I have a very very simple ups, that is working with full batteries for months now, I test the ups and it still drives my pump nanostream tunze 6055 for more than 2 days without even get to 60% of their's capacity.
I have 2 VLRA batteries 40 amperes each could put them in seerie to get 24v, but instead I use tem in parallel and have 80 amp 12v
no relays nothing...
the charger is very very weak enough for continuos charging without stressing the batteries (I believe in this way the batteries would live longer) somemthing like 13.6 V, the pump is also connect like the batteries to the same charger.
so it's working close to 14v , 6055 does work at several voltages (12-24v)
the amperes that the charger is delivering are only enough for the batteries and the pump .... if i disconnected the pump I would get more power to the batteries.
what it happens when there isn't power?
simple...
the batteries will drive the pump...it's imediate, the pump doesn't stop...
the only problem i see with this scheme is more time to charge the batteries but this is not an issue they can deliver enough power to the pump for several long days, a power outage for more than 1 day is very unlikely...
the level of the tank will drop (backflow to the sump from the return pipe) until this tiiny pump would get more "control" of the surface agitation...(more close to the surface) lots of oxigenation...when there isn't electricity.
I beelieve it's an elegant and simple way..
could watch the pics on my blog...sorry it's in portuguese.
even the charger is portuguese...
http://waterinspiration.blogspot.com/2008/12/diy-ups.html
the charger is for ups, it does control the amp towards outside, so I believe no spikes even with tiny electrical interruptions