sealed lead acid batteries float/charge voltage.

[joestue]

I recently acquired 32x 9 amp hour 12v cells (the high capacity version of the 7.2ah, 7.5ah)
and 40x 5ah 12v cells.
they were all in 48 and 96 volt battery trays.

all of them with the exception of 1 seem to be in good condition, and they all have resting terminal voltages of 13.0 to 13.2 volts, with 13.1 being the average.

so i'm looking around the internet and i find some info on APC inverters floating the battery at 13.5+ volts, and how bad that is for the battery.
well i did my own measurements and i found a Back-UPS 1500 not only floats the voltage at 13.5-13.6, but it pulls 100 ma from the battery for about 1/10th of a second, and then charges at 50ma for the rest of the time. (continuously btw) so you're looking at about 30-50ma flowing into a 5 amp hour battery (that particular ups is designed for two 7-9 amp hour batteries in series)

anyone think this isn't negligible? i don't know.
so anyhow, turns out sealed lead acid cells are only good for 2-3 years in APC products.. and this is uh, well known
I don't buy the argument that 20 discharges to zero at a 15-30 minute rate can result in exploding battery cases.

so getting back to one of the un touchable subjects here...
pulse charging..
i found a whole bunch of comments posted by John Fetter across various forums over the years.

he brings up several interesting points that i'd not heard elsewhere.

The positive active material is lead-dioxide. Lead-dioxide is a semi-conducting material. When it is housed in a lead-antimony grid, the antimony and dioxide remain firmly in contact. When housed in a lead-calcium grid, an ultra thin oxide layer of a different kind builds up over time between the grid metal and the lead-dioxide, which takes just long enough for the battery to survive a couple of years. Manufacturers add a little tin to try and prevent this problem but it is not very effective.

ultra thin oxide layer between the surfaces of the positive grids and the lead dioxide active material. This is known in the battery trade as “open circuit” or the “antimony-free effect”.

As a matter of interest, when the total volume of water that is added over the years reaches the same volume of the electrolyte, your batteries will most likely stop working. I commissioned testing on two sets of Trojan golf cart batteries, two sets of Exide and two sets of Deka golf cart batteries years ago. They all expired after using this amount of water. Two-thirds of them lasted between 40 and 60 percent longer because we had given them something to make them last longer. Their water consumption was slower. I was amazed to find the water consumption so predictive.

Very high magnification (600 to 2000) shows that lead sulfate crystals and lead metal crystals actually grow in completely different areas across the surfaces of negative plates - lead dioxide crystals and lead sulfate crystals likewise grow in different areas on the positive plates.

When a battery is in the process of being discharged, at the negative electrodes, the surface atoms of the lead metal crystals go into solution, then travel almost in contact with the surface towards the lead sulfate areas, get converted into lead sulfate upon arrival and precipitate out, adding to the lead sulfate crystals. The activity at the positives is similar.

so reading through hundreds of other comments i find some interesting things.

the manufactueres tried adding tin to stop the lead-calcium corrosion issue, but it didn't work very well.
then the price of tin jumped 5 fold (its 20 dollars a pound now) and they dropped it altogether.

lead antimony positive plates and lead-calcium negatives could work, if they formed the battery before packaging the cells together
(during forming, antimony is plated out into the negative plates and essentially you get the same water consumption as the lead antimony battery)

antimony prices went up and all the battery manufactuers switched to lead calcium at the same time.
with the exception of industrial batteries.. who's consumers know better.

but getting back to pulse charging, seems it can break through the oxide layer between the lead calcium grid and the lead dioxide active material.. but only on the positive plate, the negative does not have that problem.

so while it appears that in order to get the best life you have to constantly trickle charge..
and then the question comes up.
how much current per amp hour.

seems to me 50ma into a 5 amp cell is too much.
that's a 1/100C rate.

one thing i don't want to argue about is the float voltage.
float voltage changes with age, as the acid content rises with water removal, and then falls when lead sulphate is bound by EDTA (added at manufacture by some?)

while self discharge is strongly temperature dependant, seems to me that you could easily account for this with careful regulation of float current.

this document:
http://www3.alcatel-lucent.com/bstj/vol49-1970/articles/bstj49-7-1321.pdf
Is from 1970, and its for lead calcium cells.
It suggests float voltages on the order of 12.9 to 13.2 volts.
Float currents on the order of 45 micro amps per amp hour.

maybe the solution is to just put extra water in the sealed cells and run them as flooded?

[dnix71]

1/300C is safe for constant charging. APC branded batteries are not common sla's. They are made to specs for APC. If you substitute a cheap no-name in an APC backup it won't last as long. APC has been putting much smaller batteries in their ups's lately. The ones we have at work have batteries with less than 1/2 the AH rating they did 7 years ago. The output even in the Pro series is not sinusoidal, it says so right on the case. They are only made to back up a pc for a few minutes, nothing more.

SLA's are good when you can't tolerate acid fumes or the risk of spillage, otherwise flooded is the way to go.

The walk-behind pallet lifts we have at work have slas and the Crown rep openly admits they are only good for 1-2 years in hard daily use. He recommends converting back to the older 2v flooded stacks.

[Bruce S]

Joe;
 those packs you speak of, are they from Eaton battery backups? they sure sound like them.
Awhile back I did some training up around GHurd's area of Ohio and was taught more about why the companies are going this route than ever before.
They are building these newer units for short term loads and higher voltages thing of 48Vdc now being the lowest voltage. ALL newer units are 60Vdc and above, depending on which model. Most of the newer units are 120Vdc with stepup transformers taking that up to 208 and 220.
Reason being: the battery packs can now be only 12Vdc at 7 - 9 Ahr each. There will be say 5 batteries in each pack two pack minimum. Makes for loading new packs much easier on service guys back!! and the contacts much smaller, where the KW remains the same. 60Vdc@9ahr or 12Vdc@45Ahr.
ALL the corporations and hospitals know very well it's far far more reliable to have 15mins of battery backup and let a generator kick in than a warehouse full of batteries.
APC: Most of the older units including the matrix series, did not allow enough ventilation , so the batteries simply and unfortunately got cooked dry. I have one of the rack mount XLS units, I merely ran the battery cables outside of the unit and have not had a problem with batteries boiling dry since.
Bruce S

[joestue]

i'm certain all but perhaps one pack were APC, because they all have non standard APC brand Anderson Power Pole clone connectors, that don't mate with APP connectors.
there were 4 manufacturers of the batteries.
Kung Long, made in Vietnam
Exide Powersafe, made in India.
Vision, made in china
BB battery, made in china.

[RandomJoe]

APC branded batteries are not common sla's. They are made to specs for APC.

Perhaps they've changed in recent years (as you say they're using smaller batteries now) but all the batteries I pulled from "dead" UPSes given to me in the past seemed to be standard AGMs with an APC sticker glued over the usual manufacturer's printing.  The specs listed under the sticker were the same as any other AGM I had, nothing indicated it was a special design.  (Aside from the shape - any battery shaped to fit a UPS is suddenly worth 3-4x any other same-AH battery!  )

If you substitute a cheap no-name in an APC backup it won't last as long.

Heh.  How could you tell?  I never saw APC's batteries last very long!    Many of the UPSes I got were less than 2 years old, every one had bad batteries.  Actually, I expect "cheap no-name" batteries wouldn't last long anyway no matter what they were used for. 

I replaced the two APC 17AH AGMs in a BackUPS 1500 (or thereabouts) with two Werker 35AH AGMs.  Of course they wouldn't fit inside, just had them sitting beside the unit.  The PAIR of 35AHs was the same price as ONE of the 17AHs (also Werker, same store) shaped to fit the UPS!  Ouch...  That setup lasted about 4 years.  By the time the batteries died I had the whole-house solar / backup system (now 380AH 48V bank and Outback inverter) so no need for it anymore. 

[dnix71]

RandomJoe slas go stale after a while if unused. If I buy an APC backup on sale at Walmart I expect that it has been sitting around long enough that the sla is not fresh enough for commercial use.

There is also a big difference in quality between the older Japanese made and the Chinese made with a Japanese name. Funny, even Taiwan made stuff has come up in quality lately. Japanese made 20 years ago were some of the best. You can still get "Panasonic" slas but they just aren't the same.

The weak link in buying a ups that's ready to go is the battery should not be made or installed until the ups is ready to be used, but that's not how retail works.

You can buy direct from APC, but the price is higher and the warranty negociable. They have a commercial line. The stuff that isn't fresh enough for a big commercial customer has to go somewhere.

There are generic slas made for no particular purpose, but those are not suitable for a ups. I have an eBike that uses slas, but they can't put out the amps that the original slas did. The cheaper replacements I bought for more amp hours were made for low drain apps like electric wheelchairs. I accept that limitation because I want the range extension.