Batteries Acting Strangely

[SparWeb]

I have no idea how, but the overnight resting voltage of my 24V battery pack is hovering over 27 Volts.  It's midnight, it's damned cold (-20C) and the sun hasn't been up since 6PM.  Every evening my habit is to look in on the batteries, and after years of seeing them between 25 and 26 volts depending on usage this is quite a surprise.  Especially considering their behaviour in the past few weeks has had me worried.

About 2 weeks ago I noticed that the overnight voltage was sagging down to below 25V.  I took some detailed cell-to-cell measurements.  It seems the stack (these pile up vertically) is cold on the bottom making the voltage lower on the cells on the bottom of the stack.  No way to re-stack the cells since each pack weighs ~400 pounds and I can't separate the cells.  Well, that's what I thought the problem was, until today.

So that makes this sudden reversal of their behaviour puzzling.  The stratification effect is less apparent but it's still there.  Certainly doesn't account for an extra volt in the string (0.08V per cell).  It wasn't windy (turbine hasn't moved in 3 days) and wasn't really sunny today, either.  It usually takes several amps of trickle charge to keep them above 27 Volts.



This is a very old stack of batteries that I found discarded in 2008.  Here's a link to a post I made the first time I discharged them.
https://www.fieldlines.com/index.php/topic,130251.msg847588.html#msg847588
They are at least 20 years old.  I have been wondering when this old battery pack will quit.  Not yet it seems...

[mab]

Bit of a stab in the dark, but 'old agm' and 'stacked vertically' make me think they may be getting critically low on moisture - i speculate this might cause increased voltage sag under load, but conversely give a higher rest voltage. But it does seem quite sudden for that.

If you can place them flat and remove the valves you could try adding a bit of deionised water?

I don't experience -20c do can't guess what that might do.

[SparWeb]

After a little investigation, problem "solved"...  meh.

It turns out that what I thought was zero wind was actually zero reading from the anemometer of the weather station - which is stuck.
Looking back at my wind records, I think it's been stuck for days.  It's been dreadfully cold this week which probably has something to do with the problem.  But it's also too cold to go mess around with the Wx station tower to find out what the problem is.  So I have to live with THAT problem until it warms up.

So the batteries seem to be OK after all.  When I came out to check on them last night, a wind had just died down, and the turbine had been turning an hour before.  Float charge voltage ~27.2V was normal.

[kitestrings]

Pheeew, huh?  With all the things that can go wrong on any given day, there's nothing like a faulty gauge, or something you're accustomed to relying on for a benchmark, to be wrong, stuck or, oh how about intermittent, just to mess with you.  Glad to hear it wasn't more serious.

[SparWeb]

The anemometer is still screwed up though.
https://www.wunderground.com/personal-weather-station/dashboard?ID=IALBERTA147
And no prospect of warmer weather for another week.

[SparWeb]

But while we're on the subject of these batteries, here's what I've been seeing over the years, as I have watched the health of my battery bank.

Having 24 cells to measure, and rather quick process to put a voltmeter on them all and write down the voltage in each is rather easy, but I admit I don't do it very often.  Somewhere along the way, I realized that I could use some statistics to better judge how much "scatter" there is between these cells.
I installed the batteries in 2007.  Before I got them they had been abused, so it took some time for the scatter to come down to a rather low level.  Since then the scatter has crept up slowly.  Last year for some reason it jumped up, but hasn't gone up since then.
All of the values on the chart are the standard deviation at float charge.  I also measure at static charge (at least 6 hours since last charged) but I think mixing the numbers would confuse the picture.  This is clear enough.  I also plotted the range between maximum and minimum, because there are a few bad cells.  If they ever start to get really bad, they will drag the range up before they affect the standard deviation.  They are the culprits for the increased scatter starting in 2017.



The standard float voltage for these AGM lead-acid battery cells is 2.25, so a 0.02V standard deviation is only 1%.  Not bad for 20-year-old cells!

[Bruce S]

https://www.wunderground.com/personal-weather-station/dashboard?ID=IALBERTA147
And no prospect of warmer weather for another week.

Don't take this the wrong way,,,BUT,,,,
I wish y'all would stop sending all that cold wind thru the Rockies on it's way to Missouri  .
My knees are kinda tired of of -30F winds.

It is good to see it was only your anemometer!

Cheers
Bruce S 

[mab]

Aha! That does make more sense than loss of water - did seem like a long shot to explain 27 volts without charging. Good to know it wasn't anything major.

Interesting to see the stats on your agm - i don't do any of that with mine - but as the hydro works 10 - 12 months of the year my ageing battery doesn't normally have to do much. I just keep an eye on the ammeter in case it starts drawing more float current than it should.

[Mary B]

The anemometer is still screwed up though.
https://www.wunderground.com/personal-weather-station/dashboard?ID=IALBERTA147
And no prospect of warmer weather for another week.

Mice was reading weird... went to look... top bolt of the mount snapped in the cold(steel gets brittle at -31f!) and it is hanging at 70 degrees pointed down form the bottom bolt and the safety chain I added just in case of a failure. It will stay that way until spring, I need to move it anyway!

[SparWeb]

Shoot - everyone's feeling the cold.  Here to say a while yet, too.
The anemometer cups are spinning, so I think they ejected their little magnet at some point, possibly when the thermal contraction of the plastic got too far from the contraction of the metal.