Insulation Calculator

Quantity from coverage, savings from an explicit heat-loss model

Insulation asks two questions and they deserve two different standards of honesty. How much to buy is coverage arithmetic: area, product coverage, target R-value and depth. What it saves is a model, and this page shows its model rather than hiding it: a steady-state degree-day conduction comparison, with your climate’s heating degree days, the assembly’s present and improved R-values, and every assumption printed beside the result. Air leakage, the other half of real heat loss, is named as the thing no R-value fixes, because an estimate that implies otherwise sells insulation as draught-proofing and sets up the disappointment that follows.

Quantity
Heat-loss scenario (degree-day model)

Disclaimer: This calculator gives estimates for general information only. It is not financial, medical or other professional advice. Required R-values by climate zone, vapour control and fire rules are set by code and product instructions, which govern over any calculator.

What this calculator does

The quantity section follows each product’s own buying unit: coverage per package for batts, depth and bag coverage from the manufacturer’s chart for loose fill (because bags are rated at a settled depth and R, not a volume), sheet division for boards, and board feet for spray foam, the unit foam is genuinely sold and specified in. The savings section exists because “how much will it save” deserves better than a multiplier pulled from area: it runs the textbook steady-state conduction model over your degree days, comparing the assembly before and after, and prints the chain (heat saved, fuel saved at your efficiency, money at your price) so every step can be argued with. Neither section reaches for the house’s whole energy bill, which belongs to the HVAC and whole-house tools linked below.

How to use it

  • Depth is settled depth. Loose fill is charted at the depth it settles to, and installers blow above it initially; use the product chart’s figures for both R per inch and bag coverage, since the two come from the same settled-density test.
  • R-values add, badly installed R does not. Layers in series sum (existing plus added is the honest total the page prints), but gaps, compression and wind washing tax the sum hard: a batt squeezed into a shallower cavity delivers the compressed value on its chart, not its label.
  • Degree days come from your weather, not a national average. Utilities publish local HDD; the range across the US runs from under 1,000 to over 10,000, and the savings scale linearly with it, so this input is the scenario.
  • One element at a time. Run the attic, then the walls, as separate scenarios: the model compares assemblies, and summing two runs is valid because conduction through parallel elements genuinely adds.

The methods

  • Batts: packages = area / coverage per package, rounded up. Boards: sheets = area / sheet area, rounded up. Spray foam: board feet = area × thickness in inches; kits = board feet / kit yield, rounded up.
  • Loose fill: required depth = target R / your product’s R per inch (at settled density); bags = area / the chart’s coverage per bag at that depth, rounded up.
  • R stack: total = existing + added; the overall U-factor is 1 / total R.
  • Heat-loss scenario: annual conducted heat = area × (1/R) × HDD × 24. Saving = the before figure minus the after figure, converted at 100,000 BTU per therm or 3,412 BTU per kWh, then adjusted by the heating efficiency entered and priced at the entered fuel price.

Worked examples

The attic top-up. A 1,000 sq ft attic targeted to R-38 in blown fibreglass at R 2.7 per inch needs a settled depth of 38 / 2.7 = 14.07 in, and at the chart’s 25 sq ft per bag at that depth, 40 bags. The example’s numbers are a real product chart’s shape, and the two chart inputs (R per inch, bag coverage) are what make them yours.

The saving, modelled openly. The same attic, 5,000 HDD, going from R-11 to R-38: conducted heat falls by 1,000 × 5,000 × 24 × (1/11 − 1/38) = 7,751,196 BTU a year, about 77.51 therms, worth 93.01 at 1.20 a therm before the heating system’s efficiency takes its share. That is a conduction-only figure for one assembly, stated as such, and the next section is firm about what it leaves out.

Reading your results

Read depth as the installer’s instruction and bags as the purchase, and notice which of the two your product’s chart drives: settling products are bought by the chart, not by any universal table, including this page’s example. Read the savings as the model’s, with its inputs attached: change the degree days and the money moves proportionally, so quote the scenario with its HDD or not at all. Read the R stack for the diminishing return it shows (each added R cuts loss by a smaller slice than the last, since loss follows 1/R), which is the arithmetic behind every “enough” judgement in insulation.

Limitations

  • Conduction only. The model excludes air leakage, which in many houses rivals conduction, and thermal bridging through framing, which taxes cavity insulation by a framing factor. Sealing and insulation are different jobs and this page prices one of them.
  • Degree days are a climate summary, not your winter. Internal gains, solar gains, thermostat behaviour and the balance point all move real bills away from the steady-state figure; the scenario ranks improvements reliably and predicts bills approximately, which are different services.
  • Moisture is a design subject. Vapour retarders, ventilation of roof assemblies and condensation planes follow climate-zone rules that no quantity calculator carries; the wrong assembly insulated to the right R is still the wrong assembly.
  • Coverage charts assume installation quality. Fluffing loose fill below its settled density (more depth, less material, less R) is a known industry abuse; the depth gauge on the attic floor, not the bag count, is the delivered product.

Frequently asked questions

How much insulation do I need?

As much depth of your product as reaches your target R, bought in that product’s units: the target comes from climate-zone recommendations and code minimums for your element (attics run highest, then walls, then floors), and the quantity section converts the target into packages, bags, sheets or kits. The honest sequence is target first, product second, count third, and any seller reversing it is starting from what they stock.

Can I just add new insulation over old?

Usually yes, and the R-values add, with three conditions: the old layer is dry and undisturbed by mould or pests, the new layer does not introduce a vapour barrier in the wrong place (unfaced batts over existing attic insulation is the standard safe pattern), and loose fill over batts is fine while batts over loose fill compress it and steal its depth. The stack line on this page totals the two for exactly this common case.

Why do savings estimates vary so wildly between sites?

Because the assumptions do, and most sites decline to print them: degree days for different cities, fuel prices from different years, models that include or exclude air leakage, and before-R values chosen to flatter the after figure. This page’s defence is transparency rather than a claim of precision: every input that moves the money is a field you set, and the model is named in the method, so a competing estimate can be reconciled input by input instead of argued by headline.

What R-value should an attic have?

Current US guidance for most climate zones lands between R-38 and R-60 for attics, with walls at R-13 to R-21 cavities and floors over unheated space around R-25 to R-30, but the governing figures are your adopted code and the tables published for your zone. The shape of the answer matters more than any single number: attics take the most because the space is cheap to fill and the stack effect drives the house’s warmest air against it all winter.

Related calculators

Method: quantities from each product’s own buying unit and manufacturer chart figures, savings from the steady-state degree-day conduction model with every assumption entered by the user and printed with the result. Last reviewed: 15 October 2026.