Retrofitting an uninsulated slab for radiant
We are planning a major renovation of our 1970s ranch and would like in-floor radiant heat. The house is on an uninsulated slab in climate zone 5A. Our current plan is to insulate with 1.5″ XPS over the slab, lay Schluter Bekotec over the insulation, place the the PEX in the Bekotec, screed then tile thoughout the house.
We will be using appropriate vapor barriers and uncoupling membranes. With the XPS and Bekotec, the R-value of the floor should be 9-10. We will also be insulating the slab perimeter as much as is possible/economical, to what R-value is uncertain at this time. The goal is to run as low water temps as possible and keep our wood gasifier firings to a minimum. This plan as it is will cost us ~3.9″ in ceiling height loss. I would love to have less loss but will not forgo insulation.
Does anyone have experience with a similar project? Thoughts or suggestions? Will the 1.5″ XPS be sufficient? I don’t want to pay/do extra work to heat the earth… Thanks so much in advance!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
J Marchesi,
I general, I'm not a fan of in-floor hydronic heating. These systems are expensive, and in most cases, the money invested in this type of heating system would be better invested in improving the home's thermal envelope (reducing air leakage and adding insulation).
For more information, see these GBA articles:
Radiant-Floor Heating
Goodbye Radiant Floor
Is Radiant Floor Heat Really the Best Option?
Heating a Tight, Well-Insulated House
Martin is very anti radiant heat, but I don't know anyone who has it and would trade for anything else. Especially in a slab. If you already have a boiler its a relatively inexpensive project.
I can't speak for the design you have planned.
I do think you should try and get more insulation under the radiant
Trevor- I have radiant heat in my floor and I would not do it again. The biggest problem is that the typical system uses relatively low temperature water (compared to a typical hot water heating system.) Here's what happens to me:
I get up in the morning in winter and the house is at the 68 degree temperature set by thermostat. On a sunny but cold day in winter, I get a lot of solar gain through south facing windows. Thus, by late morning, the space (typical open plan kitchen/dining/living areas) is warm enough (70-75) that the heat goes off and stays off until the sun goes down and the temperature drops enough to call for heat. When that happens depends of course on how efficient the building envelope is. Mine is not great. So maybe the heat finally comes on at 7 PM. But in a really tight, well insulated house (like my next one) the heat may not come on until much later. In any event, if the house loses heat faster than the system can provide it, the space stays cold for a while. Thus, at 9 PM, we're maybe a little chilly. There's no point in turning up the thermostat, because radiant heat takes too long to warm the space. So we light the wood stove and by the time we go to bed, the space is toasty. Thus, the heat doesn't come on until the middle of the night.
I guess if you have a typical, not well insulated and air-sealed house, with minimal solar gain from south-facing windows, radiant heat will keep you warm and you can just set the temp and forget it. But no one is his right mind in a cold climate (I live in Maine) would ignore solar gain or not properly insulate and seal a house. For such a house, radiant heat would cost too much to install and since the boiler will almost always be oversized, the heat will be mostly off and the floors mostly cold.
If you're committed to going down this path, use 2" EPS instead of 1.5" XPS. In 50 years after the HFC134a has mostly bled off (doing it's 1400x CO2 greenhouse damage) it will have about the same R-value as 1.5" EPS of equal density. EPS is blown with pentane (7x CO2 greenhouse potential) and has already dissipated by the time it hits the distributor's lot, and has a stable R-value over decades.