I already contribute to the discussion by pointing out your error with respect to tankless water heaters.
There are “tankless” heaters with reservoir, albeit small.
I have a tankless. The incoming waterline is split and runs in smaller tubing through two heater chambers. There are a couple of 90 degree bends. Larger units have three or more chambers. The heater is definitely more restrictive than the rest of the piping.
My heater’s manual (pdf)
Restriction alone is not enough to explain what the OP is seeing. A pure restriction would result in a reduced but constant flow. There needs to be some type of “springiness” element in the system, either air or an elastic pipe. The other thing that can cause this effect is rubber wshers that expand as they heat up, but that doesn’t seem to fit the OP.
The pipe (from the heater to the faucet) slowly fills up with water in between times that the faucet is turned on. That could be your damper, right there.
If it’s an air pocket, why doesn’t it do this with other faucets? I was under the impression that it didn’t. Perhaps the OP could elaborate.
[Moderator Note]
Insulting other posters is not permitted in this forum. Don’t do this again.
If you consider posting “water is incompressible” to be intentionally baiting you, you are taking this message board way too seriously. Get a grip.
Colibri
General Questions Moderator
Isn’t that pipe always full of water?
[Moderator Note]
This is out of line too. If you have a problem with the behavior of other posters, please report the post.
Colibri
General Questions Moderator
[QUOTE=Colibri;10869723If you consider posting “water is incompressible” to be intentionally baiting you, you are taking this message board way too seriously.[/QUOTE]
Um, no. I considered this
to be an admission of baiting. He explained that it wasn’t meant that way and that’s the end of that, AFAIC.
The pressure drop across the heater isn’t a linear function. When there is no flow the pressure drop is zero. As the faucet is opened, the flow rate starts at the maximum for that pressure to the faucet, where it is limited by the size of the opening. As the faucet opening is increased the pressure drop across the heater increases until that becomes the flow limiting factor.
If the faucet were opened to some amount less than fully open that matched the maximum flow rate through the heater at the incoming water pressure there wouldn’t be a surge.
Sorry, but that’s simply not the way fluid flow works. The flow is proportional to the sum of the restrictions in the entire system (for a linear system). Assuming we are not talking about enormous pressures or flow rates. Here is a page that discusses fluid flow:
http://www.atp.ruhr-uni-bochum.de/rt1/currentcourse/node92.html
Note specifically:
Sorry, but fluid flow is just not that simple. We’re not talking about a constant flow or an ideal resistor. Note that the section on the next page of your cite: http://www.atp.ruhr-uni-bochum.de/rt1/currentcourse/node93.html talks about inertial effects.
In this case there is a “variable resistor” in the faucet, and an “inductor” in the heater. The turbulence in the heater is not a constant, it depends on flow rate. This is a problem that would be difficult (if not impossible) to model and derive a computational answer.
Yeah, yeah.
But that might take all of what, .1 second before everything reaches equilibrium?
This is not the reason for the OP’s many-second flow decrease.
If you say so. You’re the fluid dynamics expert.
Far from it, although I do design digital pressure gauges.
But, think about it - you have a spigot connected to a ten foot or so length of copper (I’m assuming) pipe, connected to the on-demand water heater, connected to a source of water. if you open the spigot, the column of water in the 10 foot pipe is going to be pushed by water flowing through the heater - it’s not going to start out at high flow and then taper off. The only way I can see this happening is air trapped in the system (could be before the heater) or elastic pipes - you see this effect with garden hoses. A system with rigid piping is not going to exhibit these symptoms without some sort of energy storage system.
Hey - I just thought of something.
There may be a valve in the on-demand heater which is responding to the water flow. What is the model of the heater?
No static restriction in the pipe could make it act as described. Try a thought experiment. You have a 100 foot long one inch diameter schedule 40 pipe with 20 psig on the pressure side and atmospheric pressure on the other. Turn on the water. Assuming you’ve got no restriction in your valve the flow will be approximately 28 gallons per minute. That’s from my Cameron’s Hydraulic Handbook.
Now, put an orifice with pressure drop of 20 feet of hydraulic head in it. it doesn’t matter where. Turn on the water again. The flow will be around 20 gpm. Does the water at first rush out more quickly and then slow to 20 gpm? How could it do that? The pipe is hydraulically full. For the water at the discharge end to rush out more quickly than the water at the pressure end can fill it is not possible. You’d be trying to pull a vacuum on it. The whole pipe flow is 20 gpm.
There are only two things that can cause the effect described. One, a pressure reservoir of some sort that initially gives a burst of flow and then slows as the pressure is depleted, or a restriction that increases the pressure drop as time passes.
I’d suspect there’s some kind of safety device in the faucet that will restrict the flow of hot water to prevent scalding. It initially offers little or no resistance, but the pressure drop across it increases when the water temperature exceeds a set value.
The restriction does increase the pressure drop as time passes. It’s not an ideal orifice.
With your figures as an example the flow might start at zero and climb to 23 gpm as the faucet was rapidly opened, and quickly drop back to 16. It never exceeds the free flow rate through the pipe at the inlet pressure but it seems like a surge because it starts low, goes higher, and drops back lower than the peak. That peak is still lower than flow through the cold water line.
It fact the inertia of the initial higher flow in the pipe downstream of the heater would be trying to pull a vacuum on the heater, and increase the flow rate through it until it slowed.
I’m assuming the city main has an infinite flow rate and no pressure loss due to flow into the house.
And that the house isn’t on a treadmill.
I’m with you on this one, Cheese, my first thought was an out of kilter valve. Turning to 100% on is actually turning it past 100% to partway off.
But since your suggestion has been entirely ignored by those with apparently WAY more knowledge about how water works that I have…I’m doubting myself.
In the system I described, that is a pipe of equivalent length 100 feet + a restriction with 20 equivalent feet of pressure drop across it the flow can never exceed 20.4 gpm. There’s no place for the driving force to come from. To send 23 gpm through the same system would take 22.9 psig pressure.
I’m confused by the inertia reference. We initially have zero flow, and 20 psig of mains pressure to drive it. When the valve is opened the flow will increase until the frictional losses in the pipes matches the pressure in. The frictional losses can not exceed the pressure driving it.
You seem to be describing a system where the fluid on the discharge side flows, pulling the fluid behind it in some kind of siphon effect. That’s not how it works. The discharge water is forced out by the pressure.