Attempt 'rejuvenating' old gel-cell lead-acid batteries with a bench supply?

[RoGeorge]

Sloppy English from my side, sorry.

Not voltage variations, I was asking about the voltage jumps of about 1.2V back and forth, like the ones seen in the lower left corner of the chart. (The smooth voltage increase up to positive voltage is what happened after disconnecting the battery)

[RoGeorge]

After more than 24 hours left in reversed polarity charging, the voltage still jumps back and forth about 1.2V, while the current is stable at 180mA, as set.  Temperature was constant all the time, the elements doesn't heat, they are between 1.5 ... 2.5 Celsius more than the room temperature.

Looking at the chart, it looks like an element (or maybe two of them) sometimes are short circuited, sometimes not, and it does that by themselves.

Anybody noticed something similar during charging a Lead-acid with reversed voltage polarity?
Any ideas why those voltage jumps are happening?

[RoGeorge]

Got bored after about 30 hours of reversed polarity charge, the voltage kept jumping up and down in steps of 1 or 2 cells (1.2 or 2.4V).

Switched the plugs to charge with normal polarity, 15V/180mA limited (2% of C), through a protection diode.  Over nigh the voltage was stable at about 7...9V, still jumping but not so often.  Will add a chart for that period, too.

Then, the 180mA current limiting was removed, and the battery started to draw about 2A at first, and after a few hours, about 1.5A.

Sometimes the battery is fizzling and popping.  Looking at the voltage with an oscilloscope, there are periods with cleaner voltage, and sometimes strong noises with an increasing in amplitude, sometimes up to 2Vpp.

The moments when oscillations suddenly stop seems to be correlated with the popping while the increasing of amplitude is correlated with the fizzling.  Strange that the audible fizzling is almost constant in amplitude, while the oscilloscope always show an increasing amplitude.   

Just for the fun of it, the random and increasingly oscillations, and another cleaner trace for the periods when the battery doesn't fizz and pop:

[mikerj]

I wasted many hours in my youth trying to rejuvenate dead lead acid batteries using various charging regimes and chemical methods (Epsom salts iirc) and the results were never worthwhile.  Some of them would accept a charge, but they invariably had low capacity and significantly raised impedance so they weren't very useful.

I managed to hurt a virtually new car battery (VARTA) by inadvertently leaving the headlights on for about 10 hours, and whilst it recovered sufficiently to start the car just that one event had clearly dented it's performance as the engine never turned over quite as enthusiastically.

[RoGeorge]

I don't hope to recover the battery and don't even need it, just being curios to log the data and see what happens.   

For this particular battery seems like 2-3 elements have an internal short-circuit, probably some dendrites between plates, since the voltage stays at about 9.4V when disconnected from the charger.

I would like to probe the voltage for each element, but the connections between elements are not visible.  Couldn't find any pics that shows where are the internal bridges between elements.  I want to drill some small holes right above the internal bridges, so I could measure each element's voltage.

Anybody have a picture of the inside of a typical 7-8Ah UPS sealed lead-acid battery, so to locate where exactly are the interconnections?

[tautech]

I don't hope to recover the battery and don't even need it, just being curios to log the data and see what happens.   

For this particular battery seems like 2-3 elements have an internal short-circuit, probably some dendrites between plates, since the voltage stays at about 9.4V when disconnected from the charger.

That certainly doesn't look good and I would chop the top off it and use it for a build study. There might be some good cells in it but some of them are certainly pooped !

Meanwhile a 22AH 12V SLA has found its way onto my bench and it measured 0.6V after an unknown period of inactivity. It's out of a portable Pelican floodlight system. Refused to accept charge from the system charger.

Siglent SPD1305X lab PSU selected for attempted recovery.
Reverse charge started some 15 hrs ago @ 2% CC.
Currently drawing ~500mA......~30mA fluctuations reported by LCD readout.

[RoGeorge]

Best luck with your battery!  Mine certainly has at least 3 bad elements.   

Charged it yesterday, and over night it dropped from 11V to only 7V.  At a closer look, some rocky gray pebbles about the size of a corn or rice grains can be seen on top of the elements.

Took one small pebble out with a pair of tweezers made out of 2 toothpicks, and try to measure if it is conductive or not.  It was very funny.  Since the pebble was soaked with sulfuric acid and certainly had some Pb in it, too, together with the tip of the probes it was making a tiny battery that completely bamboozled the ohmmeter indications.   

On the mV range, no matter where the little pebble was poked with the probes it was showing about -100..-200mV, always negative no matter how the probes were switched.   

Meanwhile, found out from some dismantling videos where under the lid those interconnecting bridges should be located :





Also found a paper about the main failure modes of lead-acid batteries:  Catherino, H. A., Feres, F. F., & Trinidad, F. (2004). Sulfation in lead–acid batteries. Journal of Power Sources, 129(1), 113–120. doi:10.1016/j.jpowsour.2003.11.003
https://sci-hub.tw/10.1016/j.jpowsour.2003.11.003

Sulfation is only one of the many other ways of failure, and doesn't seem to be the most frequent one, according to the paper:

•Loss of electrolyte.
•Electrolyte stratification.
•Hydration.
•Positive grid corrosion.
•Internal shorting.
•Passivating lead oxide film formation at positive currentcollectors.
•Agglomeration of finely divided lead in the negative elec-trodes.
•Electrolyte contamination.
•External damage to case and terminals.
•“Sulfation” (as a recrystallization effect) occurring in veryold batteries.
•Inter-cell connector failure.
•Positive electrode active material softening and shedding.
•Others (unique to recombinant battery systems)
    ◦thermal runaway,
    ◦hydrogen accumulation consequences,
    ◦lead sulfate accumulation on the negative plate

Will probably open the battery out of the curiosity to see how bad it is looking on the inside, not sure when, since it's cold outside and indoors I don't want the mess of a dismantled SLA battery.

[tautech]

I don't hope to recover the battery and don't even need it, just being curios to log the data and see what happens.   

For this particular battery seems like 2-3 elements have an internal short-circuit, probably some dendrites between plates, since the voltage stays at about 9.4V when disconnected from the charger.

That certainly doesn't look good and I would chop the top off it and use it for a build study. There might be some good cells in it but some of them are certainly pooped !

Meanwhile a 22AH 12V SLA has found its way onto my bench and it measured 0.6V after an unknown period of inactivity. It's out of a portable Pelican floodlight system. Refused to accept charge from the system charger.

Siglent SPD1305X lab PSU selected for attempted recovery.
Reverse charge started some 15 hrs ago @ 2% CC.
Currently drawing ~500mA......~30mA fluctuations reported by LCD readout.

Best luck with your battery!  Mine certainly has at least 3 bad elements.   

Thanks. 

Progress report
~48hrs reverse charge in CC mode, the PSU was displaying ~12.5V still in CC mode with a target of 14.5V
Reverse mode charge terminated and battery rested some 10 minutes where it measured ~9.5V reverse polarity.
Warm 25C day here today and battery temp was just warm to the back of the hand.

Application of a 30W 12V incandescent bulb load caused the reverse voltage to fall away completely and wouldn't light the bulb. Yet it bounces back to ~9V. Didn't measure current taken by bulb load.

Forward charge commenced.
Target 14.5V @ 400mA in CC mode.
PSU reports 0.0V @ 400mA  for a few minutes until the forward polarity becomes established then the voltage starts to count up and has already reached ~8.5V in a few short minutes.
Due to the residual heat from the reverse charge the charge rate in CC mode will be halved to 200mA to ensure the battery stays at acceptable temps.

Fingers crossed for this one ! 

[Circlotron]

^^ I don't think I'll sleep tonight!

[tautech]

I don't hope to recover the battery and don't even need it, just being curios to log the data and see what happens.   

For this particular battery seems like 2-3 elements have an internal short-circuit, probably some dendrites between plates, since the voltage stays at about 9.4V when disconnected from the charger.

That certainly doesn't look good and I would chop the top off it and use it for a build study. There might be some good cells in it but some of them are certainly pooped !

Meanwhile a 22AH 12V B.B.Battery brand SLA has found its way onto my bench and it measured 0.6V after an unknown period of inactivity. It's out of a portable Pelican floodlight system. Refused to accept charge from the system charger.

Siglent SPD1305X lab PSU selected for attempted recovery.
Reverse charge started some 15 hrs ago @ 2% CC.
Currently drawing ~400mA......~30mA fluctuations reported by LCD readout.

Best luck with your battery!  Mine certainly has at least 3 bad elements.   

Thanks. 

Progress report
~48hrs reverse charge in CC mode, the PSU was displaying ~12.5V still in CC mode with a target of 14.5V
Reverse mode charge terminated and battery rested some 10 minutes where it measured ~9.5V reverse polarity.
Warm 25C day here today and battery temp was just warm to the back of the hand.

Application of a 30W 12V incandescent bulb load caused the reverse voltage to fall away completely and wouldn't light the bulb. Yet it bounces back to ~9V. Didn't measure current taken by bulb load.

Forward charge commenced.
Target 14.5V @ 400mA in CC mode.
PSU reports 0.0V @ 400mA  for a few minutes until the forward polarity becomes established then the voltage starts to count up and has already reached ~8.5V in a few short minutes.
Due to the residual heat from the reverse charge the charge rate in CC mode will be halved to 200mA to ensure the battery stays at acceptable temps.

Fingers crossed for this one ! 

^^ I don't think I'll sleep tonight!

~24hrs with forward charge of which ~7hrs overnight was @ 200mA.
Charger displaying ~12.57V @ 400mA and the rate of voltage increase has slowed somewhat indicating good charge absorption. Battery barely warm to the touch despite a hotter day here than yesterday.
A few minutes rest and it measures 11.8V. 

14.5V target 400mA CC forward charge resumed........

[RoGeorge]

A few minutes rest and it measures 11.8V

Looks like one element is bad.   

[tautech]

A few minutes rest and it measures 11.8V

Looks like one element is bad.   

I don't draw that conclusion just yet as the recovery is a while from finished, maybe even 2 full days yet.
CC voltage was 12.57 ~2hrs ago and it's still slowly rising @ the 400mA CC charge. Currently (excuse pun) 12.62V.

The rising voltage indicates a rising SOC and while I might have the target voltage a tad high @14.5V when the battery gets to ~14V the current draw should reduce indicating a high SOC. If not we start looking for the voltage point where the current reduces which should indicate if all cells are behaving themselves.
Ideally we'd like to see a resting static voltage at a high SOC @ 2.2V * 6 =~13.2V. Anything above 12.8 I'm happy but load tests are the proof of any real success.

Others I have tried but older worn batteries have dropped their guts after a couple of days rest to measure sub 10V and they're not worth wasting more time on. 
So for now we just wait and watch.......

[Circlotron]

Others I have tried but older worn batteries have dropped their guts after a couple of days rest to measure sub 10V and they're not worth wasting more time on. 

Yep, I never had particularly good results trying to revive a worn out battery. Mine were brand new unused but had sat around for 5 years or so and had a terminal voltage of about 2V.

[tautech]

I don't hope to recover the battery and don't even need it, just being curios to log the data and see what happens.   

For this particular battery seems like 2-3 elements have an internal short-circuit, probably some dendrites between plates, since the voltage stays at about 9.4V when disconnected from the charger.

That certainly doesn't look good and I would chop the top off it and use it for a build study. There might be some good cells in it but some of them are certainly pooped !

Meanwhile a 22AH 12V B.B.Battery brand SLA has found its way onto my bench and it measured 0.6V after an unknown period of inactivity. It's out of a portable Pelican floodlight system. Refused to accept charge from the system charger.

Siglent SPD1305X lab PSU selected for attempted recovery.
Reverse charge started some 15 hrs ago @ 2% CC.
Currently drawing ~400mA......~30mA fluctuations reported by LCD readout.

Best luck with your battery!  Mine certainly has at least 3 bad elements.   

Thanks. 

Progress report
~48hrs reverse charge in CC mode, the PSU was displaying ~12.5V still in CC mode with a target of 14.5V
Reverse mode charge terminated and battery rested some 10 minutes where it measured ~9.5V reverse polarity.
Warm 25C day here today and battery temp was just warm to the back of the hand.

Application of a 30W 12V incandescent bulb load caused the reverse voltage to fall away completely and wouldn't light the bulb. Yet it bounces back to ~9V. Didn't measure current taken by bulb load.

Forward charge commenced.
Target 14.5V @ 400mA in CC mode.
PSU reports 0.0V @ 400mA  for a few minutes until the forward polarity becomes established then the voltage starts to count up and has already reached ~8.5V in a few short minutes.
Due to the residual heat from the reverse charge the charge rate in CC mode will be halved to 200mA to ensure the battery stays at acceptable temps.

Fingers crossed for this one ! 

^^ I don't think I'll sleep tonight!

~24hrs with forward charge of which ~7hrs overnight was @ 200mA.
Charger displaying ~12.57V @ 400mA and the rate of voltage increase has slowed somewhat indicating good charge absorption. Battery barely warm to the touch despite a hotter day here than yesterday.
A few minutes rest and it measures 11.8V. 

14.5V target 400mA CC forward charge resumed........

Yeee haaaa !
We might finally have a recovery !

After some 40 hrs of forward charge with 14.5V target the target voltage is now met and the charge rate has dropped to 220mA from a CC 400mA setting.
Battery cold.
After a short rest it measures 12.5V.....a bit low for my liking. 
Back on charge.

@Circlotron
I'll try some discharge/charge cycles next, what minimum discharge voltage do you recommend ?
I do have a 30A programmable load........

[T3sl4co1l]

How about zapping it with some high current pulses (say >10A charging, or >50A discharging)?  May be able to clear the short (fallen material/dendrites?), not that it'll last of course.

Tim

[tautech]

How about zapping it with some high current pulses (say >10A charging, or >50A discharging)?  May be able to clear the short (fallen material/dendrites?), not that it'll last of course.

Tim

I do have a old DIY desulfator that emits 1KHz 60V @ 7A pulses and plan to give it an overnight session at some point. It has worked pretty well in the past improving static (resting) voltage levels but it's only designed to address plate sulfation that isn't moved in normal charging.

[Circlotron]

@Circlotron
I'll try some discharge/charge cycles next, what minimum discharge voltage do you recommend ?
I do have a 30A programmable load........

Mmm... I'd say no lower than 11 volts and not more than 0.2C discharge rate. Just take it easy for the first few cycles at least. Just my gut feeling, not scientific.

[tautech]

@Circlotron
I'll try some discharge/charge cycles next, what minimum discharge voltage do you recommend ?
I do have a 30A programmable load........

Mmm... I'd say no lower than 11 volts and not more than 0.2C discharge rate. Just take it easy for the first few cycles at least. Just my gut feeling, not scientific.

It's not as good as I might have wished and after 24rs standing 12.0V ! 
Some gentle cycles and some desulphating are the next order of business.......

1 minute later.......
Nah, it's dead Jim !
Won't even withstand a 30W load !   

[RoGeorge]

Today, at about 5 days of rest after charging, the total voltage was only 5.77V, so just out of curiosity to see how each element is doing, but without dismantling the whole battery:
- drilled a 1mm thin hole above each of the internal bridges between elements (watch the video above to identify bridges location, they are on the external sides, at about 1/8 of the total battery width)
- measured the voltage and the short-circuit current of each element

Code: [Select]

Element      Open          Short
   #        Voltage       Current
(+ to -)      (V)           (A)

    1        0.8828        0.295
    2        1.1397        1.711
    3        1.8817        2.755
    4        1.8736        7.338
    5        0.0052        0.000
    6        0.0017        0.000
       
Total        5.7847

The 4^th element seems strong, it can deliver more than 7A of current when shorted, and since the wires were pretty long, most probably it can supply much more than 7 Amps.

The surprise was that none of them has the expected 2.4V or so (for a charged element), not even the strongest one, which led to the conclusion that the self-discharge rate was very high.

Would have been nice to poke the measuring holes earlier, fit a self-driving screw in each of the 5 internal bridges, then log the voltage on each of the six elements over a few days, during charging and self-discharging.
 

[CatalinaWOW]

All of this work supports the recommendation of several people in this thread.  That it is seldom, if ever, worth trying to rejuvenate these things.  It is a lot of effort.  It only "works" occasionally.  And when it does work the result is a substandard battery.

The best way to get some use out of these things is to send them to the lead recycling facility.

[Circlotron]

You must have never felt the satisfaction of bringing a battery back to life. It’s like fixing an old piece of equipment that everyone else says ought to be scrapped. But, to each his own.

[CatalinaWOW]

I have. I have restored both lead and nicad chemistry batteries.  And dis-assembled LiPos to get good cells out of the packs and build into "good" packs.  But each time the smile was wiped off my face in a few days or weeks when the "restored" battery failed again.  In a few cases I even revived those failures.

If you have any access to a supply chain to get new batteries and if restoring old batteries is not your hobby nirvana do it often enough to learn the limitations and then don't do it again.   But if restoring these batteries is what floats your boat, go for it.  You will have lots of opportunity for the joy of trying and occasionally succeeding.

[BradC]

You must have never felt the satisfaction of bringing a battery back to life. It’s like fixing an old piece of equipment that everyone else says ought to be scrapped. But, to each his own.

I can honestly say I've never felt the satisfaction of bringing a battery back to life. I have "revived" batteries sufficiently to perform a severely underrated role (Like the 7.2AH SLA that holds ~1AH @10V and runs solar lights). But I've never got anything close to rated capacity or internal resistance.

Splitting up packs of NiCD/NiMH/Li-ion and picking the eyes out to repair a pack, certainly. But again, at a reduced capacity of some sort.

There are advantages to that sometimes. I have a 12V 17AH SLA here that I let my boys play with. It'll hold ~12V with a <30W load, so they light up leds, 21W car tail lights, run DC fans and all sorts of stuff, but it won't set fire to the carpet like the battery I had as a kid that ignited an aluminium bicycle spoke I shorted it out with. It saved me building some form of current limiter and they can play to their hearts content. It even holds > 10AH, but the voltage just collapses under load. Does the battery have any real world use? Not really. It's old and tired (like me).

This is one of those topics that comes up every few years, and it always generates some enthusiastic participation, but in the end the result is universally the same unless you count "bringing back to life" as 1/10th of the original capacity, will self discharge in a week or two and has a high internal resistance.

Lead acid is a pretty well understood chemistry and construction. There really are no magic bullets.

Edit: Meant to add. Replacing the electrolyte in a flooded lead acid with pool acid (~30% hydrochloric) actually generates some neat potential with the downside of also generating non-insignificant proportions of Chlorine gas and HCl vapour. Still, I had to learn the hard way.

[RoGeorge]

Poking a cadaver might seem gross at first, yet valuable as a study 

From the last check, the 3rd and the 4th elements appeared to be in reasonable shape, one was capable of about 3 Amps short-circuit current, the other, 7A.  Both were way under the expected voltage, at only about 1.9V with no load.

Yesterday they were refilled again with water, then plugged them to charge for about 24h at 0.5A.  Not the best idea, because during charging the electrolyte overflowed a little.  Should have replenished the water after charging them, not before.  Paper towels helped to not make a mess out of this.

Now, 24h later, elements 3 and 4 are about the same voltage, 2.0V each when left with no load, but when short-circuited the max current is still slightly different, 10A and 13A.  The difference in max current is not so big now, maybe the previous 3A vs 7A was because they were not charged enough.  Will keep charge them till tomorrow.

However, until now more than 20Ah were poured in each element (the battery is only 9Ah), so by now they should be fully charged, yet the voltage is only 2.0V instead of the expected 2.4V of a fully charged element at room temperature.

At this point I suspect there was so much sulfation that a big percentage of the initial H₂SO₄ in the electrolyte was fixed in the PbSO₄ crystals, so now the electrolyte does not have enough H₂SO₄ left, and that is why the open circuit voltage can not go bigger than 2V no matter how much those elements are charged.

Should either add some sulfuric acid (which I don't have any), or maybe find a way to break the PbSO₄ crystals stored on the plate as sulfation, and make them dissolve in the existing electrolyte.  Normal charging can not break large PbSO₄ crystals.  Will see about that later.

Even so, with only 2V, there should still be about the same energy stored as it would be stored by a 2.4V element, just that the max current will be lower.  Didn't measured the C yet, so for now this is just speculation.


Not sure for how long this screws in the pic will last, but if the location is just right, self driven screws on an initial 1mm drilled hole right above the interconnection bridges can make great terminals to temporarily access each element. 

[SeanB]

With NiCd cells I made a rejuvenator for cordless phone batteries, as I used a Panasonic cordless phone ( KXT 9080A,  the version that was not type approved in many countries, because it had over 1W of RF output, unlike the 50-100mW of the others in the range, meaning it worked with a range of around 1km), and repacking the batteries every year or so because a cell was dendrited was annoying. It was simple, an old transformer that provided around 60VAC, and a single diode charging up a recycled capacitor, around 470uF 160V IIRC, that was discharged into a pair of nuts and bolts attached to the front panel at the right spacing to fit the battery contacts. A old industrial push button assembly to connect the capacitor to the terminals, and I had a quick and dirty dendrite remover tool.

Worked well enough, in that a zapped pack would not have a dead cell any more after a zap or three, measuring after each pulse with a DMM to see the open circuit voltage go up by 1V or more. Zap, check and then charge overnight, then use my discharger, a small piece of galvanised steel angle with a 10R 10W aluminium cased resistor screwed to it, with a 6V grain of wheat bulb across it, and 2 stiff copper wires to fit the battery.  Place across battery and leave till lamp is out, then charge, and if giving low voltage from phone, zap and charge again. Saved me from buying a lot of AA NiCd cells, as the most common types of tagged AA cells I could get were just too long to fit the packs.

That cordless phone lasted well, the others all eventually were killed by people, so I had plenty of spare parts to keep mine running, while the others were replaced with what we jokingly called " Same Room" phones. expensive to repair though, back when you could actually get spare parts for equipment, and Matsushita actually had a repair centre worth the cost. Lots of main boards, front flaps and flex boards, because people smash things not theirs. I must have paid their tech his salary some months.