How, exactly, does one accurately calculate the heat loss out of a slab-on-grade? Is there a pretty accurate way to determine your heating degree days by tweaking the number you use for above grade calcs? I know that the bulk of the heat loss is on the perimeter, but I don't have any equations to crank. I've seen a reference for "R/linear foot", just for this reason, but I can't find how to calc that R for my particular wall/slab insulation scheme. Any ideas? Thanks. john

The calculations that I've seen(and used for writing my own HL program) is the temperature of the earth. This temperature varies depending on location, and is the ~stable temperature below the frost line. So, if the ground temperature at your site is 60°, and your design temperature is 70°, you'd have a Delta T of 10°. On a 1,000 sqft area the loss would be: 1/R-Value x Delta T(10°) x area. This formula produces some crazy high values if you use an R-Value of 1, and you can't use one of zero.

That is not even remotely close to how the heat loss from a slab on grade is calculated.

Your vertical slab edges are exposed to ambient air. They are a major source of heat loss.

Your perimeter is exposed to temperatures much cooler than ambient ground temperature under the frost line. This is vaguely proportional to the ambient outdoor temperature as well, as the cooling effect "reaches" under the home..

John seigenthaler has the formulae you need in Modern Hydronic Heating (basically, the bible according to wetheads).

Thanks. I've read John's name before; will look for his book. j

You have to give him create for creativity! heehehe

I use Wrightsoft (Manual 'J' 8 ) for my design work, but you can do your own with SlantFin's calculator, rough but better than nothing (or making things up).

http://www.slantfin.com/heat-loss-software-get.html

Badger: Thanks. The Slantfin is for P(ieces of) C(heese) only. I have tried the calc'r at greenbuilding (and others) but it gave me more slab heat loss than my own spread sheet. Surprising. On everything else we had real good agreement. Re: "You have to give him create for creativity!" Who? John S?

Interesting. You'll need to inform the Pikes Building Department of your opinion. That is the formula that they use and they require a Heat Loss for their permits.

Rob is an RPA certified designer and uncharacteristically blunt in this case (that is my job).

Since ground temperature in the cold states varies little, it is not the major factor in heat load analysis. Outdoor (air) design temperature is first, followed by construction of windows, walls and total surface area. These are all covered in the decades old (and 'almost' universally accepted) Manual 'J' programs.

If you can't do the calculations yourself designers like Rob and I will do it for a fee.

Most competent contractors have access to heat load software (through their local hydronic distributors). If you don't have an accurate heat load, you literally "don't know what your doing".

Of all the things a DIYer (or prudent homeowner) should farm out to others, this is number one.

To get back to your original question, the perimeter is where the majority of heat will be lost (from a heated slab) and a critical area for proper insulation (determined by design conditions for your area). So you will cap the edge of the slab and install insulation in, down or even out several feet to the frost line (again, depending on your particular design conditions).

I should add for Rob's sake that the entire slab should be insulated to lower heat loss and aid response time.

Badger: Roger that. "Typical" up here is to use 2" of XPS under the slab, down 4' on the outside of the foundation wall, and then out horizontally 2' at the footer. Some people even put 2" on the inside of the foundation wall. I will likely use 3". Another slab option that some of the engineer-types have used is to go 4" thick, 4' to 8' in from the perimeter, 2" on the interior. I'd like to get ahold of the method for calc'ing slab losses so that I can play w/ various scenarios and look at pay-back times. At 75 cents/board foot for XPS, one does not want to get TOO crazy w/ it, but I am certain that 2" is a minimum. I've talked w/ a gent at the Cold Climate Research Center and they don't yet have a solid handle on calc'ing slab losses, either. The info is out there; it just needs to be dug up. This is always a crap shoot, as we can only guess what we will be paying for btu's in 20 yrs. By then we may have nuclear energy in our homes, or it will be so warm we won't need much R-value. Thanks again for the replies. john

PanelCrafters: BTW, thanks. Q = U A (delta T) gives you the hourly heat loss. You use your "heating degree days" (and other tweaks) instead of delta T to get annual losses. Our degree days number is 14,720, though with global warming I wonder if that needs updating. That works for above grade calcs, but is not entirely accurate for slab heat loss, because slab heat loss is not distributed uniformly over the slab, obviously. When it is -50 outside the perimeter of the slab, it sure is not -50 under the middle, so not as many btu's are pouring through the middle. What I envision, roughly, is a calc using your regular degree days for the peripheral 2' (4' ?) of the slab, then a different degree days for the rest. Surely there is some integratible function that gets a more accurate number, but for those of us who forget our calculus, I suspect there is a "good nuff" method that only involves arithmetic and breaks the slab into discrete chunks.

jk, are you trying to calc heat loss through a slab with only the air heated above, or with radiant heating in the slab? If the former, take a look at this long winded conversation I had with NRT.Rob last Feb. http://www.greenbuildingtalk.com/Forums/tabid/53/view/topic/forumid/12/postid/32800/Default.aspx
Basically, if you are doing space heating only above the slab, and not radiant in the slab, then anything greater than 3/4" for XPS thickness needs to looked at real hard, especially if you're not in a bitter winter cold climate like northern Minnesota. In the discussion cited, I was looking at payback for insulation under the rat slab in my crawl space with an air temp of 80F. I use the crawl space as my supply plenum.

Dmaceld: Thanks for the reply. I am looking at heat loss through a slab-on-grade, heated via in-floor radiant, though I don't see how that makes any difference. Pls explain, if brief; if it is entwined in said discussion you mentioned above, I'll read that soon. If you put btu's into a house, they run to cold in any direction, so I don't see any (significant) reason why it matters if the btu got there via baseboards or pex in the slab. Not to argue, but I just don't see it and am open to education. Re: 3/4" XPS. Man, that doesn't even register up here! I hear/have heard for decades "don't over-insulate", but I am not sure you really can (within broad limits, of course). A lot of people still think a 6" wall w/ 2" of R-tec over that is plenty, and they are only a bit off, IMHO. That is an R27 wall (if caution is taken to tape the R-tec seams, etc), and not bad, but I will be building R-40 or 41 walls. As for under the slab, see above; probably 3" of XPS. One thing that I think people do not take into account, too, is that a slab is draining btu's 24x365, though certainly not at the rate the rest of the house sees in winter. Yes, bitter cold here, like N Minn, only worse. Fairbanks. After a near-record cold snap we had an 80+ degree jump in two days and it hung at 40 ABOVE for a few days. Ice city now. My present walls, since 1980, are R39 and I have zero complaints about the "over kill" I did back then. Now, "super insulated" has caught on. BTW: I just read on an engineering site about slab loss calcs and think I have an approx method of doing so. I'll check w/ someone there and see if I assimilated the info correctly. Again, thanks all for the input/discussion. john

Panel Crafters: Pls shoot me a pm. I can not get the dang pm'ing to work on my end. thanks. j

Clicking on the envelope icon has been the only way to get PM to work in the past. PM via the poster's name link was (is) no good. Maybe that's still the case.

Dma: Thanks. I tried it, but it took me to the same page, same results. "Sens message" link looks like it activates, but nothing happens. Probably a Mac and/or Firefox issue. Oh well. Nothing life threatening. j

My concern last Feb was with the slab only. I used ICFs from the footer up so the crawl wall was insulated and I also put 2" XPS on the inside of the footer. With in-slab radiant PEX in the slab, and the slab internals, will be quite a bit higher temp than air above, and since heat transfer is a temp dependent function the higher the temp the greater the heat flow. Triple, or whatever, the heat flow per hour and you triple the dollar flow per hour. That's why it makes a difference!

Don't overlook the insulating value of snow. I presume you get a fair amount of snow there. You get 2 to 6 feet of snow piled up around your house you're going to be pretty well insulated!! Frost always penetrates deepest under parking lots, etc., that are cleared of snow.

If you can't get hold of Siegenthaler's book send me a PM and maybe I can send you a copy of the slab formula he has.

"With in-slab radiant PEX in the slab, and the slab internals, will be quite a bit higher temp than air above..." •• OK, no argument about the delta T at the floor, rad floors run about 80 degrees and an non-rad slab is, what, 66? Still, though, I think you want lots of insulation under the slab, perhaps more w/ radiant.

"Don't overlook the insulating value of snow." •• Snow is good. I've just been digging up trees (new driveway) even after a long 20 to 40 below period. However, snow near a house is bad when it melts, so I just keep it away. As now, you never know when thaw will be.

"Frost always penetrates deepest under parking lots, etc." •• Which is why my sewer line is down 11'. It runs under the driveway. Glad I was working for an engineer when I built; I would not have thought of that, being new up here then. Hell, I may not have thought of it after 35 yrs here..... Brains have a habit of going dead at bad moments.

"If you can't get hold of Siegenthaler's book send me a PM and maybe I can send you a copy of the slab formula he has." •• It is on Amazon, but I am too cheap to spend $136 for one equation/page/whatever; sorry, John! Sure, pls try to PM, but I don't know if I will be able to access it. Thanks. j

Make that "undisturbed snow". Packed snow isn't too good.

Obviously, it depends on exactly how the structure was constructed. I agree with Rob that if the edges of the slab are exposed, that's a whole different animal. Since concrete is a thermal conductor, you don't really have a Heat Loss, but a Cold Gain. Nasty scenario, and beyond the scope of what the PPBD, and most 'normal' Heat Loss calcs, are designed to compute.

jk, I sent you a PM via the link at the top of the page. I did compose message and then searched for your user id. The system says you got the message. Did you?