GEL battery drops voltage under load but not immediately.

[WoodsieLord]

Short version: One out of four batteries in series drops to 10.5V after a few amps.
----

A little background for context:
I have a bank of 4 alphacell 86A deep cycle GEL batteries that I got as new (unused) a year ago. Since I'm not living offgrid I've been using them pretty much as a super big UPS battery bank. I don't usually cycle them since I switch to solar when the sun is out and the bank is at 100% and switch back to grid before the sun sets. They're set in series to reach 48V.

I've run 5 "long" tests where I took them to 75% DoD (most of those times) and down to 59% once. Never went beyond that. Also, I did those tests at a constant 170W load (around 3,5 ~ 4A)


Two weeks ago, I replaced my PWM SCC for an Epever MPPT SCC. I read the manual before buying it so I knew beforehand that I had to configure it to use GEL charging profile (no equalization). The very next days were clowdy so I had to wait a few days until forecast was good for a proper test.
The night before the good forecasted day I started a drain test expecting to get near 55% ~ 65% DoD.
Three hours into the test and the UPS shut off. The bank had reached 42V (under load) right before the shut off.

(I collect this info through a voltage dividier then an Arduino then a raspberry pi then a home server. All DIY, so take into consideration that this might be inaccurate. Perhaps by 1 volt. It let's me see the "big picture".)

At first I did not suspect a battery was faulty so I spent some time upgrading the wires. I ran a second test with the same outcome.
I usually run these tests at night where my loads are small and very constant (load consists of a computer, wifi, security camera and some led bulbs occasionally used when going to the toilet! no motors, no surge spikes)

Yesterday, I ran a test under a slightly greater load (fairly big Lcd tv, more lights, audio amplifyier). This time, I was meassuring the batteries individually with a voltmeter every few minutes. One hour twenty minutes into the test and this single battery dropped voltage to 10.6V. (The other three were above 12.15V) I had seen enough so I aborted the test right there.

My batteries have no local warranty, and shipping lead acid batteries to USA/Mexico is out of the question. So today I stripped the label and took the vent plugs off to take some pictures.
All six cells look almost the same, I have never been this intimate with a proper GEL battery.... I don't know if there is something wrong or if there is nothing to see to the naked eye.

I saw a bunch of lousy GEL battery restoration videos. Basically they all add distilled water in. One of them extracted acid-water mix from a flooded 'donor'. I'm willing to experiment but I would rather ASK for advice first. So here I am. And here are the pictures I took:


Thanks in advance!

 

[Fred S]

Others should weigh in, but I see no other cause but a single bad cell. Could be from calendar aging on top of a defect, who knows.
Experimenting is cool, just do good homework and don't skimp on ppe. I hope you can fix or improve it!

 

[Deleted member 165]

Aren't all gel batteries sealed? How would you go about adding anything, and is distilled water appropriate?

 

[Berseker]

keenly following, i have 2 agms with same behaviour. research says the damage is irreversible,
how did ours pan out?

 

[WoodsieLord]

to be honest, a month has passed by and I still haven't added anything to the battery.
The only thing I did was check another battery to see differences. the GEL is clearly different but at the same time I can't tell if that alone is more than a hint.

Left: Good one. Right: Bad one. Different light conditions, I took the photos 1 day apart. It seems the gel in the left one is more bulky while on the right one is more "grainy". When I've touched it with the tester leads they felt about the same consistency. (I know the right one looks like crystallized moisten sugar but I think it was an effect of the led flash of my phone).


The most difficult thing that stopped me from going any further is taking readings from the cells. I tried using my multi-meter but I got very inaccurate results. I figured that depending on where I land the test leads I got different results. I disconnected the battery and tested it after it dropped to ~11V.
Most cells were pretty consistent and in good voltages. The only pattern I've noticed is with the cells in the extremes. One measured more than 3 volts and the other around 2 volts. The other 4 cells measured slightly different but all 4 over 2.2V.

Is there a more reliable way to measure cell voltages ?
Would you test the cells with the battery fully charged or almost depleted?

Aren't all gel batteries sealed? How would you go about adding anything, and is distilled water appropriate?

Yes, although they have some very tight "popping" plastic seals. They sit tight but I'm sure they're put that way in case the batteries are accidentally "overcooked" and they start gassing. That's my guess, though. About adding anything, I think it's not appropriate if I don't understand what is going on. If the battery is as good as a rock and at some point I come to a conclusion that a cell is kind of dried out.. then I'm willing to try.
Furthermore, I think adding distilled water would change the specific gravity of the electrolyte (although since it is in gel form it's not easy to read). That's why lately I'm thinking about adding electrolyte from a car battery (FLA). But first I need to know if there is a single cell failing. If so, which one. If so, should I add the electrolyte when the battery is full or not? etc.

Honestly I DON'T know what I'm doing. Before I started experimenting with solar I knew lead acid had its complexity. Now days, I realized that I was off by an order of magnitude when I said "complexity". It's quite interesting, though!

Thanks everyone!

 

[Berseker]

the little i know about gel chemistry, the gel electrolyte is in a semi solid state, and battery performance drops as soon as the gel deterioates/drys out.
conventional knowledge says once the gel drys out, its irreversible and battery failure & death imminent.

i have a flooded battery exhibiting similar behaviour. am thinking of draining the electrolyte. washing the insides very well..and topping up with fresh electrolyte. the batterys are 4 years old, guess i got nothing to loose at this stage, hould things go south

 

[MisterSandals]

batteries that I got as new (unused) a year ago.

Batteries have a shelf life based on time. Howe old are your “as new” batteries?

And, batteries can have defects that makes their life shorter than normal. As was mentioned earlier, it just takes one bad cell to get your results.

If you attempt to restore them i am sure your results, good or bad would be valuable.