Typical progression of a leaking water heater towards catastrophic failure

[Pantherkitty]

This is sort of off-topic since it's not directly HVAC-related, but the last few days have gotten me thinking... when a water heater starts to fail and leak, what kind of volume of water leakage would you typically see in the early stages of failure, and how long would you normally have before the failure would progress to a point where the pan is likely to be completely moot unless it has somewhere to drain into (or EVEN IF it has somewhere to drain into)?

My present water heater (40 gallons, electric, at least 12 years old) is sitting in a (largely symbolic?) pan whose diameter is approximately two inches larger than the heater itself, and has no drain it could conceivably empty into anywhere nearby... if the tank ever starts leaking, I think the pan could probably hold one or two gallons of water before overflowing (the house was built in 1980, before drip pans for water heaters were required by code).

Idea #1: run another 3/4" PEX line up to the roof next to the one I just ran for my upcoming new air conditioner's condensate removal, buy an additional condensate pump, and plumb the water heater's drain pan to empty into it so that if the water heater develops a leak, its (repurposed) "condensate" pump would hopefully notice and start pumping a meaningful amount of water up to the roof (vs having it uncontrollably overflow into the hallway, with catastrophic results).

Idea #2: get a (custom-fabricated) drain pan large enough to sit under BOTH the new air handler AND the water heater (say, 45" x 24" x 3"), with something to raise the water heater a few inches above the pan's bottom so it won't get wet if the air conditioner actually drips into the pan (like this: https://amazon.com/dp/B00LUU0NCY/ ), and a secondary pump to try and drain it if it ends up with water in it.

When I had installers come by a few weeks ago, they were all ADAMANT that I couldn't have a single drip pan shared by both... but when pressed for details, none of them could actually come up with any specific reason WHY... they just "knew" it "wasn't done". My (cynical) theory is that they just don't want to touch the water heater.

Is my cynicism justified, or is there some concrete technical or legal reason why the two can't or shouldn't share a single large drain pan? It seems like for a situation like mine (side by side air handler and water heater, no viable drain), it would be a fantastic solution. A 45 x 24 x 3" pan could hold almost 10 gallons of water (vs 1-2 at best with the water heater's current pan)... not enough to contain the damage if there were rapid, catastrophic failure, but possibly enough to make the difference between "a mess" and "horrific catastrophe" if the tank started leaking slowly overnight, while I was at work, or when I was away for a day or two.

I can see why there might be code objections to a water heater sitting directly in a pan shared with an air handler (water heaters rarely leak, and aren't designed to sit in standing water for extended periods of time... air conditioners SHOULDN'T drip into their pans, but nevertheless it happens a lot more often than anyone wants to admit, and often gets ignored for weeks or months when it does). On the other hand, it seems like raising the water heater above any likely condensate water level should fully neutralize such objections.

If #2 really can't be done, does anyone think doing #1 might have value? I guess it really comes down to how quickly a water heater actually progresses from "first few drops and trickle of leakage" to "total, catastrophic dumping of their contents in the hall (even WITH an automatic shutoff valve)."

[beenthere]

Some water heaters Drip for months. Others gush out almost right a way.

[firecontrol]

A water heater installed where the pan cannot gravity drain into a suitable floor drain should have a sealed pan with a water sensor in it connected to an automatic water supply shutoff.

When they start leaking, replace the water heater. A leak indicates a failure and in the case of a large vessel filled with pressurized water sitting in your home that could potentially cause thousands of dollars worth of damage is not something you really should "wait" on fixing.

[Pantherkitty]

Oh, a water heater leak of any kind definitely wouldn't be ignored. If I found water leaking from the tank for any reason at all, the water unquestionably gets shut off instantly, and I'd almost certainly connect a hose & drain it immediately.

The specific failure mode I have in mind with #1 or #2 is a relatively slow leak that begins at some point when I'm away. At that point, without some way to pump away the leaking water, massive damage would be inevitable EVEN WITH an automatic shutoff valve, because the tank would STILL have 40 gallons of water inside.

The only real question is whether a repurposed condensate pump could pump away most or all of those 40 gallons as they leaked out quickly enough to avoid having them end up overflowing the pan, or whether the tank would "give way" before then and send it all gushing out at once (in which case the pump would be futile).

As a backup plan, I was thinking about putting a tee on the hot water pipe near the washing machine, with a second (normally closed) solenoid valve along a pipe leading from that tee directly into the washing machine standpipe. Then, if water were detected in the water heater pan, I could electronically cut power to the water heater, open the "dump valve" downstairs, then close the heater's cold water supply a minute or two later (using the incoming cold water to cool off the hot coils and avoid turning residue left on exposed hot coils into steam). Hopefully at that point, the amount of water not drained away by the dump valve downstairs (the water heater is upstairs) would be small enough for the condensate pump to manage (especially if the open valve downstairs gave the water an exit path with less resistance than the leak).