Tracing radiant pex leak in slab

JaySyd

Hi all, as an adjunct to this post https://forum.heatinghelp.com/discussion/157302/radiant-pex-leak-in-concrete

We are having an issue where I'm certain we have a leak in the concrete run of some radiant heating

I've isolated it down to this loop which is under our laundry tiles and have some thermal imaging which shows the loops aren't circulating everything - just having some trouble figuring it out what would clearly identify as a leak?

Looking for some pro help from all the good folks. 

Attaching images of the thermal scans and the original floor pre slab that corresponds to it.

Thanks everyone! 

psb75

is there any insulation under the slab?

EdTheHeaterMan

Can you indicate what tubing in the before the pour picture is exposed floor, and whet is covered by cabinets of a tub on the right side of the thermal image?

there are only 4 lines showing in the thermal image but you show 7 lines in the before the pour photo.

Also, there is no insulation under the slab (according the the before pour picture). There is a lot of heat going into the ground there.

STEVEusaPA

Is this problem 6 years old? My sad recommendation is to abandon the floor heat, use the manifold to put in some panel radiators.

hot_rod

Is this room part of another room, one pex loop covering more then one room, so you do not want to abandon it?

If so, not much you can do other then chop it up, repair the leak.

Isolate that loop, pressurize it to 100 psi air pressure. that shows a pinhole leak very quickly. How quickly it drops indicates how bad of a leak. A very small leak, losing pressure slowly over a day or so, you might try one of the leak sealers made for hydronics.

Fernox has a product, that squirts in.

JaySyd

psb75 said:

is there any insulation under the slab?

So, under the slab it's the ceiling of the floor below which has standard insulation in the ceiling void for noise reduction /temp retention. 

JaySyd

EdTheHeaterMan said:

Can you indicate what tubing in the before the pour picture is exposed floor, and whet is covered by cabinets of a tub on the right side of the thermal image? there are only 4 lines showing in the thermal image but you show 7 lines in the before the pour photo. Also, there is no insulation under the slab (according the the before pour picture). There is a lot of heat going into the ground there.

You're correct here - the tubing on the right is covered by cabinets and a washer + dryer - so that would also need to ripped up if that is the route we take - 

JaySyd

STEVEusaPA said:

Is this problem 6 years old? My sad recommendation is to abandon the floor heat, use the manifold to put in some panel radiators.

Issue is that it also supplies some under floor heat to our tenants apartment on the same floor.. But I guess we could just tell them - no more free heat sorry! Hahaha 

JaySyd

hot_rod said:

Is this room part of another room, one pex loop covering more then one room, so you do not want to abandon it? If so, not much you can do other then chop it up, repair the leak. Isolate that loop, pressurize it to 100 psi air pressure. that shows a pinhole leak very quickly. How quickly it drops indicates how bad of a leak. A very small leak, losing pressure slowly over a day or so, you might try one of the leak sealers made for hydronics. Fernox has a product, that squirts in.

Thanks @hot_rod Always wondered if these sealers actually work. Do they work on pex?

As an indication, I isolated that loop, filled it to 20psi and ran the circulator + heat for around 8 hours. The pressure dropped to about 8psi within that time. Is that considered a small leak or a big leak? I'm still shocked I don't see this water going anywhere.

I don't have an air compressor, can I test with water? Should I be testing the pressure drop once the system is at 20psi, loop isolated and no circs+ heat running? Or let it circulate water and heat around? 

hot_rod

I would isolate that loop and yes you could connect a garden hose and gauge. With 60 psi or more you will know quickly if there is a small leak.

If the pressure never drops below 8 psi, I wonder if there is a leak. Is the fill valve turned off?

JaySyd

hot_rod said:

I would isolate that loop and yes you could connect a garden hose and gauge. With 60 psi or more you will know quickly if there is a small leak.

If the pressure never drops below 8 psi, I wonder if there is a leak. Is the fill valve turned off?

@hot_rod the reason it doesnt drop below 8psi is that the boiler shuts off then into safety mode as its set to do so when it reaches 4psi below the 12psi standard system pressure that's set on the unit....

hot_rod

If in fact there is a hole on the tube, it should, it will drop to 0 at some point, regardless if the boiler locks out.

The only reason it would maintain 8 psi is if the autofill valve kicks in.

Why not just isolate the suspect loop and air test it at high pressure? That confirms a leak, and confirms it is that loop.

If so, sounds like you need to remove the washer/ dryer and IR that entire floor carefully. Let that loop cool, hit it with the camera as it first starts up. The leak (hot spot) will jump out on the camera.
If the floor is a shiny covering, cover it with newspaper to get a better look.

JaySyd

Decided to walk around the house with the IR camera for **** and giggles, and find out that this loop also goes up two floors to heat a couple of bathrooms. So the plot thickens!

opened the return of the loop in the basement expecting it to come flooding out of the hose - but its literally a dribble....

Derheatmeister

Why do you think that the leak is in that area ?
It could be just airlocked and have poor to no circulation.
What kind of boiler do you have?...If it is a condensing boiler you could do a simple test by turning the boiler off, removing the condensate trap and checking if water keeps coming out of it..
BTW : 2.31 feet in elevation will give you 1 psi Static pressure ( No circulating pumps running) ..In other words if your system is maintaining approx 8 PSI to 10 PSI Static Pressure without having the autofill operational you could have a leak at approx.18.48 to 23.1 feet above the Gauge..As per some of the stories from Dans Books go you could also do the cat or the floor wiping trick

JaySyd

Derheatmeister said:

Why do you think that the leak is in that area ? It could be just airlocked and have poor to no circulation. What kind of boiler do you have?...If it is a condensing boiler you could do a simple test by turning the boiler off, removing the condensate trap and checking if water keeps coming out of it.. BTW : 2.31 feet in elevation will give you 1 psi Static pressure ( No circulating pumps running) ..In other words if your system is maintaining approx 8 PSI to 10 PSI Static Pressure without having the autofill operational you could have a leak at approx.18.48 to 23.1 feet above the Gauge..As per some of the stories from Dans Books go you could also do the cat or the floor wiping trick

Very good question - and very timely - I thought it was only that area since i understood the loop to only cover that floor - but yesterday uncovered that it covers more floors on the 3rd floor of the house - so the problem could be in a number of places - even potentially behind a wall that feeds the upper floor section of the loop.

Ran the  IR camera up on the 2nd floor too and nothing really "stands out" as a leaky spot that i can identify. 

GGross

@JaySyd

Any way to figure out how long that loop is? Maybe you just have a long loop that is either airlocked or otherwise not flowing?

JaySyd

@GGross but would that really manifest as a gradual drop in pressure after the system is pressurized each time?

ethicalpaul

These pictures are very confusing to me. If there's a blockage or airlock as suggested above, then how could 1/2 of the loop get got?

If the problem is a leak, then how could only that near 1/2 of the loop be getting hot?

How could under the slab be a ceiling when I'm looking at a dirt floor? (Or am I?)

Are the photos lying and what the IR photo is showing is the near part is in a ceiling, but the far part is over dirt?

JaySyd

ethicalpaul said:

These pictures are very confusing to me. If there's a blockage or airlock as suggested above, then how could 1/2 of the loop get got?

If the problem is a leak, then how could only that near 1/2 of the loop be getting hot?

How could under the slab be a ceiling when I'm looking at a dirt floor? (Or am I?)

Are the photos lying and what the IR photo is showing is the near part is in a ceiling, but the far part is over dirt?

Looks like dirt, but its a membrane on top of the plywood subfloor. Same membrane across the whole floor actually then thin slab/thinset for tile.

Your questions are fair ! How can some heat be getting there and some not - I have no idea! and thats why everyone on here is awesome at helping think through solutions.

The pressure drop has to be coming from somewhere - so without a true visual/audible leak, its so hard to track down!

Derheatmeister

You could also have inpropper piping techiques in the Mechanical potion..Can you please post some pix of your Boiler/Mechanical room ? Please also note if any of the relief valves are leaking !

JaySyd

edit:

@Derheatmeister Would improper piping techniques actually cause pressure to be lost when i have closed any other zones and isolated the single problematic zone?

I just replaced the PRV on top of the boiler (rated for 30psi) since the last one was leaking at the top of the NPT connection - post change to a new 30psi PRV, it seems to drip when the pressure hits 20psi - which is another weird issue - as long as i keep it below 20psi, it seems to behave. But now after returning to it after a few hours, i can see about a pint of water in a bucket that catches my PRV water so i can account for it in all this troubleshooting. Could there be something else within the boiler impacting the functionality of the PRV ? or is this new PRV also faulty?

No other leaks in the mech/boiler room that are visible to me

see image of the room here

GGross

I think you will just need to isolate that loop and pressure test just the loop to see what happens. Having a hard time seeing how that fill valve is piped into the system, it looks strangely located in that picture though almost like it is piped in between the close tees?

JaySyd

GGross said:

I think you will just need to isolate that loop and pressure test just the loop to see what happens. Having a hard time seeing how that fill valve is piped into the system, it looks strangely located in that picture though almost like it is piped in between the close tees?

Yeah the fill valve was added after the fact since the original installer didnt put any fill valve in. To fill a specific loop, i just close any valves and open the loops i need to ensure water only goes to those loops down via the manifold. works perfectly actually.

Derheatmeister

JaySyd said:

edit:

@Derheatmeister Would improper piping techniques actually cause pressure to be lost when i have closed any other zones and isolated the single problematic zone?

I just replaced the PRV on top of the boiler (rated for 30psi) since the last one was leaking at the top of the NPT connection - post change to a new 30psi PRV, it seems to drip when the pressure hits 20psi - which is another weird issue - as long as i keep it below 20psi, it seems to behave. But now after returning to it after a few hours, i can see about a pint of water in a bucket that catches my PRV water so i can account for it in all this troubleshooting. Could there be something else within the boiler impacting the functionality of the PRV ? or is this new PRV also faulty?

No other leaks in the mech/boiler room that are visible to me

see image of the room here

If your PRV is leaking it can cause air to enter your system.. When air is enters your system it can cause air locks/ partial flow conditions.
I would check the charge on the expansion tank in a relief stage( no hydronic pressure on the wet side) You may need to install a tee with a boiler drain between the expansion tank and the shutoff valve in order to perform this job correctly.

JaySyd

Derheatmeister said:

JaySyd said:

edit:

@Derheatmeister Would improper piping techniques actually cause pressure to be lost when i have closed any other zones and isolated the single problematic zone?

I just replaced the PRV on top of the boiler (rated for 30psi) since the last one was leaking at the top of the NPT connection - post change to a new 30psi PRV, it seems to drip when the pressure hits 20psi - which is another weird issue - as long as i keep it below 20psi, it seems to behave. But now after returning to it after a few hours, i can see about a pint of water in a bucket that catches my PRV water so i can account for it in all this troubleshooting. Could there be something else within the boiler impacting the functionality of the PRV ? or is this new PRV also faulty?

No other leaks in the mech/boiler room that are visible to me

see image of the room here

If your PRV is leaking it can cause air to enter your system.. When air is enters your system it can cause air locks/ partial flow conditions.
I would check the charge on the expansion tank in a relief stage( no hydronic pressure on the wet side) You may need to install a tee with a boiler drain between the expansion tank and the shutoff valve in order to perform this job correctly.

Can i simply shut off the valve connected to the expansion tank to test that out? It has a shutoff right before it. I connected a bike pump to the expansion tank to check its pressure and its sitting around 15-20psi.

Would a under/over pressured expansion tank cause the PRV to misbehave?

Derheatmeister

JaySyd said:

Derheatmeister said:

JaySyd said:

edit:

@Derheatmeister Would improper piping techniques actually cause pressure to be lost when i have closed any other zones and isolated the single problematic zone?

I just replaced the PRV on top of the boiler (rated for 30psi) since the last one was leaking at the top of the NPT connection - post change to a new 30psi PRV, it seems to drip when the pressure hits 20psi - which is another weird issue - as long as i keep it below 20psi, it seems to behave. But now after returning to it after a few hours, i can see about a pint of water in a bucket that catches my PRV water so i can account for it in all this troubleshooting. Could there be something else within the boiler impacting the functionality of the PRV ? or is this new PRV also faulty?

No other leaks in the mech/boiler room that are visible to me

see image of the room here

If your PRV is leaking it can cause air to enter your system.. When air is enters your system it can cause air locks/ partial flow conditions.
I would check the charge on the expansion tank in a relief stage( no hydronic pressure on the wet side) You may need to install a tee with a boiler drain between the expansion tank and the shutoff valve in order to perform this job correctly.

Can i simply shut off the valve connected to the expansion tank to test that out? It has a shutoff right before it. I connected a bike pump to the expansion tank to check its pressure and its sitting around 15-20psi.

Would a under/over pressured expansion tank cause the PRV to misbehave?

Yes over or under pressurized expansion tanks cause all kinds of issues..The proper way to check/Adjust it is by isolating it from the system and draining it..