All of this is making me think I'd try for 0". Press/pin the PEX/PERT into the surface of the concrete as it starts to set. Then top with thinset and tile.

Not many people want stained concrete and wood would be a real shame (efficiency wise). Net result - 85F* supply temp for a 30% saving on the heat pump electricity bill.

* - OK, I made up the 85F figure.

Lively debate,
Seems the old wisdom supports stapling to insulation, or tie to bar and mat if available and lift an inch with dobey blocks, as I understood. The loss is negligible,
I think the overall consensus is though that insulation is the key, more is better, but again depending on fuel cost there is a diminishing return.
example a wood boiler on a large parcel, zero fuel cost, 1 inch foam is reasonable. Same wit snow melt, the btu loss between 2 inch and 1 inch is not huge in that application.
I doubt the press in technique will get you far,concrete has a way of flashing.
Where the key question lies is at what point do you want to stay of the bottom, I have this unsupported number that much over 5 inches deep is a bit too deep, and then you will want to block and lift.
We have effectively placed pipe in 10 inch slabs and default to the 4" portion of slab supported with a rebar grid out door reset, and insulation.
There is some region of the north part of the country, perhaps Minnesota, or Wisconsin that has off peek electric, they require no insulation, pex pipe in a 12-24 inch sand bed, 4 inch slab on top, this creates a gigantic fly wheel, (nuke power), in my way of thinking that's just wrong. I would lay pipe in both the sand bed and the 4" slab.

Dan

I have a 6 inch slab, put the rebar/mesh grid on blocks to raise them to the center of the slab and zip stripped the PEX to the rebar/mesh. Since I was concerned about the sharp edges on the rebar/mesh I taped them all with plastic and or duct tape (dollar store stuff). It was probably an insane thing to do but I worried less about my subcontractor doing the concrete and walking on the PEX. We did glue the interior partition walls plates to the floor rather than nailing or screwing them in to the concrete. The heating system works great.

Theoretically, a suspended tube will be more responsive than a tube placed directly on the XPS, but to what end?

I see that you would like to pass this off as something theoretical, but it is a physical fact and the analysis proves it. All other things being equal, placing the tubing towards the top allows a lower supply temperature and improves system efficiency. You can argue that it is too difficult to do, but to that I would say that I did it without any trouble.

Allow me to quote John S., who I hope we can agree can't be dismissed as an "internet expert";

"In summary, tubing depth within a slab does effect thermal performance, in some cases significantly. This analysis suggests that provisions for maintaining the tubing at approximately 1/2 the slab's depth are justified and prudent."

Moreover, the engineers (who do significant radiant projects) with whom I consulted strongly concur on placement.

5-10% of "very efficient" isn't very much energy...

I guess it depends on the standards to which you want to do your work.

the vast majority of hooked mesh is, I bet, sitting right back on the bottom of the slab.

A difficult statement to prove or disprove but I've demo'ed several slabs in which the mesh is right where it is supposed to be - in the middle. Furthermore, with the slabs I've participated in pouring, once the mesh is hooked and lifted, I haven't seen any tendency for it to "sink" to the bottom. A few bucks worth of chairs can pretty much assure that, however.

Posted By Blueridgecompany.com on 19 Aug 2012 06:00 PM
Lively debate,
Seems the old wisdom supports stapling to insulation, or tie to bar and mat if available and lift an inch with dobey blocks, as I understood. The loss is negligible,
I think the overall consensus is though that insulation is the key, more is better, but again depending on fuel cost there is a diminishing return.
example a wood boiler on a large parcel, zero fuel cost, 1 inch foam is reasonable. Same wit snow melt, the btu loss between 2 inch and 1 inch is not huge in that application.
I doubt the press in technique will get you far,concrete has a way of flashing.
Where the key question lies is at what point do you want to stay of the bottom, I have this unsupported number that much over 5 inches deep is a bit too deep, and then you will want to block and lift.
We have effectively placed pipe in 10 inch slabs and default to the 4" portion of slab supported with a rebar grid out door reset, and insulation.
There is some region of the north part of the country, perhaps Minnesota, or Wisconsin that has off peek electric, they require no insulation, pex pipe in a 12-24 inch sand bed, 4 inch slab on top, this creates a gigantic fly wheel, (nuke power), in my way of thinking that's just wrong. I would lay pipe in both the sand bed and the 4" slab.

Dan

Would you really want the extra work of cutting more wood, even if it is free, just to save a few bucks on insulation. False economy. Insulation is king and the whole idea of efficiency is supposed to be cost independent (as an ideal, anyway)

to jonr: you can get an 85 degree supply temp. drop the finish floor and do stained concrete in a typical load with 12" o.c. pipe on the bottom of the slab. You'll get there. uncovered slabs really don't care about where the pipe is. remember the chart in this case is for slabs with wood on top. Only as upward resistance increases is this even a debatably interesting factor. it's a complete waste of time to consider it in a naked slab application.

as for ICFhybrid,

-If you can get a slab pour to chair mesh without rebar, you are a better man than I. I tried for ten years to get people to do that and they simply refuse. If you have rebar, then once again I agree that tying to lifted rebar probably makes sense in most cases.

-I already dissected the energy efficiency gain to be had. You can cherry pick one statement I made out of context to rebut, but the fact is even that 5%-10%, which is NOT a seasonal average gain, is REAL efficiency gain (on the real load) of about 1% to 5% if you have a COP of 2 to 4... the 5% to 10% is on the already discounted load of the heat pump. jonr's assessment of the efficiency benefit may have been slightly low but still, once you account for the COP reduction it really knocks down the improvement you're talking about. If you got that consistently across the whole season I would pursue it, but you don't if you're running a reset curve. you'll be at energy savings too small to take seriously in that case. If you're NOT running a reset curve, the impact of that is far worse than tubing position in the slab, and the forest was lost for the trees.

If you're rebar'ing already, lift the pipe.
If you're not doing wood floors over the slab, dont bother even thinking about this.
If you are doing wood floors over the slab, but you're not already doing rebar, the savings will never pay for the rebar.
If your slab guy is happy to raise mesh on chairs without rebar... refer him to all your friends and post his info on this board cause he's a rare guy.

I know Siggy personally have attended his seminars and use his software.

You will note that your quote (missing context) does not specify the thickness of the slab. Further, John qualifies the statement by noting "in some cases" and "the analysis suggests". The latter statement confirms that this is in fact, theory. The probabilities and dynamic nature of weather and building construction do not allow for certainties in the performance of any radiant panel. It takes field experience for to make reliable predictions. I am talking about the most common 4" slab here. The money you would spend (not yours I presume) to elevate the PEX tube 1.5" off the bottom is simply a waste of time and money, better spent on insulation or more effectively on tighter tube spacing, which will also lower response time, design water temperature and raise the average surface temperature of any slab.

In a 4" slab, the PEX goes on the XPS insulation, the wire or rod goes on the PEX (the only place it will have any effect) and the concrete overall. Raising the tube in the typical 4" slab would have the same effect as "Red Greening" every inch that touches the already redundant rebar...that is, very little.

You will note that your quote (missing context) ...

My quote is complete and contextual. It is an entire paragraph and is not edited. This "theory" is supported by direct modeling.

If you guys have some experimental setups which can pass scientific scrutiny that say otherwise, I'd be very happy to see them. However, all I see here is a bunch of wishful thinking propagated by guys who don't want to be bothered with pesky details. Just lift the mesh to center. It will stay there unless someone tramps it back down. If you're worried about that, throw a few bucks worth of chairs under there. Either way, it's not difficult at all.

a waste of time and money, better spent on insulation or more effectively on tighter tube spacing,

I'm sorry I have to say this again. You are free to change spacing, tube size, water temperature, whatever you want, but the tubing depth variable stands on it's own. Maybe you didn't know that chairs cost a penny or two per square foot. It would be hard to spend that on something else and get the same return and the fact still stands that you could optimize all other things and there would still be no excuse for not putting the tubing in the top half of the slab.

In a 4" slab, the PEX goes on the XPS insulation

That is the least efficient place top put it. I'm sure it is easy for you, however, and it is likely that the customer will never know the difference between what they got and what they could have had.

If you're not doing wood floors over the slab, dont bother even thinking about this.

Rob, the data presented by Siegenthaler precisely refers to uncovered slabs. It is the situation in which the effect is the greatest. If you cover the slab, the results may actually trend back towards something else.

you misread the data. I posted a link to the analysis. it was specifically for a slab with an engineered wood floor over it. this is obvious by the water temperatures noted as well which are way higher than any uncovered slab in any condition would require.. you can cook people's feet with 110 degree water or less with no floor covering on it.... the water temp differential even possible to achieve in an uncovered slab is very small indeed.

I have only ever referred to Siegenthaler.. Specifically, his text, Third Edition. I am not aware of him making any dramatic reversals on this subject, however.

He has some pictorial representations which depict the same effects with a covered slab, but the numerical data presented (in terms of actual heat output) are for bare, uncovered slabs.

"The FEA models were also run for 4-inch-thick bare concrete slabs....." The data in the referenced Figure is for 100F water, which he typically uses in his illustrations, not 110F.

ah, we're talking about different analysis. the one I referenced in the link is for a wood floor. for the one in modern hydronic heating, all I can say is 120 degree F water in a bare slab is hot enough to cause major discomfort if run continuously, so I don't know what inputs he's using in his FEA for that but it contradicts the output charts we use for slab on grade output by a fairly wide margin. I have a bare slab downstairs with tube on the bottom doing 10 BTU/sq ft at design with 85 degree water just fine.

OK, you have convinced me that this is, even with a HP, a ~3% thing and that reducing tube spacing a little would also accomplish that. Outdoor reset and a buffer tank are more important in minimizing delta T.

Actually, John S. likes to play with the round figure of 10% and my calculations show in the neighborhood of at least 8% which was similar to what the energy engineers were yakking about using different software.

AND, I have to point out again, that all other things being equal, moving the tubing AWAY from the insulation and towards the radiant surface still has an effect.

I have a bare slab downstairs with tube on the bottom doing 10 BTU/sq ft at design with 85 degree water just fine.

That's such a round number for an engineer. Are you sure it's not 9.2 BTU/sq. ft on a not-so exactly design day? ;-)

I have to point out again, this is not about what is achievable, it is about the efficiency of getting there.

you're ignoring the fact that it's 10% on a load already reduced by 50% to 75%. it may be a 10% improvement from there, but that's misleading regarding the impact of the decision. and, once again, that's only design day.

I'm an engineering type that works in round numbers because I learned long ago that overprecision leads people (most of all, the ones calculating the numbers) to believe things are more precise than they actually are!

I have no idea what your last statement is supposed to mean. my numbers, real world, contradict that FEA. I am getting performance similar to his "lifted" numbers, when corrected for likely room temperature variance, with tubing stapled to foam in a 4" bare slab. over 2" foam, over solid ledge, for whatever that's worth.

I don't think anyone will argue how raising the tube in a slab affects output. The question really is, how much. The effect is simply one of slightly raising the average surface temperature dependent on AUST, exactly the same result attributed to using closer tube spacing. After surface temperature, thermal mass must be considered, but modern construction precludes severe swings in indoor temperature. Response time is irrelevant.

When you are designing radiant floors for properly insulated buildings and design water temperatures below 80°F, tube depth simply doesn't merit special attention. Time is money, unless you have unlimited supplies of both.

you're ignoring the fact that it's 10% on a load already reduced by 50% to 75%

No, I am not. You'd have to show me how I am "ignoring" that. 10% ...is 10%.

I am getting performance similar to his "lifted" numbers

Okay...excellent for you. I am making no comment whatsoever on that. However, there is no doubt in my mind that you would be getting even "better" numbers if you had placed tubing in the top half of the slab.

you can't really improve a noticeable amount from that. I mean seriously, come on. The heat pump I'm using doesn't even have a lower temperature setting, I'd have to mix down to go any lower.

10% is not 10% when it's 2%. if you want to actually use integrity, you'd have to specify 10% on a heat pump at design day load. it's certainly not a 10% seasonal improvement. and It's not 10% of an order of magnitude to be expected by anyone who is familiar with non-heat-pump equipment. Just saying it's 10% falls in the category of technically correct and totally misleading.

you can't really improve a noticeable amount from that. I mean seriously, come on.

Of course. My mistake. I didn't realize you had achieved 100% efficiency.

no, your mistake is in your characterization of 'a noticeable amount'. If you believe you could do anything to NOTICEABLY improve my situation, I'm all ears. but I'm data logging, and I could tell you that I'd be damn hard pressed to notice any likely difference from an "improved" scenario even so... partly because I have a low load, as I noted, but also because real world this tubing on the bottom deal is not nearly as big of a deal as some numbers can make it sound. Whatever difference lifting would have made to us would never have paid for the rebar.... no question. Secondarily, no one around here will lift mesh alone with chairs. third, I don't support hooking mesh with pipe on it... I've seen that go wrong. it sucks. who gets the blame in that case?

You're welcome to be a purist, but I'm not going to sit around and let it fly about like it's a serious consideration when, as a point of fact, it's not. Even the numbers being given say it's a small effect when you finish analyzing them, and the numbers in the real world say it's even smaller than that. Siggy is great, and I think he has drawn a conclusion here that is a bit wide of the mark. It's not the first time, and it won't be the last, because we're all fallible. When observed reality doesn't match the models, I know which side I fall on.

If it can be lifted easily and safely, do it. If it can't, don't lose sleep over it.