I do plan on having an ato. How large should the return section be ideally? More than 4 gallons?
Since in reality, as with any elctro/mechanical/electro-mechanical device, "reliable ATO" is an oxy-moron, all there is, are variable degrees of risk, and variable modes of failure.
With any system if electrical, you need a mechanical failsafe, and vice versa. Problem is, the more stuff you add to a system, the more points of failure are introduced, and there are more modes of failure. I have a box of burned up pumps that prove the addage: "The more you overthink the plumbing, the easier it is to stop up the drain." Things happen. Two examples of redundancy and failure were the Challanger and Columbia explosions. Hardly reef level stuff, but just the same.
I am going to use a little
lex parsimoniae (latin) or a principle of parsimony, economy, or succinctness. It states that among competing hypotheses, the one with the fewest assumptions should be selected. This means that one should proceed to simpler theories/answers/ideas, until simplicity can be replaced by greater explanatory power. In the absence of certainty, the fewer assumptions that are made, the better, i.e. the simplest one should be selected. (Guess I am wrong 90% of the time...)
In english this is called Ockham's Razor. (Occum's) Loosely paraphrased....
A 'reliable' ATO will use a latching circuit (relay, float switches, diodes, etc) along with high water level float switch failsafe, low water level pump cutoff failsafe, along with a mechanical (non-electrical) backup for each function (failsafe.) This system works pretty well, indeed, though mechanical backup is going to be problematic, if you take a look at it a piece at a time. But a 10 cent diode blowing, consequently a $5.00 relay burns out, a little calcium deposit on a mechanical switch, we are done here, and you just lost a $300+ pump. When talking redundancy and failsafe, probability of failure does not fit in.
If we take Ockham's razor and just use the basic idea, (simpler is more reasonable) and add in the 4th Law of Thermodynamics, which states: anything that can go wrong will go wrong, at the worst possible moment (read shuttles exploding in flight,) ATO's can, and will fail, there is no certainty that they will not. Reliable ATO = oxy-moron, so we need:
A passive failsafe. This is a redundant failsafe, on top of all active redundant failsafe measures. It is the all else fails end of the line. A passive failsafe either cannot fail (no failure mode) or the only failure mode it has makes the system safer. (The air vent line in Bean's drain system is a passive failsafe.) Nothing can guard against the human factor, but that aside...explaining why something is simple, is often complicated..........
1) An ATO is a convenience, not a failsafe for your salinity.
2) An ATO is a convenience, not a failsafe for your pump.
3) Of the two, over the short term, the pump is more important.
4) Passive failsafe:
Enough volume in the return section of the sump, that will allow the pump to run for several days, without burning up, when your ATO fails. @ 2 gallons per day, 6 gallons total, above the pump inlet. (Average, 2.5 gallons per day for the unlucky ones.) Has one failure mode: Human factor. Not paying attention, or gone for four days and the ATO fails day one. Well, nothing is perfect...
Simply, cannot justify a smaller return section, because you have an ATO, because you want a larger 'fuge,' or because you think the tank will overflow because a drain plugs up (another discussion on passive failsafety.)
