Replies

1. 
   ohioandy | Sep 05, 2014 07:08pm | #1

   For what it's worth, a 15x15 concrete slab on top of ANYTHING may, at some point, crack and telegraph those cracks through the tile. An isolation membrane like Schlueter Ditra should minimize the risk. But 14" of EPS used as solely as subslab fill is the most bizarre use for a synthetic and expensive foam I've ever heard of. Even Passive Houses rarely use that much. A few inches for a thermal break is fine. If the existing lowered foundation is already concrete, why not just fill with gravel? That will never deflect.

2. GBA Editor
   Martin Holladay | Sep 06, 2014 04:50am | #2

   Andres,
   I agree with Andy that most of the depression can be filled with gravel or sand.

   I imagine that most tile contractors would prefer to see a concrete slab rather than a layer of plywood as the top layer of this unusual sandwich.

   Andy's suggestion of an isolation membrane is a good one.

3. 
   davidmeiland | Sep 06, 2014 12:02pm | #3

   What is the surrounding floor made of? If it's wood, I'd just frame it in. Use 2x8 joists with a mid-span girder set on three posts.

   If everything else is concrete, filling with crushed gravel, compacted in lifts with a machine, is probably the easiest. Tie into the existing slab with rebar dowels around the entire perimeter.

4. 
   ohioandy | Sep 06, 2014 08:13pm | #4

   Andres, it occurs to me that nobody has answered your question, "will the floor deflect?"

   It's a fact that you can actually lay tile directly onto EPS, as long as it's a high-compressive strength product like the shower pans made by Schlueter, which are rated at 48 psi. However, you're not proposing that, plus that dense stuff is really expensive. Elsewhere on this site or at the Fine Homebuilding archive you'll find articles describing how to insulate basement floors by laying down foam and attaching a plywood subfloor above it with battens. Of course it's possible to do what you're proposing--you just have to make sure the subfloor system will distribute the maximum possible live load to the foam without exceeding the maximum deflection suggested by ASTM or the tile manufacturers--something like 1/720? An engineer will have to weigh in for you.

   I just recoiled at the suggestion of filling the dead space with foam, although if you leave out the carbon footprint question (please don't), maybe a DIY rigid foam lift is actually cheaper in time and materials than framing in a floor or filling the space with gravel and pouring a slab! Crazy, if you don't mind me saying, but maybe true. Except it's a lot easier to confidently achieve the necessary floor stiffness if you approach this problem more conventionally.

   Just imagine someone moving a piano in there sometime in the future. I'd go for the concrete over gravel and 2" of foam, unless the budget is supertight, in which case I'd frame it.

5. 
   ferroxido | Sep 15, 2014 12:17pm | #5

   Thank you all for your answers and recommendations.
   After reading your replies, I realized I probably did not explain myself fully.

   Regarding ANDY CHAPPELL-DICK and MARTIN HOLLADAY's replies, I am reluctant to use gravel or sand due to weight issues. As we all know, gravel and sand weighs less than concrete but we are still talking about 110 to 120 pcf. The sunken room is too deep, therefore I will be adding a permanent dead load that the existing foundation may have not been designed for. I am afraid of cracking the existing foundation and having bigger issues later. The reason I am considering the use of EPS is to reduce weight.

   In regards to DAVID MEILAND's reply, the surrounding floor is also concrete, not wood. And yes, I am planning on tying the new slab to the existing slab with dowels around the entire perimeter (16"oc).

   In regards to ANDY CHAPPELL-DICK's last reply, I was planning on using the EPS type 2 rated at 16 psi with a 10% deformation or type III rated for 24 psi (or around that). You are correct, it is expensive, more so than concrete (per cubic yard) but the load will be drastically reduced which is my main concern. I also realize that the idea is somewhat "crazy" or "unusual" to say the least, but I did some research and found out that in Canada high density EPS Type III or IV is actually used underneath highways for insulation concerns. Of course, it is probably flexible pavement, but they have 18-wheelers driving on top of that with single and tandem axle surface loads of probably 70+ psi!!!. Your piano example is good, and I appreciate the feedback, but the pressure would be way lower than that.

6. 
   Richard Beyer | Sep 16, 2014 08:48am | #6

   Andy it's L/360 (TCNA) for ceramic tile and L/720 (MIA) for natural stone tile. Folks Schluter Ditra will perform as good as the underlayment which is provided. If it does not meet the deflection criteria it will fail.

   Andres Ferro,

   The floating foam system you may be eluding to mimicking is designed by Schluter, known as Bekotec. It is designed for hydronic floor heating systems. With this system your structural floor is installed, then Bekotec is carefully layed over the top, loosely. The panels inter lock together like a puzzle, tubing is installed, then dry pack mud (or light weight gypsum) is installed over this, then Ditra, then dryset mortar and then your finished tile. (Other details may apply)

   http://www.schluter.com/9_1_schluter_bekotec_installation.aspx

   I attended Schluter training and work with all of their systems since 2002. They make one of the best tile products on the market aside from Ardex products. Just like every building product out there, if it's not installed correctly and corners are cut, it will fail. Schluter systems should not be viewed as a cure all system for a poorly constructed structural floor.

   Remember, your building an elevated floor in a house.... your not building submarines on this floor.