SizingKitASHRAE parallel path
Insulation R-Value Calculator
This builds an assembly out of layers and reports what the whole thing performs at, which is always below the number printed on the insulation inside it. Heat takes two routes through a framed wall — between the studs and through them — and the routes are combined as conductances, because parallel resistances do not average. Every result appears as R and RSI together with its U-factor in both systems, and the page names how much of the cavity rating the framing gave back.
- 100% free
- No signup
- R and RSI together
- Framing factor visible
- Printable chart
What is being framed
The framing fraction is not the stud spacing. Plates, headers, corners, blocking and the framing around every rough opening are all wood at the full depth of the cavity, which is how 16 in on center comes out at a quarter of the wall rather than the ninth the spacing alone suggests.
ASHRAE Standard 90.1, Appendix A default framing fractions for wood-framed assemblies, which are what the standard's own U-factor tables are computed with.
25% of the area is framing. Take a real fraction off the plan for a wall with many openings.
Inches. R-6.88 of wood on the framing path at 1.25 per inch.
R-0.17 against R-0.25. Small on a wall, and a real fraction of the total on glass.
The layers, and which path each one is on
A layer that crosses the studs is on both paths and counts twice as hard as the same R-value stuffed between them. That is the whole argument for exterior continuous insulation, and this table is where you can watch it happen.
0.5 in Gypsum board — 0.9 per inch, a working figure inside a published 0.85–0.95 range · RSI 0.08
5.5 in Fiberglass batt — 3.2 per inch, a working figure inside a published 3.1–3.4 range · RSI 3.1
0.5 in Plywood or OSB sheathing — 1.25 per inch, a working figure inside a published 1.06–1.4 range · RSI 0.11
Vinyl siding over housewrap — entered from the product data sheet
The assembly R-value
Assembly R-value
R-15.67
RSI 2.76
Assembly U-factor
0.0638
BTU/h·ft²·°F · 0.362 W/m²·K
| Through the cavity — 75% of the area | R-20.14 |
|---|---|
| Through the framing — 25% of the area | R-9.41 |
| Insulation between the framing, on its own | R-17.6 |
| Everything crossing the framing, on its own | R-2.54 |
The framing gives back 22.2% of what the cavity path reaches. Add R-1 of continuous insulation outside the studs and the assembly gains very nearly the whole R-1, because it lands on both paths; add R-1 to the batt and it lands on 75% of the wall and is dragged down by the other 25%. That asymmetry is the reason exterior insulation appears in the code table at a lower number than the cavity alternative it substitutes for.
Five assemblies, nominal against actual
Read the last column. None of these is your wall; all of them show the same thing, which is that the gap between the batt and the building widens as the batt gets thicker.
| Assembly | Cavity path | Assembly R | RSI | U | Bridged away |
|---|---|---|---|---|---|
| 2×4 wall, R-13 cavity, 16 in o.c. | R-15.5 | R-11.8 | 2.09 | 0.084 | 23.8% |
| 2×6 wall, R-21 cavity, 16 in o.c. | R-23.5 | R-17.1 | 3.01 | 0.058 | 27.3% |
| 2×6 wall, R-21 cavity, 24 in o.c. | R-23.5 | R-17.7 | 3.12 | 0.057 | 24.8% |
| 2×6 wall, R-21 cavity plus R-5 continuous exterior, 16 in o.c. | R-28.5 | R-22.9 | 4.04 | 0.044 | 19.7% |
| 2×10 floor over unconditioned space, R-30 cavity | R-32.8 | R-29 | 5.12 | 0.034 | 11.4% |
Computed by the parallel-path method from the layer R-values in this file and the ASHRAE 90.1 Appendix A framing fractions. Nothing in this table is transcribed — change a layer above and every row moves.
Representative constructions, not your wall. The parallel-path method assumes heat goes straight through, which understates the loss where a highly conductive path spreads heat sideways — it is not valid for steel framing at all. It also takes no account of the thing that dominates real performance: air leakage. An attic insulated by blowing loose fill over the joists is not this calculation, because the insulation crosses the framing and is closer to a continuous layer; the parallel-path form applies where the cavity is filled to the depth of the framing and no further.
Representative R per inch at a 75 °F mean temperature, from the ASHRAE Handbook—Fundamentals building materials tables and US Department of Energy insulation guidance. Every row is a range in the source, and the typical column is the working figure the trade uses.
None of these is exact and none should be printed as a fact. Real product varies by density, facing, age and installation, and a manufacturer's published R for a specific product supersedes any of it. Installation quality moves the answer more than the material choice does: a batt with gaps, compressed round wiring, or not in contact with the drywall behind it performs far below its label, and the grading system used by energy raters (Grade I / II / III) exists to describe exactly this.
US Department of Commerce Voluntary Product Standard PS 20, American Softwood Lumber Standard: surfaced dry dimensions of nominal framing lumber. A nominal 2×4 is 1½ × 3½ in, a 2×6 is 1½ × 5½ in, and the sequence continues in ¾ in steps from the 2×8 up.
This page carries no code minimum, and that is deliberate rather than an omission. The IECC prescriptive envelope table at R402.1.2 changed substantially between its 2018 and 2021 editions, the wall column offers alternatives that trade cavity insulation against continuous insulation, and a jurisdiction may be enforcing either edition or a state amendment to one. Find out which edition your building department has adopted and read the wall, ceiling and floor rows out of that book — an R-value quoted from the wrong edition passes nothing.
How to work out what a wall actually performs at
Three decisions, and the second one is where most published figures quietly go wrong.
Set the framing before the insulation
Pick the assembly, then look hard at the framing fraction it prefilled. A quarter of an ordinary wall being wood is not a guess about stud spacing — it counts plates, headers, corners, blocking and the framing around every window and door, which is why 16 in on center comes out at 25 percent and not at the 9 percent the spacing alone would suggest. A wall with many openings, or a corner built the old way with three studs, is more than that.
Mark each layer for the path it is on
A layer between the framing counts only on the cavity route; a layer that crosses the framing counts on both. That distinction is the whole calculation. Sheathing, cladding, gypsum and the surface films cross; batts, blown fill and cavity foam do not. Get one wrong and the answer moves by more than any material choice on the list.
Read the bridging line, not just the R
Under the answer is the percentage of the cavity-path figure the framing removed. That number is what tells you whether the next dollar belongs in a thicker batt or in a layer of continuous board outside the studs — because insulation that crosses the framing is worth close to its full R and insulation that does not is worth a fraction of it.
Technical specifications
| Parallel-path method | Parallel path, as ASHRAE describes it. The layers common to both routes are summed once, the cavity and the framing member are added to each route separately, both routes are inverted to conductances, weighted by the fraction of wall area each occupies, and the weighted conductance is inverted back. Averaging the two R-values instead always flatters the wall. |
|---|---|
| Materials | 14 materials with R per inch as a low, high and working figure — batts, blown and loose fill, both spray foams, three rigid boards, framing lumber, sheathing, gypsum, concrete and brick — plus a data-sheet row for a specific product the list does not carry. |
| Surface films and air spaces | Interior film at R-0.68 for horizontal heat flow, exterior at R-0.17 in a 15 mph winter wind or R-0.25 in a 7.5 mph summer one, and a nominal 3/4 in non-reflective vertical air space at R-1.0, all from the ASHRAE Handbook—Fundamentals surface resistance tables. |
| Framing fractions | 0.25 for a wall at 16 in on center, 0.22 for advanced framing at 24 in, 0.11 for a roof or ceiling and 0.10 for a floor — the ASHRAE 90.1 Appendix A defaults the standard's own U-factor tables are built with. All editable, because a wall full of openings is more wood than any default. |
| Both unit systems, both quantities | R and RSI beside each other, U-factor in BTU/h·ft²·°F and in W/m²·K. The 5.678 factor between the two systems is derived here from the International Table BTU and the square meter rather than typed, so it cannot drift from the rest of the site by a last digit. |
| Steel framing | Refused, not approximated. A steel stud conducts hundreds of times better than the wood the parallel-path method assumes, heat spreads sideways through the flange, and the arithmetic overstates the result badly — an R-19 cavity in a steel-framed wall can perform nearer R-7. Isothermal planes or a published correction factor is what applies. |
| Reference chart | 5 common assemblies with cavity path, whole-assembly R, RSI, U and the percentage the framing takes away — computed by the same function rather than transcribed, and formatted to print. |
| Where the assembly stays | In the tab. A layer stack describes a specific building at a specific address, and nothing about it is uploaded — the calculation runs on a phone in an unheated attic with no signal, which is where the question usually comes up. |
Frequently asked questions
Why is my R-21 wall not an R-21 wall?
Because roughly a quarter of it is wood, and wood is about R-1.25 per inch against the batt's 3.2 to 3.8. Down the framing path a 2×6 stud contributes about R-6.9 where the batt beside it contributes R-21, and heat concentrates through the cheap route. The two paths are then combined as conductances weighted by area, which is what the physics requires, and a wall whose cavity path reaches about R-23 with sheathing and films included performs nearer R-17. The number on the bag was never a lie — it is the resistance of the insulation, measured on the insulation, and it becomes an assembly figure only after the framing has been accounted for.
What is the difference between R and U, and why does this page show both?
They are reciprocals, and which one you need depends on whether things are in series or in parallel. R-values add when layers stack one behind another, which is how a single path through an assembly is built up. U-factors add when paths sit side by side, which is how two paths through one wall — or several assemblies across a whole envelope — are combined. Codes and energy models are written in U for exactly that reason: a building's total conductance is the sum of every U times its area, and there is no way to sum R-values that gives the right answer. The rule that catches people is that you may never average R-values of parallel paths, because the average of two reciprocals is not the reciprocal of the average.
R-value or RSI — which one is on my insulation?
Both, and the same product carries different numbers on each side of the border. RSI is the metric form in m²·K/W, R is the US customary form in h·ft²·°F/BTU, and one R is 0.1761 RSI — so an RSI 3.87 batt in Canada and an R-22 batt in the United States are the same batt. Confusing them by a factor of 5.678 is a common and expensive error in a specification, and it is why every result on this page prints both without asking. If a figure looks impossibly good, check which system it was quoted in before checking the product.
Is exterior continuous insulation really worth more than the same R in the cavity?
Yes, and the arithmetic on this page shows exactly how much more. A layer outside the studs lies on both heat-flow paths, so it adds nearly its full R to the assembly; the same R added to the batt lands on only the three-quarters of the wall that is cavity and is then dragged down by the wood beside it. That asymmetry is why the code tables list a wall with continuous insulation at a lower total than the cavity-only alternative it substitutes for, and it is why the last few inches of batt in a deep wall buy so little. Continuous board also warms the sheathing, which moves the condensing surface out of the wall and is a moisture argument as much as a thermal one.
Why will this refuse to calculate a steel-framed wall?
Because the parallel-path method is not valid for it and would return a badly optimistic number. The method assumes heat goes straight through each path without moving sideways, which holds when the two paths differ by a factor of three. A steel stud conducts hundreds of times better than the wood it replaced, so it collects heat from the sheet metal flange across the full cavity width and funnels it through the web — the paths are not independent at all. A steel-framed wall with R-19 between the studs can perform near R-7, and getting that answer needs the isothermal-planes method or the published correction factors for steel assemblies, neither of which is on this site.
Does this tell me whether my wall meets code?
No, and refusing to is deliberate. The IECC prescriptive envelope table changed substantially between its 2018 and 2021 editions, the wall row offers alternatives that trade cavity R against continuous R, and your jurisdiction may be enforcing either edition or a state amendment to one. A minimum quoted from the wrong book passes nothing. Find out which edition your building department has adopted, read the row for your climate zone out of that edition, and note whether it is written as an R-value requirement or as a maximum U-factor — because the U-factor path is where an assembly calculation like this one is what the code actually wants to see.
How much does a badly installed batt cost compared with a cheaper material?
More than the material choice does, which is why energy raters grade installation quality separately. A batt with gaps at the edges, compressed around wiring and plumbing, or not touching the drywall behind it performs far below its label, and the grading system exists because that gap is routine rather than exceptional. Compression alone loses R roughly in proportion — an R-21 batt stuffed into a 3.5 in cavity is not an R-21 assembly and is not even the R-13 the cavity would have held. Air leakage through the assembly is a separate loss again, is not an R-value at all, and on a typical older house it outweighs everything this calculator computes.
Assembly against material, and the quarter of the wall made of wood
Insulation is sold with a number that describes the insulation. A building loses heat through an assembly. Those are different quantities and the second one is always smaller, because a framed wall is not a sheet of insulation — it is insulation interrupted every sixteen inches by a piece of wood that conducts several times better, plus plates top and bottom, plus a header over every opening, plus corners and blocking. Add all of that up by area and about a quarter of an ordinary wall is framing at full cavity depth. Heat does not distribute itself politely across the two routes: it concentrates through the cheaper one, which is what thermal bridging means, and the arithmetic for it is to resolve each route to a resistance, invert both to conductances, weight them by area and invert the result back. Averaging the two resistances is the mistake, it is easy to make, and it always produces a flattering answer.
Two conversions cause more trouble in specifications than the physics does. The first is R against U: they are reciprocals, series layers add in R, parallel paths add in U, and there is no arrangement in which you may add or average the wrong one. Codes are written in U-factor precisely because a whole envelope is a sum of parallel paths — every wall, window, door and ceiling contributing its own conductance times its own area — and R cannot be summed that way at all. The second is R against RSI, which is the same quantity in two measuring systems separated by a factor of 5.678: the batt labeled R-22 in Ohio is labeled RSI 3.87 in Ontario, and it is the same batt. This page prints both of everything for that reason, and derives the factor from the definition of the BTU rather than typing it, so it agrees with every other conversion on the site to the last digit.
What comes out is one assembly’s U-factor, which is the input to something else rather than an answer on its own. Multiply it by area, add every other assembly, add the air moving through the building, and you have the UA a heat loss calculation runs on — and from there the design-day load that sizes a furnace or decides how much propane a winter takes. That chain is also the honest answer to why an envelope is worth improving before equipment is: a wall improved once keeps paying without a compressor running, in a way neither a standby generator nor a solar array can. What this page cannot do is tell you the wall is compliant or that it will not accumulate moisture; both are code questions decided against a specific edition and a specific climate zone by somebody who signs the drawing.
Your layer stack does not travel
Every number on this page is worked out by JavaScript running in the tab you are reading it in. Nothing you type — loads, lengths, nameplate ratings, the rates your utility charges you — is uploaded, logged or kept, which is also why the calculators carry on working in a mechanical room with no signal.
The stack you build describes one building at one address, and it is never transmitted — which is worth knowing when the assembly is a client’s and the answer is going into a report with their name on it. Print or copy the layer list and it travels only where you send it. If you came here to settle an argument about units rather than to build a wall, the electrical calculator has the same habit on its side of the site: the relation is shown as it solves, in both measuring systems, with nothing assumed silently.