You raise a great point about the seasonal/lifetime fluctuations of the green roof media. In general the vegetation on trees cooperates with us as to be seasonally advantageous, but for green roofs, the growth of the media, accumulation (or erosion) of soil, etc all points toward a difficulty in modelling R performance. It seems like the most easily universalizable measure is soil depth–if soil depth was the R-value, then we’d also have an additional argument for soil depth, which equals more stormwater performance as well.

I’ve seen a bit of research that puts a generalized R-value on the green roof (between 1 and 4), but these always come with caveats as to the variability of this number due to environmental conditions. We’ve been trying to get Portland to accept green roof layers for code-insulation, with very little success (we got them to allow us to say the Portland Building green roof had an r-value of 1, meaning we only needed 3″ of insulation).

There’s a group here in Portland called Quantec LLC, that is developing a web-based green roof energy calculator using available data, which I believe could be plugged into standard energy models… they’ve definitely encountered the issues of multiple variables in their preliminary versions, if I remember correctly. I’ll try to track down where they’ve come in terms of their research and get some info to you.
JK

Where was I? In exurban SoCal where my host described theirs as “one of two houses in the neighborhood with an Obama sign.”