SLA Failed battery discharge test

[Robdog]

I acquired 2 x 12h SLA 35AH batteries from work and would like to use them in a project building a portable battery bank with 12v outlets.

I charged both batteries using an 8 stage charger. Both batteries were measured at 12.8V 24hours after disconnecting them from the charger. I proceeded to carry out a discharge test on 1 battery connecting 1 battery to a headlight out of my ute. The load was drawing 3.45A. I connected my Fluke 289 to monitor voltage. I expected 10hrs running a 0.1C load until battery voltage would drop to 11V however, I checked on it 120mins later and it was already flat.

The attached graph shows it only took 90 mins to drain the battery to 11V. Can anyone tell me why this would be? Have I interpreted discharge curves incorrectly?

Also is there anyway to determine what type of SLA batteries I have ie. AGM/GEL etc?

[Robdog]

This is the discharge cirve I was refering too

[testian]

Look up the data sheet for you battery. There it should be mentioned what type it is.
How hold are those batteries? A stamp on the case should indicate the manufacturing date.

My guess is that those are old, maybe sulfated AGM batteries with a remaining capacity of ~6Ah.

[Circlotron]

A good SLA battery that has been floated at 13.65V for a long time will stay at 13V or more for 2 or 3 weeks. How long did you leave the batts connected to the charger? I'd leave them for a couple of days and see how you go. Also, check what voltage they are being floated at just before you disconnect them.

[BradC]

Most SLAs are rated at C/20 rather than C/10, but that will make bugger all difference to your test."It's worse than that, he's dead Jim". Unless you have a reliable source, second hand SLAs are usually "mostly dead". Yours is considerably dead.

[floobydust]

12.8V after charging seems wrong, the float voltage on a charger should be higher.

During discharge testing, you might notice the battery acting like a 10V battery - with 5 good cells and one cell that has failed.
So the voltages are about all one cell lower than the curves.

[Red Squirrel]

I've found that SLAs tend to have really poor life.  After 5ish years their capacity is greatly reduced or in some cases they're pretty much flat.  I have a similar power pack I built, and any big loads don't do very well. I have two 7.2ah ones in parallel and wanted to see if I can charge a Phantom 3 battery (half decent load but nothing crazy, like around 100w)  and it lasted maybe 5 minutes before the voltage dropped to 11 and the inverter started to beep. When I took the load off the voltage went back to 12ish and when I connected the solar panel it said it was fully charged within like 5 minutes.  I'm just guessing but it might have to do with these having a higher internal resistance.

I want to build a better version of this pack eventually, and think I will use a small flooded acid battery instead of SLA.  Downside is that the pack is already heavy enough as is... lol. 

[Robdog]

The float voltage on the charger is 13.4-13.7V from what I have measured.

I have just finished testing the second battery after leaving it on float for the same period of time and the results were much the same.

[floobydust]

The batteries appear pooched, rated 35Ah and you got about 7Ah.

Notice the discharge curve (log) inflection point at about 9V, that's when a cell gets reverse voltage on it and you are damaging the battery as you went down to 7V.

I think your only option is pop them apart and add water, see if that fixes it, or toss them into recycle bin.
You can add water to gel-cells if you get in the lids.

[BradC]

I've found that SLAs tend to have really poor life. 

Batteries don't die. They are killed. Well looked after quality SLAs will last at least as long as the data sheet says they will.

Example, I have some Vision 55AH units here with a 10 year design life, and some CSB 17AH units with a 5 year design life. The Vision batteries are now into their 13th year and the CSB are into their 9th year. Both sets test out at rated capacity. I have 14 & 12 of each respectively. No early life failures, and a well looked after, temperature compensated float towards the lower end of the datasheet float voltage.

If you don't overheat them, don't overcharge or undercharge them and keep them within their datasheet cycle parameters, they'll last at least as long as they are rated to. On the other hand, I had some 17AH CSB batteries in an older APC UPS. Kept at about 35-40C and with an un-compensated float voltage right at the upper limits, and they were mostly dead in 3 years and totally dead in 5 years.

Like anything, people don't pay for quality and tend to treat batteries as a "set and forget" appliance. In those instances they won't last.

[Red Squirrel]

Hmmm most of the ones I've found that die early are in UPSes, are most consumer UPSes just crap at managing the battery then?  I guess that should not surprise me...

You'd think it would have temperature compensated charging etc though.   It also weirds me out that most UPSes don't even have fans in them.  Anything with power electronics is not going to be 100% efficient thus will create heat.

[BradC]

Hmmm most of the ones I've found that die early are in UPSes, are most consumer UPSes just crap at managing the battery then?  I guess that should not surprise me...

Yep, very few have temperature compensated charging. A lot use pretty rough high ripple bulk float charging and almost universally they are set right at the maximum float voltage to provide those extra 10/10ths of runtime at the expense of battery life.
In fact I'd say no "consumer" grade UPS looks after its battery. its only when you move up to the small business grade UPS that you start to see these features, and then only in some units.

You are also spot on with regards to fans. Better UPS have variable (or at least two) speed fan control so they keep things ticking over to minimise heat buildup in standby and then switch on the hairdryer when running on or charging the battery. Not uncommon to see a UPS idling with a battery temp of 40C. Run a battery at 40C and it'll die quickly, faster when it is still being floated at 2.3V/cell.

The Vision batteries I have here are on their second life. They spent their first 10 years hooked up to a 30KVA APC/MGE UPS. That had serious battery management with a whole host of algorithms designed to maximise battery life while ensuring they still had the capacity to deliver. After 10 years all 32 batteries tested out at rated capacity. They were in a separate ventilated chassis in a chiller room that never exceeded about 28C ambient though. When you are looking at the best part of 8 grand for a battery replacement, you don't want to be cooking them.

[MosherIV]

Hi

As someone else pointed out, there are different types of SLA batteries.
One type are 'deep discharge' - these ones are meant for cyclic charge/discharging.

The ones used in UPS are the cheaper non deep discharge type. Once discharged, they tend not to recover back to their original capacity.
I certainly would not be cycling the batteries like you are. At most I would only discharge them down to 40% of capacity to protect them.

Batteries don't die. They are killed.

Do not agree with that statement.

Well looked after quality SLAs will last at least as long as the data sheet says they will.

I agree with that statement.

Bear in mind batteries are chemical things, and the chemicals can degrade over time, regardless of what is being done to them.

[Circlotron]

Back in 1979 when I was a 4th year apprentice I helped out with a discharge test of one of a pair of telephone exchange batteries. 24 x 2v cells, 500Ah with pure lead plates. (not paste impregnated). They were 20 years old and still measured 520Ah over 10 hours. They were kept in a temperature controlled environment, specific gravity checked I think once a month, and someone regularly kept an eye on them.

[Red Squirrel]

Telco batteries are quite something when you see them for the first time.  We have 4 strings here of 4800ah(IIRC) ones.  Each battery (a cell to be exact) is 2v and there are 24 of them to make a 48v string.  Normally it's at 54v which is the voltage of the DC power supply (we call them rectifiers in telco world).  If you set the DC power supply to that voltage it keeps batteries charged without overcharging them.  Known as "float" voltage.   Lead acid is nice for that, other battery techs are more complex as you can't do that.