Hi, You'll be hearing more from me soon I hope! I'm interested in building a 24x36 (ish) home here in VT. I'm looking at shallow frost protected slabs and would like to hear what the main difference is between sfpf in heated and unheated homes is. is there a happy middle ground? i started thinking more is better in regards to my underground insulation and it hit me that I'm starting to play with the balance of the system. If i don't let enough heat out of the bottom I may unravel the whole concept. I'm looking to put 4" insulation under the slab. High density. This can't be too much insulation can it?

Have you looked at VT energy code to see what the minimum code required insulation for the type of construction you are thinking about is?

Frost heaves propgate in the direction of heat loss, which presents difference scenarios for heated vs. unheated buildings:

In an unheated building the heat loss is from the ground up into the building, so when the soil beneath the slab begins to freeze the upward forces in the middle of the slab can be tremendous, breaking the slab.

In a heated building the heat loss is from the building to the ground, eliminating the risk frost heaves center slab, but still leaves potential at the slab edge, which can be mitigated with insulation at the slab edge that either extends outward a few feet, or down a few feet. How far out &/or down depends on the local subsoil temp, soil type, and climate. But the wing-insulation needn't be as far as in an unheated building, but the higher you go under the slab, the further it needs to extend.

With 4" of XPS (R20) or EPS (R16) the ground won't freeze mid-slab in a VT climate whether the house is heated or not- the heat coming up from of the ground would be sufficient to prevent it, but the edges can have issues if measures aren't taken. If you treat the wing insulation as if it were an unheated building it'll have margin.

If you're an energy nerd into designing it for the particulars, you can grab a tool called THERM a freebie download from the Berkeley Nat'l Labs (See: http://windows.lbl.gov/software/therm/therm.html ) and drive yourself crazy with guesstimating the conductivity of the soil, or you can just punt and put down R10 of wing insulation extending from the edge of the slab to a distance equal to the 100-year design frost-depth of your area and call it "done". Wing insulation of 48" insulation would be enough protection in most low-altitude VT locations. If you went by the design freeze depth maps used for road paving standards it would likely be overkill even with R20 under the slab: http://proceedings.esri.com/library/userconf/proc02/pap1048/p1048.htm

A most-important construction detail in any SFPF is to keep it all well-drained and sending all roof runoff well away from the house: no water=no frost=no frost heave The less seasonal water percolates in under the slab & wing area, the less pressure would be applied even if it did get to freeze point, and drier soils are themselves more insulating.