Waterproofing a shower ceiling with closed cell spray foam insulation
Question about waterproofing a shower. I’m worried about creating a moisture sandwich. This shower is unique because it has a low sloped ceiling, so one of the walls only 5 ft tall. The ceiling also has closed cell spray foam insulation. (Just an FYI: above the 7 inches of spray foam is a 1″ channel, then the roofing. The channel releases condensation out of the ridge vent.) The shower is not a steam shower.
I’m using Kerdi over greenboard for the shower walls and sloped ceiling. But to be warrantied by Schluter, Kerdi must run up to the height of the shower head, which is 6’6″; 18 inches higher than my lowest shower wall (5′ tall). This means I need to install Kerdi on the sloped ceiling too. So the shower ceiling layers from inside to out will be: tile, Kerdi membrane, thinset, greenboard, closed cell spray foam insulation, 1 inch channel, roof. Will I create a moisture sandwich between the 2 waterproof & vapor proof layers (Kerdi, greenboard, closed cell foam)? I’ve read that you should never use 2 vapor barriers, because the area between them cannot dry. If, 5 years from now, moisture gets behind the Kerdi or spray foam, it needs to be able to move out, right? In my case maybe a Schluter system isn’t the way to go? I’ve heard that Schluter will grant a special warranty in certain cases. Help appreciated! Thanks.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Kerdi is not quite a vapor barrier (it is vapor semi impermeable), so you won't be creating a sandwich. If you are worried, you can furr down slightly the Kerdi on the ceiling and leave a small air path that can dry sideways to the outside walls of the bathroom.
I would use cement backerboard instead of greenboard, and then not worry about it. Greenboard is the only thing in your sandwich that is intolerant of soaking.
I am a bit surprised to confirm Akos comment that Kerdi is vapor semi permeable. I used it for a steam shower expressly for being rated to take the continuous and rather extreme vapor load. My bad for not knowing that it has a .90 permeance rating. I think the air gap idea would introduce more issues than settle them.
In any case, I strongly recommend against cement board based on what I removed from the failed steam shower installation I was re-doing. I think the original work had been done in the early 80's. The studs were covered with polyethylene sheet first, followed by cement board and then the tile. We had to sister on new stud material in most of the stall due to extensive rot. Contrary to common beliefs, the cement board had disintegrated on the bottom two feet of the wall height due to the twenty plus years of steam use. It was like a semi structured paste in those areas and fragile in much of what was above.
The plastic sheeting had not protected the studs as much as intended largely due to the numerous holes poked in it by the cement board nails and careless handling. The rotting around each nail penetration made using the old stud faces for fastening new work very risky. After truing out the stall area and seating needs, the tile setter agreed to let us use Kerdi if we set it over cement board. (My memory is that Kerdi had not yet allowed setting Kerdi over greenboard.) I was favoring Hardi board over using cement board again, but as the Kerdi would be essentially sealing off the moisture load, I conceded.
Critically, you must follow the Kerdi requirement and use non-modified thinset to be in warranty. The tilesetter wailed mightily, but we insisted. My understanding of the requirement was based on modified thinset subject to extreme moisture load being at risk of re-hydrating the modifiers, which would essentially re-soften them. I have seen the same effect occur with tile near tub/shower edges which were set on the walls with acrylic based adhesives that were not meant for water or moisture exposure. And of course, the old cement board showed what steam shower levels of moisture could do.
It is a bit of a misconception that because cement board doesn't immediately breakdown in the presence of moisture, that it is good for moist locations. Not so. One must look at the material as a potential moisture reservoir. Same with thinsets and grouts. Yes, they may all get wet and not fall apart immediately, but all will hold moisture to the degree they are exposed to humidity and bulk water and their ability to dry to air. The tiles themselves are generally not significant absorbers, but some soft clay tile types could be although one would use them in a shower.
You can see that a steam shower is a trip to hell for all the base materials and why the years of water, steam vapor and poor drying conditions between uses caused the failure I dealt with. An open shower is a much less aggressive environment with much better drying conditions. This is not to say you shouldn't seal the walls and ceiling above your shower to secure better moisture control. If you are going with Kerdi on the walls, just go all the way up and across the ceiling and be done with it. Use a setting type drywall compound on the ceiling parts not tiled and do two coats of low permeance paint primer. Make sure you have a good bath fan and use it after you are done showering.
Sharing a wall or ceiling directly with an outside wall or roof plane as you have should generally be avoided if at all possible during design, but life is life. I would not worry much about the moisture sandwich issue for your circumstance provided you keep the moisture reservoir behind your tile minimal and open to good drying time. The Kerdi system is essentially set to do that and why it can be used directly over drywall materials.
Roger, I appreciate your sharing your experience. What a bummer. I just want to make sure I'm understanding you correctly. Was Kerdi ("polyethylene sheet") behind the cement board? Was tile directly applied to the cement board? Or did you have from inside to out: tile, then Kerdi, then cement board, then polyethylene sheet, then studs?
My contractor is old school. He prefers Redgard on horizontal surfaces, and no waterproofing behind vertical surfaces (shower walls). He says it allows cement board to dry to the outside. But I've read that this method isn't recommended anymore. He doesn't usually install Kerdi, but I suggested it. Thanks.
What brand of cement board was it? I've never seen a failure like that, even with long-term immersion. The failure of the poly sheet and studs is familiar. The only thing I have seen destroy cement board like that is soaking in water with freeze/thaw cycling.
Then again, steam showers are a really tough environment. Stainless steel mesh and a full mortar base is probably the best approach to those.
Consider using Kerdi board instead of Greenboard.
I should have been a bit more clear, the poly was just plastic drop cloth level, not Kerdi. The tear out shower installation was studs-poly sheet-cement board-tile. I have no idea what brand cement board it was. It was not the newer light weight kind and I don't remember seeing the mesh layer that is bonded in to some current brands. No idea who was the major producer that long ago.
Spenceday is right about going with the full Kerdi system if you can afford it. It also further illustrates why cement board behind Kerdi is kinda silly, though I have not used it. Maybe the Kerdi board would feel spongey? The tile guy we had was still pretty tied to the past even if he set tricky tile beautifully, so cement board it was.
And yes, I was horrified as well at the total destruction of the lowest parts of the cement board. I suspect the moisture loading built up over time and gravity pulled it downward. A fairly stagnant air exchange situation didn't help, as the door to the shower would close on its own limiting drying time. Daily use of the steam feature probably didn't help either. That and twenty years.
The pan was full mortar on lead sheeting and actually was the real PITA to break out. We then went further through the slab to relocate the drain to a better location.
Please don't let your contractor do you old school dis-service by not sealing vertical walls. Red Gard is great for non-steam showers/tub setups that are tiled on the walls. I also use it extensively for floor tile in areas that might get wet on a repeat basis. Vertical walls are just as subject to long term moisture reservoir problems as floors so stopping the moisture build up to as thin a total mass of moisture receptive thinset and grout is important. That means less moisture to give back up to the air in a well ventilated bath. DO NOT expect the cement board on vertical surfaces to "vent" off moisture build up through the back unless you have very drafty stud bays. That is a good way to get a mold farm going if accepting of that idea.
Your old timer (and I am into my government supported years) is quite right to RedGard the floors in bathrooms. Too many lazy or cheap builders will tile directly to sub-flooring because no one will see until the tiles start coming loose. Particularly right in front of the bathtub apron. Wet feet, splashy children and bath mats that stay on the floor all conspire to get bulk water into the grout, then the thinset, and then be very slow to give up the moisture. I have had to pull up my share of subflooring around tubs and toilets (mostly ones with bad seals) when people only thought a few tiles had worked loose. Think of Red Gard as a fluid WRB that you can stick tile to.
Best of luck with tiling over your head. Hope the tile is small and the thinset sticky.