EEVblog #1296 - Alkaline Battery Leakage Testing 2 - Electric Boogaloo

[EEVblog]

A revisit of the Alkaline battery leakage testing.

[SilverSolder]

Use Murphy's Law to help get results?  - Install the test batteries in some expensive gear - the more expensive, the more likely they are to leak!   

[Dr. Frank]

Dave,
I think that these Alkalines only leak after years, and maybe preferably if they are not discharged, or used in low current applications, like remote controls.

That's what I experienced maybe 10 years ago, when I first found these batteries from ikea (Chinese origin), up to now. These Alkalines have about double the mAh capacity than other battery systems, and have a storage time of 5 years, the latest from VARTA AAA which I bought recently, even 10 years. Maybe they are not intended for low rated discharge, despite their long storage time.

My idea is that the liquid chemistry inside will deteriorate after years of storage, and may have a slow side reaction which creates this KOH stuff, which would not be produced by normal fast discharge.
Also, I don't think, that there is such high pressure inside, so maybe a real valve might not be used.

If you discharge these batteries quite quickly (and far before their expiry date), maybe a different chemical reaction sets in, which does not produce these crusty KOH crystals.
That's just speculation at the moment, so it might be helpful to understand the alkaline reaction path, either regular, or idle.

Then I have observed, that this alkalin stuff does not necessarily corrode metals, up to now it was a mess to wet clean the battery cabinet, but the spring contacts never were harmed.. copper surfaces maybe another story.
Here again, it would be interesting, which chemical substances emerge from leaking batteries, so to evaluate, which metals (or plastic parts) could be harmed.

Frank       

PS: Yes, these chemical processes are sketched on the German WIKIPEDIA already.

In short, there's a side reaction (called outgassing reaction) which creates Zn(OH)₄^2- and H₂, which might be the white leaking stuff, and the gas which creates a bit of pressure.
This is usually a very slow process, but may be prominent, and much faster, if the Zn has a lot of metal impurities, like Fe, if the battery is not used, or under low load conditions.
That's also called a Zn corrosion process (due to the electro-chemical electromotive chain, why this effect happens)

Zn + 2H₂O + 2OH^- => [Zn(OH)₄]^2- + H₂}

If you discharge the battery relatively fast, there will simply be no chemical stuff left for this side reaction and the leakage effect, as the KOH electrolyte and the Zn and Mn ingredients are consumed and turned into other, dry and non leaking substances.
In former times, this Zn corrosion was suppressed by addition of Hg, but as this is forbidden due to RoHS, it all depends on the availability of very pure Zn metal.

I suppose that even serious manufacturers like VARTA have still problems to get this, as they probably also produce in China, for example.


So I fear, Daves experiment on discharged batteries will not reveal any leakage effect at all, and the others might need many years before they leak.

[JustMeHere]

I used to be the IT Director for a beach resort.  We had electronic locks (~150) on the doors.  We had a bad problem with leaking batteries.  The supplier switched to Duracels from generics.   That did help a good bit.  The age (as far as time in the lock) did not appear to be a major factor.  I've tested leaking batteries that still had over 1v.   I believe heat had more to do with leaking than anything else.   Daily 38 C temps is the norm here during the summer.

[Maxlor]

Good luck with the second attempt!

I have some questions: you say the internal resistance is measured at 1 kHz, how is this measured and how is it useful for a battery that is used or DC? I just don't see how batteries and AC go together.

And another one, couldn't you just connect the batteries together in parallel, then have your electronic load slowly force the voltage of all of them down to 1V? If you start at, say, 1.3V and all the batteries are above that when you connect them, that should avoid forcing any current back into them, shouldn't it?

[darik]

I also think it would be useful to do a trickle discharge test on batteries that haven't been depleted. I've been measuring the alkaline batteries that have leaked on me and they are usually around 1.2-1.4V. They're never dead. I don't think it's a safe assumption that they have to be discharged to leak.

[Brumby]

Use Murphy's Law to help get results?  - Install the test batteries in some expensive gear - the more expensive, the more likely they are to leak!   

You were reading my mind!

I, too, seemed to think that by not considering one of the guiding principles of electronics, results could only be expected to be disappointing and (effectively) irrelevant.   

[ormaaj]

Has anyone measured a really dead battery into an ultra-high impedence / low-capacitence (electrometer or in-amp buffer, though the latter may have low input voltage limit w/ internal clamps). Maybe you can get 1000s of volts by just rubbing your socks on the carpet and passing some charge on to one of the terminals. If the charge is tiny and there's reasonable internal resistance then it might not really uniformly diffuse throughout the battery.

[rsjsouza]

In my undocumented experience, I have the impression the mechanical conditions that contribute the most to battery leakage are the confinement in the tight space of a battery holder and the pressure from the spring or possible impact (in case they leak in their original packaging).

The other factor I suspect the most is that the draining of the battery never ceases - they seem to leak more when left for weeks/months in equipment with a soft power button.

For the equipment with a hard power button (flashlights, or my kids' soap bubble machine, for example), the discharge ceases in the off position, but I suspect the harsh environment and mechanical movement/impact seems to accelerate this process.

I don't recall experiencing this with battery powered alarm clocks or other equipment that is "left alone" (wall clocks, wireless thermometers, etc.)

This is totally anecdotal, but I can't believe the manufacturers never explored all these possibilities in testing - my guess is the drop in quality and durability is entirely deliberate.

[SilverSolder]

[...] my guess is the drop in quality and durability is entirely deliberate.

It is hard not to suspect "cost optimization" of being somehow involved.

The risk is borne entirely by the customer, not the manufacturer... so there is not much short term incentive for them to do a good job. The only people who noticed are a bunch of geeks on an internet forum... 

[SpecialK]

I was disappointed the video wasn't labeled "Re: EEVblog #1296 - Alkaline Battery Leakage Testing 2 - The Quickening"

[EEVblog]

Dave,
I think that these Alkalines only leak after years, and maybe preferably if they are not discharged, or used in low current applications, like remote controls.

That's not what a paper cited elsewhere here says, plus another source I have read. More chance of leakage at the end of capacity.

So I fear, Daves experiment on discharged batteries will not reveal any leakage effect at all, and the others might need many years before they leak.

Perhaps, we'll see, that's the point of this test.

[bsfeechannel]

20 MHz is pretty much the maximum practical bandwidth you can achieve with a non post-deflection accelerated CRT oscilloscope.

[factory]

I suppose that even serious manufacturers like VARTA have still problems to get this, as they probably also produce in China, for example.

The leaking Varta battery in the video clearly has "Made in China" printed on it despite Dave saying it was German made. 

Having said that I've not used any Varta alkaline batteries as they are not sold in many of my local shops.

I've mostly been buying Panasonic batteries made in Belguim, I've since noticed some of them are now made in Poland and the more expensive Panasonic Evolta ones I got when Maplin closed down are from Malaysia.
I haven't used Duracell for a very long time due to too many leakage problems.

David

[madires]

As the family's repair shop I had several cases of leaking batteries over the years. Nearly all cases share the same combination of conditions:
- partly discharged batteries
- stored in the drawer for a long time without any use (a year or more)
- no load when device is powered off

But this is not a proven recipe for leaking batteries. Many didn't leak under the same conditions.

[Fungus]

I used to be the IT Director for a beach resort.  We had electronic locks (~150) on the doors.  We had a bad problem with leaking batteries.  The supplier switched to Duracels from generics.   That did help a good bit.  The age (as far as time in the lock) did not appear to be a major factor.  I've tested leaking batteries that still had over 1v.   I believe heat had more to do with leaking than anything else.   Daily 38 C temps is the norm here during the summer.

Maybe it's related to temperature cycling, too. Big difference between day/night, or even seasonal, could put pressure on the seals.

[free_electron]

i am never buying duracells again.
so far they have ruiened
- my apple keyboard
- my apple mouse
- my apple trackpad
- several wall clocks
- two calculators
- a multimeter
- 3 portable radio's
- several dive lights. especially when you use them almost completely and then store the light for a year or so till next trip ...

They all leak when
- not in use for a long time.
- very low power consumption interlaced with peak usage.

I have switched to Energizer. no more leaks. zip nada.
And for where it really needs to be reliable : energizer lithium cells.

[bittumbler]

First, I am not a battery expert.
But the whole battery is chemistry, including the leakage process.
Most chemical reactions go twice as fast for every 10K temperature increase.
So maybe storing the DUT batteries in a very hot environment will speed up the leakage, and at least show which brands/models are better than others.

[thm_w]

i am never buying duracells again.
so far they have ruiened
- my apple keyboard
- my apple mouse
- my apple trackpad
- several wall clocks
- two calculators
- a multimeter
- 3 portable radio's
- several dive lights. especially when you use them almost completely and then store the light for a year or so till next trip ...

Why wouldn't you use eneloops for everything non-critical?

[rsjsouza]

I don't know about free_electron, but for me the hassle of managing charging/recharging and the shortened lifetime of a single charge (due to the lower voltage per cell) is a deal breaker for some of my requirements.

[magic]

Why wouldn't you use eneloops for everything non-critical?

Be careful. I recently found a few NiMH batteries which had been unused for over 10 yeas and almost all of them were leaked, although most still work to some 50~75% capacity after cleaning. The leaked ones were non-LSD Energizers, Vartas and no-names. Two Eneloops were also found in the lot and they weren't leaked, but I don't know what would happen if I left them for another decade without charging.

[Fungus]

Two Eneloops were also found in the lot and they weren't leaked, but I don't know what would happen if I left them for another decade without charging.

Eneloops don't have any liquid in them, so... probably nothing.

[Electro Detective]

i am never buying duracells again.
so far they have ruiened
- my apple keyboard
- my apple mouse
- my apple trackpad
- several wall clocks
- two calculators
- a multimeter
- 3 portable radio's
- several dive lights. especially when you use them almost completely and then store the light for a year or so till next trip ...

They all leak when
- not in use for a long time.
- very low power consumption interlaced with peak usage.

I have switched to Energizer. no more leaks. zip nada.
And for where it really needs to be reliable : energizer lithium cells.

It may be a location and or source of manufacture/supply reason
but I have experienced the opposite of the above

i.e. Duracell ok to meh..

Enerjizzzers = Leak City, break out the nano wire brush and elbow grease and pray no creepy crawlies made it to the PCB ..

Let's see how  '..forgotten in expensive equipment'  Eneloops and Eneloop knockoffs ~perform~ after a few years   

Most likely the Gen 2 Eneloops in a few years time  'made cheaper' will drop their guts as well

Bean counting corporats just can't help themselves.. 

[SilverSolder]

I don't know about free_electron, but for me the hassle of managing charging/recharging and the shortened lifetime of a single charge (due to the lower voltage per cell) is a deal breaker for some of my requirements.

The best way to reduce hassle is to have a charger that can charge and also maintain the batteries - so when a device runs down, you just swap the batteries with the ones in the charger.  Easier than tearing open a blister pack!

[free_electron]

I don't know about free_electron, but for me the hassle of managing charging/recharging and the shortened lifetime of a single charge (due to the lower voltage per cell) is a deal breaker for some of my requirements.

Same complaint here. lower voltage, lower capacity ( or perceived capacity as devices shut off earlier due to lower voltage )

i now only use energizer lithium cells as AA replacement. they have extreme long shelf life (15+ years) and pack a lot of capacity.( 3000 maH compared to 1200..1500 for a regular aa cell. )

Or have switched to true li-ion cells where possible. i use Tomo chargers with 18650 cells. the Tomo charger eats anything ( micro usb, apple lightining and usb-c ) and can give two usb outputs, or you can simply remove the cells. convenient lcd display shows charge state and current draw. and that is a true 128x32 pixel lcd. not just a 4 line bargraph ...

https://www.tomopowerbank.com/shop/tomo-s4/

[thm_w]

Be careful. I recently found a few NiMH batteries which had been unused for over 10 yeas and almost all of them were leaked, although most still work to some 50~75% capacity after cleaning. The leaked ones were non-LSD Energizers, Vartas and no-names. Two Eneloops were also found in the lot and they weren't leaked, but I don't know what would happen if I left them for another decade without charging.

Don't buy non-LSD.

Same complaint here. lower voltage, lower capacity ( or perceived capacity as devices shut off earlier due to lower voltage )

i now only use energizer lithium cells as AA replacement. they have extreme long shelf life (15+ years) and pack a lot of capacity.( 3000 maH compared to 1200..1500 for a regular aa cell. )

Lower voltage is valid.
Capacity, you can get 2,500mAh eneloops. So any decently designed device with one of those will last just as long as alkaline, assuming their claims of discharge can be trusted.

edit: this guy measured 10%/year self discharge
https://www.candlepowerforums.com/vb/showthread.php?447670-Anyone-done-Self-Discharge-tests-on-Eneloop-Pros&p=5230886&viewfull=1#post5230886

[Fungus]

Lower voltage is valid.

They're supposed to give 1.2V for almost their entire discharge so you'll have to look long and hard to find a device that doesn't work.

If you can find one? Get a Batteroo!

[darik]

They're supposed to give 1.2V for almost their entire discharge so you'll have to look long and hard to find a device that doesn't work.

If you can find one? Get a Batteroo!

Bose noise cancelling headphones. They seem to cut off at about 1.25 V. Infuriating.

[rsjsouza]

I don't know about free_electron, but for me the hassle of managing charging/recharging and the shortened lifetime of a single charge (due to the lower voltage per cell) is a deal breaker for some of my requirements.

The best way to reduce hassle is to have a charger that can charge and also maintain the batteries - so when a device runs down, you just swap the batteries with the ones in the charger.  Easier than tearing open a blister pack!

I've tried that with a subset of the worst offenders and it is impractical for me - too many small gadgets and toys to manage, which would lead to me managing various chargers and battery sets just to keep up. Not to mention that some high power toys from my kids drain the batteries to an artifical "low battery" condition quite quickly.

[magic]

Two Eneloops were also found in the lot and they weren't leaked, but I don't know what would happen if I left them for another decade without charging.

Eneloops don't have any liquid in them, so... probably nothing.

I'm not sure if the others use any sensu stricto liquids either; IIRC from watching teardown videos they just have some slightly moist paste in them. Perhaps atmospheric water is somehow involved too. I don't know, I simply try to be careful with any NiMH and don't leave them uncharged for years.

[Dgouty]

I don't recall Dave trying to recharge them...
Guaranteed leakage, throw them on a charger.

[thm_w]

I don't recall Dave trying to recharge them...
Guaranteed leakage, throw them on a charger.

Some were reverse charged.
Charging at higher currents is something that is clearly wrong, its not failing for unknown reasons by testing that, excessive heat/pressure will be generated. The manufacturer guarantee would not apply.

[fanOfeeDIY]

I have up and down in my life as all have , and missed this video.

The biggest percentage of the reason of battery having leakage seems to be when two or more batteries are connected in series with the batteries which do not have the same remaining level left.

They suffered to find a good method of sealing electrode terminal and they used rice as one of the ingredients.

I head this when the battery was Manganese battery. The reason I remember this story is because they were talking about rise. The company was sending rise to factories in overseas. The employee of the factories were happy to have the Japanese rise for their dinner when the rise was left at the factories because it was very difficult time to find Japanese rise which is sticky short grain (which they have grown up with) in foreign countries. It is easy now, since there are many Japanese food stores now, but in 80’s and 90’s it was not the same and many places.

At least what they did was enforce tight quality control and through away of a battery which could not package in the same package from the same lot. The different lot seems to have slight variation. Having the package from all two or four batteries from the same lot for hoping to reduce the probability of different remaining level to be used in series.
This could be the reason I do not see much 3.7V soft laminated LiPo rechargeable batteries are not used in series without BMS when they want 7 or 8V.

I am not battery a person, this is purely what I have heard from who have worked at battery division long time ago