Quantitative measure of an assembly or material resistance to heat flow for a unit temperature difference and a unit area. It is the reciprocal of the U-factor. The units for R-value are ft2 °F hr/Btu (English) or m2 °K hr/W (SI or metric). As R-value increases, conduction through an assembly or material decreases for the same temperature difference. As an example of the context in which R-value should be placed, 25% to 40% of a typical building's energy use can often be attributed to air infiltration, and air conditioning loads are often dominated by solar heat gain.
The effect of thermal bridging and different framing details requires a metric more complex than just a single R-value to allow for meaningful comparisons. Five R-values have been and are used in the building industry. Oak Ridge National Labs (ORNL) proposed a number of definitions (Christian and Kosny 1995). We have found it useful to add some and extend their definitions.
- Installed Insulation R-value. This R-value is commonly referenced in building codes and used by industry. This is simply the R-value labeled on the product installed in the assembly.
- Center-of-Cavity R-value. The R-value at a line through an assembly that contains the most insulation, and the least framing, typically, the middle of a stud-bay in framed construction.
- Clear-wall R-value. R-value of an assembly containing only insulation and minimum necessary framing materials at a clear section with no windows, corners, columns, architectural details, or interfaces with roofs, foundations or other walls.
- Whole-wall R-value. R-value for the whole opaque assembly including all additional structural elements (such as double studs), and typical enclosure interface details, including wall/wall (corners), wall /roof, and wall/floor connections.
- True R-value. The R-value of an enclosure assembly that includes all thermal bridging, air leakage, wind washing, convective loops, radiation enhancements, thermal and hygric mass, and installation defects.
Each of these measures is progressively more realistic. The True R-value is very difficult to measure without field samples.