I am gearing up for my first DIY tile job, and first time using Schluter Ditra product, so I wanted to bounce my plan and a question off you all.
The project is a 70 square foot bathroom (new construction) where I’ll be installing 1/2″ natural slate stone and hydronic radiant heat over an existing 3/4″ plywood subfloor. The joists underneath are 2X10s 16″oc with a max span of about 8.5 ft (it is a circular room, so the span decreases from the center joist). I have 1 3/8″ rise from the top of the 3/4″ plywood subfloor to the top of the finished floor in adjacent room- so the height of the tile assy is of concern.
Schluter recommends a double subfloor, but I do not have the room to work with. Based on previous discussions/recommendations, I am planning to beef up the joist bays by installing 1″ plywood from underneath- glued and screwed to existing subfloor and supported by cleats from below as well. Essentially, this should give me the recommended double subfloor without adding any height to the assy.
Above the existing subfloor, I’ll install 3/8″ pex tubing 8″oc, then .020″ X 5″wide aluminum heat transfer plates (covering the pex from above). Then I’ll rip strips of 5/8″ plywood and install them as ‘sleepers’ over the top of the aluminum fins between the pex runs. Once the hydronics are done, then I can proceed with the ditra and tile.
Does this sound acceptable?
Because of the airspace between the sleepers, the folks at Schluter would like to see another layer of plywood over the sleepers to provide a solid substrate for the ditra, but they said I could also fill the space between sleepers in order to save on height. Schluter would not recommend what to fill the void with, however, so I am looking for recommendation here…it seems like this would be a better strategy, as I would get better heat transfer if there was contact with the pex and/or aluminum fins along the runs.
Any thoughts appreciated.
Replies
Why add the layer of plywood to the underside if you are adding sleepers ? The extra ply is doing nothing once you add the sleepers. I just did one by adding a strip on top of the subfloor/in line with the floor joist and placing my pex in between the strips and then adding 3/4" plywood and then my Ditra. I had to go up 2" though to match the shower pan for a handicap accessible shower.
If you beef the floor like you are planning to , you could simply grout over the entire floor in lightweight gypsum, possibly even standard mortar. The issue is a stable substrate that will allow the ditra to perform properly. The grout will add thermal mass to the floor system which will also make your hydronics perform better. It also allows you to resolve the height mismatch more accurately. If choosing this method I would suggest sistering some of the joists if possible, prior to adding the additional layer of ply and the cleating.
Edited 11/11/2009 10:53 pm ET by 911RenoGuy
use a self leveling compound after setting the pex and forget all that other plywood
Welcome to the
Taunton University of Knowledge FHB Campus at Breaktime.
where ...
Excellence is its own reward!
Piffin-
What do you recommend for a self-leveling compound, and how thin could I go with it? Since i'm trying to minimize the height, I thought the sleeper method would be best. But I like the idea of embedding the pex in compound (better heat transfer, solid substrate).
Also, would you still recommend the cleats and plywood below the subfloor if I were to go with the compound method?Shawn
You can pour it in as thin as you want. You DO want to seal places where it can leak away as it pours out thin and runs.I have used a Mapei product. Don't recall the name on the bag right now.I don't see how you gain anything with the underneathe work
Welcome to the Taunton University of Knowledge FHB Campus at Breaktime. where ... Excellence is its own reward!
Have you considered Warmboard or QuicTrak? Save you a whole lotta work and headache.
Mike Hennessy
Pittsburgh, PA
Everything fits, until you put glue on it.
They are too thick, and a little spendy...plus I'd still have the airspace issue where the pex meets the Ditra to deal with.
thanksShawn
A bit spendy, I suppose. But your space is pretty small so the extra cost wouldn't be all that much more than what you currently plan. And you put the Ditra down on thinset, which goes directly on the Warmboard/QuikTrak, so there'd be no space issue. As for too thick, QuikTrak is only 1/2", and unless your tile is really thick, I think you could even fit WarmBoard in the 1-3/8" you've got since it would function as the additional subfloor thickness you're looking for.
Mike HennessyPittsburgh, PAEverything fits, until you put glue on it.
Edit: That said, it may be more of a show-stopper that your space is round. I'd forgotten about that.
Edited 11/12/2009 9:33 am ET by MikeHennessy
You're right, that the quikTrak is basically a manufactured version of what I was planning to build on-site...but wouldn't I still have the issue of needing to fill the gap above the pex before laying down the ditra?
With the curved walls of this room, it is sounding like I might be better off just pouring a self leveling mix over the pex.
thnxShawn
Because of the circular nature of the room, track may not be a good solution (tho' not necessarily impossible). But if you can make it work, there really isn't an issue with any "air space". The tubing is pretty much level with the surface of the trak and you just thinset over it, lay the Ditra, then thinset in the tile.
Mike HennessyPittsburgh, PAEverything fits, until you put glue on it.
PS: No matter what solution you choose, remember you'll need to insulate under the floor before you do anything.
Edited 11/12/2009 7:04 pm ET by MikeHennessy
Why not install the radiant tubing under the existing subfloor, then fit some isocyanurate (sp?) with a foil face under that to reflect the heat up thru the floor. Then you could easily add another layer of ply, plus the tile and everything else.
I like Piff's idea using SLC but I like your idea best! Good flash.View Image
The Woodshed Tavern Backroom
The Topics Too Hot For Taunton's Breaktime Forum Tavern
I was surprised no one else came up with it first.
It's the nature of our work I dislike the most but I spose it's like that with anything.
Ya go over'n over'n over something in our head and then viola...you can't believe how simple it was...but the worst part is...the customer looks at it and says or thinks...big deal...thats obvious to me (them)...yeh sure...only cause you saw it done.
I hate that...lolView Image
The Woodshed Tavern Backroom
The Topics Too Hot For Taunton's Breaktime Forum Tavern
oh yeh...like the job I was looking at today (the customer wasn't home...I have the key). About four hours there figuring out all kinda little things. I talk to the customer just a little bit ago and told him I was there 4 hours and he said, "why so long...how much was there to think about??"GrrrrrrrrrrrrrrrrrView Image
The Woodshed Tavern Backroom
The Topics Too Hot For Taunton's Breaktime Forum Tavern
Just tell him you were getting the hard part of the job done. LOL
Your idea ain't bad either. I had the thought, but try to avoid trying to talk someone into too much of a change, unless it has great gains.
One mental block at a time.Does that make me an incrementalist?;)
Welcome to the Taunton University of Knowledge FHB Campus at Breaktime. where ... Excellence is its own reward!
I guess I ruled the "staple up" method out because of the heat transfer issue- wouldn't it be much better to have the pex at the surface, rather than trying to push the heat through 2 layers of plywood?
You're method would be by far the easiest and cheapest, but I like Piffin's method of embedding it in SLC....Shawn
When you brought up the height issues, yeah I agree with Shep. If it's open underneath, go staple-up. Wood has an R-value of about 1 per inch. My little rule-of-thumb is to have at least three times the R-value under the PEX as you have above it. That'll give the path of least resistance up into the bathroom.FWIW, when I built my own home years ago, I did staple up with no aluminum plates. On top of the pex is 3/4" T&G ply subfloor, then 1/2" ply underlayment, then brazilian cherry hardwood. A couple of areas (laundry, kid's bathroom) have tile. Below the PEX is 1/2" foil-face polyiso tucked up against the heads of the staples and then FG batt insulation under the polyiso.I circulate about 110-115 degree water, and it's a beautiful thing.No striping, just a nice even warmth.
Sheesh- I seem the get opposing answers whenever I ask about hydronic installation!
While we have a super tight, well insulated house, we must suffer through some brutal winter weather here in northern MN. -40F nights and highs around 0F for days on end are not uncommon. Our water is running around 90-100F on the supply side, and I've been told that staple up will be okay for "floor warming", but not for room heating in these conditions.
It would be nice to see some study or comparison heat transfer rates for a given pex size, water temp, and installation methods...until I find that, it's tough to make the call.
I also wonder about reaction time- would it take longer for the heat to get through the layers of wood in a staple up arrangement, or would the higher mass of an embedded system actually take longer?
thanks for the advice and personal experience!Shawn
Yeah, I've had several people tell me that my RFH setup can't possibly work "as is", and to get any decent heat out of it I'd need to run 140 degree water.FWIW, I installed radiant staple up in my bro-in-law's house in Eden Prairie, just outside of Minneapolis. PEX in slab in the basement but staple-up with no plates on teh first and second floors. It's been running fine for him for the past...I'm gonna say 7 years. Maybe 8. My brain is fogging over.I installed staple-up with plates in our lake house in Wisconsin. Poorly insulated, drafty, old single pane windows. But warm floors.So yeah, a staple-up with or without plates can work. In MN the plates would aid in the thermal transfer.
"I've been told that staple up will be okay for "floor warming", but not for room heating in these conditions."Regardless of how hot the supply is, if you deliver a certain number of BTUs to the floor system, then it has to go someplace.By using better/more insulation below the pex and foil delivery system than what is present in the wood above it, the heat has little choice but to move in the direction of the lessor insulation - towards the room.greater heat differentials can mean the btus supplied move there faster, but they still move.Time required for the transfer is of little importance in my mind with a radiant system, inasmuch as the entire idea of a radiant floor heat is constant rather than variable. IOW, if you are turning that thermostat up and down and up and down, you are wasting money by using radiant heat system in the first place.The best way to do radiant heat is embedded in a mass like concrete, but staple up under works fine when the principles are understood and employed in the design and installation.
Welcome to the Taunton University of Knowledge FHB Campus at Breaktime. where ... Excellence is its own reward!
"...The best way to do radiant heat is embedded in a mass like concrete,..."
Yes and no. Mostly from the performance perspective during the "in between seasons".
Years ago RFH was all high mass. Then gypcrete installations came along as a sort of "medium mass".
Now we have "low-mass" systems, staple-up, or warmboard, or quick track, etc.
In very cold climates a high mass system would likely be best.
In southern new england where I am, with a high-mass system, if we had a couple-day cold spell in the fall, the system would load the slab with heat, then after the cold snap with warmer temps , the house would overheat for a few days while the lab released it's stored energy into the house. Even with an outdoor reset. Same occurrence in the Springtime.
That's where low-mass systems can be better in some ways. They deliver the heat sooner, as there's less mass to "load", but were the power to go out in the house, with less mass for storage, they give up the heat sooner too.
But I fully agree with you, in the midst of a long cold heating season, high mass is better. In the 'tween seasons though, with temps varying up and down, it could possibly cause problems.
Ditto on all that, then compound the fluctuations with wood heat like I do. I have to anticipate temperature swings - both that of the weather, and of my wife.
Welcome to the Taunton University of Knowledge FHB Campus at Breaktime. where ... Excellence is its own reward!
In my experience, the weather is more predictable
<G>
Regardless of how hot the supply is, if you deliver a certain number of BTUs to the floor system, then it has to go someplace.
But isn't there a problem getting the BTU's out of the tubing and into the floor? If I insulate underneath, and have layers of wood (somewhat insulating) making up the subfloor, then the rate of heat transfer would have to be diminished, compared to pex embedded in concrete.
I'm not arguing that staple up can work for some (Mongo's testimonial, for one). My concern is with my scenario of extreme winter weather and low supply water temp (90-100 F). On the plus side, the house is tight and well insulated.
I calculated a worst case heat load of around 20 BTU/hr/ft2 for this room. The floor area is about 65 ft2. What I am trying to find is a way to calculate how much heat can be delivered for the different floor assemblies, given my water temp, flow rate, etc. If I choose staple up, it must require a higher water temp to achieve the same heat output in the floor, correct?
Perhaps someone has the equations for this.
thanksShawn
20BTU/hr is pretty easy with a RFH system, regardless of the type of installation. Is there a reason you're set on the 90-100F water temp and not willing to bump it up to a higher temp? I'm going to ask a few other questions here...Do you have whole house radiant? If so, with this being "new construction" why was this bath left out during the tubing installation? Or is this an addition?You're not planning on an open system with this loop are you? Have you considered electric radiant, either cables or mats?Just asking.
Yes, it is a whole house system. I am diong all the work myself, and thought it would be gentler on my wallet and sanity to do things in stages.
Two years ago, we poured a SFPF slab with pex and the building was roughed in. It is about 1200 sq ft. I heated it for the first winter with a wood pellet stove (which is now our 'backup' heat, although I use it quite alot).
I installed a 7KW micro boiler and began heating the main floor slab last winter. Even in the extreme cold weather, the boiler was only running half the day (usually it heats the slab up during the night, then goes all day with the built up heat and passive solar gains).
During the boiler install, I plumbed in 2 extra valves for future upper floor bathroom and bedrooms, which is my current project. The house is still unoccupied, but i hope to finish it this winter.
The hydronic system is closed loop. Once the house is finished, the next project is to build/install solar thermal system with large water storage tanks in an adjacent building. I have already buried a Thermopex run underground and into the new construction for feeding the solar heated water. Thus, electric heat mats are out of the question.
Why the low supply temp? I'm not sure! When the boiler runs at full load, that is all it will do. Perhaps I need to throttle down the flow rate to get the temp up...I honestly have not tried. In anticipation of the solar thermal project, however, I don't want to build a hydronic assembly that is going to require really hot water....
Any thoughts appreciated.Shawn
Sounds good. I'll try to post later. A busy day.
from the POV of heat transfer, it IS better to have the tubes on top, but you are designing an entire system, structural, insulation, and appearance, all within certain advance givens. There are compromises to made here and there.So if heating efficiency is your primary goal, then pex on top and insulation below.But if you have serious structural concerns, you need to modify that.
Welcome to the Taunton University of Knowledge FHB Campus at Breaktime. where ... Excellence is its own reward!
I'm a carpenter, not a thermal engineer <G>
You may have to add more loops of tubing to get the same heat if you put it under the subfloor. And installing foil-faced foam will reflect the heat up to where you want it. I think that with a properly designed system you won't notice whether the tubing is above or below the sub-floor.
But at least with the way I suggested, you can add another layer of ply for a good, solid base for your tile without screwing up your floor heights.
Do staple up with Al plates from below. Do you really want a 1 3/8" height diff in the finished floor levels? That seems a bit high to me. With the staple-up you can put just as much ply as you need to stiffen the floor.
We have staple up under 1" of wood flooring and it works fine. Yes it being on the bottom of the floor will mean that it will take slightly longer for the heat to travel up thru the floor and into the room but thats really a non-issue as all radiant heat systems should be run at a steady thermostat setting. Just be sure to insulate like mongo has suggested.
Actually on our system designed by NRTRob here, I have loops in cement on the bottom floor, staple up on the middle floor and rad on the top floor. Since we are only living in the middle floor (other 2 floors our unfinished-gutted-uninsulated we only are heating the middle floor and I've had to make some adjustments to the water temp in the loops to deal with the additional heating load. If it gets real cold (around here thats in the low 40's) I'll probably turn the slab on to keep the bottom floor in the mid 60's and to help out the loops in the middle floor.
I can tell you warm floors + babies is a good thing.
Daniel Neumansky
Restoring our second Victorian home this time in Alamdea CA. Check out the blog http://www.chezneumansky.blogspot.com/
Oakland CA
Crazy Homeowner-Victorian Restorer
Hey Mongo,
I'm still weighing my options...any thoughts?
Sorry, forgot all bout the thread!It sounds like you have a valid short and long-term plan. If you don't have on already, add an outdoor reset. It'll help the system perform better in your cold winters.A properly insulated staple-up with plates will give you good RFH performance plus solve your height issues. If you use heavy gauge plates instead of the lightweight in the staple-up, then thermally the staple-up with heavy plates will actually out-perform a lightweight sandwich assembly that's on top of the subfloor. The plates are fairly easy to install. Screw them up and snap the PEX into place. The heavy gauge plates are about $1.50 a running foot. Two per bay works out to $3/foot of joist bay, or about $2.25 per sqft of floor space.On top of the subfloor, since you're using natural stone I still recommend another 1/2" of ply underlayment (could be 3/8" but that's a min) for a total of 1-1/4" of ply, then Ditra, then stone over the Ditra.Were you using ceramic tile, you could Ditra or Ditra-XL right over the 3/4" subfloor. But stone needs the extra stiffness offered by the ply underlayment, even with Ditra.I'd up the PEX size to 1/2" from your proposed 3/8". Two runs of tubing per 16" wide joist bay. For 1/2" PEX I go for 250' loop max. you can go longer, but 250' is my goal. Your tubing run will be quite short, so you'll have a single loop and no pressure problems.Overall, the staple-up with heavy gauge plates will give you as good or better performance than a lightweight sandwich assembly. It'll be less expensive to install. It'll be easier to install. And it'll solve your height differential issues.
Thanks Mongo, I appreciate the feedback.
I'm feel reluctant to go with staple up, I guess because it is an unknown. But as I think about it, the only difference between this and my original plan (aluminum plates between subfloor and sleepers) is 1/2" of plywood. Probaly not worth losing so much sleep over!
Plus, the staple up method would be easier and allow for future access from below if necessary. Both good things.
When you refer to "heavy gauge plates" do you mean the .050" extruded alum plates?
thanks againShawn
No, 0.0625" thick would be the thickness of the heavier extruded plates.
Lemme see if I can find a web reference...
Try this.
Oh, and don't let me talk you out of your comfort zone or push you into an installation method you're set against. If you feel compelled to do a sandwich assembly on top of the subfloor, have at it. Thinking back to one of your earlier posts, you referenced a 90-degree circulating temp. Is that temporary, a "construction" thing? Or is that permanent? 100-115 or even 120 is a common design temp for high or low-mass systems.