A quarter of your wall is wood, and heat knows it
The number printed on a batt is the center-of-cavity R-value. It is measured straight through the insulation, in the one place on the wall where there is no lumber. It is a true number about the product and it is not a number about your wall.
Oak Ridge National Laboratory named this problem properly, in a paper called How the Same Wall Can Have Several Different R-Values. One wall has at least three: center-of-cavity through the insulation, clear-wall including the studs and plates holding it up, and whole-wall including corners, headers and every junction where the framing doubles up. Their hot-box tests came out 30 to 60 per cent below the cavity figure.
The share of a wall that is framing is the argument. Laboratory specimens run 9 to 14 per cent, and Oak Ridge say plainly that this is unrealistically low against what gets built. Two large studies measured real housing: ASHRAE found about 25 per centnationwide in 2003, and the California Energy Commission found about 27 per cent in Californian homes. So the honest default is a quarter, and this tool uses it rather than the flattering figure.
Continuous insulation is worth more than cavity insulation, point for point
This is the part that changes what you buy. Batts go between the studs, so the studs bypass them. Foam board goes over the studs, so nothing bypasses it. Every point of continuous insulation survives into the wall and roughly a quarter of every point of cavity insulation does not.
The code already knows. For climate zones 4 to 8 the 2021 IECC offers four wall assemblies: R-30 cavity, R-20 plus R-5 continuous, R-13 plus R-10 continuous, or R-20 continuous with nothing in the cavity. Look at the totals. R-30 of batts is being traded against R-23 when R-10 of that is continuous, and against R-20 when all of it is. The code is pricing the framing penalty, and it is pricing it at roughly the number Oak Ridge measured.
Those four are alternatives the code permits rather than four identical walls, and on this arithmetic they do not land on the same whole-wall figure. The tool shows each one, because if you are choosing between two of them on price you should know they are not equals.
Your climate zone is set by county, not by state
This catches people who look up "the R-value for Texas" and find two answers. IECC climate zones follow county lines in Figure R301.1, and a large state spans several: the requirement in the panhandle is not the requirement on the gulf. Zone 4 splits again into a Marine subzone that gets grouped with zone 5 for insulation purposes, which is why the table above reads "4 except Marine" and "5 and Marine 4" rather than plain numbers.
Two minutes with your building department settles it, and it is worth doing before buying insulation rather than after.
The ceiling number is the one that is hard to actually build
Zones 4 through 8 want R-60 in the ceiling under the 2021 IECC, up from R-49 in the 2015 and 2018 editions. In fiberglass, R-60 is over eighteen inches.
In the middle of an attic that is only a question of depth and money. At the eaves it is a question of geometry, because a truss heel is a few inches tall and the insulation has to thin out to nothing exactly where the roof meets the wall. That corner is where the heat is already trying to leave, so an attic that measures beautifully in the middle can still lose heat at the edge. Raised heel trusses exist for this reason, and they are a decision at the order stage rather than a fix afterwards.
Which edition applies, because the numbers moved
Naming the edition matters more here than it looks. A zone 2 ceiling was R-38 under the 2012 IECC and is R-60 under the 2021. Anything quoting an R-value without saying which edition it came from is quoting a number that may be three editions stale, and plenty of published advice still cites 2012 because that is when it was written.
Adoption is separate again, and slower. A state adopts an edition when it chooses to, then amends it, so the code in force where you are building may be several editions behind the newest one published. That is a question with a local answer and this page does not pretend to have it.
Failing the prescriptive table is not failing the code
The R-value table is one of three routes to compliance, and it is the bluntest of them. The other two are a component U-factor route and a total UA calculation, where the whole envelope is traded off against itself, so a wall that misses the prescriptive R-value can still be part of a house that passes. That is worth knowing before you tear out a wall to add foam: the question is whether the building passes, not whether one element hits one number in one of three tables. Local amendments sit on top of all three.
And the arithmetic here is an estimate by design. The Oak Ridge method is one line long, and a real thermal model accounts for the exact framing layout, the sheathing, the air films and the cladding. What one line is good for is the thing most calculators get wrong entirely: showing that the number on the packaging is not the number in the wall, and by roughly how much. A more precise wrong answer would be worse.
Common questions
Is the R-value on the batt the R-value of my wall?
No, and the gap is about a quarter. The printed figure is the center-of-cavity R-value, measured straight through the insulation where there is no wood. A real wall is roughly 25 per cent framing once studs, plates, headers, corners and junctions are counted, and heat takes the wood. Oak Ridge National Laboratory measured clear-wall R-values 30 to 60 per cent below the cavity figure in hot-box tests.
How do I work out the whole-wall R-value?
Oak Ridge publish a one-line method: for each per cent of the wall that is framing, reduce the cavity R-value by one per cent. So R-20 batts in a wall that is 25 per cent framing give about R-15. Continuous insulation is not reduced at all, because it runs over the studs rather than between them and nothing bridges it.
Why does the code let me swap R-30 of batts for R-13 plus R-5 of foam?
Because continuous insulation is worth more per point than cavity insulation. Foam board over the framing is not bridged, so all of it counts, while batts between studs lose about a quarter. That is why the 2021 IECC lists R-30, R-20+5ci, R-13+10ci and R-0+20ci as alternatives for the same climate zones. They are permitted alternatives rather than equal walls, and the tool above shows what each one delivers.
What R-value do I need in my climate zone?
The 2021 IECC prescriptive table asks R-30 in the ceiling for zone 1, R-49 for zones 2 and 3, and R-60 for zones 4 to 8. Walls run from R-13 in zone 1 to R-30 or its continuous-insulation equivalents in zones 4 to 8. The whole table is above. Your zone is set by county, not by state, so a state can span three of them.
Does this mean my wall complies?
No, and nothing here says it does. The prescriptive R-value table is one of three compliance paths in the IECC, states adopt different editions of the code and amend them locally, and the figure on a product datasheet beats any general table. This page reports what the code asks for and what the arithmetic gives. Your building department decides compliance.
Which insulation has the best R-value per inch?
Of the common ones, polyisocyanurate board and closed cell spray foam at about R-6 an inch, then XPS at R-5, then EPS at 3.8 to 4.4, then mineral wool at 3 to 4 and fiberglass at 2.5 to 4. The ranges are real and depend on density and facing, so for anything being submitted use the number on the actual product.
Why can I not get R-60 into my attic?
Depth. R-60 in fiberglass is over 18 inches, and at the eaves a truss usually runs out of height long before that. It is the commonest reason an attic that looks well insulated in the middle performs badly, because the insulation thins out exactly where the roof meets the wall and the heat is already looking for a way out.
Where do these figures come from?
Three open sources, all read directly. The code requirements are Table R402.1.3 of the 2021 IECC as reproduced by the Building America Solution Center at Pacific Northwest National Laboratory. The R-per-inch figures are from the same source. The framing factors and the whole-wall method are from Kosny and others, How the Same Wall Can Have Several Different R-Values, Oak Ridge National Laboratory. The engine tests reproduce Oak Ridge Table 1 exactly, which is how the method is checked rather than trusted.