Original data and independent reporting for the HVAC trade

How Long Does It Take to Heat a House From Cold? Hours by System, Worked Out

Homeowner checking a cold wall while waiting to see how long it takes to heat a house from cold

How long does it take to heat a house from cold? With a furnace sized at 130% of the heat loss, our model puts a house soaked at 40°F at 8 to 24.5 hours before its walls and furniture reach 68°F on a 25°F day. A 5°F setback takes 1 to 5.5 hours. The thermostat reads 68°F well before the walls do.

How many hours from 63°F, 58°F, 48°F or 40°F back to 68°F?

From 58°F, a furnace brings the whole house within 2°F of 68°F in 3 to 10.5 hours on a 25°F day; from 48°F, 6 to 19 hours. The short end is a light wood-frame house, the long end a house as heavy as one NIST measured.

HVAC Brief calculation: hours until walls and contents are within 2°F of 68°F. 2,000 sq ft house, 40,000 Btu/h loss at 5°F outdoor and 70°F indoor, soaked at the start temperature; light to heavy mass. Sun, people, defrost and duct losses ignored.
Start, then 68°F Outdoor Gas furnace, 52,000 Btu/h 3-ton heat pump plus 10 kW strips 3-ton cold-climate heat pump alone Boiler, 44,000 Btu/h baseboard
63°F (5° setback) 25°F 1 to 5.5 h 1 to 5.5 h 4 to 11.5 h 1.5 to 6 h
58°F (10° setback) 25°F 3 to 10.5 h 2.5 to 10 h 9 to 26 h 3.5 to 11.5 h
48°F (20°, a week away) 25°F 6 to 19 h 5 to 17.5 h 16.5 to 46 h 6.5 to 20.5 h
40°F (after an outage) 25°F 8 to 24.5 h 6.5 to 22.5 h 20.5 to over 48 h 8 to 26 h
58°F 5°F 5 to 14.5 h 5.5 to 16 h Never 6.5 to 20.5 h
40°F 5°F 12.5 to 35.5 h 14 to 39.5 h Never 13.5 to 41.5 h

The furnace, 52,000 Btu/h, is 130% of the design loss, inside the 100 to 140% window the ENERGY STAR National HVAC Design Report sets for heating-only furnaces. The heat pumps are ENERGY STAR listings: Carrier’s single-stage 25SCA536A**30* (AHRI 210316793) at 34,200, 20,400 and 15,800 Btu/h at 47°F, 17°F and 5°F, and Carrier’s cold-climate 27VNA036A*030* (AHRI 215679803) at 35,600, 30,600 and 25,200 Btu/h, interpolated between points. The 10 kW strips, 34,120 Btu/h, are our assumption.

“Never” means short of capacity: at 5°F the house loses 38,769 Btu/h at 68°F, and the cold-climate unit covers 65% of it.

The air warms quickly; the walls and furniture take hours to catch up.
The air warms quickly; the walls and furniture take hours to catch up. AI illustration.

Why does the air warm in 15 minutes while the walls and furniture take all day?

Because air holds about 3% of the heat the house must absorb: 288 Btu per °F in our light house, against 9,480 for drywall and furniture.

HVAC Brief calculation: heat stored per °F in a 2,000 sq ft house, from federal and national-lab property data.
What has to warm Basis Btu per °F
Air 16,000 ft³ × 0.075 lb/ft³ × 0.24 Btu/lb·°F 288
Drywall 6,000 ft² of 1/2 in. gypsum, 800 kg/m³, 1,090 J/kg·K 3,250
Furniture and contents 8 lb/ft² of floor, wood at 1,630 J/kg·K 6,229
Light house, total 4.88 Btu/°F per sq ft of floor 9,768
Heavy house (NIST NZERTF, measured) 172 W/K × 104 h over 2,700 sq ft, 12.56 per sq ft 25,118

Air density is the standard air of 10 CFR 430 Appendix M; gypsum and wood come from DOE’s EnergyPlus dataset, contents from NREL’s Building America House Simulation Protocols. Drywall at three times floor area is our assumption.

The heavy row is measured. In NIST Technical Note 1891, researchers set the 2,700 sq ft Net-Zero Energy Residential Test Facility back to about 15.6°C (60°F) on a winter night and fitted the decay: UAe = 172 W/K, time constant 104 h, cross-checked at 180 ± 8 W/K. Their product, 33,909 Btu/°F, is one house, so treat it as the heavy end.

Warming a house soaked at 40°F, 25°F outside 68°F setpoint 70 60 50 40°F 0 4 8 12 14 h walls only 62°F when thethermostat reaches 68°F first 15 min: air jumps ~10°F Furnace, light house: thermostat air Same house: walls, furniture Furnace, heavy house: air Cold-climate HP, light house
HVAC Brief model, 4,000 Btu/h·°F air-to-surface coupling. Mass from EnergyPlus, Building America and NIST TN 1891; heat pump from ENERGY STAR (AHRI 215679803).

In the first 15 minutes the model’s thermostat air jumps from 40°F to 49.7°F, since 28°F of air needs only about 8,064 Btu. After that the air can only run a few degrees ahead of the surfaces, and the climb settles to about 2.5 to 3°F an hour in the light house and 1.1 to 1.3°F in the heavy one. When the light-house thermostat is satisfied at hour 7, the walls are still at 62.2°F, which is why a recovered house feels chilly at 68°F. The AC pull-down in summer fights the same mass.

How do you work out degrees per hour for your own house?

Divide spare heat by stored heat: heat delivered minus heat lost now, over 4.9 (light) to 12.6 (heavy) Btu per °F per square foot. Heat lost now scales from the design loss in a proper load calculation.

Worked example at 55°F inside, 25°F outside: 40,000 Btu/h ÷ 65°F design difference = 615 Btu/h per °F; × 30°F = 18,462 Btu/h of loss, leaving the furnace 33,538 Btu/h spare. Divided by 9,768 Btu/°F that is 3.4°F per hour; by 25,118, 1.3°F per hour. The spare shrinks as the house warms, so the last degrees are slowest, and the normal cycling in how long a furnace should run returns only once the walls catch up.

A heat pump makes less heat as it gets colder, so recovery from a deep setback is slower.
A heat pump makes less heat as it gets colder, so recovery from a deep setback is slower. AI illustration.

Why does a heat pump recover so much slower than a furnace, and when do the strips come on?

Its output falls as the outdoor air cools while the house’s loss rises. The single-stage Carrier delivers about 24,080 Btu/h at 25°F, below the 26,462 Btu/h the house loses at 68°F, so alone it never gets back; our calculated balance point is 27.3°F. The cold-climate Carrier balances at 18.1°F, but its 5,471 Btu/h spare at 68°F lifts a light house only 0.56°F an hour.

DOE’s thermostat guidance (archived copy) calls programmable thermostats “generally not recommended for heat pumps” unless they use “special algorithms to minimize the use of backup electric resistance heat systems.” Strips make the 6.5 to 22.5 hour column possible, at resistance-heat cost; the auxiliary vs emergency heat setting decides how they share the work. Defrost is ignored, so real recoveries run longer. A cold-climate heat pump sized to the heating load closes much of the gap.

Does a boiler with baseboard or radiant floor heat a cold house faster?

Hot-water baseboard keeps pace with a furnace, because finned tube gives more heat to a cold room. Slant/Fin’s correction table at 180°F average water gives .86 with 45°F entering air against .69 at 65°F: 25% more while the room is cold.

Two things slow boilers down. Outdoor reset running cool water cuts output: at 140°F average water the 65°F factor is .38, 55% of the 180°F figure, per the Slant/Fin engineering data (distributor copy). And water, boiler and cast iron must warm first, which our model omits. DOE puts the response of steam and radiant floor systems at up to several hours and suggests starting recovery two or three hours early.

How cold will the house get with the heat off for a trip or an outage?

Cool-down mirrors warm-up: the time constant, stored heat over loss, is 15.9 hours in the light house and 40.8 in the heavy one. At 25°F outside the light house falls from 68°F to 55°F in about 6 hours, the heavy one in about 15.

HVAC Brief calculation: indoor temperature after the heat stops at 68°F. Same house, one-node decay, no sun or appliance heat.
Hours without heat 25°F outside, light / heavy 5°F outside, light / heavy
6 54°F / 62°F 48°F / 59°F
12 45°F / 57°F 35°F / 52°F
24 34°F / 49°F 19°F / 40°F
48 27°F / 38°F 8°F / 24°F

NIST’s house, at 104 hours, would fall far slower, which shows what insulation and airtightness buy. A light house left 24 hours at 25°F comes back from about 34°F, worse than the 40°F row above. The steps for holding heat during an outage buy hours on both ends, and the temperatures at which pipes freeze are the reason a vacant house should not drift this far.

When does slow warm-up mean something is wrong?

When a furnace runs steadily above about 25°F outside and the thermostat gains under 1°F an hour after the first hour. Even our heavy house gains 1.1 to 1.3°F an hour from the 50s.

The first test is the temperature rise across the furnace, supply minus return after about 5 minutes, against the nameplate range; the furnace running but house not warm guide sorts low rise, high rise and duct loss. For a heat pump below its balance point, slow is normal; a technician checks amperage on the strip heater circuit during a large call and whether the outdoor coil clears in defrost. If the thermostat starts early on its own, read the Honeywell recovery mode page. A gas smell or scorch marks at the burner are a stop line: shut the system off and have the combustion side checked.

Frequently asked questions

How long does it take to heat a house 1 degree?

Near 68°F on a 25°F day, our 52,000 Btu/h furnace has 25,538 Btu/h spare, enough for about 2.6°F an hour in a light house and 1.0°F in a heavy one, so one degree takes about 23 to 60 minutes once the walls are warm. A cold-climate heat pump with 5,471 Btu/h spare needs about 1.8 to 4.5 hours.

How long does it take to heat a house from 40 to 70 degrees?

For our 2,000 sq ft example with a furnace at 130% of its design load, 8.5 to 27 hours until the walls and furniture are within 2°F of 70°F on a 25°F day. The thermostat reads 70°F hours earlier. The range runs from a light frame house to the heavy mass NIST measured, 12.56 Btu per °F per square foot.

Does turning the thermostat up higher heat the house faster?

Not with a single-stage furnace: it burns at one fixed rate whether the setpoint is 68°F or 80°F, and the higher setting only overshoots. DOE gives the cooling version, that a colder setting “will not cool your home any faster.” On a heat pump a large jump can call the strips, which is faster but runs resistance heat at 3,412 Btu/h per kW.

Is 72 a good temperature for heat in the winter when not home?

DOE’s Energy Saver guidance points the other way: turning the thermostat back 7° to 10°F for 8 hours a day can save “as much as 10% a year,” because a cooler house loses heat more slowly. Our table shows the cost in time, 1 to 5.5 hours to recover a 5°F setback. The winter thermostat settings page has DOE’s numbers.

How we researched this

Research done 21 November 2026. We read NIST TN 1891, ENERGY STAR heat pump listings and the National HVAC Design Report, 10 CFR 430 Appendix M, the EnergyPlus material dataset, the Building America protocols, Slant/Fin engineering data and DOE Energy Saver. The two-node air and mass model (hlh_calc.py, hlh_table.py, hlh_series.py) uses 1-minute steps; assumptions are stated under each table. We excluded the top-ranking contractor and maker blogs as sources, since none cites a measured house or a rating.

Sources

  1. NIST, Omar, F. and Bushby, S. T. “A Self-Learning Algorithm for Temperature Prediction in a Single Family Residence,” Technical Note 1891, 2015. https://nvlpubs.nist.gov/nistpubs/TechnicalNotes/NIST.TN.1891.pdf. Accessed November 2026.
  2. U.S. EPA ENERGY STAR. “ENERGY STAR Certified Air Source Heat Pumps” dataset w7cv-9xjt (AHRI 210316793 and 215679803). https://data.energystar.gov/Active-Specifications/ENERGY-STAR-Certified-Air-Source-Heat-Pumps/w7cv-9xjt. Accessed November 2026.
  3. U.S. EPA ENERGY STAR. “National HVAC Design Report, Rev. 13.” https://www.energystar.gov/sites/default/files/asset/document/National%20HVAC%20Design%20Report_Rev%2013.pdf. Accessed November 2026.
  4. eCFR. “10 CFR Part 430, Subpart B, Appendix M: Uniform Test Method for Measuring the Energy Consumption of Central Air Conditioners and Heat Pumps.” https://www.ecfr.gov/current/title-10/chapter-II/subchapter-D/part-430/subpart-B/appendix-Appendix%20M%20to%20Subpart%20B%20of%20Part%20430. Accessed November 2026.
  5. U.S. DOE and NREL. “EnergyPlus ASHRAE_2005_HOF_Materials dataset,” G01a and G05. https://raw.githubusercontent.com/NREL/EnergyPlus/develop/datasets/ASHRAE_2005_HOF_Materials.idf. Accessed November 2026.
  6. NREL, Hendron, R. and Engebrecht, C. “Building America House Simulation Protocols,” 2010. https://www.osti.gov/servlets/purl/1218850. Accessed November 2026.
  7. Slant/Fin Corporation. “Engineering Data, Commercial Fin-Tube Radiation,” specification sheet 30752, distributor copy. https://www.gtsanders.com/userfiles/ad/documents/Slant_Fin_Corporation_30752_Specification_Sheet.pdf. Accessed November 2026.
  8. U.S. DOE Energy Saver. “Thermostats,” archived copy. https://web.archive.org/web/2025id_/https://www.energy.gov/energysaver/thermostats. Accessed November 2026.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *