Cold-climate heat pumps: what the spec sheet hides

You’re looking at a heat pump quote in a place that actually gets cold, and the brochure is full of SEER2 and HSPF2, which are lab-season averages. The number that decides whether you need a pile of electric backup strips on a January night is how much heat the outdoor unit still puts out at your design temperature, which is the cold hour the installer is supposed to size for. If the spec sheet only shows heating at 47°F, you don’t have that answer yet.

A heat pump moves heat. When outdoor air is 5°F, there is less heat to move, so capacity drops unless the machine is built and controlled for that. Older single-stage units dumped the job onto electric resistance (the strips in the air handler) as soon as it got unpleasant. Cold-climate models are supposed to keep a useful share of their mild-weather capacity at 5°F, and still deliver more heat per watt than a strip heater. The label on the box does not automatically mean that.

What does “cold climate” actually mean on the sticker?

ENERGY STAR has a separate Cold Climate designation on top of ordinary heat-pump certification. For ducted split systems it wants at least 8.1 HSPF2 and 15.2 SEER2. For non-ducted (mini-split) splits it wants 8.5 HSPF2 and the same 15.2 SEER2. Those are still seasonal averages. The winter test that matters is the low-ambient piece:

• Coefficient of performance (COP) at 5°F of at least 1.75. COP is heat out divided by electricity in, so 1.75 means you still get 1.75 units of heat per unit of power, which beats a strip heater sitting at 1.0.
• Heating capacity at 5°F of at least 70% of the capacity at 47°F.
• A controls verification procedure so those 5°F numbers happen with the unit’s own controls, the way it would run in a house, not with a lab tech holding the compressor at a lucky speed.

ENERGY STAR: heat pump key product criteria

If the bid says “ENERGY STAR” and does not say Cold Climate, you may be looking at a cooling-efficient machine that was never asked to heat at 5°F. Ask which row on that table they mean, and get the model number so you can look it up.

Northeast Energy Efficiency Partnerships keeps a Cold Climate Air Source Heat Pump (ccASHP) product list that utilities and rebate programs use because the ordinary HSPF2 test does not include points below 17°F and assumes electric resistance in the seasonal math. That is a real gap if your design temperature is 5°F or colder. Being on the NEEP list is not a personality contest; it is a way to see published capacity at low outdoor temperatures in one place.

NEEP: ccASHP specification and product list

Why is HSPF2 a bad winter test by itself?

HSPF2 is total heat delivered over a heating season in a defined climate region, divided by the electricity used. The season includes a lot of 40°F weather. A machine can look excellent on HSPF2 because it sips power in October and still fall on its face at 5°F. SEER2 is even less useful for this question because it is cooling.

You want a capacity table: BTU/hour (or kW) of heating at 47°F, 17°F, 5°F, and, if you live somewhere nastier, whatever your design temperature actually is. Some manufacturers publish -13°F. If the packet only has the 47°F heating number, treat it like a cooling quote with a heating brochure.

COP at 5°F tells you how expensive that remaining heat is. A unit that still puts out enough BTUs at 1.75 COP is doing the job. A unit that keeps the house warm only because 10 kW of strips are screaming will heat, and it will also blow up the January bill. The quote should say when strips come on (an outdoor cutoff temperature) and how many kilowatts of strips are in the air handler.

How should they size it for a cold house?

Installers size to a written heating and cooling load for this house (Manual J, or they should name what they used). In a cold climate the heating load is often the bigger number, so a machine picked only for summer can be short in January. Ask them to show heating capacity at design temperature, not “3-ton, that’s what we always put in this square footage.”

If they size the heat pump to 100% of the heating load at 5°F, you may get a unit that is large for July, which is a humidity and short-cycle problem unless it can turn down. Variable-speed equipment is how people try to have both. If they size to the cooling load and let strips cover the rest of the heating load, you need that strip kilowatt number and a serious talk about your electric rate.

Dual fuel (heat pump plus a gas furnace) is the other way through this. The heat pump does most of the year, and the furnace takes over below a written outdoor temperature. That only works if the control actually switches and the cutoff is on the bid. The heat pump vs furnace guide is the fuel decision; this page is whether the heat pump half of that pair can do any real work in January.

A leaky attic makes every one of these numbers worse. If the house is heating the sky, you will buy a bigger outdoor unit or more strip heat. The weatherization guide is the order-of-work version of that.

What should I look at on the spec sheet?

Bring this list to the quote:

1. AHRI certified combination, indoor coil or air handler matched to the outdoor unit. Mix-and-match indoor parts can void the rating and some rebates.
2. Heating capacity at 5°F and at your design temperature, in BTU/hour, not only HSPF2.
3. COP at 5°F. ENERGY STAR Cold Climate wants 1.75 or better at that point.
4. Percent of 47°F capacity still there at 5°F. ENERGY STAR Cold Climate wants 70% or better.
5. Backup heat: strip kilowatts, or dual-fuel furnace model, plus the outdoor cutoff in writing.
6. Defrost. Heat pumps frost up in cold, damp weather and run a defrost cycle that can blow cool air unless the control and backup heat are set up. Ask what the house feels like during defrost.
7. Sound and setback. A unit that ramps hard at 5°F is louder than the 47°F brochure. If it sits under a bedroom window, that matters.
8. Whether “cold climate” is ENERGY STAR Cold Climate, NEEP-listed, or a dealer adjective.

EnergySage’s 2026 marketplace average for a heat pump install is still about $15,393 before incentives, with ducted around $14,529 and whole-home mini-splits around $25,957. Cold-climate equipment, long line sets, and a panel upgrade move you off those averages. Use them to sanity-check, not to pick a logo.

EnergySage: heat pump cost

The federal 25C credit is done for units placed in service after 2025. Don’t let a 2026 bid subtract $2,000 of IRS money. State and utility programs that require NEEP or ENERGY STAR Cold Climate are local, and they care about the model number.

IRS: Energy Efficient Home Improvement Credit

A ducted vs ductless choice still sits underneath this. Mini-splits often publish strong low-ambient tables because the indoor heads are in the room, not at the end of leaky attic flex. Ducted cold-climate units exist too, and they keep the vents you already have. Pick distribution first if you already know the ducts are junk or excellent; pick winter capacity in either case.

What goes wrong after you sign?

The outdoor unit is a cooling-first model with a “hyper-heat” sticker and no 5°F table. Strips run all night because the cutoff is set to 35°F “to be safe.” The indoor coil isn’t the AHRI match. The load was square footage. Nobody mentioned defrost, and you think the system is broken every time the outdoor unit steams. Ducts in a cold attic steal the extra capacity you paid for.

A cheap bid that is 40% under the others is often missing winter capacity, electrical, or both. The quote checklist is the markup list; add the 5°F row to it.

How do I decide in one afternoon?

1. Write down last January’s coldest night you remember, or ask the bidder for the design temperature they used. If they don’t have one, they didn’t size the house.
2. Get the capacity table at that temperature, plus COP at 5°F if your design is around there.
3. Count backup: strips (kW) or furnace, and the cutoff.
4. Confirm ENERGY STAR Cold Climate or a NEEP listing if a rebate needs it, using the actual model numbers.
5. If two bids disagree by a ton of heating at 5°F, you are not comparing the same job.

You can heat a cold house with a heat pump. You cannot tell that from HSPF2 alone.