Charge efficiency sanity check...

[bobfandango]

So, we have a bank of 8 T-105's strung as 48V and being charged mostly by a generator through the inverter (Xantrex type).  The generator generally kicks on every 24-30 hours and has been consistently running for about 3 hours at a time.  Recently, although the discharge cycle has remained relatively constant at about 24-30 hours, the charge cycle is taking much less time, about 2 hours.  No settings in the inverter/charger have changed.  

I would expect charge efficiency to decrease over the life of the batteries (these are almost 2 years old) and not ever increase and especially not by this apparent amount.  At first, I thought that maybe one or more batteries weren't storing energy for some reason which would then require less energy to go into the bank.  But then, I wouldn't expect such a cell to maintain any kind of higher voltage indicative of a higher SOC nor would I expect the discharge cycle time to remain the same.

Is there any reasonable explanation for this behavior?  Something must have changed somewhere, but I'm at a loss for where to look...

[TomW]

bob;

I would first check the SG of the batteries.

Voltage is not a good indicator of charge altho that is likely what your generator / inverter combination triggers shut off with.

As an aside, when my old bank of Trojan T-105s finally tanked they would come right up to voltage fast then drop even faster.

The only way to tell for sure whats going on is to check the SG. Voltage definitely does not give a fair indicator of battery health when the battery is never idle as in your case where it is either being discharged or charged.

Just from my experience.

Good luck with it

Tom

[Ungrounded Lightning Rod]

Sounds like they need equalization.

 - When the least charged cell goes flat the voltage drops, starting the charger.

 - When the most charged cell goes fully-charged the voltage rises, stopping the charger.

So "discharged" is controlled by the least-charged cell and "charged" by the most-charged.  When they're far from equal charge this is essentially the capacity of the bank.  When they get out of equalization this is reduced by about the percentage of full charge that the most- and least-charged cells differ.

Note that you may have other problems as well:

 - Being undercharged for a long time can lead to sulfation in the undercharged cell.

 - One thing that can get the cells out of equalization is shedding, producing a high-resistance jumper between the plates and leakage that gradually discharges the cell.  (This usually happens near end-of-life, when the pile of shed stuff at the bottom of the cell stacks up to the bottom of the plates.)  If this is happening you'll keep going out of equalization and it may accelerate.

 - Sulfation raises the cells' resistance and may cause them to appear fully charged before they really are.

At this point I'd expect it mainly to be a lack of equalization.  I'd water 'em to make sure they're topped off and even, give 'em an equalizing charge, water 'em again (because they'll all have electrolyzed away some water, the more-charged ones more of it), and see if things are back to normal.

Then after a couple charge/discharge cycles (to mix the new water in) I'd check the specific gravity to see if they're OK.

I'd also examine the charger and see why it isn't equalizing them automatically.  (Maybe it requires manual triggering of the equalizing charge.)

If they've gone out of equalization expect them to do it again and somewhat more often as they age.  So be sure an equalizing charge now and then is part of their normal maintenance schedule.  You may find that they still have plenty of life in them if they're treated properly.

[Ungrounded Lightning Rod]

Blew an edit.  Should read:

So "discharged" is controlled by the least-charged cell and "charged" by the most-charged.  When they're close to equal charge this difference is essentially the capacity of the bank.  When they get out of equalization this "apparent capacity" is reduced by about the percentage of full charge that the most- and least-charged cells differ.

[bobfandango]

Good thoughts...  Thanks...

The funny thing is, the discharge time is staying relatively constant with approximately the same average load.  So, the apparent capacity appears to be the same overall.  Also, the cells are always topped off and the bank gets an equalization charge about every ten days or so.  

I wonder if the situation you propose could be the issue nevertheless.  The discharge rate is much lower than the charge rate and that may allow the battery with low SOC to maintain its voltage on the low end during discharge more readily than the battery with the high SOC is capable of being dragged down during charging and thus the overall voltage gets up faster...

Something is certainly amiss and, as others have suggested, we'll look at SG of all cells (not just all batteries) after the next equalization charge....  

[Ungrounded Lightning Rod]

Maybe they're starting to sulfate and it's fooling the charger.  (But that doesn't square with no change in discharge usage, either.

Deep-cycle batteries DO get better after the first FEW deep cycles.  But they should have done that a lot sooner than two years.

Another possibility:  Your genny's or charger's regulation has changed (higher bulk charge rate, earlier cutoff of finishing charge, shorter time or lower cutoff voltage for equalizing charge, etc.)

[fishfarm]

You noted that the battery charging was "mostly" by the generator. Has the other source(s) changed in some way? Also, have you checked to see if the generator fuel usage has remained the same or decreased?

[bobfandango]

Yes, mostly as in, there is also 480 watts of solar panels.  The weather, however, has been absolute crap lately and the total number of kWh into the batteries from that source is negligible.  For the panels to be the cause of this, they would have to be getting full sun during the entire charge cycle and they haven't seen full sun for some time now, much less during charging.  The fuel consumption has remained constant as well.  Anyhow, even if that were off for some reason, 25 amps into the batteries is still 25 amps no matter how much or little propane the generator is using.

The equalize settings on the inverter/charger are voltage based.  So, for this equalize, I think we'll do as Trojan advises and just maintain the equalize charge cycle until the SG stops rising in the cells...  Then, we'll know what the story is re: matching.