Heat pumps have a reputation problem in cold country. Plenty of people have heard that they "stop working" once it drops into the teens, and that the backup heat strips do the real work all winter. Newer cold-climate models claim to have fixed that.
The U.S. Department of Energy put those claims to a real-world test. In its Cold Climate Heat Pump Technology Challenge, eight major HVAC makers built prototype heat pumps to a tough cold-weather spec, and Pacific Northwest National Laboratory (PNNL) then measured them in real homes across the northern U.S. and Canada for up to two winters. The final report came out in January 2025.
The short answer: the units kept heating efficiently well below freezing, and backup heat was a small share of the total at most temperatures. But the numbers also show why a whole-house heat pump is a big load for an off-grid solar system, and how ordinary thermostat habits quietly switched on the expensive backup heat.
The study in one paragraph
Who and what: PNNL and Guidehouse, with Natural Resources Canada and the National Research Council of Canada, monitored prototype cold-climate heat pumps (CCHPs) from Bosch, Carrier, Daikin, Johnson Controls, Lennox, Midea, Rheem and Trane. 22 units finished the field validation, which ran from winter 2022 through early fall 2024. U.S. units were in occupied single-family homes; the Canadian units were in a mix of occupied homes and unoccupied lab homes with simulated occupancy. Each home was monitored for 10 to 91 weeks, giving between 898 and 3,099 hours of heating data per site.
The equipment: all were centrally ducted, variable-speed systems with electric resistance backup ("auxiliary") heat. Most were 3-ton (36,000 Btu/hour) units, plus three 4.5-ton, one 4-ton and one 5-ton. Each was sized to the home with an ACCA Manual J load calculation.
Study type: an instrumented field study published as a national-lab technical report (PNNL-37127). It is a government research report, not a peer-reviewed journal article, and the units were prototypes, not the exact models you can buy today. It did not look at utility bills.
What they found
The key number is COP, or coefficient of performance: units of heat delivered per unit of electricity used. Electric resistance heat (a space heater or backup heat strip) has a COP of 1. A COP of 2 means twice the heat for the same electricity.
| Measure (heat pump only, no backup or defrost) | Result |
|---|---|
| Median heating COP, outdoor air below 30°F | 1.6 to 2.7, depending on the temperature bin |
| Median heating COP at 0 to 5°F | 1.9 |
| ENERGY STAR cold-climate requirement (for comparison) | COP above 1.75 at 5°F |
| Median cooling COP (excluding two sites that ran cooling on cool nights) | 3.4 to 5.2 |
A few other findings matter for anyone living with one:
- Backup heat was mostly a small helper. The units ran mainly in heat-pump-only mode even at the coldest temperatures observed. Backup heat started showing up at around 30 to 35°F, but backup-only operation was minimal above 0°F. Excluding defrost, the median share of heating energy from backup heat was around 11% at 5°F, falling to 0% in milder weather.
- It got worse at the very bottom of the scale. Below 5°F, backup heat use rose sharply. Only a few sites had data that cold, and three of the four sites in the two coldest bins used more backup heat than average at every temperature, so the researchers caution that this part of the data may be skewed. In the -10 to 0°F range, backup heat was on about 25% of the time.
- Defrost was a modest cost. Between 5 and 35°F, the median share of energy spent on defrost was 5% to 8%. Units that ran their fans without backup heat during defrost blew air below 60°F on average for those periods, though no homeowner complained.
- The backup heaters are big. The first stage of backup heat was typically about 5 kW, and total backup capacity was about 15 to 20 kW per system.
- Electric draw in cold weather. Below 30°F, median input power in heat-pump-only mode was 0.5 to 1.1 kW per ton of capacity, or 0.6 to 1.4 kW per ton including backup heat. That figure is averaged across all operating time in each temperature range, including idle and off periods.
- Thermostat setbacks triggered backup heat. At several homes, a morning setback (or a bump of a few degrees at 4 pm) made the heat pump call for higher stages of backup heat to catch up. A furnace responds almost instantly to a new setpoint; a variable-speed heat pump ramps up slowly, so the controls bring in the resistance heat. The report suggests a constant setpoint, or a gradual setback if you must use one.
- Homeowners were a bit happier. Average heating satisfaction rose from 3.61 to 3.88 on a 1-to-5 scale between the pre-install survey and the second year.
- Installation wasn't simple. At 21 sites with data, 90% needed new breakers, 55% needed new wiring for backup heat, 38% a new electrical panel and 24% a service line upgrade; 76% needed some duct work. Electrical upgrade costs ran $3,750 to $8,970, and combined electrical and mechanical upgrades $3,258 to $18,400, before the cost of the heat pump itself.
What it means for an off-grid home
Most of this study was done in grid-tied homes, so read it as evidence about the heat pumps, not about running them off-grid. With that caveat, three practical lessons stand out.
1. A COP of about 2 at 0-5°F is real, but it's still a big electric load. Here is a rough, illustrative bit of arithmetic using the report's median figures (not a measured off-grid result): a 3-ton unit at 0.5 to 1.1 kW per ton averages about 1.5 to 3.3 kW in weather below 30°F. Over 24 hours that's roughly 36 to 79 kWh a day. Winter is also when solar harvest is lowest. For most off-grid homes, a central ducted heat pump as the only heat source in a cold climate means a very large array and battery bank, or a generator. Run your own numbers with the Off-Grid Solar Calculator and the Battery Bank Calculator before you commit.
2. The backup heat strips are the real danger to a battery bank. A single 5 kW stage, let alone 15 to 20 kW of total backup, can exceed what many off-grid inverters can deliver at once. If you have a heat pump on solar, find out how its backup heat is wired and controlled, and consider whether a wood stove or another non-electric source should be your cold-snap backup instead of resistance strips. Our wood stove vs rocket mass heater guide covers those options.
3. Your thermostat habits matter more than you'd think. The field data showed backup heat kicking in after setbacks. If you have a heat pump, keep a steady setpoint or use a gradual setback, and check that the switchover temperature (the outdoor temperature below which backup heat is allowed) isn't set higher than your unit needs.
There's a bigger lesson behind all three: the smaller your heating load, the more any of this works. The report itself notes that it couldn't study weatherization or homeowner education and flags both for a second phase. Insulation, air sealing and passive solar gain shrink the load before you buy any equipment. The Heat & Cooling hub walks through that order of operations.
What's still unproven
- Commercial units may differ. These were prototypes; the report says how production models will compare is unknown.
- Mild winters. Many sites had milder winters than normal, and only 8 of the 22 sites logged 10 or more hours below 0°F. Data in the coldest bins is thin.
- Ducted systems only. Ductless mini-splits, which many off-grid cabins use, weren't tested here.
- No bills, no off-grid systems. The study didn't analyze utility bills and didn't test the units on solar and batteries.
- No controlled thermostat data. The researchers couldn't see thermostat settings directly, so the setback explanation is their best reading of the data, not a controlled test.
What to use or try
- Measure before you size. An Emporia Vue 3 Home Energy Monitor clamps onto circuits in your electrical panel (the 16-sensor kit includes sixteen 50A branch sensors). Put a sensor on your heat pump and backup heat circuits and you can see for yourself how often the strips run and how much your heating really draws on cold days. Two caveats: it needs 2.4 GHz Wi-Fi with an internet connection, and panel work should be done by a qualified electrician.
- Shrink the load first. The Passive Solar House by James Kachadorian, from our Regenerative Library, covers siting, insulation and using windows as solar collectors so the house needs less heat to begin with, which makes any heat pump, stove or solar system easier to size.
Sources
- Mendon, V., Keene, K., Rosenberg, S., Rotondo, J. A., Nwe, K., Young, J., Wind, W., & Goetzler, B. (2025). Performance Results from DOE Cold Climate Heat Pump Challenge Field Validation (PNNL-37127). Pacific Northwest National Laboratory, prepared for the U.S. Department of Energy. https://www.pnnl.gov/main/publications/external/technical_reports/PNNL-37127.pdf
- U.S. Department of Energy, Building Technologies Office. Residential Cold Climate Heat Pump Technology Challenge specification (October 2021), as cited in PNNL-37127. https://www.energy.gov/sites/default/files/2021-10/bto-cchp-tech-challenge-spec-102521.pdf
- Emporia Energy. Emporia Vue 3 Home Energy Monitor product page (checked November 15, 2026). https://shop.emporiaenergy.com/products/emporia-vue-3
