Why insulation savings depend on 1/R, not on R
The heat conducted through an assembly is proportional to U, the reciprocal of R. That single fact explains almost everything counter-intuitive about insulation economics.
Going from R-5 to R-10 cuts U from 0.200 to 0.100 — you eliminate half the loss. Going from R-40 to R-45 cuts U from 0.0250 to 0.0222, eliminating 11 percent of what is left, which is 2.8 percent of the original R-5 loss. The second upgrade adds the same five R-units and costs about the same per square foot, but delivers a small fraction of the energy. This is the law of diminishing returns in its purest form, and it is why building codes stop at R-49 or R-60 in attics rather than requiring R-100.
The corollary is more useful: the biggest savings are always in the worst assembly you own. An uninsulated stud bay at R-4 wastes more energy per square foot than an R-30 attic does, so a contractor who tells you to top up an already-decent attic before addressing bare rim joists or a knee wall is optimising the wrong surface. Run this calculator once for each candidate assembly and rank them by annual saving per dollar spent before you commit.
The second factor is climate. Degree days measure how much and how long the outdoor temperature sat below (or above) a 65 °F base. The same R-11 to R-49 attic upgrade in Minneapolis at roughly 7,500 HDD saves three times what it saves in Atlanta at roughly 2,500 HDD, for identical work.
The degree-day method, step by step
Start with UA. Multiply the area by the change in U to get the change in the assembly's UA, the heat that flows through it per hour for every degree of temperature difference. A 1,200 ft² attic going from R-11 to R-49 loses 84.6 Btu/h per °F less than it did.
Turn degree days into degree hours. A degree day is one degree of temperature difference sustained for a whole day, so multiply HDD by 24. Five thousand heating degree days is 120,000 degree-hours. Multiply by ΔUA and you have the annual heat that no longer escapes.
Divide by the efficiency of the system that would have replaced it. This step is routinely skipped and it matters a lot. Heat saved at the ceiling plane is heat your furnace does not have to make, and an 80 percent AFUE furnace burns 1.25 units of fuel for every unit it delivers. A heat pump at COP 3.0 consumes only 0.33 units of electricity per unit delivered, which is why the same insulation upgrade saves far fewer purchased kilowatt-hours on a heat pump than it does on a resistance heater.
Convert to fuel units and price it. Divide by 100,000 for therms, 91,500 for propane gallons, 138,500 for heating oil gallons, or 3,412 for kilowatt-hours, then multiply by your rate. The fuel BTU content converter handles any fuel not in the list.
Add cooling separately. Cooling degree days drive heat into the building, and SEER is already expressed as Btu removed per watt-hour consumed, so the cooling load in Btu divided by SEER × 1,000 gives kilowatt-hours directly. Cooling savings are usually smaller than heating savings in most of North America because CDD are fewer and SEER is high, but in the deep South they dominate.
The degree-day method assumes the loss is proportional to the indoor-outdoor temperature difference and that internal gains and solar gains are roughly accounted for by the 65 °F base. It is the standard first-order screening method used in residential energy auditing. It is not a substitute for an hourly simulation, and it systematically overstates attic savings slightly because a vented attic in summer runs hotter than outdoor air while a sunlit roof deck in winter runs warmer than outdoor air at midday.
Worked example: blowing an R-11 attic up to R-49
A 1,200 ft² attic in a 5,000 HDD / 1,200 CDD climate currently has R-11 of settled fibreglass. A contractor quotes $2,000 to blow in enough cellulose to reach R-49. The house has a 92 percent AFUE gas furnace, gas costs $1.55 per therm, the air conditioner is SEER 15, and electricity costs $0.17 per kWh.
- U before and after. 1 ÷ 11 = 0.090909; 1 ÷ 49 = 0.020408.
- Change in U. 0.090909 − 0.020408 = 0.070501 Btu/h·ft²·°F.
- Change in UA. 1,200 × 0.070501 = 84.60 Btu/h·°F.
- Heating energy saved. 84.60 × 5,000 × 24 = 10,152,131 Btu, or 10.15 MMBtu a year at the ceiling plane.
- Gas saved. 10,152,131 ÷ 0.92 = 11,034,925 Btu of fuel, which is 110.35 therms.
- Heating dollars. 110.35 × $1.55 = $171.04.
- Cooling energy saved. 84.60 × 1,200 × 24 = 2,436,511 Btu, and 2,436,511 ÷ (15 × 1,000) = 162.4 kWh.
- Cooling dollars. 162.4 × $0.17 = $27.61.
- Total annual saving. $171.04 + $27.61 = $198.65.
- Simple payback. $2,000 ÷ $198.65 = 10.1 years, and the undiscounted 20-year net is 20 × $198.65 − $2,000 = $1,973.
Now test the diminishing return. Stopping at R-38 instead of R-49 gives ΔU = 0.090909 − 0.026316 = 0.064593, which is 91.6 percent of the saving for perhaps 78 percent of the material. Pushing on to R-60 gives ΔU = 0.074242, only 5.3 percent more saving than R-49. That is the shape of the whole curve.
Annual heat saved per 1,000 ft² per 1,000 heating degree days
| Upgrade | ΔU (Btu/h·ft²·°F) | Heat saved (Btu) | Gas saved (therms) | Dollars saved |
|---|---|---|---|---|
| R-11 → R-30 | 0.057576 | 1,381,818 | 15.02 | $23.28 |
| R-11 → R-49 | 0.070501 | 1,692,022 | 18.39 | $28.51 |
| R-11 → R-60 | 0.074242 | 1,781,818 | 19.37 | $30.02 |
| R-19 → R-38 | 0.026316 | 631,579 | 6.87 | $10.64 |
| R-19 → R-49 | 0.032223 | 773,362 | 8.41 | $13.03 |
| R-19 → R-60 | 0.035965 | 863,158 | 9.38 | $14.54 |
| R-30 → R-49 | 0.012925 | 310,204 | 3.37 | $5.23 |
| R-30 → R-60 | 0.016667 | 400,000 | 4.35 | $6.74 |
| R-38 → R-60 | 0.009649 | 231,579 | 2.52 | $3.90 |
Every figure is 1,000 ft² × ΔU × 1,000 HDD × 24, then divided by 0.92 and by 100,000 Btu per therm. A 1,500 ft² attic in a 6,000 HDD climate going R-19 → R-49 saves 1.5 × 6 × $13.03 = $117 a year on gas heating alone.
How to read the payback number
Under about 10 years, insulation is usually an easy call in an owner-occupied house, because the measure lasts as long as the building, needs no maintenance, and improves comfort and noise on top of the energy saving. Between 10 and 25 years the decision turns on how long you will own the property and whether a rebate is available. Beyond 25 years the energy saving alone does not justify the work, though comfort, ice-dam prevention or an insurance requirement might.
Simple payback flatters nothing and discounts nothing. It ignores the time value of money, which makes it look better than a proper net present value; and it holds fuel prices flat, which makes it look worse than reality if energy prices rise. Those two errors point in opposite directions and often roughly cancel over a decade, which is why simple payback survives as a screening tool despite its crudeness.
Compare against the code minimum, not against zero. Many jurisdictions on the 2021 International Energy Conservation Code require attic insulation in the R-49 to R-60 range for most climate zones, with lower values only in the warmest zones, and ASHRAE 90.1 sets parallel requirements for commercial buildings. If you are pulling a permit, the target is set for you; check the edition your jurisdiction actually adopted, because amendments are common and states lag the model code by several cycles.
Watch what happens to the payback when you change the fuel. The same physical upgrade pays back much faster on propane or electric resistance heat than on natural gas, purely because the delivered cost per MMBtu is higher. Nothing about the insulation changed — only the value of the heat it retains.
Assumptions and limits you should know about
- Conduction only. This method prices heat conducted through the assembly. It does not price air leakage, which in a leaky attic can rival conduction and which insulation alone does not fix. Air-seal top plates, chases and can lights before you blow insulation, or you will not get the modelled saving.
- No thermal bridging correction. Nominal batt R-values ignore the studs and joists that short-circuit them. A 2×6 wall with R-19 batts performs closer to R-14 whole-assembly. Enter whole-assembly R-values if you have them.
- Degree days to a 65 °F base. If your thermostat sits at 68 °F or you set back overnight, the effective base differs and the true saving shifts by a few percent either way.
- No moisture or ventilation effects. Burying an attic in cellulose without adequate soffit ventilation and a sealed attic floor can move the dew point and cause condensation. This calculator says nothing about that risk.
- Flat fuel prices and no discounting. Simple payback assumes today's rate forever. For a capital decision over ten years or more, redo it as a discounted cash flow with a fuel escalation you can defend.
- Cooling savings are approximate. Attic heat gain in summer is driven by roof solar absorption and attic air temperature, not by outdoor air temperature alone, so the CDD term is the weaker half of this method.
Where to get your degree days
NOAA's National Centers for Environmental Information publish heating and cooling degree days by station, state and climate division, and most utilities print the month's degree days on commercial bills. Use a 30-year normal rather than last year's figure unless you are specifically reconciling one season's consumption — a single mild winter can be 15 percent below normal and will make any measure look worse than it is.
Where insulation sits among the other measures
Insulation is one line in a whole-house energy plan, and it is rarely the first line. Air sealing usually costs less per unit of saving, duct sealing in an unconditioned attic often beats both, and replacing a water heater can save more than an attic upgrade in a mild climate — run the water heating cost calculator to see whether that is true for your house. Once you have the fuel savings in hand, the emissions side follows from the same numbers through the household carbon footprint calculator.
For commercial work, this screening calculation is the same one behind an ASHRAE Level 1 audit: identify the assembly, estimate ΔUA, apply degree days, price it, rank by payback. What a Level 2 or 3 audit adds is measured air leakage, infrared verification that the insulation you are pricing is actually missing where you think it is, and hourly modelling that captures solar gains and setback schedules the degree-day method cannot see.
Finally, remember what the payback figure leaves out entirely. A well-insulated ceiling raises the interior surface temperature in winter, which raises mean radiant temperature and lets people feel comfortable at a lower air temperature. It reduces ice damming by keeping the roof deck cold. It cuts the peak load, which can let you install smaller equipment on the next replacement cycle. None of those appear in dollars per year, and all of them are reasons the measure gets built.