Ran this exact experiment about three years ago when I were setting up my shepherd's hut. Grabbed four 12V 7Ah sealed lead-acid units pulled from scrapped APC UPS boxes — the sort of thing that turns up on eBay or Gumtree for almost nothing.
The honest verdict: it works, but with significant caveats.
The main issue is you genuinely have no idea what cycle history those cells have seen. UPS batteries sit on float charge for years and the capacity degrades silently. I tested mine with a proper load tester and found one was down to roughly 40% of rated capacity. Binned that one immediately.
What I'd suggest before committing to a UPS battery bank:
- Do a proper discharge test — don't just trust the rated Ah
- Check internal resistance if you can (a Victron BMV or even a cheap capacity tester will help)
- Wire in parallel rather than series where possible to avoid a dud cell dragging down the whole bank
- Keep expectations realistic — these are genuinely short-term solutions
My four batteries gave me maybe 18 months of light LED lighting and phone charging before I migrated across to a pair of Fogstar Drift 100Ah lithiums, which was a significant upgrade in every measurable way.
That said, for someone wanting to experiment cheaply, or power a garden office temporarily whilst waiting for budget, there's nothing wrong with UPS batteries as a learning exercise. You'll understand battery management, charging stages, and load calculation far better having worked with something you're not precious about.
What sort of loads are you trying to run? That'd help everyone give more specific advice about whether UPS cells are even worth considering for your use case.