Going to be finishing up my basment and before i drywall the ceiling i would like to run some pex... now i see that some people stick the tube right to the floor and run that aluminum tin over it stapled to the underside of the sub floor.  and some people use other systems like www.ultra-fin.com

What is the best way to do it? is there other system out there?  is this under subfloor system work that well, or is it not worth it?  The flooring in the house is solid walnut and the bathrooms and kitchen are slate.

heavy gauge aluminum plates are the "best" way to do it. Thermofin, Joist Trak, Climate Trak, etc. They work better than a lot of overfloor panels.

plateless, light plate, and ultrafin systems should only be used in VERY low heat load conditions. Otherwise very high water temperature requirements result, and limited output.

Ultrafin may outpeform suspended-tube, but it dramatically underperforms even UN-plated PEX staple-ups where the PEX is in contact with the subfloor at 110F or lower water temps, and barely keeps up with unplated PEX staple ups at 120F & up water temps.

It still underperforms ANY plated system, but it underperforms Rob's recommended extruded plate systems by half at temps below 130F!

See the comparative graphs in section 5-2 starting on page 60 (p75 of the PDF pagination) of this document:

http://www.chple.arch.vt.edu/CHPLE%20Research%20files/Khanna,%20Amit.pdf

(Compare, Ultrafin / Joist Space without Plates / Joist Space with Plates for the comparative differences.)

Given that it underperforms a contact staple-up, I can't imagine Ultafin being worth much whenever there's the possiblity of a full contact staple-up. If the subfloor looks like a Fakir's nap couch, and grinding 8000 nails isn't in the cards it might make some sense. If you have a condensing boiler you'll get much better condensing performance out of a fully plated system, as pricey up-front as that may be. Short of ripping up floor for a best-in-class above the subfloor system, extruded (not sheet-metal) plates are the best. Performance will depend on the thickness & type of subfloor & flooring (slate==very good, walnut== pretty-good).

Bare-tube, full-contact, staple-up will make noise and cause striping.

Plates are it.

careful linking that study Dana. the graphs are way off in many respects. check the average floor surface temp data and you may end up scratching your head about where the graphs come from... the floor surface data *seems* good but the conclusions/BTU measurement was not.

Posted By NRT.Rob on 14 Dec 2010 09:29 AM
careful linking that study Dana. the graphs are way off in many respects. check the average floor surface temp data and you may end up scratching your head about where the graphs come from... the floor surface data *seems* good but the conclusions/BTU measurement was not.

It was his thesis project- I wonder if he was actually awarded a degree?

To be sure it's not as good as other academic work on the subject, but since it directly compared Ultrafin to other methods I threw it out there without fully reading & analyzing it. (Thanks for the professional eye!)

FWIW: I first saw that thesis when it was linked to by docjenser posting on the geo forum of this site, after I expressed skepticism on his claim that he was getting 25BTU/ft² @ 120F water with Onix staple ups, through gappy plank subflooring + hardwood. (Not that the paper particularly supported his claims.) My best-guesstimate would be that something on the order of half that would be the limit of what's possible, yet he seemed to think that the gaps in the subflooring brought the average R value of the floor down to R0.25-R0.5 or something (not that the Onix could conduct through the air gap, IMHO). I'd assume even having been sanded a few times in the last 80 years the stackup would still be R1-ish or higher. I'm still skeptical.

At the risk of hijacking this thread, what's your take?  Is he in dreamland, or am I missing something?

it was basically paid for by rehau, and the purpose was to evaluate the evaluation methods for radiant floors, not to generate output charts. I believe the end of the report notes some of the lessons learned which was the real purpose of the study.

I regard Onyx as basically equivalent to a light plate pex system. Basically. If you also account for a good reflective barrier properly installed underneath... which I don't, at least not forever... but it would undeniably do its job for a few years anyway. which would mean, in scientific terms. that your claimant is high as a kite.

25 BTUs/sq ft at 120 supply with R1 on top is only achievable with slab or warmboard/raupanel/rothpanel type installs.

Nevermind the black sludge issue. I dearly wish people would stop installing that junk.

The source of my skepticism was on 2 levels- no way can the gaps in the floor reduce it's conductive R value that much, and if you do the simple-math on the condact area direct conduction through R1 isn't even getting you halfway there, assuming 3/4" wide strips of contact. That would require the non-contact surfaces of the tubing to be able to deliver at least as much heat to the cavity through radiation & convection as the contact-area portion.

Even WITH a reflective surface below and very high emissivity on the EPDM tubing, the delta-T is too low for the a very large radiant transfer with only 120F strips of tubing, and a 75-80F subfloor. I'm sure the RB does something- something even measurable, but it wouldn't likely be even half the heat transfer the direct contact area is delivering, let alone at-par with it.

docjenser seems enamored of Onix (has it in his own house in western NY) but the track record of EPDM tubing in radiant heating apps is indeed pretty sketchy.

well, the water temps are typically much higher, as they are with light plates, so the reflective insulation they specify probably does something. at 120 you aren't doing much with either onix or light joist plates.

So the best way is heavy extruded plates then.. Right? I do have some nails and it would be a bit of a pain to get them all flat. Working around all the vents and wiring grinding down the nails does not sound fun. Any construction adhesive that would work well and not make a thermal break?

Is insulation a must as well? I have these webbed joists that are like 20” or so tall. They are all 2x4 they look like this. it would be tricky to insulate them well..



http://www.timberframe.co.uk/wp-con...=300&q=100

insulation is a must.

I don't know what you think you're going to do with construction adhesive that will make a plate go through a nail? There are new graphite plates you could use that allow you to almost totally skip nail clipping... but they are pricey.

i was thinking the adhesive would make full contact with the wood... the nails that are in the way are sticking out less that 1/8"

the nails are from the hardwood install.. and the plates will not sit flat under the sub floor

grinder with a cutting wheel attachment is the typical solution. if you have a really low heat load then things like suspended tube may make sense but you'd need to seriously run some numbers to know.

I came to the same conclusion, but it is instructive for those familiar with the various emitters.

Conceivably an air nipper would cut nails off:

http://www.vesseltools.com/nippern.html

Posted By NRT.Rob on 14 Dec 2010 03:09 PM
well, the water temps are typically much higher, as they are with light plates, so the reflective insulation they specify probably does something. at 120 you aren't doing much with either onix or light joist plates.

Yet, 120F was the number he was using (being geo-sourced, and all...)

With an average water temp of 125F the light-plated zone in my house is still doing something (through ~ 2.25" of subfloor & hardwood no less), but it's nothing like 25BTU/ft.  (Fortunately it doesn't need anything like 25BTU/ft, but it could use a little bit more than it's getting when it gets below 10F, but I'm loathe to crank the temp up.)

Dry radiant floor systems allow even heating throughout the entire floor. The heat radiates from the floor and warms objects near the floor as opposed to forced hot air that tends to rise to the ceiling. Radiant floors eliminate dust, draft, and noise problems associated with forced air systems. They can be more aesthetically pleasing than other forms of heating because there are no heat registers or radiators to obstruct interior designs. Manufacturers claim radiant floor heating saves 20 to 40 percent on monthly heating bills.

Posted By greencleaning on 19 Dec 2010 09:57 PM
Radiant floors eliminate dust, draft, and noise problems associated with forced air systems.

Good points, but doesn't a forced air system with good filters (like HEPA filters) actually help reduce the overall dust?