cells when only in electrical contact with the other cells? More specifically, if each cell in the pack is physically isolated in all ways except for electrical contact, and in the event that a cell vents it's electrolyte, assume it's impossible for the venting gasses and liquids to contact the other cells.
Can any failure of a single cell induce electrical conditions in the other cells that will damage them?

It is usual practice for every individual cell to be equipped with an internal protection board, for the reasons mentioned below, and others. These provide gross/rough over and under-voltage protection, probably significant-over-current protection and usually a massive-over-current fuse. Possibly also a permanent lockout for gross under-voltage.

Sometimes multiple cells have their protection combined on a single associated board but per-cell circuit is provided as it is not safe or advisable (maybe that's "neither advisable nor safe") to use multi-cell protectors alone.
The overall battery of cells may have an added layer of protection above the individual protectors.

The fact that reputable manufacturers almost always DO equip such circuits inside every cell gives a clue as to whether it is a good idea.

While some applications and user groups favour use of unprotected cells this is often done to allow the cells to be abused beyond their manufacturer specified safety limits. The R/C (Radio Control) community has many people who do this and also "vapers".

You specified "in parallel" which is less usual and less controlled in some respects. The answer below applies more to series connected cells but has implications or parallel connection.

One possible mechanism for unprotected cells is for one cell to be lower in voltage than the others. Worst case a cell can be "clamped" in a low voltage condition due to eg prior over-discharge or charging at subzero temperatures (causing Lithium plating during charging). A charger expecting the whole pack of N cells to reach N x 4.2V in constant current mode would cause N-1 of the cells to be charged to > 4.2 V/cell if one cell was very low. If a cell reaches 4.3V it may well fail and at 4.4 V it will probably be permanently damaged (due to Lithium plating out) and is not unlikely to breach pressure vents and possibility to catch fire.

For parallel connected cells with no means of individual cell balancing the maximum charge current for N cells is liable to be set to N x I_max_safe for one cell. If one cell takes less current than others then some or all must take more. It is conceivable that one cell may accept substantially more current than the others and at least be damaged as a consequence and possibly destroyed. In this case it is not the failure of one cell damaging others but rather of a failing cell being driven more rapidly to destruction by being associated with N other 'normal' cells which cause it to undergo unusual treatment. In this case the worse it gets the worse it will get and failure may be more rapid than if the cell was operated eg in a series string.

______________________________________________ ______________________________________________

The following relates mainly to a comment on this thread. Here are makers and claimed makers of LiIon manufacturing machines or of batteries. The main lessons relating to this thread are - buy known brands with known quality - and if issues with quality cells are potentially able to cause problems, what do you think is liable to happen if many cells from lesser known manufacturers are used? For eg 18650 cells, genuine Panasonic & LG are hard to beat. For large prismatics - say 10 Ah up, more discussion needed.

LiIon manufacturers by name

LiIon manufacturers by country

LiIon manufacturers in China - this is the A.... listing !!!!

Targay - one maker of LiIon manufacturing equipment

How hard can it be :-) ?

Roadmap 201\30 for LiIon manufacture VDMA & Fraunhofer - 'huff said Alibaba !!!!!!!!!!!!!!!!!!!!!! - Some sell machines, others claim custom cells. Enough machine sellers to indicate likely market.

One example

And

Battery electrode making machone note size.

Reseller or manufacturer

And

__________________________________________________

Related:

Big makers leaving / selling ? - August 2016

Valence - LARGE models

[New tech ???(http://news.mit.edu/2015/manufacturing-lithium-ion-battery-half-cost-0623) 2015

EV manuf list 2013 Don't see "Tesla" there. Then.

Useful advice, for batteries and anything else :-):
If in need of a quality product and you've too little opportunity to do adequate due diligence, if Panasonic make it, buy Panasonic.

• There are only a few real manufacturers of lithium cells, and they provide unprotected cells for use by OEMs. There are repackagers who take these raw cells, and package them with protection. There are plenty of shoddy repackagers who do little more than wrap in heat shrink, and tell people there's protection when there isn't. Commented Nov 3, 2016 at 18:29
• @ScottSeidman I'd not say that your comment was "wrong" but that it needs qualifying. China is utterly vast. There are a few really top makers - but probably many thousands who can & do make cells. What brand did YOU buy :-). Anyone can buy a battery making machine set and set up business. Pouch cells are "easy" as long as quality not too worried about. So ... . See end of my answer. . Commented Nov 3, 2016 at 21:19
• I always overreact, but I think hobbyists can create a ton of trouble messing with lithium cells. The pros at Samsung just cost their company \$3billion Commented Nov 3, 2016 at 21:49
• @ScottSeidman Yes - lots of trouble if done wrong, and lots that can be dome wrong. But "done right" within well defoned guidelines LiIon is surprisingly well behaved in charging and use. It is FAR easier to desogn a good LiIon charging system than a good NimH one. Especially so in unusual situations such as cariable input solar. I've domne that with NimH and there is no easy + really good way to do it. With LiIon thre is. Commented Nov 3, 2016 at 21:54

Reliability of the LiPo 16850 standard cell has significantly improved.

• Perfectly matched cells for voltage and ESR can be shunted in parallel in large quantities made into modules, which are then added in series to create the require power source voltage are done with suitable protection for insurance
• The risk of internal short circuits although low can cause fires.
• Even fuses can fail, but rare.

In an electrical fire, water reacts with the lithium and can actually make things worse. Below, a 2014 Model left alone during charge operation reported Jan 2016 in Norway. The most reliable method now uses microfuses on both Anode and Cathode for such infrequent failure avoidance.

simulate this circuit – Schematic created using CircuitLab

Arrays require precision matching of Wh capacity to 0.0x% tolerance when new.

This schematic would be a 3 dimensional array, so that an open or short is isolated and the pack still survives. This is analogous to the reliability of "the Web" with distributed connections. . These parallel connections are usually done with Ultrasonic welds on SS with low ESR fuse material and PTC characteristics to accelerate fusing above a high holding temp. But normally for abrupt protection only.

• A DIY approach might be a short whisker wire with an ampacity fuse rating and DCR < cell ESR with some loss in surge peak current carefully chosen to meet both criteria of fusing and Motor start current. But soldering is a problem for SS, so ultrasonic welds are done, just like gold whisker wire inside LEDs to Anode.
• _"In an electrical fire, water reacts with the lithium and can actually make things worse." - in my experience it is the opposite - putting a burning Lipo into a bucket of water will quickly put it out. There is very little lithium in a Li-ion cell, most of the flames are caused by organic chemicals and plastics burning. Commented Dec 28, 2020 at 22:55