In my experience lead-acid batteries which where left for a while are toast. You might be able to charge them but the capacity will be nowhere near the original capacity.
I don't have much experience of lead-acid battery maintenance, but generally agree.
One question: is your 'nowhere near original capacity' including use of a charger with de-sulphation zap cycles built-in?
Going for Li-ion sounds like a better plan but the problem will be that the (remaining) charge prediction won't work at all if that prediction is based on the battery voltage.
Agree. How this works out depends on whether I can reverse engineer the UPS enough to modify the firmware. If not, then I'd just ignore the UPS's opinion on battery state, and use the BMS for that information. So long as the UPS keeps operating, who cares what it thinks.
Remember this UPS needs TWO separate battery banks, so whatever I do with batteries, I have to do twice.
However a Li-ion battery BMS is likely to have a Coulomb counter based charge gauge so if you can read that you might be able to build a lead-acid charge-voltage simulator.
See above. I wouldn't bother to fool the UPS. Just give it a 'healthy battery bank' DC voltage. It's unlikely I'd use the UPS internal battery charger, unless I can modify the firmware to do what I want.
One of the main objectives with this, is to develop my own BMS. I have some non-standard requirements, related to a different project.
Another thing to look into, is a converter from one DC supply (with more flexible voltage range) to the + & - HV DC rails the UPS requires.
The battery voltages this morning were:
12.85
12.10
11.98
11.94
4.47
0.63
Poor Mr Zero-V is currently getting a long de-sulphation cycle.