Matching of SLAs for in-series use

[Zbig]

Hi

I've got these two rack-mounted 1U UPS units. Due to their slim form factor, they use two 6V SLA/AGM batteries connected in series. Turns out that wasn't the greatest purchasing decision on my part since I've been experiencing some issues pretty much from the beginning. First was the shorter than expected runtime under reasonably low load (like <25% rated capacity). Also, during the load tests, UPS's remaining runtime estimation was falling off the cliff rapidly before it shutting down. Then, after reconnecting AC, the UPS's own state of charge estimation never goes over 89%. All of this pointed at a mismatched batteries and indeed, when I took the battery pack out, the two batteries' voltage differed. Can't remember the values, though. I took a few different attempts at balancing their SoC using lab power supply and electronic DC load to no avail. Perhaps my methodology (involving fair amount of guesswork) was flawed or the batteries were already too far gone, possibly both. So the UPS was sitting at the "89%" SoC for like a two years or so. Yesterday, I woke up to the UPS's alert with the diagnostic interface stating "No battery connected". The battery series voltage dropped to some 8V with (after taking them out) one battery sitting at over 6V and the other around 2V. I have swapped the battery pack with the other UPS unit I had in storage and the charge level keeps showing 89% so the history is deemed to repeat. I am now waiting for the delivery of two new batteries but before I hook them up, I'd like to give them my best shot at ensuring their proper operating conditions. Please note there is nothing in the form of charge balancing or even individual voltage monitoring inside the UPS device. The batteries are just connected in series and that's it: the UPS unit is oblivious to the fact there are two separate physical SLA'a in there.

So, while I wait on the delivery of two new batteries, what would be the best course of action for me before I put them in:

• Assume they were manufactured to standards, part of the same batch, hook them up and hope for the best
• Top them up individually to 6.75V each the best I can using lab supply
• Discharge them individually in a controlled manner using electronic DC load to some arbitrary voltage
• Perform a full charge-discharge-charge cycle on each of them individually to evaluate their respective capacities and somehow prepare them individually taking this into consideration
• Other?

Regards,
Zbig

[trobbins]

All of the above if you have the time and inclination.  Periodicly repeat tests if you have time and inclination.  The higher the quality of monoblocs (typically price sensitive), the more likely they will remain balanced, but not a guarantee as there may be other influences at play.

[TruslowPJ]

when you have lead acid batteries in series, there is a little bit of self balancing due to higher self discharge rates at higher voltages, but it's puny and not always enough. for them to stay in balance, you want all of them to be identical with ideally same date or batch codes if possible.
when you get them, wire them in parallel to your bench supply, charge them to 7.2v or so and leave them for as long as you can manage. that way when you take them out and put them in series, they are as close to 100% as you can get them. remove them from the power supply and let them sit disconnected (from anything else or each other) and take a voltage measurement hours later. if the open circuit voltage after a few hours of resting is the same, then they are the same state of charge. for a new 6v AGM SLA it should be in the range of 6.4-6.6v after a long rest.