The design temperature mistake that skews every heat pump quote
I had a client in the Dandenong Ranges last winter send me a photo of her thermostat reading 17°C inside while the heat pump ran flat out. Outside it was minus 2, which for that part of Victoria is not even close to the coldest morning on record for the valley. Her installer had sized the unit off "the average Melbourne winter", not off what actually happens on her specific ridge. That's the mistake I see more than any other, and it rarely gets talked about because it sounds boring. It's not boring. It's the number that decides whether your heat pump keeps your house warm on the worst night of the year, or just on most nights.
The term is design temperature, sometimes called the 99% or 99.6% design dry-bulb temperature depending on which standard you're reading. It's not the coldest temperature your location has ever recorded, and it's not the average. It's the outdoor temperature your location sits at or above for 99% (or 99.6%) of winter hours, based on decades of weather station data. Manual J in the US and AS/NZS 3500-adjacent HVAC load calculations in Australia both lean on this figure, and it's the only sane basis for sizing heating equipment. Average temperature sizing is the error that quietly undersizes a huge share of installs.
Why averages lie to you about sizing
Say your region's average winter minimum is 4°C, but your design temperature (the figure that captures the cold snaps, not just the typical night) is minus 3°C. If an installer sizes capacity off the 4°C average, the heat pump will look perfectly matched most of the season. Then a cold front rolls through, outdoor temps drop to design-temp levels or below, and the unit's output isn't there anymore. We've covered separately how rated capacity shrinks exactly when you need it most, and design-temperature sizing is the flip side of that problem: if you sized off the average to begin with, you've already given away headroom before the derate even starts eating into it. I'd argue most residential quotes in Australia still skip this step entirely. Installers price off floor area and a rule-of-thumb kW-per-square-metre figure, not a proper room-by-room load calculation anchored to a real design temperature. It's faster to quote that way, and for a huge chunk of installs on the Gold Coast or in Perth it barely matters because the climate's mild enough that undersizing by a fudge factor doesn't bite. It bites in the ranges, the tablelands, anywhere with real elevation or cold-air drainage. Lorne's coastal strip and the hills behind Mountjoy Parade are a fifteen-minute drive apart and have meaningfully different winter lows — a single regional design temperature number flattens that out unless someone actually checks it.
Where the real design temperature numbers come from
In the US, ACCA's Manual J load calculation (the ANSI/ACCA 2 standard) references ASHRAE climate design data, which is published with 99% and 99.6% heating design temperatures for thousands of weather stations. A proper HVAC contractor pulls the nearest station's figure, not a statewide guess. The US Department of Energy's heat pump systems guidance is a reasonable starting point for homeowners wanting to understand how climate zone interacts with equipment selection before a contractor even shows up.
In Australia there isn't a single universal published design-temperature table homeowners can look up as easily, which is part of the problem. Installers and energy raters typically pull historical Bureau of Meteorology station data for the nearest comparable site and apply a percentile cut, but plenty don't bother and just use a climate zone average instead. If your installer can't tell you what design temperature they sized to, or point to the weather station they used, that's worth asking about directly before you sign. The Clean Energy Council's installer guidance touches on proper load calculation practice as part of accredited installer standards, though it's aimed more at solar and battery accreditation than HVAC load calcs specifically.
The elevation and microclimate problem
Design temperature tables are built off weather stations, and weather stations are in specific places. If your block sits 200 metres higher than the nearest station, or down in a frost hollow that collects cold air on still nights, the published figure undersells how cold you actually get. I've walked properties on Falls Creek Road type terrain, steep, tree-sheltered, cold-air-pooling blocks, where the nearest airport-based weather station reading is close to useless for sizing purposes. A good installer adjusts for this. Most don't, because it takes a site visit and some judgement rather than a desktop quote. This is also where a buffer gets added or doesn't. Some installers quietly oversize by 10-15% specifically to cover microclimate uncertainty, which is defensible engineering judgement. Others oversize blind without ever checking the design temp at all, which just masks the problem until the next unusually cold stretch, and creates the short-cycling and comfort issues we've gone through in why a bigger heat pump won't heat your house faster.
What this does to your running cost estimate
Design temperature sizing isn't just about keeping warm on the coldest night, it changes your whole running-cost picture. A heat pump sized correctly to design temperature runs within its efficient operating band for the vast majority of the season and only leans on its edges during the rare extreme. One undersized off an optimistic average spends more hours running near maximum output, where COP is worse, defrost cycles are more frequent, and the backup electric resistance element (if your system has one) kicks in more often than the spec sheet implies. That backup element is the expensive part; resistance heating at the margin can roughly double or triple the per-kWh heating cost compared with the heat pump running efficiently on its own. If your running-cost estimate came from a brochure COP number rather than a real load calc, it's worth reading our piece on why the COP number on the box isn't the COP you'll actually get, because undersizing and COP inflation tend to compound each other.
Questions worth asking before you sign
Ask your installer directly what design temperature they used and where it came from. A specific number, "minus 4 at 99%", with a source behind it is a good sign. A vague "we size for your area" answer is not. Ask whether they adjusted for elevation or known cold pockets on your specific block, and ask to see the load calculation itself, not just the final kW figure. A proper Manual J output or equivalent room-by-room calc runs several pages; a one-line "4.5kW should cover it" quote has skipped the actual engineering. It's also worth asking what happens at design temperature specifically, not at the average. Does the quoted unit's rated output at that specific outdoor temperature still exceed your calculated heat loss, with some margin, or does it only clear the bar at milder temperatures? If an installer can't answer that on the spot, they probably haven't actually run the numbers that way.
Getting the ducting and circuit sizing to match
None of this matters if the rest of the install can't carry the capacity you've correctly sized for. We've written about how duct and pipe run length gets left out of quotes and how circuit sizing trips up installs before day one, and both of those failure modes get worse, not better, once you've correctly sized for a cold design temperature. A bigger, correctly-sized unit needs adequately sized ductwork and circuit capacity to actually deliver that output; sizing the equipment right and then choking it with an undersized duct run gets you back to the same cold-morning problem by a different route. If you're weighing ducted options against the load calc, products like the Daikin Premium Inverter Ducted system or Mitsubishi Electric Zuba Central publish capacity curves across a range of outdoor temperatures, not just a single rated figure, which is exactly the data you want to cross-check against your design temperature rather than a single headline kW number.
What I'd tell a homeowner doing this today
Get the design temperature figure in writing as part of your quote, ask where it came from, and sanity-check it against your actual site if you're anywhere with elevation change or known cold pockets nearby. It's a five-minute conversation that separates a proper load calculation from a guess with a brand name attached. If two quotes land on wildly different kW figures for the same house, the difference is very often this exact input, not some mysterious "efficiency" variable. Our heat pumps hub has more on sizing and running costs if you want to keep digging before you commit, and if rebates are part of your decision, the rebates hub is the place to check what's current rather than trusting a number from last year's brochure.
Anyway. Ask the design temperature question. It's the cheapest insurance in the whole quoting process.
— Priya Nathan, Heat Pumps & HVAC, IceWarm
Common questions
- What is a heat pump design temperature, exactly?
- It's the outdoor temperature your location sits at or above for 99% or 99.6% of winter hours, based on long-term weather station records. It's used to size heating equipment so it still performs on the coldest typical days, not just on an average night.
- Why not just size off the coldest temperature ever recorded?
- Sizing for the absolute record low would mean oversizing equipment for the other 99.9% of the season, which hurts efficiency and comfort through short-cycling. Design temperature is a statistical compromise that covers the realistic cold snaps without wildly oversizing.
- How do I find my design temperature if my installer doesn't mention it?
- In the US, ASHRAE climate design data (referenced by Manual J) lists 99% and 99.6% figures for thousands of weather stations, often accessible through contractor load-calc software. In Australia, ask your installer which Bureau of Meteorology station and percentile they used; if they can't answer, that's a sign no real load calculation was done.
- Does elevation really change this much for a standard suburban block?
- For flat suburban blocks near a weather station, not much. For hillside, tableland, or cold-air-pooling sites, elevation and terrain can shift the effective design temperature several degrees colder than the nearest station reading, which is enough to undersize a system noticeably.
Priya covers heat pump heating, cooling and hot water systems — sizing, running costs, and what actually changes between an Australian and a US install.
Mechanical engineering background; ex-HVAC design.
More from Priya Nathan
- The humidity myth that throws off heat pump cooling estimatesHeat pump cooling quotes focus on temperature drop, but humidity removal is the part that actually decides comfort and running cost.
- Why "heat pumps don't work in the cold" is the wrong myth to worry aboutHeat pumps do work in the cold. The real myth is believing minimum operating temperature is the number that matters for your install.
- Vetting a refrigerant handling licence before heat pump work startsBefore a heat pump installer touches refrigerant lines, check their handling licence. Here's what the card actually proves and what it doesn't.
- Why a return-on-investment number for a heat pump rarely survives real lifeMost heat pump payback calculations use one winter's gas price and a brochure COP. Here's why the real payback period almost never matches the quote.
- Vetting a ducted heat pump installer: the warranty registration trapMost heat pump warranties need the installer to register the system within days. Miss that step and the manufacturer owes you nothing.
- The buffer tank mistake that quietly wrecks heat pump efficiencyA wrongly sized buffer tank forces a heat pump into short cycling, hammering efficiency and lifespan. Here's how to spot the mistake before install day.