The two sums, and why they are kept apart
Heat leaves a house two ways in winter. It conducts through the surfaces, and it rides out on air that escapes and is replaced by cold air. The arithmetic for the first is area times U times temperature difference, where U is the reciprocal of R. A 1,600 square foot wall at R-13 has a UA of 123 BTU per hour per degree, so at a 60 degree difference it sheds about 7,380 BTU/h. Do that for every surface, add the UA figures together, and one multiplication by the design temperature difference gives the whole conduction total.
The second sum is the air. Air changes per hour times volume divided by 60 gives CFM, and 1.08 times CFM times temperature difference gives the sensible load. The 1.08 is not arbitrary: it is 60 minutes times the density of air at standard conditions, 0.075 pounds per cubic foot, times its specific heat of 0.24 BTU per pound per degree. Twelve thousand cubic feet at 0.35 air changes is 70 CFM, and at 60 degrees that is about 4,540 BTU/h.
Keeping the two separate matters because the fix is different. If the infiltration share is a third of your total, more attic insulation will move the number very little and air sealing will move it a lot. The insulation calculator covers the surface side and the home insulation guide explains where assemblies fall short of their labels.
Whole-assembly R-values, not cavity labels
The single largest error people make on this page is entering the number printed on the insulation. A 2x6 wall with R-21 batts is not an R-21 wall. Studs, plates, headers and rim joists conduct around the insulation, and in typical framing they occupy something like a fifth to a quarter of the wall area at a fraction of the cavity R-value. The assembly figure, after that thermal bridging plus the sheathing, siding and drywall, commonly lands well below the batt rating. If you enter the label value your heat loss will read low, and low is the direction that leads to a system that cannot hold temperature on the coldest night.
| Entered as | What it should represent | Common mistake |
|---|---|---|
| Wall R | Whole assembly including framing fraction | Using the batt label |
| Ceiling R | Insulation plus air films, less any compressed edges | Ignoring the compressed perimeter over the top plate |
| Window U | Whole-window U-factor from the rating label | Using centre-of-glass, which is better than the whole unit |
| Wall area | Gross wall minus windows and doors | Counting the glass twice |
Choosing an infiltration rate
This is the softest input on the page and it deserves suspicion. Natural infiltration at design conditions is not the blower-door number. A blower door reports air changes at 50 pascals, a pressure far beyond anything weather produces, and converting that to a natural rate involves a divisor that depends on climate, height and shielding. If someone hands you an ACH50 figure, do not type it here.
Absent a measurement, people commonly work in a band: a tight, recently built and air-sealed house well under a third of an air change per hour, an average existing house somewhere near it, and an old leaky house with original windows several times higher. The honest position is that the band is wide and your house sits somewhere in it. Run the calculation twice, once at the low end and once at the high end, and see how much the answer moves. If it barely moves, the uncertainty does not matter for your decision. If it moves by half, you have just learned that a blower-door test is the cheapest thing you could buy. The home energy audit guide covers what a test actually reports, and the air changes calculator handles the volume-to-CFM conversion on its own.
Where this stops, and why that boundary is real
A full residential load calculation is not this arithmetic with better inputs. It is a different exercise. It works room by room rather than whole-house, so it can tell you how much air each room needs rather than only what the equipment must produce. It accounts for the orientation of every glazed opening and the shading over it, because a south wall and a north wall of identical construction do not behave alike in either season. It separates sensible from latent load, which is what determines whether a system dehumidifies. It adds internal gains from occupants and appliances, subtracts nothing for wishful thinking, and applies the design conditions published for your specific location rather than a temperature you picked.
Equipment selection follows from that calculation. The reason this matters more in cooling than in heating is that an oversized cooling system satisfies the thermostat on sensible temperature long before it has removed enough moisture, so the house ends up cold and clammy, and the compressor short cycles its way to an early death. Oversizing is the most common error in the trade and the most expensive one to live with. If you find yourself reaching for a rule of thumb in square feet per ton to check this page against, that is the moment to get a real calculation instead: those rules were derived from housing stock and climates that may have nothing to do with yours.
Once you have a defensible load, the sizing pages downstream are the mini-split sizing calculator for ductless equipment, the AC BTU sizing calculator for the cooling side, and the radiant floor loop calculator if the heat is going into the floor. For what the result costs to run, the heating cost guide and the heating cost savings calculator take it from BTU to dollars.
Questions people ask
Can I use this instead of a Manual J?
No, and it is not built to pretend otherwise. This page sums conduction and infiltration from numbers you supply, which is two of the terms a real load calculation contains. It does not model orientation, solar gain through glass, shading, internal gains from people and equipment, latent load, duct losses in unconditioned space, below-grade ground coupling, or your location published design conditions. Those terms are not rounding errors; solar gain alone can dominate a west-facing room on a summer afternoon. Use this to understand where your envelope loses heat and to compare one improvement against another. Use a proper calculation to select equipment.
Why is the summer number labelled incomplete?
Because it is only the conduction and sensible infiltration part of a cooling load, and in most houses that is a minority of the total. Sunlight through windows, heat from occupants and appliances, and the latent load of removing moisture from the air are all missing. A cooling system sized from the partial figure would be badly undersized on a sunny day and, more insidiously, could be badly mismatched on the latent side. The page shows the number because it is useful for comparing envelope changes against each other, and labels it so nobody carries it to a supplier.
What temperature difference should I use?
Use your area published winter design temperature for the outdoor figure, not the coldest reading you remember. Design temperatures are set at a percentile so that the system is sized for nearly all hours rather than for the worst hour in a decade, which is deliberate: sizing to the record low guarantees an oversized system that spends the rest of the winter short cycling. Indoors, use the temperature you actually keep the house at. The difference between those two is the multiplier for everything on this page, so it moves the answer proportionally and it is worth getting from a published source rather than a guess.
My infiltration share came out very high. Is that wrong?
It might be right, and if it is, it is the most actionable result the page can give you. In an older house with a well-insulated attic and leaky walls, rim joists, penetrations and a chimney chase, air leakage can genuinely rival the surface losses. The check is to look at the ACH you entered and ask where it came from. If it was a guess, run the calculation at both ends of a plausible range and see whether your decision changes. If it came from converting a blower-door number by dividing by some factor, be aware that the divisor is climate and house dependent and the conversion is approximate.
Does the calculator handle a basement or a slab?
Only crudely, and you should know where the crudeness is. Below-grade walls and slabs lose heat to the ground, and the ground is neither at outdoor air temperature nor at a constant one; the loss depends on depth, soil, and the perimeter rather than the area, and it lags the season by months. Entering a basement wall as an above-grade surface at outdoor design temperature will overstate its loss substantially. If your house has significant below-grade surface, set the floor area to zero, treat the result as the above-grade portion, and get the below-grade term from a method that models it properly.