Wood floor underlayment over radiant heated floor
Hi,
I am building a 2 story house wood frame house(with basement) that is heated with radiant hot water system installed in the floor.
The joists have been blown in with loose fill cellulose (for both sound and to keep the heat up) and a plywood floor is on top of the joists. R5 foam board is installed on top of the plywood and the PEX tubes are installed to the foam board. They will be pouring a self leveling concrete on top of the PEX and then gluing down engineered wood floors.
My contractor told me that wood floors on a light weight cement will feel very hard and he recommended installing a type of foam product that would go between the concrete and the wood floor. he says it will soften the floor.
I am worried that this type of foam will act as an insulator and I will need to raise the water temperature of the system.
Does anyone here recommend putting this type of under layer down and if so can you recommend a product that would soften the wood floors but not have too high an insulating value?
thanks
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Anything you put there will behave as an insulator- there's no free lunch here. The softest-floor way to get there with low water temps is to skip the cement, and go with with sheet-aluminum heat spreaders with foam insulation between the tubing, and nailing the flooring to sleepers laid out between the foam sheets.
There are commercial versions (eg: Warm Board) using routed out aluminum-clad plywood subflooring, but it's not rocket science to do your own using tightly fitting sheet metal heat spreaders designed for under-the-subfloor approaches to radiant floor, and routing suitably sized channels in your foam layer.
Using 1.5" EPS and 2x4 sleepers works pretty well, or 3/4"EPS and 1x4 sleepers, but with the thinner foam it's better to glue the EPS to the subfloor as you go since even with half-inch PEX you'll need to rout down to less than 1/8" above the subfloor to accommodate the ~0.625" outer diameter of half inch PEX plus the thickness of the heat transfer plate.
Some people just DIY-it by directly routing the subfloor (no foam) eg:
http://www.builditsolar.com/Projects/SpaceHeating/SolarShed/House.htm
There are good/better/best versions of the sheet metal heat spreaders- you want to use one with some spring to it that grips the PEX tightly so that it evens out the temperature gradient a bit, limiting the amount of temperature "striping" across the floor. With crummy no-grip versions the heat transfer isn't as effective, and the striping becomes dead-obvious. With the better ones the water temp requirements are about the same or lower than lighweight concrete, as long as you don't have anything more insulating than 3/4" thick hardwood flooring between the aluminum and the room. These are pretty good, but there are several comparable vendors:
http://www.eagle-mt.com/downloads/Heat_Transfer_Plates.pdf
Using the 2-channel versions of the plates means you need higher precision when routing out the EPS, but it'll have fewer issues with striping.
Dana, I have pex heated floors installed now for twenty years.
Heat striping, if looked for can be felt. The idea of looking for it fades. Long term living with a heated floor the occupants do not notice the floor heat striping. Focus on the floor and striping and it can be found. Get back to living life and it disappears.
I do not endorse Eagle Mt but they provided my heating system and I used their 8" OC aluminum tubing plates. The groove is tight enough that 1/2" PEX pops in and stays in. But, you need grooves at 8" OC to allow the tubing groove to sit below the plane of your subfloor. I ripped sheathing into 7" wide strips but, other than time/labor, I do not see why you could not route grooves in the foam board (a little tricky at the loop ends though).
I agree with AJ that heat striping is no real concern. You will feel heat differences but it is not an issue for me.
HI,
At this point it is too late to reverse course on the concrete - it is ordered and it's being installed on Monday. The question is really does the hardwood floor need anything underneath it for "softness"
Not all PEX heated floors are the same. In this instance we are talking PEX above the subfloor, in contact with the finish flooring, which is something of a worst-case condition. Staple up radiant with the PEX under but in contact with the subfloor has fairly obvious striping without the plates, but is barely noticeable with plates (any type). Suspended tube PEX has even less striping, but takes considerably higher water temps. To achieve radiant-slab type water temp requirements without the slab requires an above-the-subfloor approach, as well as some amount of aluminum to transfer heat from the PEX into the finish flooring.
Without the heat spreaders striping can be pretty severe in high load 40+ BTU/ft^2 -hr applications, especially when the PEX is above the sub-floor. WITH heat spreaders it becomes a "Princess and the pea" sort of thing, especially in high-R homes where the peak heat emittance of the floor is 15 BTU/ft^2 -hr or less. Without knowing the load, it's hard to say what is/isn't worth it, but it's not a huge up-charge for the dual-channel heat transfer plates, and it's just fine to go a bit wider than absolute minimum on the cut-in channels in the EPS for the PEX.
It's not clear to me what has/hasn't already been installed, but it's written in the past tense:
"R5 foam board is installed on top of the plywood and the PEX tubes are installed to the foam board."
Cutting up some 3/4" EPS &/or 1x sleepers to fit between the already-installed PEX would probably be easier than ripping out an re-installing anything that's already in place, and a layer of sheet aluminum under the engineered wood flooring can act as the heat spreader, though it's not quite as good as a heat transfer plate that grips the PEX. It' would be similar to this approach:
http://www.infloorheatingsystems.com/subray_radiant_system/index.html
The sleepers could even be 2" wide strips of OSB through-screwed to the sub-floor with the appropriate widths of EPS glued to the XPS between the sleepers & PEX. Having solid layers foam or foam + sleeper between the engineered wood and the structural subfloor adds quite a bit of support for the engineered wood, which is sufficient to make up for not being applied directly to a subfloor or lighweight concrete.
Sorry for not being clear
we have, starting from the ground up
Plywood Floor
R5 Foam board
PEX attached to the foamboard
Lightweight concrete poured on top of the PEX (Pex should end up in the center of the concrete)
The question is now... what goes on top
Engineer wood floors glued to the concrete or some type of foam underlayer and then the engineered floors.
OK, I'm still not 100% clear about your use of English tenses.
By " ...we have..." (as opposed to "...am building..." ) you mean that every layer of that stackup is already installed?
If yes, there is no underlayment with any "give" for the engineered wood that would not increase your water temperature requirements (at least a little bit.)
You can't just glue wood flooring to concrete in a radiant floor. The dimensional changes of the heated wood & concrete don't match sufficiently and the adhesive is guaranteed to eventually fail. The floor has to be floated,with a slip-surface underlayment between the wood & concrete. Maybe a thin ~50-60 mil low-durometer silicone rubber or neoprene serving as both padding & slip surface would be enough? Beats me, just don't make it too thick- 1/8" tops.
What exact product and thickness was the contractor recommending? And how was the flooring to be attached?
Dana standard 1/4" foam pad used below engineered floor will work fine. Temperature difference will not add up to a $100 bill annually. Huge benefit to wood layed on foam pad.... drop the noise level of walking on the floor in shoes dramatically.
I have in my early days done quite a bit of radiant under flooring. Important to give up on changing the the tstat too often as temperatures can not be quickly changed including sun over warming issues. To cool a sunny overheated room.... open a window... a door.
AJ,
I have a 2 story house with a basement. In the basement is a gas furnace and on the first floor I have a anthracite coal burner. I use the coal burner more than I use the gas furnace, however I have cold floors because of the cold basement. I am installing new hardwood over existing hardwood and I was thinking of adding a layer of 1/4" rigid insulation under the new hardwood. Do you think this is a good idea? Could I do this?
Neal
I agree with AJ and Dana. The foam pad will require you to run a slightly higher water temperature, but, depending on your heat source, that won't matter much. Assuming you are heating the water by burning some kind of fuel, you have the capability to make the water way hotter than you need anyway. If the foam is similar to insulation foams, 1/4" will be about R-1. If your heating output from the floor is (10 BTU/h)/square foot, that would mean 10 degrees F higher water temperature. That's not negligible, but it's not going to matter much...unless you are heating the water with a heat pump, in which case that gives you lower COP, or it exceeds the capability of your heat pump. In that case, it might be worth working on finding a padding material that conducts heat well.
At my peril , I would hope that someone designed this in floor radiant heating system . That someone should be quite capable of entering the information in a computer program or on paper and determining based on the R value of the foam how much the SWT must be raised if at all . Do not use adhesive to glue the engineered product to the slab , the result WILL be undesirable . If you must use this type of floor since you are already obviously engaged I suggest 1/8" solid rubber padding , it has an R Value of .3 and will act as further mass due to it's properties .
Richard,
What is the undesirable effect of gluing the product to the slab?
Off gassing , adhesive longevity , most engineered floors are meant to have the ability to move independently from the subfloor . You are as stated , already into this assembly so I won't confuse the issue . What flooring are you using and what are manufacturers recommendations ? There seems to be a bit of indecisiveness , do you want a solid floor or one that is soft and moves ? Who helped you make these decisions that you are this far into it and still having questions ? I would possibly halt all further details until you sort through the details .
Coal Burner: A quarter inch of rigid insulation is only about R1, and isn't going to make an appreciable difference in floor temperature.
If it hasn't been done already, air sealing and insulating the foundation walls & rim joist could potentially raise the floor temps a few degrees or more, and would be well worth the trouble. During the winter stack-effect forces pull air out of the top of the house, pulling cold air in at the lowest leak points of the house, and leaky basements can hit temps quite a bit cooler than the deep subsoil temps. An insulated air sealed basement will run air temps somewhere between the conditioned space air and deep subsoil temps, and will usually stratify, with the warmest air at the subfloor level of the room above.
What is your location, and how cold does the basement actually run in winter?
I am surprised to hear glue down is not advisable! I am in a similar situation and my engineered flooring supplier (who does offer a warranty) does allow glue down but only when using a prescribed glue - which is considerably more expensive that the regular glue. They also require the max. temperature sensor pads to be installed. Also my underlayment supplier insists I use a Primer to ensure the Cement bonds to the plywood sub-floor - would this not pose an identical issue of separation due to expansion mismatch?