Heat pump sizing · Canada

Heat pump sizing for Canadian homes: Manual J explained

The single most common reason a heat pump fails to keep a Canadian home warm in February is sizing. The unit was specified using the contractor's quick rule-of-thumb instead of a Manual J load calculation, was sized for the milder US climate the manufacturer marketed to, or was sized for cooling load instead of heating. This guide covers what Manual J actually calculates, the Canadian design-temperature traps that catch US-trained installers, how to read a load report, and the sizing rules-of-thumb that work if you don't have time for a full calculation.

Why sizing matters more in Canada

Heat pump sizing matters in any climate, but Canadian winters punish a mis-sized system in ways that don't happen in milder regions. An oversized air conditioner in Texas runs short cycles and feels muggy. An oversized heat pump in Winnipeg runs short cycles, fails to dehumidify in summer, and short-cycles its compressor in -30°C weather where the compressor needs continuous operation to maintain low-side refrigerant temperature. A 20% oversized cold-climate heat pump can shorten compressor life by half.

An undersized heat pump is even worse. It runs at 100% capacity continuously below your balance point, fails to keep the house warm, and triggers backup electric resistance heat — which is the most expensive way to heat anything in Canada outside of propane in remote communities. A home with a heat pump that's 25% undersized for the climate ends up with January electricity bills that exceed what the previous gas furnace cost, and the homeowner concludes (incorrectly) that heat pumps don't work in Canada.

The federal Greener Homes Loan eligibility criteria explicitly require sizing to meet 100% of heating load at the provincial design temperature without resistance backup. Loans for systems sized below this threshold can be clawed back during post-retrofit verification.

What Manual J actually is

Manual J is the residential heating and cooling load calculation procedure published by the Air Conditioning Contractors of America (ACCA). It's an industry-standard methodology that takes a detailed home audit and outputs a number: how many BTU per hour your home loses to the outside at the design temperature (and gains in summer at the design cooling temperature).

A Manual J calculation accounts for:

  • Exterior surface areas by orientation. North walls lose more heat than south walls because the south wall has solar gain to offset losses during the day. East/west walls vary by sun-hour exposure.
  • Window and door specifications. Single-pane vs double-pane vs triple-pane, low-E coatings, frame material, glazing area by orientation, infiltration rate around the frame.
  • Insulation R-values for walls, attic, basement, slab, exposed floors.
  • Infiltration rate. Typically measured with a blower-door test. Pre-1980 Canadian homes commonly come in at 8-15 air changes per hour (ACH50). Modern code-built homes are 2-4 ACH50. Net-zero ready homes are below 1.5 ACH50.
  • Internal gains. Occupant body heat, lighting, electronics, cooking — usually adds 600-1500 W of constant background heating.
  • Design temperatures. Indoor setpoint (Canadian standard: 21°C) and outdoor design temperature (varies by location — see table below).
  • Duct losses for ducted systems, particularly significant when ducts run through unconditioned attic or crawl space.

The output is a room-by-room BTU/hr requirement and a whole-home total. The whole-home heating BTU/hr is what determines the heat pump's required capacity at the design temperature.

Canadian design temperatures by city

Design temperature is defined as the 99% winter design temperature — the temperature your area falls below only 1% of winter hours. Heat pumps are sized to provide 100% of heating capacity at this temperature, with backup heat handling colder excursions. The 1% threshold is the Canadian building-code standard, drawn from National Research Council Canada (NRC) climate data.

Approximate 99% winter design temperatures across Canadian cities:

  • Vancouver, BC: -7°C
  • Victoria, BC: -5°C
  • Halifax, NS: -16°C
  • Charlottetown, PE: -19°C
  • Saint John, NB: -18°C
  • St. John's, NL: -16°C
  • Toronto, ON: -18°C
  • Ottawa, ON: -22°C
  • Montreal, QC: -23°C
  • Quebec City, QC: -25°C
  • Sudbury, ON: -28°C
  • Thunder Bay, ON: -30°C
  • Winnipeg, MB: -32°C
  • Saskatoon, SK: -34°C
  • Regina, SK: -33°C
  • Calgary, AB: -27°C
  • Edmonton, AB: -30°C
  • Fort McMurray, AB: -36°C
  • Whitehorse, YT: -41°C
  • Yellowknife, NT: -41°C
  • Iqaluit, NU: -43°C

Notice how broad the range is. A heat pump sized correctly for Vancouver (load at -7°C) is undersized by 60%+ for Saskatoon (load at -34°C) on an identical home. This is the failure mode when a contractor uses a generic rule of thumb without referencing local design temperature.

When rules-of-thumb work and when they don't

Industry rules-of-thumb exist because Manual J is time-consuming and not every quote justifies a full calculation. The most common Canadian sizing heuristics:

  • 20-30 BTU/hr per square foot for code-built modern homes (post-2010) in continental climate zones (Ontario, southern Quebec, southern BC interior).
  • 35-45 BTU/hr per square foot for older homes (pre-1980) in continental climates.
  • 45-60 BTU/hr per square foot for older homes in cold-continental climates (prairies, northern Quebec, Maritime interior).
  • 15-25 BTU/hr per square foot for modern homes in mild coastal climates (south coast BC).

For a 2000 sq ft home, these ranges produce required heating capacity of:

  • Modern Vancouver home: 30,000-50,000 BTU/hr (2.5-4.0 tons)
  • Older Toronto home: 70,000-90,000 BTU/hr (5.8-7.5 tons) — note this exceeds typical residential heat pump capacity, suggesting the home needs envelope upgrades before heat pump retrofit
  • Modern Calgary home: 50,000-70,000 BTU/hr (4.0-5.8 tons)
  • Older Winnipeg home: 90,000-120,000 BTU/hr (7.5-10 tons) — again, likely needs envelope work first

Rules of thumb are useful for initial conversation but break down at three places: very old homes (where the actual heat loss is highly variable), very new homes (where modern construction can be 40-50% better than the rule of thumb suggests), and any home where windows occupy more than 20% of wall area (window losses scale linearly with area and dominate the calculation). For any of these, a real Manual J calculation is the right move.

How to read a Manual J report from your installer

A complete Manual J report should be 8-15 pages. A one-page sheet that says "Heat Loss: 48,000 BTU/hr" is not a Manual J report — it's a guess. When your installer provides a load calculation, look for:

  1. Indoor and outdoor design temperatures on the first page. Indoor should be 21°C. Outdoor should match your local 99% design temperature from the table above. If the outdoor design is off by more than 2°C from the table, ask why.
  2. Room-by-room breakdown. Every conditioned room should appear with its own BTU/hr loss. The total of the rooms should equal the whole-home total. If only a whole-home number is shown, the calculation isn't a real Manual J.
  3. Window inventory. Each window listed with size, orientation, glazing type, and U-factor. Windows are usually the largest single loss component in a Canadian home.
  4. Infiltration rate in ACH50 or CFM50. If the home has had a blower-door test (required for EnerGuide audits), the measured value should be used. Otherwise, the report should state the assumed value and why (typically based on home age + construction quality).
  5. Final whole-home heating BTU/hr at design temperature. This is the number that must match (not exceed) the heat pump's rated heating capacity at the design temperature.

Manual J reports are most often produced using Wrightsoft, Elite, or HVAC-Calc software. Any of these is fine. Excel spreadsheets with hand-entered formulas are usually fine too if they cover all the elements above. PDF "load worksheets" with just a few line items aren't real Manual J.

Cold-climate heat pump sizing rules

Cold-climate heat pumps (CCHPs) have a particular sizing twist: their rated capacity is measured at +8°C, but their actual delivered capacity drops as outdoor temperature falls. A 4-ton CCHP rated to deliver 48,000 BTU/hr at +8°C might deliver only 36,000 BTU/hr at -15°C and 24,000 BTU/hr at -30°C.

Your installer should provide the manufacturer's capacity-vs-temperature curve for the specific model. Look for the BTU/hr delivered at your provincial design temperature. That number must equal or exceed your home's Manual J heating load.

Example: a 3000 sq ft 1990s Ontario home with measured ACH50 of 6 has a Manual J calculated heat loss of 56,000 BTU/hr at the Toronto design temperature of -18°C. The installer recommends a 5-ton ducted CCHP (rated 60,000 BTU/hr at +8°C). Checking the capacity curve: this model delivers 42,000 BTU/hr at -18°C — which is 14,000 BTU/hr short. The correct specification is either a larger CCHP (5.5-6 ton with capacity-derated of 52,000-56,000 BTU/hr at -18°C) or a dual-fuel setup with the existing furnace handling the gap below -10°C.

Get a quote with sizing done correctly

Every installer in our directory runs a real Manual J calculation before quoting — not a square-footage rule of thumb. If they're proposing a dual-fuel install, the balance point is explicit. If a CCHP is proposed, the capacity-at-design-temperature is verified against the Manual J number. Quotes include the load calculation report.

Get a free quote within 24 hours


Sources

ACCA Manual J — Residential Load Calculation 8th Edition. National Research Council Canada (NRC) — climate design data for Canadian cities. CSA F280-12 — Determining the Required Capacity of Residential Space Heating and Cooling Appliances. ENERGY STAR Canada Cold Climate Heat Pump Specifications. AHRI Directory of Certified Product Performance — capacity-temperature curves. Provincial building code amendments where they alter the 99% design temperature methodology (Ontario Building Code Supplementary Standard SB-12, BC Building Code Part 9). Last full source check: 2026-05-20.