Water velocity and pipe life - should I be concerned?

[Rossn]

I am looking to use a push-button on-demand hot water re-circulation system with a dedicated recirculation return line as part of my remodel.

I've read in a couple posts on this site that high water velocity can lead to pinholes in copper, and in the same post, some persons said that is not true and that it is only water chemistry. I've had pinholes in a different home with normal flow rates.

The pump options I'd be using could flow up to 8 or 18 GPM, and I was leaning towards using 3/4" copper for a variety of reasons (though PEX could be possible on the hot side).

Should I have any concerns with this type of water flow rate on copper or PEX pipe and fittings, from a longevity perspective?

http://www.gothotwater.com/hot-water-systems/components-specifications

 

[Reach4]

The pump options I'd be using could flow up to 8 or 18 GPM, and I was leaning towards using 3/4" copper for a variety of reasons (though PEX could be possible on the hot side).

An 18 GPM recirculation flow seems incredibly high. I expect the flow head is such that you never get to that part of the curve. This system would be energy efficient. Any erosion effects on pipe should be low due to the very low duty cycle.

 

[Sponsor]

 

[hj]

You ONLY need a flow rate that keeps the water warm/hot at the faucet. ANYTHING greater than that WILL cause erosion, especially at turns in the piping.,

 

[Reach4]

You ONLY need a flow rate that keeps the water warm/hot at the faucet. ANYTHING greater than that WILL cause erosion, especially at turns in the piping.,

Do you worry about that erosion for filling tubs or taking showers?

 

[Jadnashua]

While you may consider the Copper Institute's handbook biased, at least a majority of it is based on proven facts and science...on hot water, they recommend <=5fps flow rates. The actual volume depends on the ID of the pipe used. On 1/2" copper, that equates to about 4gpm, and 8gpm on 3/4". Water chemistry can compound problems, and is one reason why the hot velocity is lower than that of cold...heat accelerates most chemical reactions. Not reforming the pipe end after cutting it with a typical pipe cutter can also lead to turbulence, and increase the erosion of the insides of the pipe after a joint, too.

Given two systems, one with higher velocity will tend to exhibit problems more often than one with the velocity within the design parameters.

As stated, your least likely situation would be to adhere to those guidelines.

As to a tub/shower...most 1/2" valves have a maximum flow velocity of about 6gpm, and considering some of that is cold, you probably would not exceed that 5fps rate on the supplies. The outlet of a typical tub spout is much larger, and they do recommend a min/max distance from the valve, which probably helps, too.

A recirculation system, over the lifetime of the system, would be running far longer than the time the line to the tub would be used in most implementations. Even one where the pump may run constantly, a well-designed system will have valves to cut the flow once hot is achieved at the sensor locations, and even then, it often isn't full hot (but can be close). Most of the engineered systems utilize a very small pump, so velocity is not an issue, and the flow stops even if the pump may stay on. There is also an annoyance factor with running the water through the pipes at high velocities...it tends to start to make noises.

At 18gpm, you're over 200% of the recommended maximum velocity with a 3/4" pipe and 450% with a 1/2" line. The standards do have some margin for error in them, but 200% is probably pushing it, 450% is just plain asking for problems!

 

[Asktom]

Be sure to ream the copper or it will create turbulence which will erode the inside of the pipe. The higher the velocity of the water, the more important it is.

 

[FullySprinklered]

Should you worry about your pipes wearing out? Depends on how old you are.

Since I turned 65 last year, I've been offering a lifetime warranty on all my plumbing work.

You do the math.

 

[Rossn]

Really great discussion here -- thanks for the education!

For those suggesting low velocity, I'm trying to serve an area that will have about 100' of pipe, and was gravitating towards the push-button recirculation pump due to the associated energy savings and variability in use of the further areas. So, the faster the flow rate, the shorter the wait.

Haha -- nice on the lifetime warranty. I'm probably looking to be in this house for somewhere in the 20-30 year range.

I probably have about 8-9' of actual head from ground to 2nd floor, and it sounds like the S3 pump would be ideal and keep within reason on the copper velocities on the medium setting, at about 6.5-7 GPM, which is around the recommended max of 5fps for 3/4" copper < 140degF.

Now, head loss equivalent, I need to figure how to calculate. One-way (not the full loop), I'm looking at around 100' of pipe through a contorted path, with around (12) ell's, (1) 22.5, (5) straight run tee's (I think that's what it is called). The pump manufacturer recommends wide/long radius ell's.

Given the above, would I be significantly better off looking to PEX, and does PEX have the same issues with max velocity around 5 fps?

 

[Jadnashua]

There is a similar PEX installation handbook for whatever pex manufacturer you happen to choose, but pex does allow a higher velocity than copper in general.

I'd be more inclined to use a switch (could be in addition to a timer) to enable the system when needed, and not worry about trying to minimize the wait times or the potential erosion of the pipes. Occupancy switches are fairly easy to install, and may give the system enough time to get the hot water where you want it verses having to remember to turn the system on. The pump on mine is relatively miniscule, draws only about 20W, and probably only runs a total of 20-minutes during a typical day. With that system off, it can take nearly a minute to get fully hot water. With it running, the first time when the timer turns on, it might take nearly 2-minutes since the flow rate is slow, but after that initial run, and periodic runs throughout the day as needed and sensed by the aquastat, there's hot water everywhere within seconds. Running moderately warm water back to the WH also means less cold water from the supply overall is coming in, minimizing the energy use. The pump is so small that it is hard to tell if it is running even when standing right next to it.

In a closed loop filled with water, you don't need a lot of energy to move the water as whatever height you have to raise it, a corresponding amount is falling back down and pulling it along with the aid of gravity. Pumping water say from a well is a very different technical problem that requires a lot more power. You don't have that here.

 

[Rossn]

Thanks, Jim. That is exactly why I'm looking at the D'MAND system, so as to use a switch or motion sensor, instead of just a timer. This is in an area that may be used regularly sometimes, and not at all at other times.

I'm not concerned about the energy use by the pump... that is small. However, the energy loss due to 2+ gallons of hot water continually being heated and cooled (sitting) is a lot of energy to throw away when running on a timer. Additionally, I have to think about it kicking on the 230k BTU boiler more often if re-circulating a few hours per day, and that is a real energy concern.

 

[Jadnashua]

Two gallons of water, about 16#. Raising it from maybe 80-degrees once the system is up and running to about 105, 25BTU*16#=400BTU. 400BTU=about 117W maybe 10x/day, not even 2Kw, or about 4cuft of gas. Your boiler is probably in the order of 80-100K BTU, how often do you think it will need to run to offset a 400BTU inermittent load? The hysterisis on most WH controllers would indicate probably no extra boiler run-times most days caused solely from recirculation. I almost never see my boiler come on during the day except after the morning rush of bathing, or maybe if washing clothes or dishes.

A well-designed circulation system shuts itself down when the sensor temp reaches about 105-degrees, and even if it doesn't shut the pump off, running against a closed valve, essentially idling, draws very little power. Mine shuts the pump off. Been running for about 16-hours (enabled, not actually running) now for about 13-years. Insulate your lines, keep them out of unheated spaces, and enjoy.

 

[Rossn]

The boiler is pretty huge... 232k BTU, cast iron. I figured more like 1100BTU's to raise the temp from 65 deg to 120deg for the volume of water entering the pipe (since the water itself is heated to at least 120deg). I figured with the boiler at 85% heating efficiency, and the water tank also having losses, we might look more like 75% efficiency, or net about 1500 BTU's. That doesn't account for de-rating the boiler for 5000' elevation, which probably knocks it closer to 1700 BTU's to heat up that line when someone pushes the button. Of course, if on a timer, there is less energy lost in the line, but if you are sucking cold water from the unheated side of the re-circulation line or from cold water supply, you're still using the same energy to heat the water.

 

[Reach4]

Now, head loss equivalent, I need to figure how to calculate. One-way (not the full loop), I'm looking at around 100' of pipe through a contorted path, with around (12) ell's, (1) 22.5, (5) straight run tee's (I think that's what it is called). The pump manufacturer recommends wide/long radius ell's.

100 ft round trip or one way?

https://www.pexuniverse.com/pex-tubing-technical-specs has some pressure drop numbers, and you would add the effect of fittings.

http://www.pressure-drop.com/Online-Calculator/ is a pipe calculator.

 

[Jadnashua]

The thing is, the water you'll be pushing back to the WH won't be anywhere near 65-degrees, especially if you have insulated the lines except maybe the first run of the day. At the sensing point, most systems try to maintain that at 105-degrees, and the water behind it is hotter still. So, you'll not be heating 65-degree water. A BTU=1# of water, 1-degree. The numbers I listed are actual energy to the water, so yes, your boiler efficiency would play into it, but at 75%, you're still talking about 500BTU verses the 400, and as I said, at least in my system, it only runs less than 20-times a day, for a total of probably less than 20-minutes. WHen you are actually using hot water, that limits the number of times the recirculation would need to run since it would have recharged the lines with hot in the process.

 

[Rossn]

Ok, thanks -- I'll check out those sites.

100' is one way. I would be putting the pump at the far end, so as to only push heated water half-way and not waste the heat energy on the return. That's also why the water returning to the boiler will be closer to room temperature. The alternative would require as much or more energy to heat the pipe the full length of the loop, and in the long run, more energy would be lost, regardless of frequency upon which the pump is run, right?

 

[Jadnashua]

The water on the return side will still be a lot warmer than your estimated 65-degrees. Assuming it is in conditioned space, and you maintain a nominal 70-degrees there and the pipes are insulated, it shouldn't be anywhere close to 65-degrees most of the time. Then, you don't really need the point of use location to be the full 120-degrees, but close to it is more than enough. Most packaged systems aim more for about 105-degrees. As long as the sensing point is at the point of use, it doesn't really matter where the pump is...and, the pump can continue to run without a lot of wasted energy, if the sensing point shuts the return line off. The RedyTemp system I have has an adjustable aquastat and shuts the pump off. I have it stop when the last POU location on the loop is warm rather than hot. I use the cold water as my return since running a return line would have been a major pain, but the system would work in the same manner if I had. If I flush my 1.6g toilet, it takes all of about 2-3 seconds to flush the warm water pushed into the cold line and get full cold water there after the toilet has refilled, so that warm water really didn't fill anywhere near the full path.

A recirculation system doesn't have to be an energy hog if you insulate your lines and don't go wild.