@RobMcC or others... I am unclear what caused so much damage, since I have not had the misfortune of 3 days with equipment down. Were the Tunzes not enough flow compared to what the corals were used to? Some systems have no sump with return and no skimmer. Also, can't corals live for 3 days without light? Did something die, like snails, and set off a chain reaction due to water chemistry? Did CO2 build up quickly and make the pH plunge?
Sure. Good question. Flow in my tanks is not just the Tunzes (I also run 3 MP40s in the big tank, which are not connected to battery back-up). Also the back-up battery is 12V, whereas most of my tunzes are set using the 24 v jumpers, so the flow of those on backup is approx half of when supplied my the power grid. True, some tanks are designed without sumps, but that is the normal case for them and is what the inhabitants are used to. For example, I'm sure zoas and other softies would be fine even in reduced flow to normal, but acro colonies not so much if they experience drastically reduced flow for 3 days. The skimmer is also the primary source of aeration in my tanks (as for many). Hard corals, and big acro colonies in particular like stability and need flow and oxygen and without it they die very quickly (which is why they do not ship well). Basically, as I mentioned my system was designed to run for a day or so on battery back up before requiring the back-up generator. Indeed, this worked fine for the first round of power cuts three weeks ago, as I wasn't able to start the generator until about 18 hours after power went out, but it would seem the limit on my systems is somewhere between 18 and 72 hours. The big tank was hit much harder than the small one, so I presume its limit is shorter. This may be because of the higher fish load, but it could be for any number of reasons (it is in a windowless room for example, so had no light either).
I doubt the chemistry of the water changed much, as no fish died. Likewise snails and other mobile inverts are very hardy (alk was increased very slightly in the small tank and by 1 dkH in the big tank, but that is expected given the dosing pumps resumed dosing, but there were less acros and/or stressed acros, hence growth and therefore alk/cal uptake is reduced.
Regarding lighting, as I mentioned, they didn't come back on when the power was restored (it appears they default to 'off' after an extended period without power), so the lights were out for about 6 days until I returned home. This may or may not have caused some of the issues, as it is confounded with the reduced flow and colder water, but it probably wouldn't have helped.
To be honest, I'm surprised more corals didn't die. Some acros colonies had half stripped, but appear stable now. Usually when that happens they are done for. I have two nems, they also look stressed, but at least they haven't started wandering, so I'm sure they'll be fine.
In terms of lessons for the next outage, in the short term, I think I'll add another deep cycle 100 Ah SLA to each tank and use 12V to 36 volt transformer for the return pumps. I have a DC to AC 500 W inverter that I could use for skimmers, but given this will drastically shorten the back up time, and I'm not often away from home, then I won't pursue this option.
Long term solution, power wall and auto-isolator/transfer switch. If I didn't have neighbors close by, I definitely go with a Generac whole house back-up generator, but they are just so noisy it isn't an option for most folks in a suburban area.