diy solar

diy solar

Telecom Industrial AGM Batteries (Why Not)

Cool - just know that they have now gone in cyclic service - and as even the manual shows, they need to be returned to 100% SOC at least every 10 days. And for AGM's, whether big or small, that means a LONG time in float AFTER absorb has finished. At least 8-10 hours or more or you'll be walking them down in capacity with hard sulfation.

Hard to do with just solar psoc. So once in awhile, they need a no-load recharge - like a whole day to themselves while you watch frustratingly low float current pass.

Not realistic? So if you have a source of AC for a charger at least bi-weekly, then use it overnight.

An interesting thought is a "donor" smaller maintenance LFP battery (charged by solar perhaps) just to serve as this bi-weekly float with a dc-dc charger for the agm's. Sure you could do an LFP "add-on", but it still doesn't relieve you from that pesky agm maintenance.

Interesting thought using LFP as a "maintainer" for a large Pb bank once every week or so....
 
I don't see where your definition of 100% charge needing 8-10 hours of float after absorb agrees with the Powersafe charging instructions and charts. It also would not agree with AH counting as they describe. Float should be just to maintain 100% charge, NOT to complete a charging cycle. If your setup needs 8-10 hrs of float to reach 100% charge then its likely the absorb stage is completing prematurely. I spoke with a rep at Powersafe (Enersys) and he told me charging is complete when the absorb reaches around 3-4 amps at 2.40 volts per cell. That is exactly how my charging behaves. I also confirmed absorb voltage and float voltage I uses with them.

You may be detailing generic AGM procedures but on high end batteries like this its the best bet to go with the manufacturers recommendations.
 
Cool - not many solar users of TPPL, which was my first love before LFP. So I know you know based on the above.

That 8-12 hour recommendation for float was indeed for the "conventional" recycled lead agm, and for those who put a 5 watt panel on their 100ah batt, with an SCC that times out to float after an hour or two. :)

Either that, or they found an Odyssey or Optima shelf-queen previously abused in the garage somewhere...

However, the specs are always based on NEW, not used. In this case, instead of striving for a specific end-current value, which is valid when new, when aged/aging, I tend to watch for a stall in the CV/Absorb current that doesn't change for more than an hour as to when to stop (or switch to float if needed). So for me it isn't hard-coded and allows for aging / IR when watching for a stall. I'd only eyeball this once a year or so and make adjustments if needed.
 
It's interesting to learn more on the cyclic capabilities of these Powersafe batteries. It may be I can consider getting a little more functionality out of them than I expected. Not sure what though!
 
An interesting thought is a "donor" smaller maintenance LFP battery (charged by solar perhaps) just to serve as this bi-weekly float with a dc-dc charger for the agm's. Sure you could do an LFP "add-on", but it still doesn't relieve you from that pesky agm maintenance.
I was thinking along the lines of combining a smaller LFP bank with his existing lead to take most of the day to day cycling load off the AGM's, it would probably prolong their useful lives by quite a bit. That being said, even with DIY a 100 Ah 48V LFP bank will cost a fair bit.
 
However, the specs are always based on NEW, not used. In this case, instead of striving for a specific end-current value, which is valid when new, when aged/aging, I tend to watch for a stall in the CV/Absorb current that doesn't change for more than an hour as to when to stop (or switch to float if needed). So for me it isn't hard-coded and allows for aging / IR when watching for a stall. I'd only eyeball this once a year or so and make adjustments if needed.

You got me thinking. These Powersafe batteries are rated for 10-15 year lifespan under float conditions. Why would the manufacturer publish recommended charging/discharging specs that would only apply when the batteries are new (new= under 1 year?). They are only new until the day they are installed, basing specs on that would be pretty worthless. Going further, they do have a warranty so its in the companies interest to publish specs that will not hurt life expectancy or performance. I would assume the published specs are good values for field use for at least half the lifespan of the batteries if not more. Agreed that these are not your typical AGM battery intended for marine/golf cart/auto or similar service.

Its worth noting that the fellow I got these from has a high end conductivity tester, the preferred method of judging a batteries condition these days. The batteries all tested well over 90% of their new spec even being 5+ years old.

I have noted that my battery/inverter setup behave much like you describe. At the end of the absorption/cv stage the current ceases to drop (around 4-6A typically for my 340AH array) and remains at that level for at least a hour before toggling over to float mode. If there is a notable drop in battery voltage due to load the inverter reverts back to CV mode again.
 
Last edited:
Do you have a shunt meter to keep track of your actual depth of discharge?
 
Do you have a shunt meter to keep track of your actual depth of discharge?
I had one but it would increment on discharge and decrement on charging so it wasn't much good. The inverter has a internal shunt that reads battery current and power used so that works pretty well for my purposes. I have on occasion gone off line and only recharged the batteries to compare the power required for recharge to what I believed was actually used, the numbers compared favorably allowing 110% replenishment power.
 
On a practical level, I think you'll be fine. Some battery maker spec sheets will provide both an initial end-current value, OR a stalled timeout value, whichever comes first to account for aging.

But we're really reaching a level of esoterica that won't affect 99% of the casual user.

For instance, Enersys has a white paper discussing the narrow window of float service voltages for TPPL long-term UPS standby, otherwise degradation occurs. Ie, there is a difference in 10 years whether you float them at 13.4, 13.5, 13.6, and 13.7v. So the industrial user had better get this right - which Enersys published to help some realize why there were only getting 10 years instead of 13! :) But did your telco company float at too low or too high a value before you got them? The paper is out there, but I can't find it now. Enersys should have it certainly.

Ideally, once you notice the beginning of the stall in current, that's when you adjust either your current sensing logic, or shorten up your CV timer.

BUT, and this may seem esoteric, CV is not actually efficient chemically. The old "get that last 1% charged" or you may suffer capacity walkdown in cyclic service, it may be prudent as the bank ages to float after the initial CV stall as much as you can. And with solar, that might just be an hour or two before the sun drops, but it helps. The float may be super slow, but with aging, the efficiency is just a tad higher to try and nail that last 1%.

Again, 99% of the users will have moved on to other things or changed their bank because their needs have differed or aren't as exacting like us tppl propeller-heads like myself for this to even be an issue. :)
 
I saw some Leoch industrial batteries for sale around 200 dollars a KWH with 4400 cycles at 50%. 15 year design life So I would say 20-25 is possible with light use. Carbon based so less sulfation. LRC Leoch batteries.
 
Back
Top