No question saltwater presents challenges that require a different risk assessment. I've been there. I should really have mentioned that its only freshwater I'm considering here. My bad.
I'm totally with you on the consequences of an ATO failure. Almost certainly its either bad or really bad. That's why this is interesting - we all know it must not happen. Ever. Its how we weigh (or distort) the risk and select what feels like the appropriate effort to avoid it that I think is oddly mis-calibrated.
First, my personal example of a single simple float valve fed from a reservoir by gravity, siphon or pump (if you must) makes my point. There's a great deal at stake, all depending on that one little $10 plastic device. Its so easy to point right there and declare it could never be trusted. Single point of failure. Too easy to imagine the snail, or leaf, or hard water scale, or any number of little things jamming those moving parts. Ok, but now spin that argument around. That mechanical simplicity is a strength. How many failure modes does it have? Two. If that. Lets just focus on those. Keep it clean and test it once in a while. Rig up a simple barrier to surface debris and/or tank life forms. Keep foliage away. Check there's no debris in the feed setup. None of this is remotely difficult and the resulting system is just not gonna fail, its not, over the spans of time between your regular, or random, mtc sessions.
Second, lets step back for a second. Why are we doing this ATO thing? For large tanks its absolutely justified. Go nuts. For less than say 120g (I chose that only because I've had several 120s and know the evap challenge during a dry Canadian winter) lets admit the ATO is simply a rather nice convenience. No pitchers or buckets, set and forget, steady water level and chemistry, more fun gear to setup and play with. But the true cost you bear is very high. It is the introduction, willingly, of this entire big scary flood anxiety. So, to address this new problem, we add more complexity, naturally. I'm saying hold on, try some healthy skepticism here. All that risk, all the cost, all the compounding complexity justifies a moment to rethink. The alternatives are many: simple robust float setup as above, or maybe rig your RO reservoir or setup so that its effortless to manually hit a switch or ball valve and stand there for 2 minutes. Or quickly easily fill a small jug and pour. It can be made trivial. Show the family. Set things up - design it - for the humans. You can usually find a few laying around.
If you're worried about those week(s) long vacations, many low tech low risk temporary setups can be made ranging from a fixed quantity drip feed to the valve arrangement above. You're gonna need a house sitter anyway. I assume you are here because you run a planted system. Then you well know they require regular attention.
I'm therefor suggesting that the value proposition of the "A" in "ATO" may often be far less than assumed and directed at an exaggerated problem.
Originally Posted by CrypticLifeStyle
The failures are more prone in saltwater setups, plenty of examples out there to google. Freshwater, I'd say there is so many factors in everyone's water quality, build up is inevitable for some, even if it wasn't for you or someone else. Some people are also lazier then others, some are on-top of obvious husbandry or preventative maintenance or just lucky.
Failures however do happen, and when it happens it can cost a lot of money. That is the concern, there is a actual risk percentage that may or may not happen. Not everyone is ok with wait to see mentality. If you rent or own a condo as example you have the costs of repair to other areas as well.
How does one weigh the risks, that is a personal decision with how acceptable you are of, risk.