Canadian heating design temperatures by city

Why this is the number that matters

A heat loss calculation is, at its core, a temperature difference multiplied by how badly the building leaks. Every envelope assembly, every window, every air change — all of it scales with the gap between the indoor setpoint and the outdoor design condition. Get that gap wrong and the whole calculation is wrong by roughly the same proportion, no matter how carefully you measured the walls.

Canada makes this unusually unforgiving. Victoria’s heating design temperature is −1.3 °C. Yellowknife’s is −39.2 °C. Against F280’s 22 °C indoor heating setpoint that is a design ΔT of about 23 K versus 61 K — a factor of roughly 2.6 on the same house. A calculator that offers you “cold” and “very cold” as climate options is discarding the one input that does the most work.

What “2.5%” means, and why it is not the record low

The 2.5% January design dry-bulb is the temperature the location stays above for 97.5% of January hours. It is colder than the site is for all but roughly eighteen hours of that month, and it is deliberately not the coldest night on record.

That choice is not laziness, it is the whole point. Sizing to a record cold snap gives you equipment that is oversized for every other hour of the winter. An oversized furnace short-cycles, swings room temperature, and runs its blower in bursts instead of the long steady cycles that actually mix a house. An oversized heat pump is worse, because it also gives up the part-load efficiency you paid for. The design condition is a deliberate bet that the handful of hours below it are covered by the building’s thermal mass, the setback recovery margin, and — on a heat pump — the supplementary heat that is already sized in.

HDD18 is an input, not trivia

Heating degree-days base 18 °C measure how far and how long a location sits below 18 °C over a year. In CSA F280-12 this is not just a climate statistic to quote in a report — it feeds the infiltration method, so it changes the calculated load directly.

It is also the fastest way to compare two places honestly. Toronto and Winnipeg differ by roughly 2,000 degree-days. Two identical houses, one in each city, do not use “a bit more” heat in Winnipeg; they are in different energy regimes, and the equipment, the ventilation strategy and the payback maths for a heat pump all move with it.

Design conditions by city

Sorted coldest first. The station column is the NRCan weather station that the listed postal code actually resolves to — see the note below the table on why those sometimes differ from the city name.

Why your postal code may resolve to another town

There are 403 stations with published HOT2000 design data and vastly more communities, so every forward sortation area is mapped to the nearest station that has real numbers. Most of the time that is invisible and harmless — design conditions change slowly across a region.

Sometimes it is visible in the table above. Hamilton resolves to Burlington Piers, across the bay. Gatineau resolves to Ottawa Intl, across the river. Vancouver’s downtown FSA resolves to West Vancouver. In each case the station is close enough that the substitution is defensible.

Where it deserves a second look is anywhere terrain changes fast — a valley bottom versus the plateau above it, a lakeshore versus twenty kilometres inland — or anywhere in the territories, where a single FSA can span an area larger than most provinces. Nunavut is the clearest case: one FSA covers communities more than 1,500 km apart, so a nearest-station match there can be a long way off. If the resolved station is not obviously your climate, use the design temperature your AHJ or the provincial code appendix publishes for that municipality instead.

Using these in a calculation

These are inputs to a load calculation, not a substitute for one. The heating design dry-bulb sets the ΔT for conduction and infiltration; the cooling dry-bulb and its mean coincident wet-bulb split the cooling load into sensible and latent; HDD18 drives the F280 infiltration method.

If you want to see them applied rather than tabulated, the free CSA F280 calculator runs a whole-house block load from a postal code in the browser and shows you which station and which temperatures it used. For the full calculation offline — blower-door airtightness, HRV and ERV recovery, window U-factor and SHGC overrides, and a report you can hand over — that is the CSA F280 app.

FAQ