We are mid-installation of our radiant system, and the tubes are running throughout the joist space in the basement ceiling. Because of some frequent cross-strapped beams, we can only use aluminum plates on about 1/2 of the tubing. So i was thinking of putting an additional insulation layer underneath all the tubing (with the regular insulation below that being R-0 Roxul Safe n' Sound). For the additional insulation layer (a) do we need it, (b) would foil or foam or other be better? Any product recommendations would be great. Thanks.

We don't recommend under-floor hydronic radiant heating systems because they are the opposite of high efficiency/performance. Only having 50% plate coverage won't help either.  As such, you will be depending heavily on less efficient thermal convection in lieu of more efficient thermal conduction to create the heat transfer through the floor into the room.  Thermal convection likes increased air movement to gain increased heat transfer efficiency, which is not what you will have.  So more plates would be better, however, you will still have a lot of floor R-value that will work against you in achieveing good hydronic radiant emitter performance.

Thanks, so foil or foam? Specific product to recommend?

If you have to pick between them, use foam. Or slightly better, use foil faced foam. On the other hand, once you have pretty good insulation in place, don't expect a big improvement by adding more to a basement ceiling.

The effect of underfloor tubing on efficiency depends largely on your heat source.

Foil faced Thermax foam would be my recommendation.

"THERMAX™ Sheathing is a non-structural, rigid board insulation consisting of a glass-fiber-reinforced polyisocyanurate foam core laminated between 1.0 mil smooth, reflective aluminum foil facers on both sides. The reinforcement, along with chemical modifications, contributes to fire resistance and dimensional stability. THERMAX™ Sheathing can be installed exposed to the interior without a thermal barrier and is especially appropriate for hourly rated assemblies."

http://building.dow.com/na/en/produ...athing.htm

How about this product - rigid foam with foil facing. So would have pex with some aluminum plates, then this aluminum-faced insulation, then roxul safe&sound.

http://www.lowes.ca/products/view.aspx?sku=9662264&

Any thoughts would be appreciated.

Seriously!

Thermax is a wonderful product. I have it on the walls of my basement and specify it for many additions we consult on. I have never used it to insulate a sub-floor radiant heating system since it is serious over-kill. In fact any reflective insulation is a waste of money since a dirty reflector is an inefficient reflector and anything you put under floor boards in new construction or old will get dWEirty/dusty. It is effective, though very time consuming to install the appropriate XPS on the cold side of any radiant panel. The standard for sub-floor insulation is 5 times the resistance of the floor above. I used R-13 batts in my own remodel with perfect performance and typically use a 13 - 19 depending on the spec.

True staple up, something I learned not to do in the 100+ year old four-square, is fast but inefficient requiring the maximum water temperature and will cause serious striping, which may be indicated in the floor covering, never a good thing. Even the more accepted, but still not right, "suspended tube" sub-floor systems will require high water temperatures in most climates and may show deficiencies in response time, output and of course overall efficiency.

The answer then is in the use of extruded aluminum heat transfer plates. There are many on the market but only one original and the only one we use. You may opt for the cheaper, lighter, formed plates if the heat loads indicate they will be adequate. My software program models in all three accepted design criteria following decades of research by Wirsbo/Uponor in collaboration with Wrightsoft, one of the serious radiant software companies. The use of extruded aluminum plates will increase the heat transfer rate of bare PEX by 75% and the potential output by 50% in many cases.

The key to insulating below a radiant sub-floor or sandwich system is making the joist space below air-tight, most especially at the end of each floor joist space and of course the rim joist where applicable. This is where 2# foam should be sprayed to 2"-3". Don; over do it. 2# foam can be a fire hazard when sprayed too thick in one application as it is an exothermic process. This is true also for foaming over PEX, something we generally don't recommend since it can melt the PEX if set too thick.

Some of us will market over-priced products and advocate highly energy inefficient design/installations. However, some of us are actually green builders and staunch environmentalists too:

Water Watch Annual Celebration of Oregon Rivers

Please don’t hold this against us and please be reminded that this is a green building site after all…

FWIW,I am just beginning to install my boiler etc but have all the radiant plates under the subfloor installed. I used Joist Trak heavy extruded aluminum plate based on recommendations on this site. My TECA designer took this into account. He specified R8 floor insulation which I upgraded to R13 Roxul (R8 is hard to find and R13 is cheap enough anyway). Rim joists are insulated with R20 Roxul AND 1" foil faced polyiso spray foamed around the edges for air sealing. I have yet to commission the system but think the levels of floor and rim joist insulation combined with heavy plates will perform very well. (I too am in the Vancouver lower mainland and heat loads here are relatively modest). Due to beams and other structural elements you will never get 100% plate coverage. I was ending plates 16-18" back from the rim joist in order to respect the loop bending radius and get all my rim joist insulation in. End-to-end gaps typically 4". Not sure why you are having joist bridging issues because you should still be able to slip your plates in above the "X". (I actually went with 2x10 joists and 2x8 solid blocked flush with the ceiling below. This gave me a 2" gap at the subfloor to feed the radiant components through.) Anyway, PM me if you are interested in detail, photos or just want to discuss. John7

Once your hydronic radiant floor heating system is installed and operating, it can be used to access the performance of both your building design and your hydronic radiant emitter design.

With regard to building design performance and with an indoor temperature of 70 degrees F, we consider a hydronic radiant floor surface temperature of 70-75 degree F as excellent performance (i.e., a low-load, energy efficient building), a floor surface temperature of 75-80 degree F as average performance, and a floor surface temperature greater than 80 degree F as poor performance (i.e., a high-load, energy inefficient building).

With regard to hydronic radiant emitter design performance and with an indoor temperature of 70 degrees F, we consider a 10 degree F or less temperature difference between the hydronic radiant floor heat source supply temperature and the hydronic radiant floor surface temperature as excellent performance, a 10-25 degree F temperature difference as average performance, and greater than a 25 degree F temperature difference as poor performance.

One should always keep in mind that you have to live with the performance of both your building design and your hydronic radiant emitter design forever.

The aluminum panels are up and boiler almost installed. I have been told by two plumbers that the Roxul Safe n' Sound (No R value published) + drywall our contractor was planning on placing underneath will be adequate insulation. The other solutions proposed here would involve material cost / labour so I am leaning towards either going with their recommendation of the Safe n' Sound alone, or upgrading to Roxul Comfortbatt (R-14). Would the comfortbatt also provide adequate fire & sound insulation (family upstairs, suite downstairs) and fire protection? Thoughts on this?

I like Bagers spray foam on the perimeter, clobber air and cold infiltration. Then I would use R19 itchy foil face up,
Sound and insulation in one.
Dan

Posted By Vancouver123 on 25 Nov 2014 05:45 PM
The aluminum panels are up and boiler almost installed. I have been told by two plumbers that the Roxul Safe n' Sound (No R value published) + drywall our contractor was planning on placing underneath will be adequate insulation. The other solutions proposed here would involve material cost / labour so I am leaning towards either going with their recommendation of the Safe n' Sound alone, or upgrading to Roxul Comfortbatt (R-14). Would the comfortbatt also provide adequate fire & sound insulation (family upstairs, suite downstairs) and fire protection? Thoughts on this?

Both the Roxul Safe N Sound and the Roxul Comfortbatt are made from the same non-flammable material - stone wool.  The primary difference is the Safe N Sound is only 3" thick while the Comfortbatt for 2x4 cavities is a full 3.5" thick.  I suspect the difference here is that the thinner Safe N Sound doesn't completely fill the 3.5" stud cavity which aids in uncoupling the sound vibrations on one side of the drywall from the other side.  In your case where neither of the insulation batts wil complete fill the joist space, I would opt for the thicker 3.5" Comfortbatt to get slightly more R value.

Blocking sound is a whole subject by itself and to do it well, you want something more than just insulation in the cavity. For example, hat channel + two layers of drywall with Green Glue between them.

Got your PM but couldn't figure out or get it to reply so here I am again. For the rim joists I installed R28 Roxul vertically. Then 1" foil faced polyiso vertically. The polyiso was cut 1/2" undersized and spray foamed around the edges. This gives me R34 rim joist insulation, an air & vapor barrier, and a radiant reflector (whatever that's worth). The polyethylene vapor barrier on the wall below was acoustic sealed to the bottom edge of the polyiso completing the barrier. It is important the rim joist insulation is barriered or condensation will rot out the framing. PS I am in Vancouver, BC. John

So update I called Roxul and learned (a) they don't publish the R-value for safe n'sound but it is R11, (b) the sound-dampening ability of the 3.5" R14 Comfortbatt would be the same as that of the Safe n' Sound, and thicker would be better. So I'm going to use the 5.5" R22 roxul comfortbatt insulation throughout the ceiling below the aluminum plates. I don't think there's a reason to go one step further and put Reflectix foil or something in between.

John - Comfortbatt says it "does not store or transfer moisture, and it's completely resistant to mold, mildew, rot and bacterial growth" - would that be adequate on the rim joists, or would just using that run the risk of rotting out the framing?

Just because it won't grow mold on the rock wool itself doesn't mean it will protect the rim joist from air-transported moisture adsorption. Rock wool is air & water vapor permeable, and will thus always be able to "...transfer moisture..." as water vapor from the conditioned space to the rim joists. It won't wick liquid water quickly, but it won't block water vapor. It has no "magical" properties that somehow stops water vapor, but not air.

By putting air-impermeable closed cell spray foam between the fiber & rim joist you create a condensing surface that is a higher temperature than the rim joist, and a layer that is fairly vapor retardent too. The ratio of foam-R/fiber-R that's needed to keep the amount of liquid forming at the foam/fiber interface well bounded is climate dependent, so it matters where you are.

Rock wool's inability to store or wick water isn't necessarily a good thing- it's just a characteristic of the fiber. Cellulose is extremely hygroscopic, redistributing moisture freely without damage or loss of function (until it's fairly saturated), which means it will share the moisture burden that otherwise ends up the structural wood. In really leaky assemblies where the cellulose MIGHT get saturated it might be better to go with rock wool, but making the assembly air tight on both sides and insulating cavities with cellulose offers substantially more moisture resilience than rock wool does. In assemblies with no cold wood & condensation/adsorption issues that come with cold wood rock wool is preferred due to it's rapid drying capability and it's extreme fire resistance.

We can't find anything like 2# spray foam for the rim joist. Sub-floor insulation covering aluminum heat transfer plates is dictated by the heat loads above and below and the consequent design water temperature driving the radiant floor above. R-19 batts, as Dan suggests, would be more than adequate as it is working perfectly in my own system here in Minneapolis.

Dana has educated us all on the many benefits of Roxul but it seems a bit over-kill for sub-floor insulation. Jonr has a very good point on double sheet rock as mass is the major factor in sound transmission.

Dana is spot on about the barrier issue and Roxul (or any other fiber insulation) permiability. However in case it wasn't clear my rim joist assembly is 2x10 rim joist / R28 Roxul / R6 foil faced polyiso board. The polyiso board is sealed to the joists, subfloor, and wall poly below. My condensing plane is on the condition space side so I don't have "dew point mid-wall assembly" issues. Additionally, the rim joist insulation and the general floor insulation serve two different purposes. The rim joist insulation is reducing your heat loss to the exterior (this is real heat loss). The general floor insulation (between conditioned space floors, not a slab etc) is there to limit back losses so you can better control zoning etc between floors. There is no need for barriers here. Lastly spray foam was mentioned. The spray foam I was referring to was the $6 a can "G...Stuff". I did research true 2# and the deal breaker for me was you cannot apply it to cold framing so its a summer job. I do have other concerns like encasing the mudsill with the top surface of the concrete foundation wall. Concrete will always transport some water upward. I do have the proper capillary break under the mud sill but that doesn't address the remaining 4.5" of concrete wall. Sealing that up with the mudsill is not something I am comfortable with. When applied to new construction there are likely gaps when the framing dries to MC equilibrium and shrinks 3-5%. I also didn't relish the idea of trying to spray 9x14 bays overhead with all the mechanical stuff and wall framing complicating access. For me 3 sheets of board insulation was much easier to execute. (The only cellulose they do in this region is attic blown stuff. I did want dense pack in the walls but nobody does that work here. How exactly do you use loose cellulose on a rim joist or floor?) John

thanks for all the great advice. It sounds like for rim joists, foam-board insulation (with spray foam around the edges) is a relatively simple solution. Also sounds like the order should be rim joist, then roxul, then the foam board, then joist space. I have no idea how to seal the bottom of the foam board to the drywall below, is that necessary?