UPS ruins batteries every three months

[davidebaldini]

Problem description:

My UPS requires new batteries every ~3 months; in this time frame, an average of about 5 to 10 power flickers or short power interruptions (< 2 minutes) occur. After an average of 3 months, a new pair of batteries begins to fail to keep the UPS up during any flicker or power interruption. The load it powers is a single computer/workstation running 24/7 with a continuous consumption between 200W to 300W.

UPS model:
The Atlantis 1001 is a non-inexpensive (~120€ back then) home/office offline UPS, rated 1000VA, 600W, designed for two 7Ah 12V SLA batteries in series, as per its user's manual. The UPS doesn't run periodic battery tests, and just keeps a float charge while on standby.

Troubleshooting:

I opened the UPS while plugged to the mains, powered on, and powering the computer. The float voltage is 26.99V across both batteries, partitioned in 12.85V on the left battery (2.14V per cell), and 14.14V on the right battery (2.35 per cell, see picture). While I took the measures the UPS was not recharging the batteries, these are the float voltages.

The left battery has a corroded negative terminal.

The temperature inside the UPS case feels about 30°C (86°F), approximately. The case has a fan, but it is normally off and spins only while on battery power, so the inside remains always warm.

Interpretations:
The UPS is designed for 600W, or 600W / 24V = 25A on the batteries (~3.5C). Even with half the load (300W, 12A, ~1.7C) the current on the batteries is quite high. I always bought cheap batteries from eBay, so I imagine this is an overload regime. However, very few power losses occur during the 3 months that a new pair of batteries endures, so I imagine this doesn't justify the short lifespan.

AFAIK, the float voltage of 26.99V across both batteries is a good value per se, but badly partitioned across the batteries, so that 12.85V causes sulfation on one battery, and 14.14V causes electrolysis on the other.

The warm temperature also doesn't justify the extremely short lifespan of the batteries.

[stj]

Re: UPS ruins batteries every three months

i could be overcharging them,
or they could be shit batteries.

those batteries come in different types.
you have ones designed to be trickle-charged forever and only discharge ocasionally at reasonable current - for ups/ backup lights etc,
and you have ones designed to be constantly charged and dumped at high current for use in electric vehicles and such.

i think if you use the type intended for golf carts, wheelchairs etc, they wont like the constant charging.

[eccerr0r]

Re: UPS ruins batteries every three months

I also have a 1000W UPS that ate its batteries in a year.
I didn't think to do battery tests when they were working, but need to make sure that the UPS does not charge the batteries past 20.7V, which would be 27.6V for your system. It is okay to go higher but only for a short period of time.

But back to the reason why for doing battery tests - since you have more than one battery pack, the packs need to be identical and balanced. If one is weaker than the other, then the whole set is as weak as its weakest cell. Agreed that it's possible you have one poor quality pack that ruins it for both. Ideally there's a balancing circuit just like for lithium ion packs... but for lead acid, they don't explode as violently as lithium ion when they're abused...

[davidebaldini]

Re: UPS ruins batteries every three months

Further tests:

I unplugged the two batteries from the UPS; in open circuit, the corroded battery (float-charged at 12.85V) reads 2.5V across its terminals, the other battery (float-charged at 14.14V) reads 12.8V. No doubt, the corroded one is broken.

I pried open the good battery; looking thru the cell caps, there's a white sponge inside each cell (fiberglass?) so these are non-flooded SLA batteries, which require a slightly higher float voltage than the flooded ones. The sponge looked dry so I added ~80ml of tap water across all cells, or 13ml per cell.

Now I'm charging the battery with a bench PSU at 18V, 0.6A to gassify the liquid; I read that forcing gassification can clean the lead plates and increase conductivity.

So, these batteries dried up too quickly in just three months, with an apparently correct float voltage and an acceptable temperature inside the UPS case.

[eccerr0r]

Re: UPS ruins batteries every three months

I've had very poor luck dealing with these AGM/SLAs after they've failed. After watering them they last a little while longer but will fail again soon after, usually shorting out a cell.

I gave up on my UPS for now after tossing the battery packs, and really need to think about
a BMS for SLA packs as well...just for longevity.

[eccerr0r]

Re: UPS ruins batteries every three months

I'd also have to add corrosion at a terminal of a SLA is a bad sign...

[brethin]

Re: UPS ruins batteries every three months

Id love to find a case that uses 18650 that allow cell replacement with bms etc same size to replace my 12v lead acid in most ups, i still have tons of 18650s i need to find a use for.

[sam_sam_sam]

Re: UPS ruins batteries every three months

Id love to find a case that uses 18650 that allow cell replacement with bms etc same size to replace my 12v lead acid in most ups, i still have tons of 18650s i need to find a use for.

You can do it but you have to find the right BMS protection board for this to work right

I have done this myself but I have not found the right BMS board but I did find a work around

But here is the problem I was using 26650 batteries that the voltage is 3.5 volts fully charged which is only 7.75 ( which means that I used 7 batteries ) batteries this works however the cutoff the BMS protection causes a problem in that if battery runs down to BMS cutoff voltage when power comes back on the UPS dose not come back on you have to power up BMS board manually before it will work again

You do not need a case for the 18650 batteries but I would recommend that if you can find cardboard battery sleeves you use this type because I do not really like the plastic sleeves because they tear very easily if you use hot glue on them

If you do use a metal case for the 18650 battery cells make sure that the case fits into the UPS batteries case

But yes you can use 18650 batteries but it is very important that you use the right BMS protection board and you do some bench testing to make sure that it functions correctly

I would recommend that you run theses test many times to make sure that your batteries do not over charge / under charge that you are not over currenting the batteries when the UPS is at maximum current and when the UPS turns off because of low voltage at the batteries that when you turn power back on that the UPS turn back on and the batteries are charging


One final comment but a very important one about this type of battery and BMS protection board

You MUST make sure that you use batteries that have the same milliamperes testing results and the same run time if possible because if you do not do this step you are going to run into this problem

Your run time is going to be not very long because ——>

The battery that charge up first to the voltage cutoff will turn off the charging current to other batteries
The battery that discharge first to the cut off voltage will turn off the BSM controller

DO NOT BE TEMPTED to use this type of battery with out a BMS protection board
This will become very dangerous to life and you could burn down your house

[CapLeaker]

Re: UPS ruins batteries every three months

maybe print a case with a 3d printer.

[sam_sam_sam]

Re: UPS ruins batteries every three months

maybe print a case with a 3d printer.

This would be a good idea if you have access to one

Do you have someone or do you have one

[eccerr0r]

Re: UPS ruins batteries every three months

There's not really a good way to replace SLA packs with Li-ion (cobalt/manganese) directly without some circuit hacks. Likely you have to get the more expensive Li-ion iron phosphate cells.

I need to get more 18650s Li-ion cells to play with...

[stj]

Re: UPS ruins batteries every three months

why not use car batteries in a sealed box with a hydrogen vent hose?

[petehall347]

Re: UPS ruins batteries every three months

why not use car batteries in a sealed box with a hydrogen vent hose?

thats what i do .longer run times too so long as the ups isn't programmed to shut down after x amount of time . i find ones with fans seem to stay on ok .

[eccerr0r]

Re: UPS ruins batteries every three months

For the OP, two car or motor cycle batteries will do the job as replacements. For me, not acceptable as I want for cyclic use.

[eccerr0r]

Re: UPS ruins batteries every three months

BTW I wasn't thinking about replacing the batteries in my UPS. Mainly because my UPS uses an 18V battery system and that is rife with pitfalls if you don't do 3x6V, sorry if people were thinking that I should re-cell my UPS.

I was thinking of a different project

[petehall347]

Re: UPS ruins batteries every three months

have a 48v one the batteries died . still need to get 4x car batteries for it . handy thing it is as 2.8kw output

[stj]

Re: UPS ruins batteries every three months

it must be huge - is it a rack module?

[petehall347]

Re: UPS ruins batteries every three months

it must be huge - is it a rack module?

similar to a large computer case

[Curious.George]

Re: UPS ruins batteries every three months

similar to a large computer case

I have a pair of rack-mountable 2200VA units each with an "expansion battery". This was the easiest way to address that load as I could use two 20A 110V branch circuits to power it (without having to go to 220V). It also makes them a lot easier to manage (physically) as each of the four "boxes" (2 UPS + 2 expansion) can be lifted by a single person (under the 70lb limit).

I sited them in such a way that I can make a SHORT cable run through the wall and into the garage to install a larger battery pack, if I ever need longer up-times.

And, by sticking with 48V batteries, I can charge/power them directly from the 300W solar panel I rescued. (I'm ever vigilant for other such panels so I can eventually power the UPSs directly without having the utility involved!)