How would one test possibility AA battery leak

[RoadRunner]

Hallo,

I am thinking of doing a AA battery comparison. I want to test which AA battery/brand is more prone to leak.
Is battery leak just random or we can deliberately create test conditions for battery to leak and then compare them with one another?


Regards

[CaptDon]

I don't think any form of forced leak is a good indicator? They leak for several different reasons. Years ago the case of the battery was consumed during discharge and so a fully discharged battery now had a thinner 'cylinder' eroded from the inside out. If some part of the case was consumed more quickly it would result in a hole in the container. Some leak at the seal due to unusual circumstances like trying to charge an alkaline that was not designed for recharging. Sure you could force them to leak / explode but not sure if that data would have any real meaning? Made in China would be a good indication of a turd battery. Like the 9900mah pieces of shit to replace 2300mah cells. The Chinese trash made about 880mah and certainly not the 9900mah claimed!!!

[james_s]

Dave already did some tests like this. In my experience the most reliable way to get a battery to leak is to put it into something expensive and then put that away on a shelf for a while.

[Terry Bites]

Not being used for long periods increases the chance of leakage as every flashlight owner knows.
Slow self discharge generates hydrogen gas that builds up over time and pushes out the satanic electrolyte ooze.

[Conrad Hoffman]

I used to have an answering machine with an absolutely brilliant way of making batteries leak. It would work fine until the battery got low, then a small LED would turn on so you knew to change the battery. Of course the LED would deplete the battery completely in short order and then the battery would leak. There should be awards for designs like that.

[coppercone2]

measure the quiescent current of your primary cell devices and you might find out why they are leaking. Sometimes they have a silly off -current draw .

[rsjsouza]

As others have said, the root causes vary and several discussions around here, as well as Dave himself, tried to explain the phenomenon and narrow down the root cause. Quite fruitless.

One thing I am thoroughly convinced is that quality of batteries of yore were significantly better. My last short video (in portuguese, sorry) shows three sets of batteries completely forgot in remote controls that were subjected to 10~70% RH and temperatures between -2°C to +38°C (in my garage). The batteries are 2008 Gold Peak (GP), Supergard 2009 and 2015 Energizer Black Top. The Supergard is at complete zero volts, while the others have quite the capacity still remaining.

We as consumers have zero recourse.

https://youtube.com/shorts/y9VQt0xsBV0

Dave already did some tests like this. In my experience the most reliable way to get a battery to leak is to put it into something expensive and then put that away on a shelf for a while.

Not only expensive, but rare or irreplaceable will do as well. And I agree: that is the only sure way to force a battery to leak.

[Ian.M]

I think there may also be an element of shipping damage.  If the carton of battery cell blister packs has been dropped on one end, the cells banging together could possibly crack the seals, or the whole shipping container full may have been roasted on top of the pile transiting the tropics.

The days of self-healing bitumen mastic seals (as found in vintage dry cells) are long gone.  Not that they were any better - they just leaked *through* the Zinc can!

You could *TRY* putting cells on discharge (simple resistor across them) in an environment chamber and thermally cycling them to  try to get early failures, but if its quick it wont be representative of real life, so you'll need some way of monitoring them automatically.  Maybe put them all standing on their negative terminals (the end with the seal for alkaline cells) on a thin strip of aluminum foil as an alarm circuit loop.  When significant electrolyte escapes, it will corrode and break the foil, opening the loop and sounding the alarm.

[rsjsouza]

I think there may also be an element of shipping damage.  If the carton of battery cell blister packs has been dropped on one end, the cells banging together could possibly crack the seals, or the whole shipping container full may have been roasted on top of the pile transiting the tropics.

That is one of my anecdoctal experiences as well. When batteries are installed in mechanically hostile environments (toys, flashlights, etc), I noticed there seems to be a concentration of events there as well.

https://www.eevblog.com/forum/blog/eevblog-1296-alkaline-battery-leakage-testing-2-electric-boogaloo/msg2987604/#msg2987604

I remember those bitumen seals in red eveready and yellow ray-o-vac batteries. On those days, alkalines were promised to almost never leak while Zinc were a disaster. Look at where we are now. 

[merkatorix]

Slow self discharge generates hydrogen gas that builds up over time and pushes out the satanic electrolyte ooze.

I just had an experience, which might fit this topic:
I just found a battery in its original packaging, which had one of the worst leaks I've ever seen. However, it was at least 8 years old.
Then again, Google says they have a self-discharge of 2-3% per year. Having a shelf life of 7-10 years.
I should check if the date 2014 was the "best before" date or the production date. If it was the production date, the shelf life should be the life until it is sold and not the life until it starts messing up its surroundings, shouldn't it?

[SeanB]

2014 is the best before date, though I have come across many batteries still in the blister pack, with a best before date 3 years in the future, which are leaking in the blister pack.

[merkatorix]

Dave already did some tests like this.

Has there been any conclusion?
A quick search shows this video showing a cell covered in a lot of crystals EEVblog #1296 - Alkaline Battery Leakage Testing 2 - Electric Boogaloo

When googling a little, I noticed that my Duracell in its original shrinkwrap are probably not worth photographing Duracell batteries leaking
This Duracell leak thread "DuraHELL" mentioned the original test of DAVE.

All this cells, which already leaked in their original package, reminds me of the video from Big Clive pre-exploded surge protection strip
.

[jonpaul]

avoid duracell

[Kleinstein]

The reasons behind the leakage are likely quite complicated. If this would be an easy topic chances are the batteries would not leak anymore.

There are quite some parameters that can effect leakage: mechincal stress (including transport), thermal history (have the batteries been out in freezing cold),  kind of discharge, temperature variations and just random production variations. So it would need an awful lot of samples from different lots to get a clear picture.
With multiple mechanisms it is no clear of elevated temperature is a valid way to accelerate the test. So the tests would take a long time. Chances are that when done, the new batteries produced by than may not be identical to the ones tested.  So what is gained of you know that brand A from 2000 is really shitty and Brand B is good, when you no longer can buy them.

A more practical way to work on the leaking battery problem may be collecting cells that have leaked and do a careful analysis on what actually failed. Chances are this is what some of manufactuers already do with quite some effort.

One could also do a distributed tesing: keep the batteries that are already empty and instead of directly disposing them, keep them in a safe place to wait until the actually leak. So no need to buy extra samples and many cells could be tested.

P.s.:
If really interested:  maybe get in contract with a site for old battery collection / disposal to get access to a really large number of old cells to do statstics with larger numbers.

[Sal Ammoniac]

Hallo,

I am thinking of doing a AA battery comparison. I want to test which AA battery/brand is more prone to leak.
Is battery leak just random or we can deliberately create test conditions for battery to leak and then compare them with one another?

For recent DuraHell batteries, the only conditions needed for them to leak are, well, nothing. They leak even sealed in their packaging..

[oz2cpu]

i got en idea for leak detector;


wrap each cell in one layer of gaze or similar light woven material that is non conductive,
then wrap one layer alu foil,
now measure ohm / voltage / conductivity DC or AC, to either + or - of the cell to the alu foil,
when you see value change, you got bingo..

some cells take 10-20 years to puke thru
while others do this 1-2 years old

so prepare the test setup well, so each DUT is easy to remove from each socket

[Haenk]

I think the quality of the seals have gone south quite a bit, probably due to very aggressive market pricing. And, correct me if I'm wrong, e.g. Duracell now being a cost-optimized brand owned by Warren Buffet. As we all know, optimizing costs almost always mean crap quality.
So it would be interesting to compare some 90s Alkaline to current designs. Any bet the seals were way more massive back then.