Resistance and transmittance are the same fact stated twice
R-value measures how hard it is for heat to get through an assembly; U-factor measures how easily it does. They are exact reciprocals: U = 1 ÷ R. R-19 insulation has a U-factor of 0.0526, and a window rated U-0.30 has an R-value of 3.33. Nothing is lost or approximated in the conversion.
Which one gets quoted depends entirely on the product. Insulation is sold on R because a bigger number sells better and because insulation is bought by the layer. Windows and doors are specified on U because a window is a manufactured assembly of glass, gas fill, spacer and frame whose performance cannot be decomposed into layers, and because energy codes set maximum U-factors for fenestration. Whole-assembly requirements in codes appear both ways: the same wall may be described as R-20 continuous or as U-0.051.
The reason U is the more useful quantity in engineering is that it is what appears in the heat flow equation. Heat transfer through an assembly is Q = U × A × ΔT, and the UA product summed over every assembly in a building is exactly the term that drives a design heat loss, as it does in the Manual J load calculator. R-values are what you buy; U-factors are what you calculate with.
The last quantity in the family is conductance, C, which is the transmittance of a single layer without its surface films, and conductivity k, which is the conductance of a material one inch thick. A material with k = 0.27 delivers 1 ÷ 0.27 = 3.7 R per inch, so a 3.5 inch batt of it is R-13.
Add resistances in series, add transmittances in parallel
This is the rule that decides whether an assembly calculation is right or badly wrong, and it is the one most often broken.
Layers stacked through the thickness of an assembly are in series, and their resistances add. Siding, sheathing, cavity insulation, drywall and the two surface air films all sit in the path of the same heat flow, so R-0.62 + R-5.0 + R-13 + R-0.45 + R-0.85 = R-19.92 and the assembly U-factor is 0.0502. Convert to U only at the end.
Paths side by side across the face of an assembly are in parallel, and their transmittances add, area-weighted. A wall that is 23% framing by area with U-0.081 at the studs and U-0.050 at the cavities has an average U of 0.23 × 0.081 + 0.77 × 0.050 = 0.057, which is R-17.5, not the R-20 the cavity alone would suggest. Averaging R-values instead of U-factors gives 0.23 × 12.35 + 0.77 × 20.0 = R-18.2, which overstates the wall by 4%, and that overstatement grows as the conductivity of the bypass grows. That is the mechanism behind thermal bridging, and it is why continuous exterior insulation buys so much more than its nominal R suggests: it lies across every path.
The air leakage side of the envelope obeys a different equation entirely. No R-value describes air moving through a hole, which is why a well-insulated leaky wall performs badly, and why loads are calculated with a UA term and a separate 1.08 × CFM term.
Worked example: a 2x6 wall converted four ways
Take a 2x6 stud cavity filled with a batt whose conductivity is k = 0.25 BTU·in/h·sq ft·°F, and convert it into every unit you might be asked for.
- Layer resistance. The cavity is 5.5 in deep. R = t ÷ k = 5.5 ÷ 0.25 = 22.0 h·sq ft·°F/BTU. Equivalently, R per inch is 1 ÷ 0.25 = 4.0.
- Add the air films. Still interior air with horizontal heat flow is R-0.68; an exterior surface in a 15 mph winter wind is R-0.17. Total 22.0 + 0.85 = R-22.85.
- U-factor, IP. 1 ÷ 22.85 = 0.04376 BTU/h·sq ft·°F.
- RSI. 22.85 × 0.1761102 = 4.0241 m²·K/W.
- U-value, SI. 0.04376 × 5.678263 = 0.2485 W/m²·K. Check it the other way: 1 ÷ 4.0241 = 0.2485. The two routes agree, as they must, because 0.1761102 × 5.678263 = 1.
- Heat flow. Over 400 sq ft of that wall with 50 °F across it: Q = 0.04376 × 400 × 50 = 875 BTU/h, which is 256 W.
Now suppose the same wall is framed 23% by area, and the framing path is 5.5 in of softwood at k = 0.80, giving R = 6.875 plus films = R-7.725, U = 0.1294. The area-weighted assembly U is 0.23 × 0.1294 + 0.77 × 0.04376 = 0.0298 + 0.0337 = 0.0635, which is R-15.7. The cavity says R-22.85; the wall delivers R-15.7. The framing occupies less than a quarter of the area and carries 47% of the heat.
Reading the numbers against code and against reality
Energy codes state envelope requirements two ways, and the two are not interchangeable without care. The prescriptive route gives an R-value for the insulation only, ignoring framing and films: R-20 cavity or R-13 cavity plus R-5 continuous, for example. The performance route gives a maximum assembly U-factor that must be computed with framing included. A wall that meets the prescriptive R-value can fail the equivalent U-factor if it is framed heavily, which is why the two paths are listed with separate tables.
Window figures need particular care. An NFRC label U-factor is for the whole product, including the frame, the spacer, the glass and both surface films. Converting it to R and then adding air films double-counts them. That is the single most common mistake with this conversion, and it is what the warning on this page checks for.
Insulation R-values are measured at a mean temperature, conventionally 75 °F for the ASTM test methods. Some materials, notably certain closed-cell foams and loose-fill products, perform differently at winter mean temperatures, and dense-packed cellulose and fibreglass both lose effective R when air moves through them. A labelled R is a laboratory value for the material lying flat, undamaged and dry; an installed assembly with compression, gaps and moisture will not match it.
Once you have assembly U-factors, they multiply directly into load calculations. Multiply each by its area, sum, and you have the UA that the furnace size calculator ultimately converts into an equipment rating.
R-value, U-factor, RSI and U-value side by side
| R-value (IP) | U-factor (BTU/h-sq ft-F) | RSI (m2-K/W) | U-value (W/m2-K) |
|---|---|---|---|
| R-1 | 1.0000 | 0.1761 | 5.6783 |
| R-3 | 0.3333 | 0.5283 | 1.8928 |
| R-5 | 0.2000 | 0.8806 | 1.1357 |
| R-10 | 0.1000 | 1.7611 | 0.5678 |
| R-13 | 0.0769 | 2.2894 | 0.4368 |
| R-19 | 0.0526 | 3.3461 | 0.2989 |
| R-21 | 0.0476 | 3.6983 | 0.2704 |
| R-30 | 0.0333 | 5.2833 | 0.1893 |
| R-38 | 0.0263 | 6.6922 | 0.1494 |
| R-49 | 0.0204 | 8.6294 | 0.1159 |
| R-60 | 0.0167 | 10.5666 | 0.0946 |
These are material or assembly resistances without surface films. Add R-0.85 for the standard interior and winter exterior films before converting an assembly to a U-factor for load calculations.
Key terms
- R-value
- Thermal resistance in h·sq ft·°F/BTU. The temperature difference needed to drive one BTU per hour through one square foot.
- RSI
- The same quantity in SI units, m²·K/W. Numerically about 0.176 of the IP value, so R-19 is RSI 3.35.
- U-factor
- Thermal transmittance, the reciprocal of R-value, in BTU/h·sq ft·°F. Always for a complete assembly including its surface films.
- Conductivity, k
- The conductance of one inch of a material, in BTU·in/h·sq ft·°F. R per inch is 1/k.
- Conductance, C
- The transmittance of a specific layer at its actual thickness, excluding surface films. C = k/thickness.
- Surface film resistance
- The resistance of the boundary layer of air clinging to each face. R-0.68 for still interior air with horizontal heat flow, R-0.17 for a 15 mph winter exterior.
Mistakes that corrupt the conversion
- Averaging R-values across parallel paths. Framing and cavity are in parallel. Convert each path to U, area-weight the U-factors, and convert back only at the end.
- Adding air films to a window U-factor. NFRC labels already include them. Doing it again makes the window look about 10% better than it is.
- Confusing RSI with R. RSI 3.5 is R-19.9, not R-3.5. A metric label read as an IP number understates the assembly by a factor of about 5.7.
- Using nominal insulation R as assembly R. A wall with an R-19 batt is not an R-19 wall once framing, sheathing and films are accounted for; it is usually nearer R-15.
- Converting a c-value or an R-per-inch as if it were an R-value. Rigid foam is often quoted at R-5 per inch. Two inches of it is R-10, and the number on the board is not the number for the layer.
- Ignoring the mean temperature of the rating. R-values are measured at a standard mean temperature. Some materials perform differently in the cold, and none perform to label when compressed, gapped or wet.
Where these numbers go next
An assembly U-factor has exactly one job in HVAC: it becomes part of the UA that sets the building's design loads. Multiply each assembly's U by its area, add the infiltration term, multiply by the design temperature difference, and you have a heating load; that is the whole of the arithmetic in the Manual J block load calculator. Every dollar spent on the envelope shows up as a smaller UA and therefore smaller equipment, shorter duct runs and lower operating cost.
The same conversions underpin condensation analysis. The surface temperature of the interior face of an assembly is the indoor temperature minus the fraction of the total resistance that lies outboard of that face, times the temperature difference. Compare that surface temperature against the indoor dew point from the dew point calculator and you know whether the assembly will sweat. A window at U-0.30 in a 70 °F room at 0 °F outside runs an interior surface near 55 °F, which condenses whenever indoor relative humidity climbs past about half.
Finally, note what R-value cannot tell you. It says nothing about air leakage, nothing about moisture transport, and nothing about the thermal mass that shifts a cooling load in time. Those need the tools alongside this one, not a bigger number on the batt.