That "thick headed engineer" was right and understood the scope of the problem. You on the other hand are still guessing. Nothing new, it's been a long source of amusement to me and other engineers to listen to tradesmen try to explain concepts they don't fully understand while they expound on all the engineers being idiots. (We learn to listen to the symptoms for clues, and to ignore most of the "analysis" offered.) My favorite example of this sort of blue collar analysis was the fellow in charge of maintenance for the water dept. in another town long ago explaining why all the leaks were finally showing up when warm weather arrived after a very hard freeze/blizzard: "The warmth is pushing all the cold deeper into the ground and busting the pipes." I was in my first year of engineering school at the time but already knew he was full of it... Coulda been your long lost brother.
It is a "Strange Day Indeed" when plumbers start agreeing with their senseless ramblings...
Stranger indeed is the day when you actually understand the "why."
It's plain and simple Expansion Tanks Work!
That they do, as long as the bladder remains intact and they are maintained. (Remembering our well house tanks as a kid, those bladders have a limited service life.) With those caveats they will even work when you have a system that wouldn't have failed your test in the thread you linked to. (As in the system I observed and measured before changing out toilets...that were acting as pressure relief.) Any time the valve pops it might not reseal--that includes people "testing" them.
Instead of the tank going through pressure cycles and being subjected to metal fatigue and an early failure, the tank pressure remains constant.
Gee, based on that astute analysis there must be a million or more water heaters outright rupturing each year from metal fatigue and blowing houses down! (Because thermal expansion tanks don't seem to be standard fare yet...they are becoming so but until I had one put in I had never lived anywhere that had one.) Most of what I've read and seen of water heater failure is corrosion rather than fatigue. Leaks, drips, pinhole sprays seem to be the most common failure mechanisms. Most likely a rupture is going to occur because of a double failure: gas control valve sticks open and T&P fails (or plugs.)
One thing I remember my family talking about decades ago was a neighbor's water heater relief sticking open while they were on vacation. Steamed the house thoroughly.
Neither device protects against runaway heating, that is the job of the BTU rated T&P valve. With the price being equal and one having the ability to waste large volumes of water while subjecting the water heater to life shortening pressure cycles, it is like a no-brainer to me but that doesn't always work with engineers.
Good luck with that. I've done the relief valve cases for worst case scenarios (heater input as well as fire and/or known reactions). And I've seen what happens when a properly sized (by another) relief valve plugs on a process vessel due to a case outside the envelope... It's a whole lot more impressive than your little water heater rupturing. Engineers don't like to rely on relief valves for that very reason. By the time we get to the relief valve a helluva lot of other things have already gone wrong.
And having dealt with many relief valves in many hazardous services, I'll point out that any time you open a relief valve there is a fair chance that it won't fully seal. That is actually one of the best arguments for the thermal expansion tank, much better than "metal fatigue". A few cc's of water in a pan every day is no big deal, but if it were ever to stick open or erode...that would be a problem. The temps in a water heater are much milder and they are somewhat buffered as opposed to the condensate I've worked with in low, med, high pressure industrial boilers, but I don't discount the corrosive characteristics of hot water on carbon steel when exposed to air.
Testing T&P's seems to be a Catch-22. The services they are most likely to leak or plug in are ones with sediment/scale. If you don't have sediment/scale build up they are more likely to work. If a person jacks with one of mine without asking first, they'll get to pay for the replacement if it sticks open. And unless you test to see if the valve really opens at its set temp and/or pressure you really don't know if it will protect the tank or not...making the "test" highly questionable. That's why in industry we have relief valves on a regular test schedule so that periodically they are pulled and bench tested/reset/rebuilt. Of course, when they come back they sometimes leak...oh well.
jadnashua explained the reason for the growing need of expansion tanks well. His experience matches mine.
