We had a hot water recirculation system installed a few years ago, which has a pump (on a timer) that circulates water from the hot-water line at the back of the house to the cold-water inlet at the water heater. There is a one-way valve that prevents the cold water from flowing back into this loop. This valve has needed replacement twice because it became clogged.
But the main problem is that, if the pump is turned on while a significant amount of hot water is drawn from the back of the house (a couple of showers), a fairly large temperature drop in the water occurs. The drop is much more than when the pump is off.
My theory is that when the pump is turned on while hot water is drawn from the water-heater, and thus cold water is entering it, increased mixing of the temperature-stratified water in the heater occurs, thereby reducing the temperature at the top of the water heater.
Is this correct? Is there a fix?
Thank you for your thoughts.
Replies
the way i understand the system you describe is: when a hot water valve is opened in the back of the house the recirc valve senses via thermostat the water in the pipe has grown cold and it kicks on to send cold water back to the water heater thereby drawing all but a trickle back to the water heater for a few seconds untill temp comes up and it shuts off and water is hot at the place valve was turned on much faster than before system was installed.
is this what you have, a retrofit recirc system?
the method for discovering the problem (troubleshooting) is to list the "sequence" of operation as i did above, and then one by one make sure that each step the system is acting like its supposed to.
if you aren't sure of exactly what is supposed to happen when, then you will have a tough time finding the problem.
if you aren't sure how it is supposed to work (sequence of operation) you need to find that out first.
if my description of sequence of operation was correct, your problem is with recirc pump thermostat.
if it is not a retrofit recirc system like i described above, but a system where a pump was installed close to the heater (you already said its on a timer) and its coming on because the timer kicks in right when you are in the middle of a shower?
i have heard of problems of hot water not lasting long enough because of the recirc pump was recycling it "too fast", so if you have ball valves ahead of and behind the recirc pump (like it should have been installed IMO) you can just turn the valve slightly to reduce the amount of water passing through the valve.
you want to restrict the flow on the "downhill side" this will not harm the pump and actually use less electricity. (the pump doesn't know its restricted, it just turns slower and draws less juice, hydraulics is way cool)
lastly, along with the timer is there a thermostat? could that be what is causing the pump to kick on when you don't want it to kick on? if so your thermostat may be out of calibration, or it may not be working correctly.
if it is out of calibration, but still working, what that means is that the temp setting is NOT what it says on the dial. ignore the numbers on the dial, and turn it up or down and see if that helps. note: be methodical, you are a scientist conducting an experiment, carefull observation and documentation. in other words, turn it up and leave it for a few days and see if it is worse or better, if its better but not quite good enough, turn it up some more. if its worse, go the other way, but be patient! the worst thing is to go back and forth chasing your tail.
tell us how its going, you have piqued my interest in this puzzle, i need to know how it comes out!
Segundo,
Thanks. The system has no thermostat, just a loop with pump on a timer. The loop pumps water from the hot water pipe at the far end of the house, into the cold water inlet at the entrance to the water heater. There is a one-way valve between the pump and return line to the cold water inlet to the water heater. The pump and valve are located close to the water heater.
Laseruomo
Don't know much about it but let me just say a few things: The cold water inlet into the tank deposits cold water in the bottom of the tank, right? And the hot water is pulled from the top of the tank. So, my guess is that when the pump kicks on while the hot is being used, or has recently been used, the water that is entering the tank should be pre-heated, except possibly the first gallon or so (however much the return pipe holds). So I'm thinking that it would be not so much that the drop in water temperature at the shower is due to cooler water coming from the water heater tank but rather because of maybe a drop in pressure/flow rate of the hot coming from the tank due to the fact that the pump is pulling hot water from where it is needed. (hope that made sense)
I'd be interested to see if the temperature dropped significantly if the hot water only was running and the pump kicked on.
Again, I don't know anything about this, but I wonder if some kind of pressure switch could be installed that would prevent the pump from turning on when the water was other than at full pressure - ie, when the water was being used.
matt's absolutely right, it won't be stratification, it may be reirculating too fast.
I believe that recirculation pumps are normally quite low volume. Maybe the pump is just oversized? Again - just throwing out ideas.
i am not sure of what volumes are, if recirc pumps are fast or slow, but i am positive that a problem of running out of hot water too sson was solved by restricting the flow on more than one occasion.
we don't have a lot to go on with this one so far, hopefully he is under the house working on it now and will get back to us soon!
Segundo,
No, I'm NOT under the house working on it. :-) It has been this way for several years. Some people think that the return line of the recirc system should enter the drain of the water heater, not the cold water inlet. Do you have any thoughts on this?
Thanks.
Laseruomo.
i believe it is correct returned to the cold inlet downstream of the check valve.
no thermostat, you are recycling to much hot water, restrict the amount of water going through the recirc system (slow it down) and you will have solved your problem.
you should have a valve somewhere (if you ever need to replace or service the pump you should have valves upstream and downstream of the pump so you don't have to shut off the water to the house) i prefer ball valves.
just turn the valve downstream of the pump to slow the recirculation flow, and send me your address so i know where to send the bill!
You have hit the nail on the head. I have seen this many times with recirculating systems. It doesn't really matter much if the return water enters the drain or if it enters the cold water supply, the return water still ends up pretty close to the bottom of the tank.
If you have a pump that is moving too much water, it will shoot the cold water that is coming into the tank from the supply, right up to the top of the tank and cool the water you are using.
The solution, as you said is to valve down the flow of water from the pump or use a much smaller pump.
The very best way is to design the system so there are no heat traps and use natural circulation without any pump.
Another problem associated with too much pump flow is that you will circulate sediment or droppings from your anode rod into the system and clog your filters on washing machines and dishwashers.
Segundo,
Thank you for your help. I will try restricting the water flow in the hot water recirc system. There is a flow valve in there now, that I always leave wide open.
Laseruomo
One elegant way to restrict this flow would be to use a Grundfos comfort valve at the far end of your re-circ system rather than the simple check valve you have been replacing (probably because it is catching air pockets and corroding) The Grundfos comfort valve has a check valve with a thermal sensor (a wax expansion cylinder) that allows water colder than 90 degrees to return to the water heater but throttles down the flow rate as the temp of the water gets to 90 degrees and stops it al together as the temp exceeds 90.Attached are two images, the first shows the set-up pre-assembled prior to installation in the master bath vanity. It is intended to be installed using the cold water supply as the re-circ return so it has a hot water supply connection on the left with a riser to the vanity top left the thremal check mechanism in the center then a cold water riser on top right and cold water supply connection end right. We are using a dedicated return line so we have to cap the cold water riser as it is not needed. We connect the cold water supply connection to the return line. You cannot just put a 1/2" cap on these valves nor can you us a 1/2" FIP to PEX adapter as it's a ceramic thread that is not compatable with FIP threads. I have already written Grundfos about the inconveniance of this with no effect. The other image shows it installed in a recent home we plumbed. the bottom supply in the cabinet is the cold, middle is hot and top is return. The valve is mounted upside down to ease the connections, the supply line that is hanging straight down is the one that is capped off at the end and would have gone to the cold inlet on the vanity faucet had we been using the cold water line as a return line. I hope this is helpful, I have used this in homes with double master bedrooms where we have a single Taco 006 circulator pump pressurizing the hot water and two of these Grundfos comfort valve set-ups regulating the return flow in the two master bathrooms.
That is a functionally elegant solution for sure! Hot water just circulates to the supply legs that need it, no more, no less. Too bad Grundfos won't help with the cosmetic elegance. Could the "ceramic threads" on the valve be screwed directly to their ISO flanges? (The ISO flanges are what is used on their pumps). I know that ISO flanges are available with pipe threaded centers, and could be put on the end of the braided flex lines. Then the expansion valve can be installed ISO to ISO flange.I haven't studied your arrangement to know if that is any advantage, just a thought...
I'll look into that more but the guys at Grundfos never thought to suggest it. I've been using Taco pumps for a long time so I'm not familiar with those ISO flanges.
Question to all:
What is the impact on one's water heating bill for these reticulating systems - be it pump driven, or just convection?
Sure you are saving on not wasting money and time waiting for the hot water to become available at the spigot, but I'm guessing that amount of water might be in the tenth of a cent range whereas having your water heater kick on periodically would be a lot more significant $s. If a convective loop type setup is really a good thing, why is this setup not more common, or even standard practice?
Not trying to rain on anyone's parade - just trying to get the facts.
I think a set time for the pump to start would not work in our house.
Our shower use is to varied. Seems like a unit under the farthest sink with a push button to start it would be the most economical.Have looked at Grundfos and chilipepperappliance.com. Both seem more appropriate for our household. You can even get a remote so you can start the pump before you even get out of bed.. How cool. Warm when you get to the shower.Pete
Yes, depending on how well insulated your piping is you will definately use more power (but much less water) I have a "no one home" shutoff switch for the whole system near my front door but I only shut it down when I'm on vacation. Some folks use timers, some use a motion detector switch in the master bath. Some wire it into the master bath light switch.
It's pretty hard to come up with figures to answer that question. There are so many variables. I have electic water heating in my home and I didn't notice any increase at all with the loop system compared to my last home without it.
You must consider that many people wait 30 seconds or more for hot water to appear at the faucet. All of that wasted water has to be heated. Not only that, but in our area sewer bills are two to three times your water bill.
Even if there is a small increase in cost, it is worth it in my opinion, just for the convenience of having instant hot water at the faucet.
Most people make the mistake of moving way too much water through the circulation system. All you need is a very small flow to keep the water hot at the fixtures.
In addition, you should insulate all of your lines very well with a high quality foam pipe insulation.
I would never build a new house without a circulating loop. You should build it so there are no heat traps in the line and it will probably work just fine through natural flow without even having a pump in the system. Mine has no pump and works perfectly.
regarding pipe insulation and convection reticulation loops, I read that the return part of the loop should be uninsulated. Do you concur? I plan to build another house for my wife and I one day and am more interested in a convection loop than some device that requires power, maintenance, and extra installation costs.
PS: convection reticulation loop - I guess that is what you would call it? Would the idea be to install the hot supplies higher than the cold return? Like maybe the hot in the ceiling and the cold in the floor? Or is that even necessary?
I have my whole system insulated. There is no need to leave part of it uninsulated. The water is going to cool down more the farther you are from your water heater, so it will just naturally flow in the tubing back to the heater. What you want to do is go to a much smaller tube after you pass your last fixture.
You can return with a 1/2" or 3/8" line.
My return line enters the tank through the drain valve. The supply simply comes out the top of the water heater, follows along the bottom of the floor joists and returns to the water heater also following along the floor joist (at the same elevation).
I snake it around the house in a big loop so it is very close to each appliance or faucet. It takes about a second or so of water running to have hot water at any faucet. People are always amazed when they come to our house and turn on a hot water faucet!
What you want to avoid is a heat trap, which will stop natural convection dead in it's tracks. This is a situation in your piping where the supply pipe rises sharply, travels a short horizontal distance and then drops sharply back to it's original level.
You also must make sure that you don't have a heat trap built in to your water heater.
If you have a two story house, you will probably have to pump the water.
The trick with any type of recirculating system is to keep the flow to the absolute minimum required to keep hot water at the faucets. Anything else is just wasting energy.
What controls the start/stop of the pump. It doesn't run continuously does it??
Yes the pump, a little Taco 006 drawing slightly less than 60 watts of power, runs continuously and the valve adjusts the flow to keep the water temp in the MBR from exceeding 90 degrees.
The reason I asked is that I have seen these pumps get very hot because they are running constantly when the water is not moving through the system to keep them cool.
I would think you would have a very short pump life if those pumps run for an extended period of time with the Grundfos valves shut off completely.
<<The reason I asked is that I have seen these pumps get very hot because they are running constantly when the water is not moving through the system to keep them cool. I would think you would have a very short pump life if those pumps run for an extended period of time with the Grundfos valves shut off completely.>> No doubt that they do get hot, but I have one in my house that has been going for four years with no failure. That's one of the things I like about the Taco pumps and their soft plastic impellers, they seem to be more abuse and grit resistant. Also the Grumdfos valves don't exactly shut down completely, more like a trickle, you can actually hear them hissing a little as the water squeezes through the choked down valve.
Thanks Matt, I understood what you said, and no, it is not a drop in water pressure, I think. There seems to be plenty of that. To clarify my theory:
Warm water is less dense than cold water. When the pump is ON and we're taking a shower (for instance) and warm water enters the top of the tank (at the cold water inlet), the slightly warm water that enters the tank now floats up in the water heater and destroys some of the natural stratification that exists when only cold water enters at the bottom.
Laseruomo
Why does the recirc line tee into the tank cold water inlet? It is supposed to connect at the tank drain.
i respectfully beg to differ, thats the way we used to do them, (hook them into the drain) but it is supposedly better to route in to the cold line just downstream of a check valve to prevent stratification in the water heater.
now i could be wrong.....
this was not my idea to do it this way, but i was convinced that it was a superior method.
since it was not my engineering i will not argue the point, that being said, many times i do not know why, but i know what to do!
I need to correct my earlier statement that the recirc return should only connect to the tank drain. According to the pump manufacturers, the return can connect to tank drain OR the tank cold water inlet, always using a check valve to assure one-way flow. The pump must be located below the top of the tank water level, to avoid air in the pump.Poster said hot water was depleted sooner when pump was operating. I agree with throttling the pump and/or checking elements if HWT is electric.Poster said "check valve was replaced twice" due to sticking open. Since the recirc return is to cold water inlet, I wonder if the check valve is located above tank and is corroding from air pockets?
If it's an electric water heater the lower element might be bad.
Laseruomo:
There are a few plumbers around here that might be able to help you with your problem.
Edited 2/10/2007 7:35 pm ET by Matt
One thing to test is measure the tempature of the return line.
Measure it at the shower or at the vanity. And again at the pump. If water is still flowing from the shower to the tank. Then it will be colder at the tank.
Also measure the tempature from the tank hot water line. If the tank is not supplying enough hot water then that tempature will drop.
I can't remember the details, but I remember one person that was having some kind of problem with temp changes on a recir system. Finally found that the pump was installed backwards.
I don't believe in the stratification problem unless the pump is circulating a H*LL of a lot of water.
You should be able to draw off multiple showers or other users at the same time without a drop in temp until the tank runs out.
The amount of cold water in and hot water out running 2 showers should not be more than one shower and the pump.
One problem would be a bad dip tube.
.
A-holes. Hey every group has to have one. And I have been elected to be the one. I should make that my tagline.
before i got out the thermometer i'd take a close look at the pump and see if it has any little arrows on it. those little arrows are a good indicator of what direction the water is traveling, of course the individual will still have to determine by him/herself if that arrow is pointing in the correct direction, which apparently some people (even plumbers) find difficult.
the odds are in your favor however, even if you never bother to consider what direction the pump should be mounted, just pick it up and slap it in, you have a 50% chance of getting it right.
this simple mathematical fact should greatly increase the odds for anybody that actually looks at what they are doing.
Let's look for the obvious first.
Like post 34 & 35 check direction of pump flow ( it's probably correct) but never hurts to look.
Check for one way valve direction, they are called check valves FYI.
But like Mr Hartman stated this sure sounds like a dip tube problem.
If the tube is broke or has a hole towards the top, then the cold water coming in short circuits the tank under usage.
Too high of flow, is a rare thing, but let's check to be safe, what size pipe is the circ, size & hp of pump. How long is the circ line?
“Well, the election campaign in the country is picking up speed... All the candidates are talking about health care now. Don’t they realize that it’s their campaign speeches that makes us sick?” —Bob Hope
Too high of flow, is a rare thing, but let's check to be safe, what size pipe is the circ, size & hp of pump. How long is the circ line
Bill, I don't agree with that at all. I think it's a very common thing. Most people I know that have had trouble like he's having have cured the problem by cutting back on the flow rate in the piping.
That was exactly what I was experiencing on the first system that I installed and I have seen it numerous times since then. I went out on a job last week to see a guy about something else. He had a brand new house and told me he liked the house but was having trouble with his hot water circulation system. The water was cooling off too much when he ran the pump.
He described it as "running out of hot water". I shut one of the ball valves down to about half open and its working fine now.
:-)
I should have stated "rare to me".
I have seen it as a problem, but with todays tract houses & Mcmansions, bigger pumps usually means more money. I see more get it in cheap as possible situations.“Well, the election campaign in the country is picking up speed... All the candidates are talking about health care now. Don’t they realize that it’s their campaign speeches that makes us sick?” —Bob Hope
I think you've been working on too many of those 50 story high rises! ;)
i had the same exact problem with the same exact solution, and a neighbor who was a city of san francisco water treatment engineer (expert on hydraulics) suggested the cure for my problem.
"He described it as "running out of hot water". I shut one of the ball valves down to about half open and its working fine now."Maybe it was.I wonder if these loop are big enough and so much water is being circuilated that it is basically a radiator and 1/2 the heat is being lost to the radiator.One way to test these vs the suggestion that the velocity is keeping the tank water mixed up.Change the loop to about a 12" loop, 6 out and 6 back, but with the same pump. Vs a 120 ft loop it has hs now.If it is strickly the velocity then you would have the same problem.If it is caused by heat lose then you won't and also the untility cost would go way down.BTW, something to help this and also to save power for the pump is to install an aquastat on the return line.When the return line is hot then the pump shuts off unitl the water in the return line starts to cool and then it turns back on..
.
A-holes. Hey every group has to have one. And I have been elected to be the one. I should make that my tagline.
Bill, That is the problem with the way some of these systems are set up...it does act like a big radiator.
I should have been more clear when I talked about shutting down the ball valve. I meant that I shut down one of the ball valves on the return line, not on the supply line. That simply slows the flow of return water through the system and doesn't affect the supply of water to the faucets. That is often all that it takes to correct the problem he is describing. If you are pumping too much water around that system, you will definitely have trouble with cold return water being shot up into the upper part of the water heater. Keep in mind that while you are running hot water from the faucet, you are also getting cold water dumped in to the bottom of your heater, right next to where you are pumping.
This is one of the benefits of returning your circulated water through the cold water line. It will temper the inlet water by mixing it with the warmer circulated water.
As for your idea of the aquastat....If I have to use a pump, I use a pump with a built in aquastat and locate it past the last fixture on the run. This is the best way to use a pump. If the water temperature drops, the pump will come on.
I still favor recirculating into the tank drain when possible. Since the cold water line enters the tank bottom via the dip tube, the "cold end" of the recirc loop effectively tempers the incoming cold water anyway as both enter the bottom of the tank and mix there.Plus, it puts the end of the recirc line at or near the lowest point of the system, making it easier to drain the recirc line when needed. Air pockets in the pump and check valve might be avoided, too.Perhaps a disadvantage (or an opportunity if you see it that way) is the need to remove the crappy plastic drain valve from the tank and replace it with a brass ball valve.To save some space, time, and just because, I used "The Isolator" manufactured by webstonevalves dot com. See pic. Versions with different fittings available.Has anyone mentioned that too high of circulation volume increase water velocity in the (typically 1/2") lines resulting in pipe erosion and pin-hole leaks? This is a pretty well documented concern. Most of the systems I've seen have no aquastat, timer, restriction valves, or thermostatic valve to avoid unnecessary operation.Was use of type L (thick wall) copper pipe for the recirculation lines suggested yet?
pic of recirculating pump isolator valve
I would like to add that the design of the recirc loop is critical as far as energy used. I recently saw and example in a seminar where an uninsulated loop which ran 24 hours/day consumed more hot water than the actual water usage!
Gary Klein from the California Energy Commission has done some great work in this area. My guess is that you can google his name and get some of his papers to read.
Bill