Battery cell going bad

[SparWeb]

My big stack of batteries have really stood the test of time.  They were found discarded in 2007 after they were removed from a building's battery backup (think grocery store freezers, but I don't know the details).  They may have been made in the early 2000's, making them about 20 years old now.  Not bad for lead-acid, but these are a premium brand (GNB Absolyte 90A-11).

Recently, my semi-regular checks of their condition started to reveal the bad news.  There's always one bad cell that dies first.
It happens to be #8, as you can see in the trends below:



These batteries have been the heart of my RE system, and I would not have committed so deeply to this hobby if I hadn't found them.  I tried to give them a good home, and for many years they've been superb. 

I haven't taken anything part yet.  This photo is from moving day, 2013.  Just to give a better idea what they look like, and how they are mounted in the casing.


It may be possible to open the containers and pick the best.  They are currently wired in 2 parallel strings of 6 cells.  Each cell is 2V, making this a 24V system.  I could choose the worst 12 to discard, and keep the rest for a single string that would still give me 24V.  Removing them from their cases will be difficult.  Old batteries always swell a bit, and these cells, good and bad, are tightly squeezed against the walls of their cases.  I would have to cut the steel case from the outside, then shut it again somehow, after mixing-and-matching the remaining good ones.

[SparWeb]

The other curious thing in that chart is a "bad day" that happened in the spring of 2014.  It was a surprisingly hot and windy day, and the insulated box was still closed (had been all winter).  A lot of amps passing through when they couldn't lose the heat - and there may have even been a bad setting on the diversion charge controller.  When I checked on them I could tell something was wrong - the temperature inside the box was about 40C. 

There didn't seem to be any long-term bad effects from that mistake.  Cell #8 has always been the weak one, and even before that day, it was usually the lowest.

[Mary B]

(Attachment Link)

My big stack of batteries have really stood the test of time.  They were found discarded in 2007 after they were removed from a building's battery backup (think grocery store freezers, but I don't know the details).  They may have been made in the early 2000's, making them about 20 years old now.  Not bad for lead-acid, but these are a premium brand (GNB Absolyte 90A-11).

Recently, my semi-regular checks of their condition started to reveal the bad news.  There's always one bad cell that dies first.
It happens to be #8, as you can see in the trends below:

(Attachment Link)

These batteries have been the heart of my RE system, and I would not have committed so deeply to this hobby if I hadn't found them.  I tried to give them a good home, and for many years they've been superb. 

I haven't taken anything part yet.  This photo is from moving day, 2013.  Just to give a better idea what they look like, and how they are mounted in the casing.
(Attachment Link)

It may be possible to open the containers and pick the best.  They are currently wired in 2 parallel strings of 6 cells.  Each cell is 2V, making this a 24V system.  I could choose the worst 12 to discard, and keep the rest for a single string that would still give me 24V.  Removing them from their cases will be difficult.  Old batteries always swell a bit, and these cells, good and bad, are tightly squeezed against the walls of their cases.  I would have to cut the steel case from the outside, then shut it again somehow, after mixing-and-matching the remaining good ones.

Sand along the top sides of the battery cases, where the top lid flap folds over the side. Bet you find some spot weld you can use a spot weld cutter on, or careful grinding with a dremel to remove the spot weld. Top should lift off then. That is how I opened a set fora friend, he cleaned out an old telephone company switch building and it had a stack of similar batteries. Most were bad but we managed to put together 2 strings he used for his ham station.

[noneyabussiness]

40c shouldn't be a problem,  lol, we get to 40c ambient temps quite regularly over summer, haven't had an issue yet, i have a ex forklift battery with very similar cells... sounds more like age, aka " cell shedding " , normal for lead acid... just a thought...

I would definitely pick the best 12 to continue using, scrap the rest and use said money against a new bank....

[Scruff]

Impressive Sparweb.
And that inverter is running as long I take it?

That's the wrong kindov switch for an inverter it doesn't react fast enough and the inverter caps are probably blowing the contacts to pieces every time you connect it.

It it possible to reprogram that Trace to 22V?

[SparWeb]

Mary, they seem to be seam welds along the entire lid, and steel dividers making a compartment for each cell.  But basically that process - just way more work than you and your boundless optimism would have me believe.

Scruff,
Count on you to try something wild!  I don't think the Trace would let me go that low - or it would kick off with low-voltage disconnect if I pulled much current out of the system.  Yup, it's an antique, too.  I bought that used about 2010, after I got these batteries, and it was plainly obvious that RV inverter didn't do these batteries justice.  Pretty sure it was made in the 90's.  Doesn't have the stickers on it for a code-compliant grid connection these days, but suitable for my isolated system.

The switch you point out is not suitable for disconnecting current.  Despite being red it's not for emergency use.  I would not disconnect current with that switch.  It's for days I'm opening the case of the inverter, cleaning it, checking connections, etc.  I've only turned it a handful of times, and only when the Trace was Off, first.

[Scruff]

That there is wan oh them rare don't-make-'em-like-they-husta-machines! I'm not at all surprised it's still truckin'
Not many like it. You can taste the quality from the hernia you get lifting it.


Worth a shot SparWeb;



It'll only cost you a few button presses an two jumper leads to find out.
{Change Inverter LVD too}

That switch isn't for current aye. The inrush from the capacitors of that machine is well in excess of welding current.


PS. you don't have to break the cage to rewire the pack.


For my next trick I'm gonna grid tie a campervan. 
...did someone say 3 ton micro-inverter? 

[Scruff]

[SparWeb]

Hmm
Stop it before you convince me to do something crazy.

Yes, it has occurred to me that I can make some heavy jumpers with welding cable and fat crimp-lugs.  I'm also leery of keeping an disconnected cell in the pack with the rest of them.  It will be flat, the rest will be happily charged, then someday it gets really cold, it freezes and barfs acid over EVERYTHING.

There's also the corresponding problem with the charge controllers.  They definitely CAN'T be convinced to do anything lower than 25V.

You found the last of the "good" Trace manuals.  Later versions of that manual were re-written by Schneider and completely illegible in any language including English.  See page 5 for the health recommendations.

[Scruff]

Rewiring a half pack is about the same work as 50% cull to be honest except breaking the cages won't cost you anything and you get scrap metal.

I hadn't thought of the freezing...but they're agm...and the stuffing would be frozen so hard and not runny...meh...solar battery heaters bro..they're the latest thing! 

What are your charge controllers? All of mine can be tweaked.
Wanna buy a diversion controller off me?  I have 60A and 45A flavours for sale. Zero load hours on both. Special Mod discounts available.. I'll even preprogram them for 22v and include a free remote temp sensor.

Trace are unrivaled. They're the Toyota Land Cruiser of inverters.
Schneider are good when it comes to commercial & industrial. I won't touch them for Off-grid they're too institutionalised about things like acceptable quiescent currents, electronic noise pollution in the audible frequency spectrum and ticky box stuff.
Off-Grid hardware designed in an office, no thanks.
Besides they assimilate a company every month...they're not interested in small markets.

[Scruff]

D'oh.
I can't help with the Xantrex other than offer you a mate's rate TS45.

I can send you a custom profile for that TS-60

[SparWeb]

You're a champ, thanks for offering Scruff!
I'm just conservative about some stuff, I guess.  If the program's all weird, then I don't know what the "happy" numbers are any more when I check on the equipment every evening, neither will anybody else if they have to touch something, and I'm old enough that, 5 years from now, I'll have forgotten enough details about what I did that there's no way to adjust something by then.

That Trace inverter.  I should have bought 2.

[Scruff]

Thuther option being scrap a ~half ton of working lead.

I'd go with reprogramming and some labels. I don't really use Voltage for SOC anyways. 12v makes sense to me, 24v kinda makes sense but I'm too lazy to multiply everything by 2. 48V doesn't. 59v LFP kindov used to it now, 12V LFP good > 12.8V > bad.

No being a pushy sales type but it would seem that if you invested in another TS-45, a RS232 to USB dongle and a cheap battery SOC monitor

You'd be about 20kWh better off than the alternative and the extra hardware can always be sold later if you revert to a more trad setup, or sell the Xantrex or have a second diversion controller.

Yurp. Those Traces are getting rare alright.
Studer are about the only thing close these days but they'd want €400 for an interface etc..in keeping with the Toyate reference you'd be comparing a Land Cruiser to a G-wagon.

[Scruff]

Another super messy option is keep a good 24V battery and make a 12V with leftovers.

Then MPPT charge the 12V from the 24V with constrained LVD ~26V
Then to go back up you can you could use an inverter -> charger LVD ~12.4V.

Allov which makes me wince and think shenanigans new battery time.

I've seen cells drop off with age. There's seemingly random failures after ~7 years...it doesn't mean the whole lot are on the way out.

[SparWeb]

Not far from an option I'm thinking about.
I can split the bank into a 24V for one location, and a 12v for another location.

[kitestrings]

There's no chance of getting more cells of same make, config, and similar vintage I don't suppose?

[Scruff]

Even if there was the legacy cells got age on them.

It's like replacing one tyre on a car ...well worse.

I dunno Mr Web Sir one battery is always better than two....for lots of reasons...I could name a few more

[mab]

My understanding is that you have two 24v strings in parallel? As long as the bad cell doesn't go short cct I'd be tempted to leave it as is, maybe add a couple of ammeters so you can monitor charge and discharge balance between the strings.

I've got 3 strings of AGMs for 24v and my worst cell is down at 2.03v at rest, and whilst that string takes more charge than the other 2 and contributes less under load, it still seems to provide useful storage, keeping my depth of discharge to a lower %age of capacity, so i judge it's better to keep it than removing the whole string from the system.

[SparWeb]

I'm somewhat optimistic that I can get many more years out of the rest of the bank.  I wouldn't want one bad cell to harm the others before their time.  Isn't there a chance that it could go very suddenly?  One cell passes a point of no return, imbalances the strings, and then the other string drives more and more current until they get all heat-stroke.  Or something like that.

It's a long way to go if I want to visit any scrap dealer, just to see if *maybe* they have some cells such as these around.  Not that I haven't seen a stack of somewhat similar AGM's at a scrap hauler once (and I've only bothered going maybe 3 times in 10 years) but finding a matching cell is pretty long odds.  If I'm going to visit the scrap dealers, better to go prepared to make a deal to buy a full set if they have one, then see if it can be used or rehabilitated as a set.  Just move on from there, if I can.

[Scruff]

Combining good with bad means the bad slack off the good are overworked and powering self-discharge and often the cell voltages drift so the battery might be 12.8V but the cells might be 1 x 1.75V and 5 x 2.2v.

There's no reason to say the remainders won't last another 5 years.

The only time I've given a battery heat stroke was charging a 12v at 28v followed by a three week absorption ~300% capacity. You'll smell "rotten eggs" long before meltdown.

In terms of string health especially 48V the middle are always healthier and the outer ones by the terminals are always harder working and effectively older and tireder. It's to do with terminal heat and internal resistance. I periodically swap inside out and vice versa. With 6volters in 12v config you just swap them.

You notice these things with flooded batteries and hydrometer checks.

You might get more mileage going 12P12S instead of 2P12S. It'll balance better once you partner up suitable couples.
If you want to replace just one or two cells that'll work fine if you put a balancer on them.

 

[Bruce S]

Sparweb;
You can gently rewire this to use one bank while doing the repairs.
One of the things you can also do: Remove that 2V unit and turn it over.
From working with TomW when he first got his massive NiCads was to move them about.
From the way they are currently stored they may have gotten stratified.
Scruff works with Pb much more than I so he may tell me I'm way off.

BUT! At the very least I would by pass that whole bank keep the good bank going.
That one little 2V bank goes south and starts boiling the rest of the bank  just trying to get them up .

Best
Bruce S

[Scruff]

AGM don't really stratify, that's more to do with a wet system and denser mass sinking in solution. AGM electrolyte apparently doesn't move.
I don't see too many stratified cells. Most of what I deal with is mounted to a mobile vessel.
There are automatic stirrers for stationary flooded batts.


I do wonder about side stacking....the vents are designed to be upright.
I've done it, more than once, handy thing. I just wonder.

[clockmanFRA]

I had a lot of AGM batts when i started this lark.

All were second hand from mobile phone masts back up PSU. After about 8 to 10 years or so they all failed.

All of them failed when one cell would short out internally.  When i examined them, with wood Chesil and hammer, PLEASE DONT DO THIS FOLKS. The plates had swollen about 5% and upwards about 5% and that was sufficient to short out a cell plates on the internal busbars.   The 10mm standard clearance was just not enough.  No wet acid inside just matt with impregnated acid between the plates. 

[Scruff]

Weren't they the infamous yellar tops CM?

[clockmanFRA]

AGM's ,The First were US made Grey cased 110ah Marathons, then the same but made in Europe, then some large 230ah blue ones made in Spain, but they all split there sides as they shorted out, and finally the yellow 105ah Hawker types.  Remember no wet acid inside so no dripping.

All were in matching strings of makes/types and ah to give 48v.  Individually the lightest weighed 34kg.

Shame really if a better stronger case design and better internal clearance, i might have got another 5 years out of them.

But honestly constant checking each one nearly every day became a nightmare, and getting replacements was fast becoming expensive as the wheeler dealers got involved.   Infact, a second hand AGM had become the same price as a new LUCAS 110ah ultra deep Marine/golf cart battery.  With the wheeler dealer saying, "But New this AGM was over $350".

Anyways I am now working on the PLANTE lead acid battery.

  Exide still make PLANTE but the plates are thin.
 
  I am R&D a 2v multi plate cell battery at about 60kg at about 500ah each basically what one man/woman can lift on there own.
 
 Polypropylene is readily available and can be made/moulded to any size and thickness, and the plate separators is a woven polypropylene material made into a sock.   I am using rolled thick lead sheet that is roughed scratched surface, this roughing needs further research, for the correct oxide to form.   There are a lot of variables here even with the thickness against supporting its own weight, case support without shorting out on any sediment etc etc. What i am trying to do is for a cross section of the case is to pack the multiple plates tight in the case with their socks on and therefore keep the Acid content minimum.  Then there is expansion of the plates etc etc, anyway you get the jest .

It takes a minimum of 40 charges and 40 full discharges to build up sufficient oxides on the plate for the PLANTE battery to become efficient, so the more you use it the better the ah gets.  Life of the battery is about 30 to 40 years dependant on the plate thickness.

And yes i have built some test multiplate cells, but for some reason every bloody time i get the positive and negative the wrong way around, although technically it shouldn't matter with a PLANTE as what you decide positive then charge and discharge with one being positive the cell will remain.

I will continue the works slowly in my retirement very soon.!    Then i might thrash out another book 'How to Make a PLANTE battery' .     

[Scruff]

Go into production CM I can't get batteries anymore. All my suppliers are behind the Brexit curtain and the Paddies are Rip-Off merchants.
Charging €175 for 200 cycle engine batteries I used to be able to get 1200 cycle golf carts you can push waaayy harder for the same.

As regards battery failure I inherited 9 AGMs from an emergency lighting system for a theatre.
The system was "scrapped" due to warranty replacement.
They were tested once a month that they could deliver a 4-hour runtime for 5 years.

I put them on float on a MS PWM for 4-5 years

I put them into service recently. 1 was bad when I got them.
2 failed since.
The remaining 6 are displaying 100% rated.

I think the flow battery is the one we're all waithing for, that'll be the actual "game-changing, linchpin, missing link dam burster" if we can find enough vanadium.

[Bruce S]

I think the flow battery is the one we're all waithing for, that'll be the actual "game-changing, linchpin, missing link dam burster" if we can find enough vanadium.

I'm still diggin the LMB (Liquid Metal Battery) concept.

[Scruff]

I like that one a lot.

"If you want to make it dirt cheap, make it out of dirt...locally sourced dirt".

Can I have one that fits under the driver seat?