Help with varying pressure on radiant floor system

infloorradiantheat

Our infloor radiant heating system is about 30 yrs old and has always run at 20 psi. The pressure gauge did not change whether running or not and was the same hot or cold.

Recently we had an expansion tank rust out. That has been replaced with an new RX-30 to avoid a future rust out. The tank was pressurized to 20 psi prior to installation. The system was flushed which found alot of air in the tubing due to it running with a hole in the expansion tank. The floor system was pressurized to 20 psi.

When running the pressure rises to 25 psi and then cold is 20 psi. We supected a faulty or dirty pressure gauge so changed it. No difference. We added a hose connector style pressure gauge to one of the drains and the pressure really is changing by 5 psi when just filled and 10 psi after it has been running a week (see below).

Another problem is the cold pressure drifts down by about 1 psi per day cold. It is always 25 psi hot. We've added water twice when it has gotten down to 15 psi cold. Can it still be air working it's way out of the system?

My husband just isolated the expansion tank, checked the air pressure and it was 22psi with cold water in it but isolated, not drained to zero. He lowered it to 20 psi but that is with water in the system and the tank isolated but not relieved of all pressure. The system heated to about 24 psi from a start of 15 psi cold (this is about a 1 psi difference from before). Could the 24 psi be from an over inflated expansion tank? Everything I read says it should be pressurized prior to installation to the system pressure. It showed 18 psi when we got it and we raised to 20 psi with a bike pump. Maybe we should have left it at 18. The installer said they never adjust it.

We have had this system for a long time and the pressure always stayed rock solid at 20 when cold or hot. Any ideas as to why it is going to 24 now when hot? We don't understand why this expansion tank increases beyond 20 psi when it was pressurized to 20 psi. It does this when the cold pressure is 20 psi and when the cold pressure is 15 psi.

We are also chasing the pressure drop. We have a plastic bag on the pressure relief valve outlet and there is no water leaking there. We have tried closing the air bubbler and it continued to drop when cold but always goes to 25 when hot.

There are no leaks that we can find in the copper plumbing or the manifolds so maybe something in the gypcrete floor but I'm hoping were still purging air from the system. Any thoughts on this would be appreciated.

Jamie Hall

First of all, it is more or less normal for the pressure to rise when the system heats up. How much it will rise depends on the size of the system relative to the size of the tank -- but a rise is normal.

If it wasn't rising before you changed the tank, that may be been that air in the system was aiding, as it were, to hold a more constant pressure.

However, I have to ask: is there a pressure reducing valve connection between the heating system and, if so, what is it set for?

Also, a pressure drop when sitting is not normal. Somewhere you have a leak. That really does need to be tracked down.

infloorradiantheat

The pressure at the fill valve is at 20 psi. However we usually fill and then turn off the fill valve so as not to flood the house if there should be a break in a pipe. Our system has always functioned properly with the expansion tank holding things at 20 whether on or off so that is why we are confused by the 5 psi (from 20 psi) -10 psi (from 15 psi) rise when the system heats up.

We are actively trying to track down the reason for the pressure drop. I agree not normal but we've never had a leak in the system that caused air to be distributed through out the hoses so originally we attributed it to air in the lines which we did find when we flushed them.

Alan (California Radiant) Forbes

This happens often on my jobs and I have no explanation for it. I think hydronic systems have a life and don't like to be messed with. If you do mess with them, they get upset and their (blood) pressure goes up and down.
I think the pressure changes you are seeing are mild, i.e. they are not exceeding 30 psi. Give it a week or two and it should calm down.
If they persist, check the x-tank pressure with the system pressure at zero.

infloorradiantheat

Thanks. It has been 2 weeks since we last filled to 20 psi. We've been letting it drift lower to see if it ever stabilizes. Do you continue to fill when cold back to the design pressure of 20 psi? Or do you just leave it be?

Given that it is going up to 25 psi when hot do you think the tank is over or under pressurized? I've been trying to work out the logic and can't figure it out.

Jamie Hall

On the expansion tank pressure. The expansion tank needs to be pressurised with air to the desired system cold pressure -- in your case, 20 psi -- when the connection to the system side is open to the atmosphere. The tank is drained, and the drain left open. This gives the maximum volume in the tank to absorb the change in volume of the water as the system heats, and thus to keep the pressure change to as small a value as possible.

Then the tank is reconnected to the system and the system brought up to cold pressure.

Now. If this is done correctly, that does mean that if the system fluid volume becomes less -- that is, if there is a leak -- the expansion tank can't stabilize the pressure, and the pressure can drop quite a bit, even with a very small leak.

If the tank was pressurised properly in the first place, but the pressure drops, water should be added to the system to bring it back to cold pressure -- with the system cold, of course.

hot_rod

How low does the pressure drop if you leave it off for several days? Does it eventually drop to 0?
Maybe with heating season upon us you cannot do that test.
Some contractors charge the tank 2 psi less than fill pressure. This allows some water into the tank. If air does purge out for a coupled days/ weeks you have a bit of a safety seal.

infloorradiantheat

Jamie,
We did pressurize the tank prior to install. Pumped it up with a bike pump to 20 psi and then used a digital tire pressure gauge to reduce it a smidge until it was exactly 20 psi. It was only just today that my husband decided to take 2 psi off it to see if that changed the maximum pressure the system was getting to. We heated it to 140 F feed (after heat exchanger but before mixing valve) and it was 24 psi instead of 25 psi so it may have helped a bit. Tomorrow morning we will check it again to see if there is a change in the maximum pressure as the system temp. should be higher when it is running longer to heat the house.

Another worry I have is that we developed a leak in the tubing. In Sept. when this whole thing started we had the heat on one night and then it was really hot the next day. We turned the air conditioning on but forgot to turn the heat off so one of the zones was going for hours trying to heat the space. We have a mixing valve that should have prevented the water from getting too hot but we have never run the system against an air conditioner like that so hopefully the pex did not get too hot.

hot rod,
The pressure has not dropped to zero. We refilled it 2 weeks ago to 22 psi cold and it got to 15 psi yesterday (cold) and then was looking like 13.5 psi today (cold). If there is a leak it is a slow one.

When my husband took 2 psi off the expansion tank he refilled it but only to 15 psi. I wanted to go to 20 psi but he wanted to change just one variable at a time.

One theory is that as the weather has gotten colder the cold temp of the system is also lower and the pressure is dropping because of that. Another theory is that there was alot of dissolved oxygen in the fill water and it is coming out with the accumulator but that doesn't explain the pressure drop when we had the vent closed. Whereas the first theory does.

The process has been going on for months because after the tank rusted out and was replaced we had pressure dropping. But after doing some research we figured it was due to all the air in the lines so we did a system flush. We both watched the Caleffi video on how to properly purge and fill the system. Thank you so much for that video!!! We both watched it several times to make sure we were doing things properly

I don't think this had ever been done properly when we had our boiler replaced. We did get alot of air out of the lines. We also had several instances of dirty water that some how was hanging out in the lines which makes me wonder if some of the loops were blocked. Most of the water was blue which we think was from the Fernox chemicals.

We have a Caleffi fill valve and did not know what the grey dial on the bottom was so we learned that also.

In researching this we learned that the Amtrol tank the contractor used was likely to rust out again because we have non-oxygen barrier tubing. So we bought the one for non-oxygen barrier tubing.

Then when the system heated up we found at 160F (pre mixer temp) there was a leak at the pipe connection just before the expansion tank. We had the contractor out who replaced that section of pipe and all seems dry now. That was about 3 weeks ago. First week lost 5 psi. Refilled to 22 psi. Two weeks later down to 13 psi. Seems to be about 1 psi per day loss.

hot_rod

yes it sounds like it could have a pin hole leak if it continually need more water added. It can take weeks for a system with a lot of radiant tube to get 100% air free. Air comes out of solution every time you raise the temperature. It takes a bit for those tiny bubbles to fill the air separator and vent out. If it has been 3 weeks to a month, time for plan B.

I'd not worry about the pressure increase. If you knew the systems volume you could calculate the expected system pressure increase. Amtrol has an online calculator for that answer.
If you do have a high temperature operating boiler, say a cast, and you mix down form the radiant the math is a bit fuzzier. In reality a low temperature radiant system does not need as much expansion capacity as one would think

Troubleshooting an assumed pinhole can be a bear. You end up isolating manifolds then individual loops. To do that you need isolation valves everywhere and they need to be 100% shutoff. That rarely happens.
Let the system cool then buy or borrow an infrared camera, fire it up, that can be the best way to find an in-slab leak, even a pinhole will almost immediately show up on the camera.

You will discover a lot other things with an IR camera, like where all your infiltration is, bad window and door seals, settled insulation, the temperature of your pets, etc.

infloorradiantheat

Thanks. I have an IR camera and was using it today to check the floor from underneath. I think the heat had been on too long so there were lots of "puddles of heat". I was able to verify that the tubing going into our shower comes through the shower door and not under the wall. We had the shower replaced this summer and I was worried that the work on the framing might have involved a long screw that penetrated a pipe. But using the camera found no tubing went under any wall that was framed.

I will have to do it again when the system just heats up. Checking from above is also hard because the tubes run under walls, under the shower base, under the tub etc. But I don't have much choice until the system stabalizes so we will continue to search. Also the tubes seem to run very close together in some areas so that makes it challenging as they show up as hot spots. We saw this when we first got the camera years ago. At that time the system was rock solid on pressure so no leaks.

Zman

What type of tubing is it?
How tall is the house above the expansion tank? 20 PSI is a little high for a typical home.
Adjusting the pressure in an expansion tank which is not open to the atmosphere does not work. The tank must be completely isolated from the system and open to atmosphere.

infloorradiantheat

Zman,
Our markings are Infloor heating system tubing are:
3/8" (cts-00) 100psi @ 180 F ATSMD-3309 NSF-RFH PB2110 SDR-11.

We have it at 20 psi because the original design used that and it worked well. We are at altitude in Denver Co. Our house is a 3 story home with radiant heat on all three stories. The taco circulator on our sidearm has a note that to prevent cavitation you have to run it at a higher pressure when it is at altitude so they may have taken that into account.

Another question is what temperature is the pex tubing ok to handle? The design documents show 130 F as the design temperature. But our mixing valve sends 145F when the smallest loop is on that has the most insulation on the tubing. This is our upstairs which is only 600 sqr ft and has a medium plush carpeting. At this point the temp across the heat exchanger is about 180F. The boiler high point is set to 190F.

This was the adjustment on the mixing valve prior to flushing the lines and heat exchanger to try and get the oxygen out. I think the rusting out tank and the junk that was still in the lines had clogged up the heat exchanger as I had not seen the temp across it get to 180F prior to this. I just lowered the boiler high point to 185F.

Before we had the boiler replaced we always ran the old one at 180F but when we had it replaced in 2015 the house was really slow to heat up. We blamed it on the heat exchanger and changed the high set point to 190F and all was almost back to being as good as before. There is another problem with how they piped it that has the master bedroom not getting much water flow until the west and upper zones are satisfied but we have just been living with that.

The old system was a DIY by the original home owner with some help from a DIY shop and frankly it was so well designed compared to the professionally installed system we replaced it with when the boiler started to rust out. Turns out with the non-oxygen barrier piping we have to use alot of higher priced components and a heat exchange for the cast iron boiler so that is what we did in the redesign.

Zman

Hmm,
It's sounds like you have 2 expansion tanks with 2 closed systems.
I assume the one you are referring to is the boiler loop?
What model boiler?
What Taco circ has that note?
Raising the boiler temp could be the cause of the pressure change.
I seriously doubt you need more than 12 psi on the boiler side and 15 psi on the radiant side.

infloorradiantheat

Zman,
Forgot to mention the pressure adjustment. We did adjust it open to the atmosphere prior to installation so we know it was at 20 psi. The differential that my husband did was to take it down 2 psi. We do have a way to drain pressure off the top of the expansion tank without draining the system but there would still be some small pressure from the water in the tank, just no system pressure. Otherwise we have to disconnect it which probably doesn't buy us anything since we did do the initial adjustment prior to installation and with it reading 22 isolated and cold meant that the tank was probably holding pressure.

This morning the system was running at 180F at the tank and the pressure didn't go above the 24-25psi mark. I think I did hear a couple of bubbles come out the air bubbler so it is possible with our mild temperatures that we haven't heated the water enough to get all the air out. This is the first time I have ever seen it at 180F across the heat exchanger. Due to seeing it so high I lowered the boiler high point from 190F to 185F.

We have a triangle tube TTP3-20 flat plate heat exchanger on the floors. I looked back in my notes and we have had problems getting 140F across the heat exchanger with all 3 floors going which is why we raised the set point on the boiler. Maybe this flush degunked it enough that we will get better performance? Hard to say. The triangle tube TTP3-20 is also on our snow melt and we can't get much heat across that one either.

Here are my notes from when it was installed:
"We have the same triangle tube TTP3-20 flat plate heat exchanger on the floors to the house. Interestingly enough it also cannot support 140F across it when running all three floors. Makes me wonder if the heat exchangers are damaged or undersized. During the winter I run the boiler at 190F because the triangle tube heat exchanger can't deliver the heat for the floors at 3-4 gpm unless I run the boiler a little hotter. This does not include the load from the basement. If the basement runs than it is putting about 8 gpm across the flat plate heat exchanger. Given that the performance is fine with 1-2 gpm but tanks with 4 gpm I don't think it can support the basement and floor load combined. At this point I am not addressing this issue. "

Zman

Have all the issues been on the boiler loop? It seems like any leak would be visible if it is in the boiler room.
With a heat exchanger setup, you have 2 completely separate closed loops. There should be 2 air separator, exp tanks, fill valves, ect.

infloorradiantheat

Zman,
The boiler loop is holding pressure fine. We actually have three systems. The boiler loop with a side arm water tanks is one. Then a heat exchanger to our floor loops which are an upper zone, east zone, west zone and basement zone. The reason we have a heat exchanger between the floors and the boiler is that we have oxygen permeable tubing on the floors. The pex tubing was installed in 1994.

We have a second heat exchanger to the garage loops and the snow melt. This one has glycol in it.

All three systems have their own expansion tanks, air purge, and fill valve.

The problem is with the weird pressure rise in the floor loop when hot and the ongoing pressure drop when cold in the floor loop. Given that I might have heard air purge this morning when it was at 180F I am somewhat hopeful that it is air purging. So I am on a wait and see on that one. However I don't understand why the expansion tank isn't holding things steady at 20 psi.

Zman

Ok, I get it.
Aside from the operating conditions being different then before, the pressure changes are not cause for alarm. If this was my system, I would set the system pressure to 15#, that change alone will probably stabilize the pressure as a larger tank is required as pressure increases.
The potential leak is another matter. If it does not go away as the air escapes, you may need to troubleshoot with infrared, high pressure air or helium detection. If it is a very small leak that is not presenting anywhere, there is little downside to ignoring it. It's not like the oxygen will be an issue.

infloorradiantheat

Zman,
Sorry for the delay, had to get the Taco instruction sheet. This came with our circulator in 2015. It is labeled for "00" Cartridge Circulators. This circulator feeds our side-arm water tank.

Caution: Installations at higher elevations over 5000 ft must have higher fill pressure of 20 psi minimum to prevent pump cavitation and flashing. Premature failure may result. Adjust expansion tank pressure to equal fill pressure. A larger size expansion tank may be required.

Zman

I have never seen that note before and have seen 100's of Taco "00's" at > 9,000' running at 12-15 psi. I suspect taco reconsidered the label after realizing the implications. The NPSH requirements of "00's" is nowhere near the atmospheric pressure of Denver. It is far more important to position the circ with the shaft horizontal and "pumping away" from the exp tank then it is to run it at a specific pressure.
Out of curiousity, what model is the circ?
Perhaps @SteveThomas of Taco will weigh in...

infloorradiantheat

Here is the link to the pdf on taco's web site.

https://www.tacocomfort.com/documents/FileLibrary/102-054.pdf

hot_rod

The 20 psi and increase is not a problem, seems you are more interested in finding a leak.
A pro might come and put 100 psi on the system to troubleshoot it , especially if it is isolated from the boiler with a HX.

Zman

The Taco note is interesting. I sure appears to be a solution with no problem in site.
The 20 PSI is likely contributing to the pressure spike but in the end, neither the pressure spike or the 20 PSI is an issue at all

infloorradiantheat

Thanks hot rod and zman.

If it continues to fall after one more refill we will try and find someone to trouble shoot it. It sounds like you are recommending we drain the floors and put 100 psi on it. Do you think we can locate a leak in the gypcrete that way?

We did find one copper joint in the feeder piping to the floors with some white powder on it today. Had a plumber look at it and they didn't think that it was a leak. My husband is going to clean the joint and see if we can spot any leaking from it.

Just trying to figure out why the system used to stay at 20 psi and with a new same size expansion tank doesn't.

Zman

You might have success looking for the leak with an inexpensive IR camera. Turn the system on from a cold start and look for the blob. Leaks within a continuous run of pex are pretty rare, my money would be on the slab on grade basement slab. Concrete crews are often abusing to the tubing.

Putting high pressure to the loop and isolating them one at a time would help determine where the leak is. I am not convinced you will be able to hear it without sophisticated equipment.

Helium is an excellent troubleshooting tool which allows you to pinpoint the location.

These guys specialize in both listening and helium testing and would be able to help you out. https://www.americanleakdetection.com/

infloorradiantheat

Thanks zman! I was wondering who to call for this so the recommendation is really appreciated.

Your thought on the basement slab is a good one. We did zone off the basement and the pressure continued to fall. When we zoned off the east zone it stopped falling (this was prior to running heat and prior to the flush so the data isn't good any more as there was a ton of air in the east zone).

Unfortunately for all but the basement manifold we don't have the ability to isolate the loops without changing out the manifold. There is a rubber gasket on the flow control for each loop that has hardened so if we twist the knob it leaks. It stops when we turn it back to it's original position.

We had the basement manifold replaced with a shut off valve on each loop both entering and leaving the loop. At that time we had a pressure loss that stopped when we isolated the basement. When we had the new manifold put in it looked like one loop had no water in it so we isolated that one when we recharged it. There was no pressure loss so we thing we found a leak. Or it could have been just an airlocked loop as we did not know about that at the time. Now it has some clear pex in the transition from the old tubing to the manifold and there appears to be water in there so who knows why. There also appears to be an air bubble on top of the water.

To my knowledge we have never opened those valves. But that doesn't mean someone working on the system didn't do it. I once found my valve from the boiler to the hot water tank turned when we were having the master water valve to the house replaced. No reason anyone should have turned that valve but it had been shut off. Maybe they just shut off all the valves in sight when replacing the main and forgot to put this one back on.

I do have an infared camera. We bought it back when we thought there was a leak in the basement. It is pretty cool to see the loops but I can't really see the leaks. I've been trying with this current one but there are so many blobs of heat once it gets going because of how the tubing goes around in circles and runs close to itself or though doorways that I never found anything. Probably a more experienced eye would be able to see it.

infloorradiantheat

We decided to try and fill just using auto fill. We have a Caleffi auto fill set to 20 psi. Left it on several hours yesterday while the system was running. Shut it off over night and then left it on while the system was running today. As the day heated up the floors shut off and the pressure dropped to 15 with the auto fill set at 20 psi. It has held there for 5 or 6 hours with the auto fill on. Does this make any sense? Shouldn't the auto fill back fill to 20 psi?

Zman

They tend to much up and do that over time. Try tapping on it with a screw driver handle.

hot_rod

might be might be time for some pics or a drawing. Where is the Gaul you are reading located? It should be at the expansion tank and the fill valve there also. That is the point of no pressure change. The circulator should be pumping away from that point. Could be the gauge you are reading is not what the fill valve is seeing? How old is the Autofill? Sounds like it has been working if you add water often?

infloorradiantheat

Thank zman,
It is confusing. We had to turn the boiler off today to work on our fresh water hot water pipe that had developed a pin hole. As a result the whole system cooled down. A leak on the boiler loop at the magnetic removal device developed. This is actually not uncommon. For some reason the turn style connectors of this device seem to need to be tightened whenever the system is worked on and cools all the way down. My husband had waited until he was done sweating the other pipe to tighten it so a little water needed to be added to the boiler. The same fill valve is connected to both systems. So we opened the valve to the boiler and at 20 psi the water shot in. In fact I shut it off super quick since it will easily fill over 20 psi on the boiler. That is why I was so surprised it did not back fill the pressure loss on the floors.

infloorradiantheat

Hot rod,
Here are some pics. The fresh water fill for the floors comes in right above the expansion tank. This wall shows the expansion tank and the connection to the upper, west and east zones. The extra temperature dials are strap on's. I used these to try and get the mixer dialed in and to double check the built in temp. gauges. We had a lot of part failures when the system was installed so had to do alot of trouble shooting.


This picture is around the corner and shows the fill valve (just above the boiler). It fills all three systems (boiler, snowmelt and floors) there are shut off valves and back flow preventers on the floor and snow melt systems as well as the fill valve. Don't know why the boiler doesn't have a second back flow preventor as well.


One reason we are so confused is that prior to the old grey amtrol expansion tank rusting out we did not have any weird pressure rise or the continual falling. We thought maybe the rust it put in the system hung up or damaged the pressure gauge. So we changed that out. There was something like a large rust flake hung up on it. However we replaced it with a brand new one. We also added a Watts hose attachment pressure gauge to one of the drains to double check the new one when the same problem continued.

These wierd pressure changes occured with the brand new grey amtrol tank that the installation company installed and the brand new red amtrol tank that we bought and they installed. We checked the charge pressure of the red one (and upped it to 20 psi) before it was installed.

Zman

One Fill serving both sides is not a good thing. Make sure you always have one closed so the 2 sides don't become one.

infloorradiantheat

Thanks Zman,
We keep all the fill valves closed. Only open one set at a time. There is a backflow preventer on the glycol. Also one on the floors. The third is on the fill valve. Until I took the picture I thought there was one on the boiler system also but there is not.

hot_rod

That is an odd setup, feeding multiple systems from a single Autofill. If any isolation valve are leaking, you get the highest pressure trying to equalize. Although it looks like every connection has a back flow device.
A tight system as you know early should not need much if any additional fill water on a continuous basis.

Does your Autofill have a tapped port to add a pressure gauge? We added that plugged port option a few years back. It would be good to add a top quality gauge at that point. You could read all system pressures on the same, accurate gauge.

infloorradiantheat

hot rod,
We have a knob in that position but it looks to be solid metal, not something we could screw a gauge into.

Normally the autofill is off and we only turn it on when we need to back fill because of a repair. Generally the systems have never lost pressure on their own. In fact something like changing a circulator or pressure gauge doesn't usually even require an adjustment to pressure because so little is lost. The circulators all have shutoff valves on both sides and the pressure gauge can be isolated enough that a quick removal and replacement doesn't involve much water loss.

The auto fill, if left to it's own devices will fill the boiler beyond 20 psi when set to 20 psi. Currently the floor system seems to only fill to 15 psi when set to 20 psi. I remember when we were bringing the system up after it was rebuilt that I turned the pressure up beyond 20 to get to 20 psi but that was because we were impatient. I don't know if it would have stopped at 15 psi on it's own.

Currently we are at 21psi after filling it (by cranking to 27 psi) yesterday. So we will see where we are in a week. Hopefully it stabilizes and turns out to be air in the system.

I wish we had the old expansion tank that rusted out. I would like to be able to test what it was charged to since it kept the system so stable at 20 psi.

In terms of them all equalizing I don't think they would do that as they would each have to get through 2 ball valves and in the case of the snow melt or floors a back flow preventer. They can all handle 20 psi so that wouldn't be a problem.

If the ball valve from the street pressure failed and the autofill failed the street pressure would still have to get through another ball valve. I kind of like having so many ball valves and back flow preventors between the systems and my potable water.

However I didn't realize the boiler didn't have back flow preventer like the floors and the glycol do. So for boiler water to contaminate potable water there are two ball valves and one backflow preventer. Does the caleffi fill valve also have a built in back flow preventer? I don't see an additional drain valve for it so I'm guessing not.

For general operation all ball valves around the autofill are closed and we only open up one pair at a time to fill so putting a pressure gauge on it wouldn't be of benefit for our system.

Jamie Hall

OK. Now I'm badly confused. You have three completely separate systems? No flow in between them, except when valves are open? What transfers heat between these systems? Heat exchangers? Separate boilers?

A more ordinary system would have one fill point, with an autofill valve on it set to -- say -- 20 psi? Then the water would circulate among the three systems controlled by mixing valves or flow valves as needed -- and thus, since it is all connected together, it has one expansion tank...

Can you unconfuse me?

infloorradiantheat

We used to have what you describe. However when we replaced the boiler to make the warrantys valid we needed to isolate the floor and snow melt with separate heat exchangers because we have non-oxygen barrier tubing. So we basically have three systems. The boiler and side arm water tank together. Then a heat exchanger to the indoor floor system. There is another heat exchanger to the garage and the snow melt that has glycol.

hot_rod

The Autofill should maintain exactly what you set it at. Did the clear plastic top ever pop off by chance? If so it may not have been put back in the same position and pressure and setting may not match the actuals.

Also it is a fast fill valve all the time, no need to set the pressure higher to fill faster, it fills at 5.2 gpm as long as you have 30 psi or more feeding it.

If it overshoots the desired fill pressure then stops, I suspect it was assembled a bit off, or in the field it was tinkered with.

It it continues to fill above the setting and never stops, it probably has some crud stuck in the valve mechanism. Very easy to disassemble and clean or replace entire cartridge.

Although I would replace it with one that has the gauge option.

Notice also the grey knob on the bottom is a shutoff valve, needs to be wide open CCW to fill at 5 gpm. The valve is also handy for troubleshooting. Fill to desired pressure, turn off and observe for pressure drop. Although sounds like you have done that.

As for pressure, .433 psi to lift 1'. so if its drops to the same number always, say 7 psi, never lower, without heating or changing the fluid temperature, look for the leak
at that level above the fill point, about 16' (.433X16= 6.92 psi).

Any manifolds located in areas on the upper floors? That is a good place to find slow seeping leaks sometimes, at the fittings or air vents.

You can buy small accurate 1/2 water meters online, be interesting to know exactly how much water is getting added. Too much fresh water will start scaling the boiler and all the components, quickly!

infloorradiantheat

hot rod,

Lots of good info.

I did not see the plastic top pop off during installation. I don't think the installers fiddled with the interior components and I was present for the installation.

I did not know about the shut off valve at the bottom until we watched your video. We did open that all the way yesterday or the day before. If having it partially closed (about one full turn) affected the fill pressure we did not know it. It is now full open. I don't think the installers knew what that knob did either. We had never touched it as I didn't even know it was a knob. So it was left in the position the installers left it in. If it were designed more like a ball valve it might be more obvious. Or if it had a label.

The only think I can say about it's fill capability is that leaving it set at the same number, 20 psi, will over fill the boiler instantly beyond 20 psi and will underfill the floors even when left for 6 hours. The boiler is physically below the fill valve and the floors are above.

We have never let the system get much lower than 15 psi and then have refilled. We were worried about damage to the system running it at a low pressure and worried about pulling oxygen into the system again which is what we think happened when the expansion tank rusted out. Flushing took a full day and alot of water. Do not want to do that again.

We have one manifold on our upstairs and two on our main floor. We have inspected them all repeatedly for leaks and they are bone dry. They are very old and we cannot turn the flow control knobs without causing a leak because the rubber gasket has hardened into a position.

I've been working hard to try and figure this out before it gets too cold. When it is milder we are able to shut the ball valves for a whole zone and watch the pressure. But some times it takes a few days to see any drop as the gauge is not very fine. So not sure I can do the whole zone isolation for a few days to a week in the winter.

Eek don't say that about the boiler. We seldom have to add water but I don't even want to think about another boiler replacement. We have only done one big flush of the boiler which was to back flush the heat exchanger in an effort to get the snow melt to get better heat transfer. I don't remember if we back flushed the floor heat exchanger at the same time. We have replaced the circulator to the snow melt heat exchanger a couple of times to try and get more flow through it. So far not impressed with the triangle tube heat exchangers.

hot_rod

Guess I'm confused by this statement?

The only think I can say about it's fill capability is that leaving it set at the same number, 20 psi, will over fill the boiler instantly beyond 20 psi and will underfill the floors even when left for 6 hours. The boiler is physically below the fill valve and the floors are above.

The Autofill is just a pressure reducing valve. If you had a gauge at or on it it should read close to what the number on top show. Anything below that point will see a bit higher pressure, any thing above would see a lower pressure.

If you get to a number, say 20, shut the grey knob off, which does have a label on it, but hard to see. Now the only way pressure increases is with heat or pump delta being applied. The only way it decreases is a leak, air venting out, or temperature decrease. Close off all air vents to eliminate air leaving.

Say the boiler pressure gauge is 6 feet below the fill point 6 X.433 = 2.6 so that gauge would read 22.6. Anything above the fill point gauge would read a lower pressure, that elevation thingy again That is why an accurate gauge at one point would be the best to document pressure changes. Gauges do lie at times, especially some of those dual boiler gauges.

Circulators running or not can also show a pressure increase, if gauged on the discharge side you will see that., as will any change in temperature.

Of course the reverse is true, as the water cools, expect pressure to decrease. With circulators off expect a decrease, depending on where the gauges are located in relation the pumps and expansion tanks.

With heating season upon us, it will be tougher to do your tests at one temperature, no boiler running.

I'd be tempted to get a meter like this on it, and get a handle on how much water the system is actually losing. Watch it for a month to be sure the system is completely air free. As any air leaves the system pressure will drop, or fill water will enter if the valve is open.

Could a heat exchanger be leaking across, do you have the same pressure on both sides of the HXers?

You just don't have enough accurate data to start panicking, IMO.

Jamie Hall

OK. I'm still slightly confused. Your three systems are completely hydraulically separate, with heat transferred to them by heat exchangers? If that is so, a pressure change in one of them will have no effect on the others at all. However, all three need expansion tanks, sized to the system serviced by the tank and set to an appropriate pressure for that system. They also need separate pressure gauges and separate fills. Is that what you have?

infloorradiantheat

Thanks hot rod,
We have the auto fill valved off all the time. Yesterday was the first time in 20 years we have run the system with the auto fill valve open. We were trying to see if it would fill back to the auto value but it stayed at 15 psi so it was a failed experiment.

The fill valve is isolated from the floors by a backflow preventer so a pressure gauge at the fill valve location would not show the pressure of the floors.

I appreciate the explanation of how it works. The boiler pressure gauge is about 1.5 ft lower than the fill valve so that may be what is happening for the boiler. The floors pressure gauge is about 3 ft higher than the fill valve. Maybe in that instance the height of the expansion tank is affecting it. Good to know the reason why I have to adjust it when filling.

I know the heat exchanger is not leaking because there is no corresponding pressure drop on the boiler loop when it drops in the floors. Not even when cold. We have two pressure gauges on the boiler loop. One on the boiler, which is down low and really hard to read, and one I had them place at eye level so I could give it a quick glance when walking buy.

The eye level one on the boiler does go up and down with temp. and flow from the circulator and has since it was installed. I didn't worry about it much. The reason the same behavior in the floors caught my attention is that the floors had never done that before. They were always rock solid.

Which brings me back to the expansion tank. The floor one likely was not adjusted on install because our installers didn't believe in doing that. The boiler one was undersized at install and had to be replaced. We may have asked them to adjust that pressure to be sure we had a working system. If so it is at 20 psi, just like our new floor expansion tank. So I guess it makes sense that they both do the movement with temp. That really makes me wish I could check the old floor one and see what it was set to since it held the pressure completely steady at any temp. and and flow.

I am reassured to hear you say that it can take 3-4 weeks to get the oxygen out. We will keep watching it. In the mean time I've been gazing at the basement ceiling and other ceilings looking for leaks. So far no sign. Also have closely inspected the manifolds and none there either. Although I did find a pinhole leak in our hot water recirculation line that we fixed yesterday. Also found some white crustiness in some places that we are going to keep an eye on.

We did try a period with the air vent closed but probably messed up the experiment when we would open it to check if a bubble would come out.

I have a clear ziplock back on the pressure releaf valve outputs and so far nothing is leaking out due to an over pressurization.

I will report back in a week if we have the same type of pressure drop (about 1 psi/day). It will be a big party here if we don't as we are both tired of trying to figure out this issue.