Hello,
We have had on and off problems with our boiler/radiant floor heating for a few years now. Having 3 or 4 different plumbers out, we had the circulating pumps replaced last year and the 2 zone valves and a fuse replaced this past March. Everything seemed to be working well after the last visit. However, now that winter is back our floor is not heating up again. The pressure gauge reads about 16-17psi and the appropriate pipes are hot so I thought it was possibly an issue of airlock in the pipes. However, when I looked at the thermostat on the boiler tonight, it reads 190F. The return thermostat reads 60F. 190F sounds a bit high to me. I know there have been a few threads on bleeding/purging the system, but how do I troubleshoot if it's air or if its a boiler issue (or something else)?

Thank you kindly,
David

 

Hi David,

With radiant floor heating there should be a 'mixing valve' that tempers and limits the water temperature sent to the tubing in the floor.

The boiler will (and should) still operate at a higher temperature, up to it's limit setting (of which 190 is not unreasonable), and the water sent to the floor is 'mixed down' with some of the return water to provide a lower temperature to the floor.

Can you take some pictures of the installation? We'll take a look and see if there's something the techs are missing.

 

Thanks NJ Trooper. The layout of my system is a bit of a mess hidden behind the hot water tank in a small boiler room, so photos of the complete system are difficult. If you'd like to see more detail on anything, please let me know and I'll try get some better shots.

 

David, I have no idea what's going on there... that is a very bizarre mash-up of piping and parts!

I find that water supply line into the system cantilevered like that especially attractive.

May I ask who built this system? i.e. not you, or a relative, or a friend? I want to make sure I don't personally offend you before I continue.

 

You won't offend me; we essentially bought the place like that (and any changes were done by licensed plumbers). It makes me dizzy anytime I go in there.

 

OK, good... in that case I don't feel bad about saying that it appears to be a total clustermess. I've seen worse... but not often!

One of the first things I would suggest is to get some kind of support on that water feed line. It looks like it's just 'propped up' in place. Even some stout wires from the ceiling down and around the pipes. You don't want that thing falling and breaking the water line...

How exactly is that connected to the boiler piping? I can't tell if there's a tee fitting there, or what?

At this point I can't suggest much more because I can't make a 'mental piping diagram' out of the pictures.

Where is the boiler itself in relation to this? Can you tell me what make/model the boiler is?

If the boiler water temperature is consistently up around 190, I would suggest that you save a few bucks on fuel and turn that down.

Can you take more pics that show a wider angle? or is the room too small for that?

I always suggest that the expansion tank be checked for proper air charge because it seems that nobody ever does this. Not even so-called 'good' techs, and it is the cause of MOST relief valve openings.

See what you can do for more pics and I'll look and try to figure something out.

 

This is in our rental property, so I'll grab some more photos and post them tomorrow. The boiler is just to the left of the hot water tank (at the bottom of that silver ducting seen in a couple of photos). It is a Super Hot Mini-Gas boiler.

I'll also secure the expansion tank; it is just held in place by that one tee fitting.

The boiler isn't always at 190F, but probably about 160-170 (we cranked the heat to test if it made a difference).

Thanks

 

I've attached a few more photos for you NJ Trooper. I did try to check the PSI of the expansion tank with my bike pump, but it's either really low or my pump just doesn't read below 10PSI very well. I'll try with another gauge tomorrow.

In the first photo, A1 and A2 are hot; B1 and B2 are cold to the touch.
Attachment 21683Attachment 21685

(sorry about the sideways picture; it doesn't want to flip for me).

 

Note: To calibrate the pressure tank, first you need to (turn off the furnace) and depressurize the system, usually done by draining out some water. You need to read a gauge that is not on the pressure tank although the gauge can be on the pipe coming out of the pressure tank.

Next, set the pressure tank (add or bleed air) to about 2 PSI under the desired system working pressure according to a gauge on your air pump. The system gauge should remain near zero during this process; drain additional water to achieve this.

Finally, add water and turn the system back on. You will need to bleed the system and add small amounts of water over the next few days to maintain the desired system pressure.

Are there any (manual) bleed valves at first floor level? One or more may be hidden under hatches in closets for example.

Are the supply tubes or pipes hot where they go into the floor slab and are the return tubes cool where they emerge from the floor slab?

 

Allan, I don't have another pressure gauge in the system asides from the one on the boiler. If I don't connect to the schrader valve on the expansion tank, how would I connect a gauge to a pipe off the expansion tank?

I have never seen any bleed valves on the floor (its a bungalow with no basement).

The supply tubes were both cold. Just tonight I closed a valve (the valve on the vertical pipe running between A1 and A2 on my previous photo) and the temperature gauge instantly started to go up; the supply tubes also started to heat up a bit. I had no rationale behind closing the valve other than "what happens if I close this one".

Thanks

 

You need a gauge on your pump or at the schrader valve because the boiler and piping should be depressurized (close to 0 PSI) while you are calibrating the pressure tank. The bladder in the tank keeps the air cushion separate from the portion of the tank the water occupies as needed to stabilize the pressure.

Inability to pressurize the tank while the rest of the system is depressurized means the bladder has ruptured or failed necessitating tank replacement. While there are a few obscure circumstances where a pressure tank with a ruptured bladder will still work, it will never work unless the connection with the rest of the plumbing is underneath it.

Water was short circuiting from the supply side to the return side without taking the "long" route through the floor slab radiant heating pipes/tubes.

By closing that valve you stopped the short circuiting.

I don't have enough 3D mental visualization power to figure out from the pictures you have given us so far exactly what the water path should be which in turn will lead to the proper valves to be closed or open for proper operation.

Vocabulary: Manifold -- A section of piping typically just outside the furnace or boiler, perhaps constructed with a number of T fittings joined together, where several tubes or pipes to/from the rest of the house branch off or come together.

 

I think 'calibrate' is a rather strong word to use in relation to an expansion tank. It connotes that it's some kind of precision device. It's not. Let's just say 'charge' the EXPANSION tank, shall we?

ALL expansion tanks LOSE about 2 PSI per year, so why would one start with a deficit of 2 PSI? "...working pressure..." ? What is that? If you said "Cold Fill Pressure", I wouldn't have to ask.

Set the air pressure to the SAME PRESSURE as the COLD FILL PRESSURE.

Tiger, read this, there are much easier step by step instructions to charge an expansion tank:

http://www.doityourself.com/forum/bo...sion-tank.html

As for the pressure gauge, read this:

http://www.doityourself.com/forum/bo...ure-gauge.html

I'm going to study your new pictures and see if I can't make heads or tails out of that piping.

I don't recommend opening and closing valves that you aren't sure what they are doing or for... at least to not leave them in that position...

 

Can I get a closer, better focused, and brighter look at what's going on at the bottom of this photo? Use a lamp or something to get more light down there.

I believe that the pipe that heads down from the left hand pump is coming directly from the hot boiler supply, is that correct?

That round handle valve that 'bridges' between the two pipes above the two pumps... do I see a 'capillary tube' leaving from that valve and heading down to that small white thing that's strapped to the pipe just to the left of the temperature gauge?

 

In this picture, I need to know if the three red tubes toward the wall join up with the other three that are joined on the pipe with the drain valve on it. Is there a tee behind the tank that joins all six together in that vertical pipe that disappears behind the tank?

Are there TWELVE tubes altogether? (6 supply, and 6 return)

Is that vertical pipe behind the tank the one that is coming up and turning left into the brass thing (which is an air removal device) at the upper right in this picture?

 

David, I'm still going to want to verify how I believe that your system is piped, so the following doesn't negate getting more pictures... and answers to any questions I've asked, but I think I have it figured out.

You seem to have intuitively ALMOST done the 'right thing'... but you don't want to close it ALL THE WAY!... more on this later.

Take a look at the valve with the round knob on it to see if there is a make or model number.

Are there any temperature markings on that valve? If so, what temperature (or setting) is it currently at?

 

Here is a diagram of how I think your system is piped. Hopefully you didn't desert us...



This could have been more easily with a 3-way thermostatic mixing valve piped to a pair of closely spaced tees on the boiler loop, but perhaps when this was built the guys didn't know about them?

The way this system is suppposed to work:

When there is a heat call, both pumps run and boiler is fired.

The thermostatic valve is monitoring the temperature in the radiant loop and should be adjusted for around 120°F on the temp gauge.

However...

Without SOME restriction at valve B it is unlikely (and proven by observation) that there will be little exchange of water between the boiler loop and the radiant loop.

So in order to cause hot water to be transported to the radiant loop, valve B must be PARTIALLY CLOSED... NEVER FULLY CLOSED! in order to cause exchange of heated water.

Using a BALL VALVE at B is a BAD choice. Ball valves are VERY difficult to 'throttle' for this application. It should have been a good quality GLOBE VALVE

Valve A should be OPEN at all times. It would be closed for certain servicing procedures.

Bottom line is that valve B and the thermostatic valve need to be 'fiddled with' by trial and error until a good balance is obtained.

Valve B should be closed NO FARTHER THAN NEEDED in order to cause enough hot water to be transferred through the thermostatic valve to maintain the 120F setpoint of the thermostatic valve.

 

If I were adjusting this, I would start with B fully open and the thermostatic valve set to maximum temperature.

With boiler firing and water hot, I would slowly close B until hot water started flowing through the thermostatic valve.

Then adjust the thermostatic valve for 120F on the temperautre gauge.

I would probably then remove the handle from valve B ... being a rental property and all...

 

Thanks for the replies and sorry for being M.I.A. for a couple days. I've been away for work. I will reply back with details tomorrow.

Cheers,
David

 

So after being out of town for work for a couple of days, we booked a last minute trip to Mexico for a wedding last week. As we had a new tenant move in this week, we ended up having to call a plumber.

New thermostatic valve and new expansion tank and all was well...until the weather got cold on Monday. NJ Trooper, I did use your diagram and followed your recommendations for making small adjustments to valve B yesterday, but still couldn't get the temp gauge to read above 85F. The only time the floors got really hot was when we popped off the cap on the old thermostatic valve and pulled the pin(?) out a bit, allowing more water to flow through. I talked to the plumber again today, and now he thinks we may need to replace the pump going into the floor. I'm pretty confused by this system to start with, but if the pipe on the right hand side of the thermostatic valve isn't getting that hot I'm not sure how the pump comes into play. The pump is only 2 years old and the motor is working. It may have been more fruitful (and cheaper) to stick with you guys, but time was a factor and I thought it'd be one and done. Lesson learned.

Anyway, thanks again for all the help.

 

Tell me the make/model of that thermostatic valve (TV).

You started with the TV all the way to the highest temp, correct?

Something you should know about ball valves... I said this before, but it bears repeating... they 5uck! for 'throttling' applications, which is what you are trying to do. They are very 'touchy' to adjust because 90% of the flow change occurs in about 10% of the valve opening. Since the valve goes from full open to full closed in 45° of travel, it means that you have about 5° of travel to 'play with'.

All this means that when properly set, that valve will be ALMOST closed. When set correctly it will be between 8 and 9 PM position.

The reason you don't want it ALL THE WAY closed is because when ( IF ? ) the TV sensor is satisfied, it will begin to close the TV and you need someplace for the water that the left hand pump is pushing to go. You don't want to 'dead head' that pump, or cause it to force water through the TV.

I don't recall if I explained the way this setup is 'supposed' to work, but it should help if you have an understanding of that.

There are two pumped loops. One around the boiler, and one around the system.

Those two horizontal pipes, the one with valve A and the one with the TV form a 'bridge' between the two loops.

The boiler loop is going to be HOT! The system loop, not so much... I would say 120°F is probably right for radiant floor heat.

With valve B fully OPEN, there is no reason, no 'pressure differential' between the two loops, for the water to 'cross the bridge. No matter how far the thermostatic valve is opened, without some reason for it to cross, it won't. This reason comes from partially closing valve B. This will redirect a portion of the HOT water across the bridge to mix with the cooler water in the floor heat loop.

The amount of water crossing is governed by how far open the TV is, and that is governed by the setpoint of the TV.

The amount of HOT water that crosses the bridge will be equal to the amount of COOL water from the returns of the floor loop that will travel back through valve A to the boiler to be reheated.

================

No, there's nothing wrong with the pump, don't spend any more money changing that.

Try again with balancing the B valve. Turn the TV to the hottest setting.

Close valve B down until you get HOT water crossing the bridge. You should see the temperature gauge start to go up.

If that temperature starts to go over 120F, dial the TV down and give it some time to react and keep closing it until it's set at 120 and regulating the temp at that gauge.

If you see the temp start to drop, close B a smidgen and see if you can't find a balance point where the temp gauge stays at 120F.

=================

Though not the best way to achieve what this system is trying to achieve, it probably will work, it's just going to take some fiddling with the valves to get it where it needs to be.

This would be SO MUCH EASIER if it were piped with a THREE WAY mixing valve... if I get a chance, I'll modify the drawing so you can see what I'm talking about.

 

There is one other possible 'solution' to 'fix' this setup.

It would involve some less invasive piping changes than I mentioned with a three way valve.

That would be replacing valve B with a "Differential Bypass Valve".

This bypass valve would AUTOMATICALLY vary it's opening based on how far the TV was open.

The bypass valve and the TV would work together. When the TV opened, the bypass would close. When TV closed, bypass would open.

 

Thanks for the detailed description. I will go by tomorrow and try to play with valve B and the TV again. When I did this yesterday I didn't see much movement, but I'll see how high I can get the gauge to read (thus far it's only hit about 85F). If that doesn't work, I'll look into the Differential Bypass Valve.

The TV is a Danfoss; I'll have to check the model.

I opened up the spigot by the return line last night, just letting about 3-400ml of water out, and the water was a dark brown from all the crud. I doubt it's an issue, but is it possible a very dirty line could impact water flow?

 

One thing I probably should have mentioned... make sure that BOTH zone valves are open when adjusting the valves. You want to be able to maintain the 120F in the loops under worst case conditions... which of course it's not the dead of winter yet...

Did you say that this tubing is buried in a slab? Or staple up under wood floors?

Either way, you will need to give the floor assembly time to warm up. Remember that when first starting up, that slab will be COLD, and the water coming back from the floor will be too...

This is what a 'standard' implementation of a 3 way mixing valve would look like:



If you decide to make changes and use this, make sure that the mixing valve you select can handle the GPM required for both zones.

This would probably be the easiest to adjust, simply dial the temperature knob until you get 120F out of the mixer and done.

 

And this is the differential bypass version:



Thing is though, you should be able to get what you have tweaked enough so that it works... it's not ideal, but it should work.

 

The tubing is under concrete slab. With an overnight low of -30C (-22F) and -39C with windchill (-38F), we're definitely approaching worst case scenario. Thank goodness for the fireplace and space heater!

I'll try get the heat up as suggested now, and look at making changes to the system once it starts to warm up a bit. Thanks.

 

Uhhhhh... you live north of the Arctic Circle? Holy Cow! That's freekin COLD!

I am REALLY surprised that anyone even TRIED to use radiant to heat that place.

Please tell me that you are confident that the perimeter and underside of the slab is heavily insulated with foam board insulation! If it's not, you will be losing a TREMENDOUS amount of heat to the ground. You might not ever be able to get the return temps on that floor up...

 

Based on the heating system and plumbing, I'm not really confident about anything.

I had the TV set to 5 (highest) and valve B was about 7 o'clock this morning. By this evening, the temperature (and heat in the house) hadn't really changed. I have now moved valve B to 8 o'clock. Maybe a bit too much of a move at once, but if it gets too hot (unlikely) I'll dial it back a bit. I understand it'll take time for the floor to heat up, but when fiddling with valve B and the TV, should I see changes in the temperature gauge quite quickly, or would that change also take some time?

 

I don't think so... I probably would have started at 8...

I should think that you would see something on the temp gauge below the pump almost immediately.

Is the pipe hot on both sides of the thermostatic valve?

Still with the TV at the highest, and both zones calling for heat, close B until you feel the TV pipe getting hot.

From 8 to 9 it's going to be very 'touchy', because that's where the most change with the least handle movement occurs.

If you see the temp gauge going over say 125 (I'm going to revise UP in temp based on your conditions) then dial the TV back a bit.

The ultimate goal is to get the TV set at around 125 with B closed enough to pass enough HOT water to the floor heating loops.

It could take HOURS (days even? at your outdoor temps and possibly nil insulation around and under the slab) to get that slab up to temp!

 

There wasn't much change in the gauge after moving the valve to 8. I watched it for about 10 minutes.

The pipe is warm on both sides, but not hot. I'll check it out again in the morning and see if it's increased any.

 

You said earlier that when you CLOSED it completely that you immediately saw the temp start to come up, correct?

Close it some more now... 8:15 ...

I have a feeling you are going to end up with that valve almost all the way closed.

You still have A wide open, correct?

 

Yes, when it was closed all the way that's when I saw the temperature rise and the floor was hot (and also when the pressure release valve looked like Old Faithful). I'll increase it to 8:15 and see what happens.

Valve A is wide open.

 

My tenant reported that the floor felt a bit warmer this morning, but the temperature gauge is still just under 80F. (To confirm, we are talking about the temp gauge on the loop and not the boiler temp gauge, correct?). The pressure gauge on the boiler read 20PSI.

I have closed valve B to about 8:15. There was no immediate change on the temp gauge. I'll check again this afternoon.

Temp gauge


Boiler gauge


Valve position

 

The pipe that the TV is on should be as hot as the ones around the boiler.

Keep squeaking that B valve down until you see that system loop temperature start to go up.

 

The tenant reports that the floors are much warmer and the pipes around the TV are hot. However the temp gauge still hovers around 80F. The boiler pressure has reached max of 21 (that I've seen) and over a 24 hour period the pressure relief valve blew off a couple ounces of water.


Valve B is now about 8:45pm. How high can the pressure safely go without damaging or overheating the boiler? Is it ok if the pressure relief valve regularly blows off a small amount of water?

 

I checked the boiler again this morning - temp is still about the same and pressure only read at 12PSI with valve B set about 8:45pm. However, there was at least 23oz of water released from the pressure relief valve overnight (I keep a container under the pressure relief valve). I'm reluctant to close the valve anymore with that much pressure building up.

 

No, it's not OK.

Pressure relief valve opening is not related to the closing of valve B.

Your pressure problems are a separate issue.

They ARE however VERY marginally related. Here's how:

Your expansion tank needs to be sized such that when ALL OF the water in the system is heated, there will be capacity for the tank to receive that expanded water. If the system loop remains cold, it means that you are not heating ALL of the water...

You had a new tank installed. IF that tank is large enough for the system, AND it was properly charged with air when installed, THEN you should not be having a problem with the pressure relief valve opening no matter the position of valve B ...

 

I've spent the week trying to tweak Valve B to get the temperature to increase, but without success. The highest I've hit on the temp gauge is 85F and heating in the condo is just bearable. If this didn't work as a temporary solution, I'm not sure if changing the valve as you previously mentioned would work. Thoughts?

Thanks again.

 

And what happens when you close it all the way?

Has this system EVER worked properly to your knowledge?

Start again from fully closed, and then just crack the valve open...

 

The system did used to work properly, but we've had issues with it for about 4 years now - heat comes and goes. The last time it got really hot about a month ago...I'm sure that's when I had it closed all the way for a few hours.

My plumber told me the boiler will overheat if it's closed too much, so although I've gone to very near closed this past week, I'm a bit hesitant to get close to fully closed for fear of doing damage to the boiler itself. But I'll try your suggestion this week.

Thanks

 

The boiler itself has a 'high limit' control on it so it can't really overheat.

There could be a problem with the boiler pump itself though since in some cases it would be deprived of flow. Even though the boiler water that flows through it is HOT, it is still the primary 'coolant' for the pump. 'Dead heading' a pump is never good for it... short term, no problem really, but long term operation can damage it.

Thing is though... by closing the B valve, you aren't really dead heading that pump UNLESS the thermostatic valve also closes. Given that there is a sensor downstream that tells the valve to open or close, as long as that sensor does not hit a high enough temperature to actually close the valve, then it won't [close].

Look at the diagram again and follow the water flow.



You will see that even with B closed fully, you will have flow through the TV as long as it is open, and that flow will go through the floor and come back to the boiler through valve A.

With B closed the only problems will occur when/if the TV starts to close.

This is where the 'balancing act' comes into play with valve B.

Let's say that you want 120F water in the floor loop, and that the boiler water is at 180F.

What is supposed to happen is that a PORTION of the boiler water travels across the TV "bridge" and mixes with the cooler return water coming back from the floor tubing in order to satisfy the 120F setting of the TV.

I suspect that MOST of the boiler flow is going to want to cross the bridge, but B must be open enough to allow SOME flow.

If you close B all the way, there is no choice but for ALL the hot boiler water to cross the bridge and in this case you should see the temp gauge go way up... and then the TV should start closing, depending on it's setpoint.

So starting with B closed, and then opening as required might be an easier way to adjust this setup.

I do feel that if you wanted to keep the TV and the system the way it basically is now, swapping out valve B for the differential bypass might be the best way to go. Doing this basically 'automates' the setting of B... once the diff bypass is set correctly, as the TV closes to regulate the floor temp, the diff bypass will open to allow the rest of the pump flow to pass back to the boiler... and vice versa.