Slab over pour/Gypcrete nuances
Alaska zone 7ish. A lightly insulated former 20×30 garage slab is becoming the part of the foundation to support a two story addition. Slab is monolithic 4-6” in the field with 2” of EPS underneath. “Floor drain” will be decommissioned. Slab edges will get 4” of new EPS plus 4’ frost wing just below grade. The slab set up is not ideal, so I’d like to convert it for radiant floor heat by laying Pex and doing an over pour, probably of Gypcret -thin slab concrete stuff is out of my league. The Gypcrete has the added bonus of leveling the floor which has slope for the former floor drain.
I assume that the gyp needs to be decoupled from the concrete, correct? EPS could do this, but I have concerns about keeping the foam down during the pour, there are some floor height issues, and have it in my feeble brain that the mass of the old slab below will somehow be benficial as a radiant heat source (but the gyp is relatively low density…?)
Then there’s the matter of attaching the pex to the slab, bare or otherwise. A few hundred tapcons through clips? Hot glue? Turns out Rehau makes a Velcro mat that is a decoupling air/VB and matching Velcro pex tubing for this. I’m $ure it$ very nice, but open to other ideas.
The existing slab is only exposed at the building edge on two sides- the other sides have adjoining framed floors and another small slab, properly done. This is my most complex retrofit yet and I would run screaming the other direction, except- it’s our house this time!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
I see building season is upon us.
I guess my biggest question/concern is whether or not to put at least some foam under the gyp-crete, or use the mass of the existing slab to our advantage-the power goes out often in winter.
And maybe somebody has a good argument for a completely different approach? Thanks!
It would be best to have a decoupling system of some sort but since the gypcrete will be covered and it's non-structural, it's not critically important. A layer of 6-mil poly would work.
I'm not sure how you think the existing slab will become a radiant heat source. It's not the source of any heat, it's a heat sink that will absorb energy from the heating system. If you turned off the heat, the mass of the floor would keep the space warm for a short time due to its "thermal mass" (determined by its density, heat capacity and volume) but not long, a few hours at most. You might be thinking that the existing slab would reflect heat up into the space, but that would require a reflective material and an air gap. If you can afford the space for a layer of foam insulation, that will make a big difference on performance because you won't be heating the existing slab and the soil.
I'm not sure how to fasten the tubes to the existing concrete; I have only installed in-floor heating using routed plywood channels or staples over existing wood floors. Installing Tapcons would be very slow; I'd consider Ram-set nails.
Thermal mass and radiant heating or often thrown together as a good thing. This is not the case, all the heat capacity does is slow down system response, which is the opposite of what you want from a heating system. It makes it slow to heat up and react to temperature changes, this is especially the issue for a place with solar heat gain.
I think the simple/cheaper option would be in your case is to install sleepers and run the tubes through it. You wan rip the sleepers to take out the slope of the existing floor.
This will add a bit more height but now you can use standard flooring which is much easier on the joints.
You can look at systems like Ultrafin for under floor or go for warmboard or quick tracks over the subfloor. A layer of foil faced polyiso between the sleepers over the concrete would also be a good idea for a bit of extra R value.
For smaller spaces with tight height, I've installed quicktracks directly over rigid on the slab. The panels were tapconed down and tile installed over 1/4" cement board right over the quicktracks.
Thanks of focusing my thinking on this, really appreciate the input. Yes, I was thinking that the mass of slab would be beneficial, but what you guys say makes sense. I know the slab has 2" of EPS below, but that's really not much up here in Zone 7+, even if I go all in on a thick frost wing and wall. But, if I treat the slab as only a footing, You could argue I'd want at least 4" of EPS on top, room which I do not have- there are already floor height challenges in this retrofit. Stuck in the grey area, as usual. So middle ground it is?
Now I'm thinking that at least some EPS fastened to the slab below is the direction to head. Pex stapled into the foam above (foam staples), plus gypcrete above, may be the middle ground. Already thinking Gyp for upstairs (lots of bedrooms, small spaces), and need it for another project as well, which would offset some of the cost.
If I had room, I'd be tempted to run sleepers with foil faced EPS between, another layer of transverse sleepers with pex or some radiator plate between, subfloor on top. Need to find 3-3/4" somewhere...
I've been thinking 1/2" pex. "That's what we always use"... Perhaps this is short sighted. I like the quick track idea, and could make it easily, but the approach lends itself to a 3/8" or 5/16" tubing. Getting in over my waders here, so could use some clarification on this, too.
I know nothing about radiant in the ceiling, but can also see that if I dedicate the height I have spare to insulation above the slab, and put radiant in the lid, I could heat the upstairs and downstairs potentially- and skip the gyp? Lots to learn.
Foam plus stapled pex would work. You can also look at the preformed EPS panels made for this like Isorad or ampex, bit more cost but simpler tubing install.
For upstairs, a simple staple up install would also work. Ultrafin is a good option for this as it is a pretty simple install and keeps the tubing well away from any current or future nails through the subfloor.
Unless you are using the specialty floor panels, I would stick to standard 1/2" pex.
Thanks! I'm strongly considering an A2W heat pump for primary heating as well. Nothing like that going on around here, but willing to move the needle. A2A stuff is now trickling north, but as usual Alaska is mostly 20 years behind. Unclear if a radiant floor would run a different temp water than a ceiling? Would certainly need back up sources (fuel oil, birch, and spruce, are our options). Learning so much, I'll keep digging in, awesome resource!