I assume there’s a factual explanation, but after googling it for a while I can’t find an answer that makes sense to me, but I’m not a plumber. Here’s what happened.
I live in a house with an on-demand water heater and I’m taking my morning shower. For the past 3 months that I’ve owned the house there has never been a problem with the hot water, other than it takes a while to heat up when I first turn it on.
I’m five minutes into my shower and the bathroom lights go off. Before I can utter oh $%^#, they come back on. The power glitch was no more than a second or two. I wouldn’t have thought any more about it except within10 to 15 seconds the water starts cooling off, and after a minute or so it is unbearably cold. I shut the water off, get out, and dry myself. Thinking it has something to do with the water heater I wondered why it would go from hot to cold so quickly because of a short power glitch?
I went to the bathroom sink, turned on the hot water, and ran it for a few minutes, no problem. I check the other sinks in the house and the hot water also seems fine. I went back to the shower, turned on the hot water, and it is now working again. Somehow, the power glitch caused the shower to temporarily lose hot water even though there was hot water in the lines running throughout the house.
What could have caused the shower water temp to drop so quickly, but only for a short period, following a power glitch? Any guesses?
Because the heater has a controller board that had to “re-boot” after the power failure.
The board measures the temperature of the down-stream water and adjusts the power to achieve the desired temperature. When the power failed, the board got reset, and probably had to go through a bunch of self-tests before it would start working again.
I suppose that might explain it, but when I tried the bathroom sink after just getting out of the shower there was hot water there almost immediately. So the hot water that was heading to the shower was switched over to cold water while the water heater ran the self-tests?
Presumably, the shower takes a lot more water than the sink. Maybe that’s the difference.
But, it’s hard to know the reason without seeing how everything is plumbed.
Could the tampering valve ( what prevent you from being scalded if someone flushes the toilet while you’re in the shower ) of the shower head have something to do with it? Maybe suddenly getting only cold water messes up with it.
I think on demand water heaters have a pressure or flow sensor. When the pressure drops, it knows there is hot water demand. If the water was running during a power cycle, the sensor may not have reset and re triggered. The other water outlets would sense the drop, as they did not get interrupted during water flow.
What actually was the time between shutting off the cold shower and testing
the bathroom sink ? That could make all the difference. Could you try an
experiment with a tap nearer to your water heater: run some hot water, switch
off the power to the heater for a couple of seconds and monitor the tap situation.
I’d say it was between two and three minutes after getting out of the shower before I turned on the bathroom sink. The laundry room sink is closest to the water heater, but the master bathroom shower and sink are only a few feet further away. Hot water comes fastest to those three places, and it takes the longest time to reach the kitchen sink, which is annoying when I want to wash dishes and have to wait two or three minutes for hot water to arrive. I’m looking at installing a mini tank water heater under the kitchen sink to take care of that problem.
The furthest sink away is in the guest bathroom, but it takes longer for hot water to reach the kitchen then it does the guest bathroom. That is most likely due to how the hot water lines are routed in the crawl space.
Hot water going flowing to your shower. Previously you may have used hot water at your sink. The hot water supply pipes (probably the same until it splits between the sink and shower are full of hot water. Power goes out, hot water heater now has cold water flowing through it. The WH resets itself and starts working. In the meantime you quit showering.
So the hot water pipes have a hot water, slug of cold water when the WH wasn’t working that came out of the shower head, then hot water after the WH started working. The sink supply pipe still has hot water in it. If the slug of cold water passed the tee connection for the sink no cold water would make it to the sink.
With the slug of cold water in the shower supply pipe all other sinks would have no cold water in the hot water pipe.
Did you leave the shower running in its cold state while you went and checked if other hot water taps had hot water? I think not. You shut it off. Then the pressure reset. Now the On Demand system could see that there was a pressure drop or change in flow rate when you opened the other taps…
When you “reboot” electronics, the caution is to turn the power off, count to 10, then turn it back on. (Some say 30 seconds). Devices with a power supply (what doesn’t?) use capacitors to smooth the rectified current. Leaving the power off for a few seconds allows the capacitors to drain, so when power is restored, it starts from 0V.
Not an electronics expert, but this is my understanding… Most electronics chips (what isn’t using a microcontroller these days?) have a “reset line” that tells the chp to restart. This is tied to an RC delay circuit that takes a while to go to “on” (voltage transition signals restart). So, the chip does not execute “reset” startup routine until the power has had time to get to steady full voltage. If the power goes off and back on again quickly, there is a chance the processor is stuck in an unsual state and does not get a signal to reset.
You can freeze computers sometimes by quickly flicking the on-off switch, except nowadays that’s often not a physical switch, just a signal to the power supply. A power bump, or a momentary blink of the lights, can sometimes freeze a computer.
Why your unit eventually reset - who knows how it’s wired…
I think if I was as designing such a heater I would not be letting it start up in a condition where it was heating from the get go.
Too many nasty corner cases where it could behave badly and unsafely.
Internally it probably just has a simple programmable logic device implementing a small state machine. There is no code in the manner we think of processors running stuff. It just has logic that responds to changes in input through a fixed number of states. I would not have any transitions that energised the heater until the system had reached a known stable state. And that state sure would not include flowing water through the system
State machines allow for very robust and verifiable designs. There a very good reasons they remain a popular paradigm.
I think the point is so many circuits nowadays unless they are ridiculously simple as just as easily done with a programmable controller. Code revisions are a lot easier than scrapping and replacing circuit boards to change components. Quite often the program is either eeprom, or more permanently sometimes you program the procedure by “burning in” the code in the PROM, essentially the zeroes are set with little fuses on the chip being burned.
(I remember the early days of network hubs, we would get circuit boards with over a dozen fine wires soldered onto the board to revise the hardware without having to use a whole new printed circuit board. )
As someone mentioned too, perhaps the heat trigger is the flow going from none to on. But then, you would think a program would allow for the situation where there was an interruption in water flow, and a few other situations. (I.e. did not start at full pressure with a drop for flow) Imagine all the situations, these contingencies make for a complex state machine, i.e. a processor. Especially, the important thing in the design is to avoid a situation like dry “tank” that burns out the element.