Hot water backflowing into cold supply

pjb999November 14, 2006

OK, I'm rephrasing my question, as what I'd asked before wasn't very clear.

The hot water tank in my place - I moved into the house in May- is around 4 years old.....I'm not 100% sure if this problem's been there all along, I suspect it's new but I'm at a loss to explain it.

When there's a high demand for cold water, hot water is flowing back up the cold supply (cold in, and hot out are both on top of HW tank) and mixing with the cold supply. I first noticed the longer you ran a cold tap, the warmer the water got....and verified what was happening by the cold inlet pipe on the tank becoming hot, very hot, and the heat travels right up to the T with the cold pipe.

The humidifier supply tap, one of those horrible pipe-piercing screw on types (clamp the thing round the pipe, screw the 'tap' shut and it pierces the pipe, and supposedly seals itself....) had been leaking and I replaced it - in hindsight I should really replace the tap with a proper one, it's a horrible system....

Apart from anything, how can hot water backflow if there's no air coming into the system? I wonder if the humidifier tap's allowing air in, after all, the old one was leaking somewhat....or is it too close to the HW tank? I certainly wouldn't have put it where they did.

Here in BC I think we're supposed to have some sort of vacuum breaker thing on the hw tank supply, and there is a brass gadget- a plug arrangement with a slot in it with a metal pin visible inside (I know it's not the pressure relief valve with the lever for testing/clearing, it's on the side) made by Watts Canada, it's a "No. 36A"

Could this thing be failing and letting TOO much air in? Can't say I've noticed a leak around it, or hissing or anything at all. It might be as new as the HW tank, it could be older. I can't see it on the Watts site, but since it's the only thing between the hw tank and the cold supply, I assumed it may be responsible.

So, I wonder, is it a design/configuration issue with the plumbing? Should I just put a backflow preventer on it? When I moved from Australia recently I had to sell all my pipe soldering kit, so I'll have to buy new stuff, otherwise I would have just redone this setup - it's messy, with a few joins etc from fitting the new hw tank I guess.

Any suggestions? I really don't want this backflow situation, it wastes gas, and water from hot water tank's not really good as drinking water - too much muck and apparently copper ions too.

Thank you for reporting this comment. Undo

It reads a lot like the original thread, I am just as lost. But I can speak to this new question: "Apart from anything, how can hot water backflow if there's no air coming into the system?"

I wouldnt think air matters any more than anything else you have described including the funky gadget, which I think you will have to identify before getting a response. So that leaves us with, " how can hot water backflow  ?"

And the answer is: the pressure is higher on the outlet than on the inlet. First thing, examine the hook-up. Is the cold water actually running to the COLD connection? LetÂs say, it is.

So, now you turn ON the COLD tap. Are you sure it is the COLD? Is the faucet plumbed backwards? LetÂs say, it is normal. Water flows and behaves as it does in most homes, with HOT water; it gets warmer as you draw it.

You didnÂt really say this, but I am assuming that the HOT water behaves as expected. So, all other things being normal, I would say that you have a condition that gives you two HOT supplies instead of H&C. This indicates that there is a cross-over. (People with a water softener or other inline conditioner can imagine the by-pass conditions.)

In other words, you have a mixing valve somewhere that is allowing water to flow thru it and not out the spigot. Try to isolate such valves or faucets, by turning OFF the line stops, where available. Start with the laundry. The last attempt should be at the bath or shower.

FWIW, I simply canÂt understand: " a brass gadget  a plug-arrangement with a slot in it with a metal pin visible inside." You also described it: "Just where the cold water pipe enters the hw tank, there's an odd brass valve-type thing with a slot and a pin/plunger type arrangement you can see."

That does not paint a clear picture. But if it were a vacuum breaker  and only in Canada would they want one in a household WH  it would do one of two things: either prevent backflow (which is does not) or leak to the atmosphere (ditto.)

Your problem is a cross-over.


    Bookmark   November 14, 2006 at 7:44AM
Thank you for reporting this comment. Undo

Thanks, Pinoke.

I had not considered the notion of a cross-over as you describe, however, it'd only work if the taps were open at the time. True, the bathroom tap where I've noticed the warm water the most has two taps and one spout, so it's a mixer, but it runs normally 'cold' if it's the only thing on. Flush a couple of toilets or any other cold-only apparatus and then the warm water starts....if you feel the cold water pipe that runs INTO the HW tank, you can feel the warm water moving up the pipe...starts cold then the warmth literally rises til it hits the main cold pipe...this convinces me it's backflow, although your idea is intriguing. I'll have to try all cold-only taps and see what happens.

It strikes me that this ought to happen all the time - the pressure drops in the cold line due to demand, of course the water will want to flow back out of the hot water tank....although if I had a hot tap on as well, I doubt it would happen as there'd be a pressure drop on the hot side, so the water would want to flow in the correct direction, cold water to cold water inlet, hot water out of the hw outlet. How is this normally prevented?

I wonder if part of the problem is the cold water pipe is directly over the hot water tank, and since heat rises, there's a sort of a natural venturi effect as well, as the cold water rushes through the t over the hw tank.

However, the interchange ought to be very slight one would think, as this is in essence creating a vacuum on the hw tank....unless air's getting in. I don't understand the purpose of the vacuum breaker anyway, except that maybe it's allowing it because otherwise, trying to draw hot water back through the cold inlet shouldn't work, since all hot taps are closed, there's no way for the water to be sucked back, it's like a straw with a closed end. Hot water flows out the hot end, normally, because the cold water is there to replenish the tank and push the hot out, when a hot tap's turned on....and if you turn off the cold supply, the hot water will stop flowing.

Maybe the setup's always worked this way, but I'm not so sure....I think it's a new development, and maybe this valve thing is to blame. I'm going to email the company and ask them what the thing is and maybe they can shed some light....I'll post back.

    Bookmark   November 14, 2006 at 4:13PM
Thank you for reporting this comment. Undo

Based upon the model number and physical description of the device in question it is a vacuum breaker however it is a physical impossibility for that device to be causing the problem.

While the vacuum breakers are seldomly required on water heaters in the USA they are commonly required on heating boiler water supply lines. The purpose of the vacuum breaker is to admit air into the line when a NEGATIVE PRESSURE condition exists to prevent the possiblity of syphoning the water out of the boiler or water heater vessel. (If the water were to be syphoned out the vessel could be damaged or even more importantly, as the internal pressure of the vessel drops the boiling temperature of the water decreases proportional to the pressure drop.) The only time this negative pressure can occur is when the house "Water Main" is closed and the water distribution system is being drained.

We must also keep in mind that all pressure in the water distribution system originates solely at the "prime mover" which is either the municipal main water pressure or the home well pump pressure. You then have one single line from the prime mover source to the Tee at the water heater therefore the pressure going into the water heater must by necessity be equal to the pressure going into the cold water line at that point. It would be a physical impossibility to have a positive pressure on the cold water line and a negative pressure on the hot water line.

It is remotely possible that there could be a slightly higher pressure on the hot water side during flow as a result of water standing in the hot water lines to a higher elevation creating a positive downward vertical head pressure equal to .434psi per foot of vertical rise. The solution would be to install a backflow preventer on the cold water supply line between the Tee and the Water Heater.

    Bookmark   November 14, 2006 at 5:07PM
Thank you for reporting this comment. Undo

Wouldn't the pressure on the hot water side be actually LOWER when the hot tap's on? ie it will make cold want to flow into the tank to replace it; You are saying that device I have is a vacuum breaker, ie if there's an excess of suction on the hw tank, eg through a leak, the valve will allow air into the vessel (thus emptying it faster, I would have thought)

Anyway, it's not that issue, it's the negative pressure in the cold line - this would in effect be suction, and could it not be drawing air through the valve? Otherwise, I fail to see how the hot's drawn off when there's no air or water rushing into the tank to allow the backflow....

I agree that it would seem a backflow preventer's the answer, I am sort of curious as to why the question (why it's needed) has arisen....and if, like I suspect, the condition didn't exist when I first moved in, why does it exist now?

Ok looking at the level of the hw pipe and the cold pipe, directly above the tank, the hot is an inch or so higher than the cold pipe. I suppose that's enough to make it want to flow the other way...especially when there's suction on the cold side, when there's heavier demand on it. On the other hand, the highest point in the pipes is probably the kitchen, and of course the hot and cold are more or less at an equal level....although the situation closest to the hw tank is probably the most critical, and it's here that the hot's a bit higher.

Perhaps this is why in Australia, hot water pipes are always 1/2" and cold are always 3/4" - this would make backflow less likely, as there'd always be better cold flow, and I imagine simultaneous demand for cold will always be greater than demand for hot.

This leads me to a semi-related question - I've heard you can reduce the effect of showers going cold when another hot tap is opened, by adding a return line to the hw service from the furthest hot tap, back to the tank, thereby creating a circuit or means for the water to flow both ways. I've also heard of house wiring being done that way, I think in the UK, so each power circuit forms a loop (not in series but active loops back to active etc)

    Bookmark   November 14, 2006 at 7:08PM
Thank you for reporting this comment. Undo

I sent an email to their support people and someone from Watts actually phoned me - now there's service!

He couldn't absolutely say why I'm having this problem and agreed it wouldn't hurt to replace the anti-vacuum valve (and confirmed that's what it is, although from the sound of it, they don't even know much about their old product) but he did suggest that I could install a "check valve" which is essentially a backflow preventer but sort of a lower-grade one, whereas a true backflow preventer is designed to be 100% effective, it's not really what I need - the check valve should do what I want it to, prevent most if not all of this backflow, and keep my cold water uncontaminated by the hot, and perhaps reduce my hw (and thus my gas) consumption.

I'm still interested in theories as to why it's happening, though. As far as I know there's no leaks or other faults in the system.

He did ask me what temperature I'm getting from the HW as this check valve's only rated to 180 deg (hmmm I wonder if he meant farenheit) and noted that excessive temperature in the hw system might cause expansion and therefore a tendency to backflow somewhat - I'll run down and check it but I may have turned the hw up as we were seeming to run out of hw, what with three adults in the house. I might try turning it down in the meantime.

    Bookmark   November 15, 2006 at 6:11PM
Thank you for reporting this comment. Undo

Well, thats a turn of the worm. You have the temperature up and a vacuum breaker. What you really need in this system is probably an expansion tank. It will allow the excess pressure created by expansion in the heater to be absorbed by a head of air.

If you were to put the expansion tank in the COLD line the HOT water would not be inclined to reverse direction, flowing to COLD abnormally. In fact, your supply pressure may be low enough that the effect of expansion in the WH is greater for awhile, allowing its water to serve the open tap.

I believe this device would be sufficient. It would certainly perform a needed service. And if it were to be insufficient, the introduction of a check valve would assure that water only flows forward from COLD to HOT to outlet.

Frankly, I think this is your reward from spending too much time in the Outback. The poor folk there are not only burdened with all the ailments that life dishes out, but they dont have words to describe their problems. Imagine wanting to go into a closet and having to call it a "cupboard". (Another thread)

So, there you are Peanut Butter. Mystery solved. Thankfully, you didnt get a release from the T&P valve to spoil all this sleuthing. Then again, if you do some service, this would be a good time to change that item, too. I suspect its not up to the task.

Gotta go, now. Mwallabies are a'beggin.


    Bookmark   November 15, 2006 at 8:11PM
Thank you for reporting this comment. Undo

LOL - strewth mate...

Your point about the water supply pressure made me think - I actually have a pressure regulator/reducer on the main supply near the shutoff - and the water pressure's not spectacular, not as high as I had in Australia - but similar to what we had in Ontario. I wonder if it's set right.

I suspect I'll end up with a water pressure meter, I've seen them incorporated with check valves etc and I suppose if you have an expansion tank, they might have them too.

An expansion tank's a possibility but I suppose a check valve'll be cheaper - but less stress on the hw heater with an expansion tank....but I'm fairly pressed for space in that room. All good solutions, but my sleuthing mind still wants to know why I need it (ok maybe the temp up and the vacuum valve) and whether it's done this all along, and I'm just the first fussy person...although I'd say the hot water making the plastic tube on the humidifier supply might put anyone off...

    Bookmark   November 16, 2006 at 2:17PM
Thank you for reporting this comment. Undo

I have figured out the problem however it also brings to mind that I made a serious mistake in a previous post which I am suprised that no one has caught.


Up to this point we have all approached the problem in terms of pressure however understanding that all pressure originates from the cold water line it is physically impossible for the pressure in the hot water line to be greater than what is being supplied from the cold water line.

The problem is not pressure, it is volume. Water reaches maximum density at 39degF and when heated or cooled from that temperature it expands. Water is not compressible so as the water expands the additional volume must go somewhere. Under normal circumstances the additional volume is absorbed by backflow to the municipal main however it has now come to light that this installation has a PRV(Pressure Reducing Valve). Pressure reducing valves have an internal backflow preventer therefore the code requires that when a PRV is installed we are required to also install an expansion tank on the water heater cold water supply line to absorb the thermal expansion of the water as it heats.

The last post confirms that there is a PRV but no expansion tank, therefore as the water expands the increase in volume must be relieving either by blow off through the T&P valve or backflow into the cold water line when a cold water faucett is open. This is a very dangerous situation and an expansion tank should be installed immediately. Do not take this lightly as it is very dangerous to operate that water heater under these conditions.

    Bookmark   November 16, 2006 at 3:59PM
Thank you for reporting this comment. Undo

Thanks for this. What's the principle risk of having a check valve and no tank? Just an excess of pressure? I guess the pressure relief valve is purely for emergencies and shouldn't be relied on for everyday use, I can see the expansion tank'll do the job of the pressure relief valve in this instance.

Do I need to remove or relocate the vacuum breaker? Currently it's right at the cold water entry to the hw tank. Cold water inlet runs six inches to the side then straight up over a metre (4 feet) or so to the cold line.

Not sure if I've seen an expansion tank but I suppose it just goes in the line....

As far as I know, an expansion tank isn't an absolute requirement, although I guess it's my choice to deal with the hot water backflow, certainly it produces an intolerable situation with the humidifier, I had a minor flood again last night because the hot water softened the humidifier supply tube enough to leak...

    Bookmark   November 17, 2006 at 12:55PM
Thank you for reporting this comment. Undo

I doubt very seriously if the failure of the humidifier supply tube is a result of heat.

First of all, let us consider the temperature of your incoming cold water supply. The mean average of the earths crust is 55degF (geothermal temperature of the crust). Soil is a rather poor insulator so the surface soil then cools proportional to the ambient air temperature. While the crust is 55degF the actual surface of the soil will freeze for a short depth which we define as the "frost depth". While the frost depth is the deepest point at which freezing commonly occurs it must be understood that from that depth the temperatures increase gradually from freezing temp(32degF) to the natural geothermal temp(55degF). Typically geothermal temp occurs at approximately twice the frost depth.

In order to protect our water lines from freezing the codes require that we bury the line 6" below the frost line thus when the dirt is frozen to the frost depth the average temperate at the water line may be only slightly above freezing (34degF+).

Fresh water reaches a maximum density of per gallon at 39degF. There is a very coplex formula to determine the rate of linear expansion but for the sake of illustration let us use the maximum here. When water is expanded from its maximum density of 39degF to the boiling point it expands by approximately 40:1 thus we could say the physical volume of a 40gal. water heater expands by one gallon. To put this in perspective consider that a 1/2" pipe has a volume of one gallon for each 98 linear feet.

We must now keep in mind that water is not compressible therefore as the volume expands it must go somewhere. Normally the increase in volume is compensated by physically pushing some of the water back to the municipal main or to your well pump pressure tank however if we install a check valve in the line there is no place for the additional volume to go. In this case the excess volume would by necessity be relieved by discharging through the water heater Temperature & Pressure Relief Valve (T&P valve) BUT let us consider another potential problem here. Imagine for the moment that your T&P discharge line discharges out through the wall of your house. The discharge end of the line is then exposed to atmospheric temperature and in a worst case scenario the outside end of that line could freeze shut. At this point there would be no place for the additional volume to be released and it could result in blowing out fittings or valves or rupturing a pipe or the water heater vessel.

Now keep in mind that the T&P valve is a spring loaded valve that is set to release when the pressure reaches 150psi and most likely the plastic line to your humidifier is failing slightly before the internal pressure is sufficient to release the T&P valve. On the other hand, if you did not have that plastic line there is a very strong likelyhood that the increased volume could be relieved by blowing out a compression fitting at a fixture angle stop, a pipe joint might fail or premature failure of Bibb Washers in your faucetts.

The proper solution is to install an expansion tank on the water heater side of the check valve to absorb the increased volume.

Installing an expansion tank is quite simple. You simply install a Tee on the cold water line with a male thread adapter on the Tee stub out, then screw the expansion tank onto the thread adapter.

    Bookmark   November 17, 2006 at 4:26PM
Thank you for reporting this comment. Undo

Thanks - for some reason I assumed the expansion tank would have a connection top and bottom but of course only one is needed as it serves as entry and exit depending on demand. I priced one at Home Depot yesterday, it'll cost me around $60-$70 cdn.

Does the tank need physical support? I assume there'll be some sort of hanging bracket or something, but you're saying it can go anywhere on the cold line? I want to make sure there's no more warm water backing up, so can I put it on the cold branch that supplies the hw heater? If I was to put a check valve into that line, it'd have to go on the supply side of the expansion tank, obviously.....but would it be recommended? Maybe the expansion tank'll obviate any need for it.

I did notice hot water creeping up the supply line when there was no cold demand, so quite clearly it's not just when there's a pressure drop in the cold, but just plain expansion in the hot water, with nowhere to go.

I will replace the t&p valve just for fun, although when I use the test lever on it, it flows freely.

You may well be right about the main issue with the plastic humidifier tubing being pressure, although it definitely softens, I then find the little brass insert is quite loose in it.

I've already seen a better valve for this purpose, all compression fittings - and I think I'll go with copper, although not til I've put the expansion tank in of course.

    Bookmark   November 19, 2006 at 5:38PM
Thank you for reporting this comment. Undo

Hi there,

I know this post almost two years old but thought I could add some info to it. One of my clients had the same problem. Upon investigation, it turns out somebody had installed a 'Y'hose which connected both the hot and cold supply to the hot water connection for the dishwasher. Problem solved by removing the 'Y' connection and just connected only the hot water supply to the dishwasher as it should have been done in the first place.

    Bookmark   September 3, 2008 at 11:48AM
Thank you for reporting this comment. Undo

I've been having the same problem - warm/hot water in my cold taps when I first turn it on, and also lack of hot water; it's kinda cold. So I went down into the basement, and felt the temp. of the various pipes. The cold pipes where warm, so I turned on the cold at my utility sink, and the pipes cooled off. Then I could feel them warming up again - but the heat wasn't coming from my hwt, but from the cold pipe on my shower connection. So what I suspect is that the water mixer on my shower. Hears what I think is happening; hot water is leaking thru the temp. mixer on the shower valve, even when it's off. It's the only explianation I can come up with, but I have no idea as to how these single value hot/cold mixers work, so I don't know HOW it's happening. Anyone care to explain?

    Bookmark   January 20, 2009 at 6:18PM
Sign Up to comment
More Discussions
Water pressue and other plumbing issues (MULTIPLE ISSUES - LONG)
I have a couple plumbing issues in my house. Some background... Approximately...
Rainfresh water filter leaks from top.
I have a Rainfresh water filter. It seems to be leaking...
Plumbing Kohler Ceiling Tub Filler K-922
Anyone have experience installing this laminar flow...
Inspection Hell (or why not to DIY)
Hi everyone, First time poster but long time lurker....
Does reverse osmosis water corrode copper plumbing? The answer ..
The answer is YES. I did a very unscientific study...
People viewed this after searching for:
© 2015 Houzz Inc. Houzz® The new way to design your home™