So yo want tile over that huh....

[fcfc]

Three years ago a GC convinced a homeowner that he could add a layer of wood to the original sub-floor, then routered in floor heat channels and run pex lines. Over this he convinced the homeowner, the store that sold the material, and the installer that it would be a good ideal to then put the tile directly over this in floor heat system with no CBU. Just tile directly to the wood and directly over radiant heat pex lines.

Not knowing the specific system that he used, or whether it actually matters, or the depth of the various routered spaces, what type of wood was used as the filler, or the actual structure of the subfloor itself leaves some questions in the air. This just seems like it was a bad idea all around from the beginning.

The floor has failed in numerous places, cracked tile, loose tile, disintegrating grout joints throughout. The floor is now a total fail. Has anyone out there heard of such a technique or seen something similar? The only documentation I have found on systems like this all say to tile 'as normal with cbu etc' over a comparable type floor system.

I am hoping to salvage the structure, pull the existing tile, float or grind it flat again, install CBU and new tile. Once I do a full inspection next week I will add more detail.

 

[rugaddict]

we need to know floor joist size and spacing----also number of and location of support columns----lots of builders like to use the plywood i-beams 2 foot centers and forty foot runs with no additional support---gives you a wide open basement---also gives you a bounce house upstairs----movement means cracked tiles

 

[Ernesto]

Happens quite frequently. I see some thinset specs say you can go directly over wood, although I'm certain they mean on walls not floors.

 

[fcfc]

I saw that TCNA included a spec a few years ago specifically for installation over wood subfloors with rather specific requirements for thickness and type of subfloor along with proper structural support, however I do not recall any mention of in floor heat systems. It just struck me as damn odd that putting tile directly adjacent to an in floor heat system supported solely by wood was to anyone a good idea.

 

[Floorist]

I saw that TCNA included a spec a few years ago specifically for installation over wood subfloors with rather specific requirements for thickness and type of subfloor along with proper structural support, however I do not recall any mention of in floor heat systems. It just struck me as damn odd that putting tile directly adjacent to an in floor heat system supported solely by wood was to anyone a good idea.

He got paid and it probably outlasted any warranty he had. That's all some installers care about.

 

[fcfc]

This floor has been down 4 years total. The tile installed was a heavy wear residential grade ceramic tile 18 x 18. Dotted throughout the floor were various insets custom cut into a pattern.

Existing Issues -
Grout has failed in approximately 50% of all joints in floor
Schluter in areas adjacent to carpet has failed and is freely moving with pressure
At least 12 tile are currently broken in places throughout the kitchen, dining room and walkway areas
Several tile show movement from normal traffic and rock back and forth freely within their grout joints
Grout was unsealed and shows discoloration where not damaged by movement
Carpet to bedroom was not ramped to tile area and represents a toe stubber
Tile was installed over a routed-in pex line in-floor heat installation whereby 5/8" plywood was added to existing 3/4" sub-floor. There was no additional CBU between the wood with pex lines and the tile.
Joists are 16" on center, location of bearing wall below indicates maximum runs in affected area to be roughly 14 ft.
Tile on bathroom wall installed to match pattern of floor. Tile installed directly on top of said floor. Extremely doubtful that the floor can be demoed without causing damage to the wall tile or relaid to exactly the same spot while raising the floor height and or with changing the pattern of the floor, will most likely have to simply pull and relay the wall as well.


All in all this is somewhere around 500 ft of flooring/wall work.

The thing that bugs me though is that so many of the tile have simply broke in the middle of the floor. There is 1.25 inches of actual plywood underneath it, though I am unsure if the breaks are simply where possible voids from the pex line channels happens to form or some other reason. I also was not able to determine how the two layers of plywood were actually attached to each other. I do not know at this point if the floor heat was on during or shortly after the install.

 

[fcfc]

There is one thing though that I did note, there is a smaller hallway/entryway that seems to be free from the majority of damage. And I am almost certain, free from the actual in floor heat system as well. This space represents less than 10% of the total floor, but did not appear to have the rampant grout joint damage/breakage as the rest of it.

 

[DarisMulkin]

I'm not a tile guy so just asking questions from precious posts I've read here or there.
Is it possible that the heating system was turned on two quick drying out the grout and thinset? Also is there movement in the floor?

Daris

 

[Ernesto]

Schluter in areas adjacent to carpet has failed and is freely moving with pressure

So the tile is or is not set directly to wood, or over top schulter ditra? Or are you talking about schulter trim? Sounds like a huge mess to me.

 

[fcfc]

Its the schluter trim pieces that are loose. Normal traffic causes it to deflect completely as though it is not supported from below at all. The whole thing is a hot mess. Given the time frame during in which it was originally installed, I would suspect that the floor heat was not on during installation, but was turned on shortly afterwards.

Tile is set directly to wood. The layer of wood the tile is on has the routed pex lines intalled in it, so there are places where tile is in almost direct contact with pex line in floor heat.

This thing is a hot mess. I figure it will take me almost 2 weeks to fix this mess depending on how touchy it is getting the tile up and not damaging anything.

For the fix we are planning on adding a layer of ditra to the subfloor after grinding/scraping off all the old debris (and adding slc where needed) to find some kind of flat to work with.

 

[highup]

Its the schluter trim pieces that are loose. Normal traffic causes it to deflect completely as though it is not supported from below at all. The whole thing is a hot mess. Given the time frame during in which it was originally installed, I would suspect that the floor heat was not on during installation, but was turned on shortly afterwards.

Tile is set directly to wood. The layer of wood the tile is on has the routed pex lines intalled in it, so there are places where tile is in almost direct contact with pex line in floor heat.

This thing is a hot mess. I figure it will take me almost 2 weeks to fix this mess depending on how touchy it is getting the tile up and not damaging anything.

For the fix we are planning on adding a layer of ditra to the subfloor after grinding/scraping off all the old debris (and adding slc where needed) to find some kind of flat to work with.

How hot is the water in the tubing? There are limits on how hot the water can be for flooring. Does the consumer flip the thermostat off and on like a light switch whenever they leave the house? I'd suspect with tile, it would be best to keep the heat on all the time instead of cycling it like a forced air furnace.

I will assume that floor deflection doesn't change by adding ditra.
What is the span where the tile is cracked the worst?
Has anyone actually measured the deflection?
https://www.tcnatile.com/faqs/30-deflection.html
http://www.johnbridge.com/vbulletin/deflecto.pl

 

[Kman]

I wouldn't touch that mess with a ten foot pole.

I'll post some more when I'm not on my phone, but for now, run. Run fast, run far.

 

[Kman]

There are so many potential problems with that installation, but let's assume for a minute that the things that we don't know about have been done correctly. For instance, let's assume that the joist structure is adequate, the subfloor is oriented and attached properly, and the thinset used was appropriate for a direct-bond to plywood.

When installing a second layer of plywood, it has to be oriented the same as the subfloor. It should be an exterior glue plywood with a face grade of C or better, it should installed with the joints offset from the subfloor, screws or ring shank nails should be used at the appropriate shedule, and the only time glue should be used between the layers is if it's a full spread with a trowel. If all those requirements are met, then a person could theoretically install tile directly to that second layer using a modified thinset rated for such an installation. I wouldn't do it personally, because there is too much risk involved, and the only savings is the time and money it takes to install an appropriate tile backer or membrane.

In this case, that top layer that the tile is attached to has been cut up in order to install the tubing. Each of those grooves is a weak point in the substrate, not to mention a "hot" weak point. I don't know that the heat itself would cause a problem, but I could see it being a factor when it's installed as described.

There is nothing that is going to save this floor. Any attempted "fix" is going to be a bandaid on a gunshot wound. At a bare minimum, it should be take down to the subfloor.