Warm Water in the Cold Water Line

Alan (California Radiant) Forbes · January 25

This Grundfos hot water circulating pump attaches to the water heater and has a thermal bypass under the furthest fixture, usually a lavatory or sink. The pump pushes hot water out of the heater to the fixture with the bypass. The cold water ahead of the hot water gets diverted into the cold water line until the water temperature reaches 100F when the thermal valve inside the bypass closes.

The problem that one of my customers has is that when they turn the cold water on, it's warm and stays warm for a couple of minutes. Is this normal or is there some leakage through the thermal valve?

Jamie Hall · January 25

Typing from a woeful lack of knowledge here, @Alan (California Radiant) Forbes -- but I wouldn't think that a leak in the thermal bypass valve at the fixture would do it. All that should happen is that the recirculating flow should stop (and, hopefully, the pump?). However, I can see two other possibilities. One is that the hot water line and the cold water line are run close together or even touching. That could do it. The other is that when the pump is running there is backflow to the cold water line feeding the hot water heater -- but I can't really envision how that would happen, given the various pressures involved.

Let us know what you find1

Larry Weingarten · January 25

Hi, It's not uncommon to experience just what you're seeing. Is there any possibility of running a separate recirc return line... maybe in 3/8"? That would fix the un-cold problem for certain.
Yours, Larry

Lyle {pheloa} Carter · January 25

I have seen a number of the shower valves that depend on a diaphragm in the balancing portion.. The diaphragm would fail and a bypass would occur at that point in the shower valve. During a time when the temperatures are normal. I would turn the pump on and feel the various waterlines in the house to find one that changes temperature when the fixtures are not being used..

EdTheHeaterMan · January 25

What you are describing is a way to use existing hot and cold water pipes to have a faux recirculation pump to maintain hot water at all the fixtures without a dedicated recirculation line from the furthest fixture to the heat source (water heater). If properly applied this can work in theory. I have found that it may not actually work in all applications.

The hot water that is being pumped from the DHW source to the furthest fixture may start out at 120°F and be pushing 60° (or cooler) water ahead of it to that furthest fixture. Where is the cold water going to go when this happens? back to the DHW source thru the cold water pipe. As long as there is a clear and defined separation of the hot water and the cold water in the pipe from the DHW source to the location of the thermal valve, the idea will work. We will call that location fixture Z (end of the line)

The problem happens when someone is using a hot water fixture at a location different from the thermal valve location. We will call that fixture B. When Fixture B is in use, the space between fixture B and the furthest fixture Z where the thermal valve is located the water inthat pipe can drop over time and cause the pump to activate and the thermal valve to allow water to circulate from the cold water at location B and take the less resistant path to fixture Z and back to fixture B. In some cases the pressure adjusting systems in some of those fixtures may allow the cold water from location B to reduce the water temperature to location Z and the thermal bypass will get a constant flow of 85° or 90° water for enough time to fill the cold side with that hotter water.

Eventually fixture B is closed and the less resistance path is cut off and the 120° hot water from the source gets to the thermal cut off valve and recirculation stops. By that time the damage is done. If it happens too often, then the only recourse is to find some way to run a 3/8" PEX tube from location Z back to the hot water source and abandon the cold water faux return for a dedicated return.

As an idea you may apply in some rare cases. If you are operating this faux recirc line thru the cold water to address only one area like a bathroom that is far from the rest of the plumbing in the home. You may get away with this idea i used for my own home in Brigantine NJ.

There was a normal plumbing system installed in this 1960's built home where the water heater was very close to the bathrooms and kitchen sink. so the wait time for hot water was less that 20 seconds. no big deal.

The previous owner added a powder room on the opposite side of this home and ran the hot water pipes up and over then back down to the hot water fixture for one lav. That took over 70 seconds for the hot water to get from the source to the lav in the powder room. it was was annoying to say the least. especially in the winter when the crawlspace temperature can get below 40° and all you want to do us use about 30 seconds of warm water to wash your hands.

I installed the system you have described and connected the motor to the light switch in the powder room. When you entered the powder room to use the facility, the pump would move all the cold water from the hot side to the cold side in about 70 seconds and the thermal valve would stop the circulation. That is just about enough time to do what you came to the rest room to do. Then you turn on the lav fixture and find that the hot water was already there. Viola! When you leave the powder room, you turn off the light switch and the pump would not operate until someone else turned on the light switch.

As I said, limited usage. This was not for the 8500 SqFt McMansion, but that place should have a dedicated insulated recirc loop for DHW. This is for the rare case where a homeowner thought it was a good idea to add an outdoor shower near the pool and a small powder room 100+ feet away from the DHW source.

Teemok · January 25

With thermal bridge systems I started using a digital timer for hot pipe times with a plug in recycle timer piggy backed on it. I test for the time it takes to heat the bridge from full cold and then set the recycle timer's on time 20% below that. Then I assess the insulation and pick an off time base don how fast I think it cools, generally between 15-40 minutes. The short on time is the key. Sometimes only 2 minutes on for every 20 minutes, depending on the pump and piping. I haven't had a thermal bridge complaint since and have fixed a few repeat failures. Yes if you happen to use cold water right after a heat pulse it will be warmer than normal cold but just for a second or two. That's the cost of not having a true return. Once explained most aren't bothered, if they notice it at all.

Intplm. · January 25

@Alan (California Radiant) Forbes

This is doing just what the system is meant to do. It sounds like it is working properly.

One thing you can try is to put a second bypass valve under a different sink in the house.

I did it and it cut the hot water from the cold side down quite a bit.

I did one house with three of these things under fixtures.

Depending how the piping is configured, and some trial and error, you can get some great results reducing hot on cold side timing while getting an even better response of hot water circulation to the baths.

Jamie Hall · January 25

Thanks guys. Learn something every day -- never occurred to me to use the cold water line as a false return line.

hot_rod · January 27

Teemok said:
With thermal bridge systems I started using a digital timer for hot pipe times with a plug in recycle timer piggy backed on it. I test for the time it takes to heat the bridge from full cold and then set the recycle timer's on time 20% below that. Then I assess the insulation and pick an off time base don how fast I think it cools, generally between 15-40 minutes. The short on time is the key. Sometimes only 2 minutes on for every 20 minutes, depending on the pump and piping. I haven't had a thermal bridge complaint since and have fixed a few repeat failures. Yes if you happen to use cold water right after a heat pulse it will be warmer than normal cold but just for a second or two. That's the cost of not having a true return. Once explained most aren't bothered, if they notice it at all.

Could a timer and a thermostat on the cold at the tank add some benefit. Adjust the thermostat to a low temperature 80F for example?

Teemok · January 27

@hot_rod "Could a timer and a thermostat on the cold at the tank add some benefit. Adjust the thermostat to a low temperature 80F for example?" The place for a pump switch thermostat would be at the bridge. Wires. By the time the cold makes it back to the tank at 80F lots of water has crossed the thermal bridge and will annoyingly be found in the cold line. When new, the bridges work well to stop hot saturating the cold line but if the pump runs for 15 min. or as I find them, continuously pushing at the thermal valve for hours. Any tiny leak makes the cold line hot. The recycle timer allows for more frequent short pulses, less stress on the thermal bridge and small leaks don't result in the undesired. No wiring and simple adjustable logic.
A wireless sensor at the bridge that reported to a pump logic controller that auto tuned a time clock and the off and on pump pulses for each individual case would be a superior control. Another retirement funding idea given away. LOL

Intplm. · January 27

Something to look at again.
There is a timer on the pump pictured above, and thermostat is provided on the heat source.

Teemok · January 27

What the tank or source thermostat does is not the issue. The timer is set by a human or not in 15 min increments or hours. The thermostat is a valve in the crossover bridge independent of the pump control. The problem is often hot water being push passed the thermal bridge valve after it's closed heating the cold water line.

dko · January 27

https://aquamotionhvac.com/hot-recirculation-systems/

Aquamotion has great systems, including cartridges that also fit Taco. Patents on their thermal bypass valves.
Won the innovation award at AHR Expo 2018

https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/11713562
Bi-metal coil instead of the paraffin wax Watts/Grundfos uses.

They also have on demand push button kits, motion detection kits, pumps with built in aquastat and/or timer. Both undersink and at tank/tankless systems. Great company to support in Warwick, RI that manufactures most of their own stuff.

B&G undersink system with built in temp control and timer that is also very popular.
https://www.xylem.com/en-us/products--services/pumps-packaged-pump-systems/pumps/wet-rotor-circulators/hot-water-circulators/ecocirc-b-23-5-act-undersink-pump-for-potable-water-systems/

Intplm. · January 27

Teemok said:
What the tank or source thermostat does is not the issue. The timer is set by a human or not in 15 min increments or hours. The thermostat is a valve in the crossover bridge independent of the pump control. The problem is often hot water being push passed the thermal bridge valve after it's closed heating the cold water line.

I understand.
But what you will not undo the under-sink bypass valve. Piping arrangements in homes are different in each home.
So, Would what you are suggesting work? Sure. Maybe, but to what end? The bypasss valve is gonna bypass.
Would it reduce the hot on cold time frame? Maybe reduce it to lesser than a minute? Two minutes? It's hard to say.

Alan (California Radiant) Forbes · January 27

The problem is often hot water being push passed the thermal bridge valve after it's closed heating the cold water line.

That's the problem here. It takes them an annoying amount of time to run the cold tap before getting cold water.

I've not had good luck with this system. The thermostat in the crossover valve is set to 100F and my customers have complained that it's not hot enough.

dko · January 27

If they only care about quick hot water and not about wasting water...

connect the hot out into the drain! like RO filtration systems..
though quite a risk if the bypass valve stopped working and didn't close anymore...
so better for on demand systems.

Intplm. · January 27

Alan (California Radiant) Forbes said:
The problem is often hot water being push passed the thermal bridge valve after it's closed heating the cold water line.
That's the problem here. It takes them an annoying amount of time to run the cold tap before getting cold water.

I've not had good luck with this system. The thermostat in the crossover valve is set to 100F and my customers have complained that it's not hot enough.

Have you replaced that little grunfos/watts valve?
When something new comes out, (and this has been around for maybe twenty years.)
I test it on my own home. I tested this as well. I found that some of these little under sink creatures will react differently just from swopping one from the other.
@Alan (California Radiant) Forbes what @dko has posted above might be of some help. (The aqua flash).
I haven't tried that one yet but will in the future.

dko · January 27

also like @Intplm. said using 3 of the bypasses, the aquaflash says up to 5 per pump. Maybe think of it like venting a steam system?

Intplm. · January 27

dko said:
@Intplm. also like you did using 3 of the bypasses, the aquaflash says up to 5 per pump. Maybe think of it like venting a steam system?

Absolutely. And I gotta tell you, when I tried the three under sink, there was nothing written if you could or could not do that.

The trial and error still comes into play.

So five per pump on the aqua flash. Good to know. If the aqua flash is anything it looks to be. It will have a good future.
Thanks for sharing.

Teemok · January 27

I read, "use three thermal bridges", as three stacked together in the same location to prevent the bleed through, not in three different locations.

Connection of potable water with no air gap to a drain is not great advice. I assume that's a joke.

Aquaflash looks promising as a better valve. Still don't like pumps running dead headed.

Using the cold line as a return is a less than ideal solution. A compromise from the start. You won't consistently get very hot water and there will always be some warm in the cold at certain times. When done right it's not but a few seconds of warm in the cold. When it is sold with that in writing and well explained verbally with the right timer set up I've had no complaints. I tell them to remember the cost of the proposal I offered to install a dedicated return pipe as the warm water comes out of the cold. The rich can pay the cost of closer to perfect. The rest are forced to make compromises.

hot_rod · January 27

Connection of potable water with no air gap to a drain is not great advice. I assume that's a joke.

Especially below the weir of the trap

Teemok · January 27

@Intplm. So, Would what you are suggesting work? Sure. Maybe, but to what end? The bypasss valve is gonna bypass.
Would it reduce the hot on cold time frame? Maybe reduce it to lesser than a minute? Two minutes? It's hard to say."

Easy for me to say. I've found it does work when tuned well. Skepticism is healthy and it's also easy to find for yourself it won't work when you don't believe it will.
The by-pass is intended to close not continuously bleed. The design meets a market need and has flaws. Valve leaks and pumps pushing at should be closed valves with poor pump control is not great. BUT it sells because people want to believe.

dko · January 27

hot_rod said:
Connection of potable water with no air gap to a drain is not great advice. I assume that's a joke.

Especially below the weir of the trap

Changed to tailpiece!

Wasn't a joke Teemok. Do people add checks for the most commonly installed non-air gap RO systems? They just tap a saddle into the drain tailpiece. Can make it possible with a backflow.

But yes, it's not an advisable solution. But a solution nonetheless for those who don't care about anything but the end result. Not sure why i'm arguing for it

Was just a passing thought, pay no mind.

dko · January 27

Teemok said:
The rich can pay the cost of closer to perfect. The rest are forced to make compromises.

I can agree with you are trying to say. But I deal with a lot of "rich" folks, and they are the more "cheap" than those with less. Guess that's how you become rich!

Yes imagine the systems you can furnish without a budget restriction.

Teemok · January 27

The rich "CAN"

Teemok · January 27

From Aqua flash instructions:
"Since the cold water supply line will be used to return the cooling hot water to the hot water heater you will experience some warm water from the cold water line. Opening the cold water faucet fully will flush the warm water from the cold line in a short period of time." Hmmm.... More reliable I hope.

I like that they have created an application for tank-less add on with an at pump flow by-pass. The thermostatic valve looks capable of higher bridge flows, meaning faster line heating. All that with a pre-connected wireless push button.

Intplm. · January 27

@Teemok

When I used the three valves at different points in the system it helped to bring the hot water to other places on the hot line lessoning the time from hot on cold at the previous single point of use.
It branched things out and shortened the time of hot on/in cold. That's all.

I discovered this through trial and error.
The three thermal bridges as you call them are not stacked together. They are installed at different further most fixtures.

Dare I hazard a guess that @Alan (California Radiant) Forbes has a bad recirc-valve? Possibly?

Teemok · January 27

@Intplm. I figured that's what you meant after considering it. I have also used multiple bridge locations with success but now always with the recycle timer. The first one I used the factory clock. I corded out of the clock box with a plug for the recycle timer and plug back to the pump. A digital battery back up timer is far better. The extra independently opening and closing bridges provides more time without the pump pushing against closed (supposedly) bridge valves. If the pump is set to constant on, the valves will eventually start to leak a bit and the cold lines get hotter and hotter. Shorter pump on pulses, I find, takes pressure off the valves. They don't tend to leak as much and if they do leak a little, it's much less leak-by than a the factory time clocks minimum 15 min on cycle. When the bridges are open it doesn't take long for the line to heat up and they attempt closing against pressure. The little seal and wax cap do their best but the smallest flaw/debris makes a small leak that erodes and grows till it doesn't seal well at all. The last sentence is my conjecture.

Larry Weingarten · January 27

Hi, Maybe it's only for the rich who are willing to part with the money, but Metlund systems work: https://gothotwater.com/ They have a sensor in line to detect when the water begins to warm up. Once it does, the pump is turned off and a valve closed, separating hot from the cold line. It's demand activated, with a button or motion sensor. With this system, the cold water stays pretty cold.

Yours, Larry

Teemok · January 28

@Larry Weingarten The trouble with the Metlund besides price: power needed under the sink, clutter under the sink and noise. It also does only one distance point. That said, I've put a few in over the years and they are good. Each situation is a factors read.

Warm Water in the Cold Water Line

Comments

Categories