First off, I'm new on these boards and came here to learn from the experience of others. I'm GC'ing my own house next year in Durango, CO. It will be a 2600SF Timber Frame home, covered w/ SIP panels, and built using an ICF stem wall & 4" slab first floor. First floor is 1900SF. My question concerns the installation of hydronic radiant heating in the first floor slab. I know there are multiple schools of thought on what level within the slab that the PEX tubing runs should be located. I see posters talking about stapling the tubing to the XPS board underneath the slab, with re-mesh & slab above, effectively putting the heat at the bottom the slab. I also see posts advocating tying it to the re-mesh so that the tubing is raised into the center of the slab along with the re-mesh when the concrete is poured. I have also received multiple opinions from contractors I have spoken with. What are people's opinions on the board as to which way they prefer & why? Thanks in advance.....JKZ

When I am looking at pipe depth in slab I am looking more at application and thickness of slab and insulation.
Points to consider. assume 4 inch slab, height is not going to make much difference with proper insulation, 2 inch rigid.
You now have options, staple to foam, tie to bar, tie to wire 6x6 mat.
If you are not required to have rebar or wire mesh then stapling 1/2 pex is easy, fast effective.
If you are required to have mat, use road mat 7.5'x20' flat sheets. Now you have to be mind full, each cut wire should be seen as a knife or sharp edge that will damage your pipe when the flat worker steps on it un knowingly while placing the mud.
So pay attention, tie pipe so it will not be able to slide in to the sharp point. easy. now you can lift the mat an inch or so.
You can also place the mat on top of the pipe, but again, sharp edges have to be observed.
Tie to rebar is the same.
If you are doing a 6 inch slab I would lift system to the 2-3 inch level over insulation.

Dan

Put it in the top 2" of the slab for efficiency of heat transfer. Put it lower and you are just wasting energy.
I like #3 bar in the slabs and you can just ziptie it to the rebar grid. I shoot to get the rebar in the center of the slab and try to tie the PEX on the top of the bar.

Morgan, Rob, Dana,
Do any of you have the actual data on gain regarding slab location, As I recall, the gain on slab pex pipe location was marginal where the pipe is placed on a 4 inch slab but I am not certain,
ICF do you have the actual numbers?
Thanks,
Dan

Posted By ICFHybrid on 19 Aug 2012 01:35 AM
Put it in the top 2" of the slab for efficiency of heat transfer. Put it lower and you are just wasting energy.
I like #3 bar in the slabs and you can just ziptie it to the rebar grid. I shoot to get the rebar in the center of the slab and try to tie the PEX on the top of the bar.

I can't see any use in keeping the tubing close to the top of the slab EXCEPT for slight increase in response time during big temp changes. Of course, slab heat systems are supposed to be steady state heating systems, as much as possible.

The heat is going to stay within the slab regardless when there is good insulation levels and especially edge insulation so it will become pretty even throughout the 4" slab over time regardless of the height. I have taken many pics with my Flir camera over the years ( or seen others) where there is  a higher temp variation for tubing closer to the surface then at the bottom of the slab.It is a more even slab temp underfoot.

Also, when pouring a slab before interior walls are installed I feel a lot happier if I know the tubing is down at the bottom of the slab. Those pesky Ramset nails can bite you in the butt. If the tubing is close to the surface, it is better to use adhesive.

It is different when the job is a suspended slab or commercial work where we fasten EMT and other things to the ceiling. Attach it to the rebar near the middle if possible. I guess that is old school now with all the steel pan forming available.

Steel is about 40x better at conducting heat than concrete. So I would attach the tubing on top of the mesh/rebar to get better distribution and transfer. But I have no data to say that the effect is significant. This homeowner reference says shallow is critical:

http://inspectapedia.com/heat/Radiant-Slab-Heat-Mistakes.htm

Thx for comments so far. As before, there still appear to be differing opinions on this question.

JKZ

ICF do you have the actual numbers?

Yes, I do. I modelled it using finite element analysis. But you don't have to try and read my data. John Siegenthaler did the same and turned it into lots of pretty pictures which he published in his book "Modern Hydronic Heating".

The heat is going to stay within the slab regardless when there is good insulation levels and especially edge insulation so it will become pretty even throughout the 4" slab over time regardless of the height.

That is a generic statement which tends to true if there is a LOT of insulation. However, reality often is that there is LESS than that and the placement makes a difference even with code-required levels of underslab insulation. In heat transfer, gradients exist everywhere. You may not "think" it makes a difference, but it does. I have seen some horrendous high-end heating systems in which the heating guy didn't "think" that underslab insulation was important or that it was okay to bury the tubes right down by the ground.

It is a more even slab temp underfoot.

Frankly, I am pretty tired of hearing that argument. Even heating "underfoot" can be achieved by manipulating other variables as well and we all know it is easy to achieve "comfort" at the expense of "efficiency".

Those pesky Ramset nails can bite you in the butt.

It takes about 10 minutes for the tubing guy to mark safe avenues in green paint on the wall and saves hours of boo-boos later. Outside the green you have to use adhesive.

The heat ... will become pretty even throughout the 4" slab over time regardless of the height.

I hope not. The top side is radiating/convecting heat into the building. The bottom is (ideally) not losing any heat. So the top side must be cooler. Thinking of concrete as insulation (albeit poor) helps.

Siggy did detail this in article and such. the upshot is, in a typical 4" slab situation, there is a small increase in downward heat loss with tubing at the bottom. It's pretty small, but from a purist standpoint lifting the pipe has positive effects in that regard. there is also a very small boost in output. Neither effect would be considered significant or noticeable by most people, even people doing math.

lifting the top 2" is a potential for snafu too. if the MESH is 2" lifted, the the pipe, assuming 1/2" nominal 5/8" OD, is only 1-3/8" from the surface. control cuts are usually 1/3rd of slab depth, or 1-3/9" (1.33") deep. not much margin for error there. You can do shallower cuts, of course. but I've seen tubing lifted too far and cut. not pretty.

thicker than a 4" slab and you should lift off the bottom at least. I don't really think lifting to anywhere near control cut depth is a good idea though. and even when mesh is present I am skeptical that lifting really does much. If you really want a lift to be sure, you have to lift on rebar. If you put the rebar OVER the pipe and lift it, you have a very safe and predictable lift. that probably looks like mesh > pipe> rebar on top.

and who does that? no one, really.

As a heating contractor, I HATE going back and digging up and repairing a leak that a framer caused or when the owner changes the interior wall location (which happens often enough) so it IS ONE acceptable reason for keeping the tubing lower down. I add in these things to keep the potential problems to a minimum.

I don't like having to talk about where the heat is greater or lesser but i used to get it often enough and I had to explain it to homeowners. It is simple enough, if the tube is lower down the temp will be more even on the surface.

I won't do a radiant slab with any less than 2" of XPS under and around it so I start with the assumption that I can define the surface temp to a reasonable extent. I don't screw the tubing to the XPS. I always use mesh so in general the tubing will never sit on the XPS. It is a moot point because if there is no place for the heat to go other than up, it will go up. I missed a comma above and it sounds like I believe you don't need insulation which isn't true. If there is reasonable insulation, the slab will be a nearly homogeneous body of heat. Yes there are gradients but they are minimal.

the slab will be a nearly homogeneous body of heat.

I'm sorry, but that is NOT what the data says.

if there is no place for the heat to go other than up, it will go up

Heat goes through insulation, too, just at a different rate. This notion, that if you insulate it you don't have to worry about it any more, is a step up, but it is not the last word in the analysis of these systems. Look very carefully for the data that shows what the supply temp has to be in order to deliver a particular rate of heating from the bottom of a 4" slab and from the middle of a 4" slab.

Neither effect would be considered significant or noticeable by most people, even people doing math.

I'm afraid that's exactly what math does do; it separates the facts from the opinions.

I've seen tubing lifted too far and cut. not pretty.

That's why I like rebar inserted into a precise end location and held up by dobies of the appropriate size. Wouldn't want to bet the system integrity on some arbitrary lifting factor. I would also point out that control cuts don't have to be 1/3 the depth of the slab, either.

It is simple enough, if the tube is lower down the temp will be more even on the surface.

And reaction time will be slower, heat loss out the bottom will be higher, delta T (which decreases efficiency) will be higher, # of nail punctures will be lower.

ICF: I've seen and done the math. it's not a large enough effect to even get past rounding errors in a design in most cases. I don't think it's worth worrying about. Siggy's numbers on this don't change the facts.

if you're doing rebar anyway, then fine, you have a lifting method that is sure and potentially safe. if the tubing is under the rebar, you're in perfect mode. most people aren't. but even with rebar, no one is ever going to have noticeable differences in efficiency by worrying about this detail if the bottom of the slab is properly insulated and the slab is not ridiculously thick.

I've seen and done the math. it's not a large enough effect to even get past rounding errors in a design in most cases.

Then you have seen the differences in supply temp for tubing located at the bottom vs the middle of a slab. Supply temperatures can and do contribute significantly to the efficiency of a system.

I don't have Siggys book. Could someone post a graph? I find it hard to believe there is that much difference between top and bottom of a 4" slab WHEN the system is designed for optimal liquid temps. ie: 21C ambient and 28-30C water temp. Slab should only be 25-27C. Remember this is only at design temps as well.

Maybe if it is an inefficient system and you need 40C+ to meet the design temp but I would like to see a graph.

for a typical load, you might see a 5 degree rise in design water temp. maybe. and that's on design day, so on average you are literally talking about maybe 2 or 3 degrees... well inside the margin of error for anyone running math on these systems.

if you're in a higher load situation, you can and should compress the pipe spacing and get nearly the same improvement as moving the tubing up (compare lower depth peaks to raised depth valleys and note how close the two are) http://www.pmmag.com/Articles/Column/31ed5e1f0bfc7010VgnVCM100000f932a8c0____&sa=U&ei=_xsxUKH5G6Pr6gGZsYHYDQ&ved=0CBQQFjAA&sig2=AJeuoHcZ8njkGkA0NZX3Ng&usg=AFQjCNFgpayLYJcPl5CaZlmjSduRWXlsug

and this all presumes wood floor on slab. on true slab? how often do you see that? skip the wood floor and this conversation becomes completely moot... your supply water temps in an uncovered slab will rarely break 100 degrees no matter what your load is or where the tubing is placed.

Again, if you are already doing rebar, sure, lift it. I would advocate for rebar on top if possible, but the controllability of a rebar lift is pretty good with chairs so it's not required. But I would not add rebar or mesh to make this happen nor would I even consider tubing depth in a design calculation except maybe in a high load, high-r floor covering situation.

OK, take a heat pump producing 100F from 50F well water vs a heat pump producing 95F from 50F. About a 10% difference in heat pump efficiency (and the cost to run it), right?

Rob is right;

Output and performance of any radiant panel is mainly dependent on the temperature differential of the slab, the heat transfer fluid, the AUST, and convective currents. There is no good argument, theoretical or practical, for suspending a PEX tube in a 4" slab. The typical slab-on-grade building will use a small percentage of the potential output of any give slab during design conditions. The evidence of this is proven in our driveway snow melting systems here in Minneapolis, where typical outputs are in the 100 to 200 btu/sq.ft. range. Yes, the tube is on the XPS and wire over all, is accepted practice. On the other hand, as previously noted, placing the tube off the bottom, certainly invites saw cut disasters and when one considers the specific heat of concrete (0.18) the argument of suspending tube weakens quickly.

Theoretically, a suspended tube will be more responsive than a tube placed directly on the XPS, but to what end? Response is also depend on load. Low load, low response time, no worries. The potential output argument is the same. If the thermostat is satisfied and the people comfortable, the response time is perfect, even if it is not "optimized" by perfect tube placement. Well insulated buildings protect the interior from severe weather change and the required response time is diminished. This is most particularly true if outdoor reset is employed.

Suspending tube in a 4" slab is a poor use of manpower and certainly qualifies for the law of diminishing returns. It does make for good grist if you are hammering out your "thesis" or trying to impress others with your singular grasp of radiant panels, but overall it simply distracts the layman from more important aspects of good radiant heating practice.

Funny, as I look back through all of Ziggy's published work, I can find precious few of his designs, or pictures of installed radiant panels, sitting on wire chairs, but still have to smile when "internet experts" challenge decades of installation experience with a perversion of an otherwise instructive and occasionally useful observation. I just tell my customers that suspending PEX in any slab is for experienced professionals and is certainly not the standard or accepted practice in the field.

If you run the heat pump at a fixed temperature, you already did FAR worse to your efficiency than we are talking about here. You're probably about right on the impact in this case jonr.

but problems with that:

1. on a reset curve, which no one should be running heat pumps without, you only get that much improvement on a small amount of the yearly BTUs... because most of your curve will be closer. as I said, the average difference, not the peak difference, will only be 2 or 3 degrees. they start at the same temp, and diverge more as it gets colder. so now you're at a theoretically seasonally adjusted max of MAYBE 5%. maybe. for a heat pump.

2. 5-10% of "very efficient" isn't very much energy. heat pumps are already very efficient. small adjustments to efficiency are not very noticeable... half to a quarter as noticeable as with regular COP1 equipment. 10% might normally be noticeable. with a heat pump, not. and a heat pump is the only thing that would care about water temp changes at this temp range of this size... at all.

3. you're also presuming that normally reset curves are dialed into exact perfection. the vast majority are not... you'll be within five to ten degrees of reality, maybe, but unless you are a homeowner who is aggressively tweaking your curve to match reality... or, unless you have an indoor feedback control... most people would never ever actually hit "reset perfection".

again: if you have rebar going in, great. If not, the energy savings will never pay for the rebar, and you are never going to get mesh only pex to a midpoint. I smile at badger's characterizations. I campaigned for raising pex for years. until I finally watched a slab pour in progress. without rebar and chairs, this discussion is moot. hooking is not going to position pex with any reliability and the vast majority of hooked mesh is, I bet, sitting right back on the bottom of the slab.