OK guys, after days/weeks of modelling I think I have some interesting results. At least I hope you think they are interesting.
There is so much I've had to split it into lots of sub-articles.

This is the starting point:

Test Series 2: Introduction to Modelling Concrete Foundations with THERM

(Yay, I've sussed how to do links on here :) )

I love this subject. You really need to know the site specifics to make the best decision. The soil type and water table will influence heat transfer every bit as much as the outdoor temperature. In southwest Montana, with well drained soil, we would finish a slab edge with 2 inch xps down two feet and skirted out two feet. Under slab will have less temperature difference and 1 inch xps is more cost effective than 2 inch. This is for a heated slab on grade application. For a 4 ft. foundation wall it would be 2" xps for the full height of the foundation wall and 1" under slab.
I am currently collecting data, from a superinsulated home near Helena, Montana, that will allow me to analyze the overall heating and cooling performance of the WHOLE HOUSE. We included sensors to determine heat flow from specific parts of this building.(slab, ceiling and walls). While studies have been done by NREL and ORL, they were quite short (days) and did not account for seasonal transitions of below grade heat loss. The water table is 60 feet from the surface and the site has excellent drainage. This information will help us to fine tune the next project, much like a nascar team fine tunes the WHOLE race car and not just the engine.
Site specifics determine the set up for every project.

Hi Dteltech,

It sounds like you'll find this paper interesting if you've not seen it:

[b]con76: A Four-Year Site Measurement of Heat Flow in Slab-on-Ground Floors with Wet Soils[/b] H.A. Trethowen and A.E. Delsante
http://www.branz.co.nz/cms_show_download.php?id=78

Notice the time scale!

I also love this subject, and agree with you entirely. It seems to me that very few people (here in NZ at least) even consider their site conditions.

Have you checked the deep soil temperature for your area? I understand it varies from place to place, and obviously has quite a large effect on heat loss through the slab.

"At soil depths greater than 30 feet below the surface, the soil temperature is relatively constant, and corresponds roughly to the water temperature measured in groundwater wells 30 to 50 feet deep. This is referred to as the “mean earth temperature.” Figure 2 shows the mean earth temperature contours across the United States. In Virginia, the mean earth temperature ranges from 52ºF in the northern Shenandoah Valley and Winchester area to 62ºF in coastal Tidewater."

Earth Temperature and Site Geology
http://www.geo4va.vt.edu/A1/A1.htm#A1Sec2

ARCHIVED Deep Soil Temperatures at select NDAWN stations
http://www.ndsu.edu/ndsu/ndsco/soil/index.html

Steve

PS I hope you are going to publish your results. I'd love to compare them with my modelling. It's the only way to know if the 'theory' is right.

I will be publishing this study and am working on web based access to the the project. I am also using this to calibrate the modeling software that I have been using for many years (25). If you would like to send me an email, I would be glad to exchange info.
As to ground moisture and heat flow, I am also a very experienced pit chef. "I cook stuff in the ground" It is a very good way to put your knowledge of heat transfer and ground conditions to the test. When you have 50 to 200 people waiting to eat the result of your efforts, you have to get it right. NEVER try this in saturated soil. :(

We like to use 2" of styrafoam and a product called insultarp witch is R-11 for a total for R- 21 we then tape the seams for a vaper barrier .heat seaks cold 50 degress or 10

In March, Meyer Enterprises, the manufacturer of Insul-Tarp, settled a complaint brought by the FTC in federal court: http://www.ftc.gov/os/caselist/0823062/index.shtm The FTC alleged that R 7.5 was more like R 2.