Heat Pumps Explained: How They Work and What Affects Running Costs

A heat pump can lower a home’s heating bill or leave a household disappointed, and the difference is usually not the marketing label. It is the match between the equipment, the house, the climate, and the price of electricity. This guide explains the mechanics in plain language, separates efficiency ratings from bill outcomes, and shows how to estimate annual running cost with your own numbers. The dollar figures are illustrative, not a quote or a forecast.

What a heat pump does, and what it does not do

A heat pump uses a refrigerant cycle to move heat from one place to another. In heating mode it extracts heat from outdoor air, the ground, water, or waste heat, then amplifies and transfers that heat indoors. In cooling mode it reverses the flow and moves heat out of the home. The U.S. Department of Energy describes heat pumps as an efficient alternative to furnaces and air conditioners because they transfer heat rather than generate it by burning fuel.

That is why a heat pump can put more heat into a room than the electricity it consumes. The International Energy Agency notes that the output is normally several times greater than the electrical energy used to run the compressor and fans, with a coefficient of performance (COP) around four for a typical household heat pump at a given condition. The IEA also describes current models as 3 to 5 times more energy efficient than gas boilers. Energy efficiency is not the same as a lower bill, however; fuel prices and real operating conditions decide that.

Most homes use an air-source heat pump, which looks like an outdoor air-conditioning unit and transfers heat between outdoor air and the home. Ground-source systems use buried loops and can be more efficient in some conditions but cost more to install and require suitable land or drilling. Energy Advice England notes that heat pumps work best with good insulation and lower flow temperatures, and that they do not suit every property without changes to radiators, pipework, or insulation. In a U.S. home, the equivalent point is that ducts, radiators, and the building shell matter as much as the outdoor unit.

Heating, cooling, and hot water are separate jobs

A ducted air-source heat pump can both heat and cool, which is one reason it is common in the United States. The Department of Energy notes that heat pumps provide cooling in summer as well as heating in winter. A heat pump water heater, however, is a different appliance with its own efficiency rating and its own electricity use. Pool heaters and clothes dryers are separate loads too. When you compare quotes, keep these jobs separate so you do not credit one appliance with savings from another.

Some homes use hybrid systems that combine a heat pump with a gas furnace. The International Energy Agency notes that heat pumps can be combined with other heating systems in hybrid configurations. A hybrid can be useful where both fuels are available, but it adds control complexity: the controls must choose which source to run, and that choice should reflect local fuel prices, comfort, and maintenance needs.

The efficiency numbers that matter

Power, energy, and heat are different units

Power is the rate at which energy is used or delivered, measured in watts (W) or kilowatts (kW). Energy is power multiplied by time, measured in kilowatt-hours (kWh). A heat pump’s capacity is a rate of heat delivery, often expressed in British thermal units per hour (Btu/h) or in tons; one ton of capacity equals 12,000 Btu/h. Its electricity draw at a moment is measured in kW. Annual consumption is measured in kWh. A 3-ton unit is not a 3 kWh unit; it is a machine rated to deliver 36,000 Btu/h under specified conditions.

For comparisons, keep these conversions handy: 1 kWh of electricity equals about 3,412 Btu of energy; 1 therm of natural gas equals 100,000 Btu; 1 million Btu (MMBtu) equals 1,000,000 Btu. A furnace’s efficiency is often stated as annual fuel utilization efficiency (AFUE), which is the share of fuel energy that becomes useful heat over a test season. A heat pump’s efficiency is stated as COP at a condition or as a seasonal rating such as HSPF2. Those numbers are not interchangeable.

COP: efficiency at a moment

COP is heat output divided by electrical input at a specific operating condition. A COP of 3 means that, at that condition, the heat pump delivers 3 kWh of heat for each 1 kWh of electricity. The IEA’s example of a typical household COP around four is a useful scale for understanding why heat pumps can beat electric resistance heating, which has a COP of about 1. The U.K. government’s public guidance gives a similar plain-language scale: around three units of heat for every unit of electricity. But a single COP is not an annual average. It changes with outdoor temperature, indoor temperature, defrost cycles, fan power, and how the system is controlled.

HSPF2: a seasonal heating rating

In the U.S. market, HSPF2 is a heating seasonal performance factor expressed in Btu of heat per watt-hour (Wh) of electricity over a standardized season. Higher is better. To convert HSPF2 to an approximate seasonal COP, multiply by 1,000 to get Btu per kWh, then divide by 3,412 Btu per kWh. In short: seasonal COP ≈ HSPF2 × 0.293. An HSPF2 of 9 is about COP 2.64; HSPF2 10 is about COP 2.93; HSPF2 8 is about COP 2.34. Those are arithmetic conversions, not claims about a particular model.

Cooling has a separate rating, SEER2. A heat pump used year-round has both a heating and a cooling rating, and the bill impact depends on how much heating and cooling the house needs. Do not compare a furnace AFUE of 80 or 90 directly with a heat pump COP of 3 or 4. They describe different quantities over different test conditions.

Where running costs come from

For space heating, the core calculation is variable electricity use multiplied by the price of electricity. A practical version is:

Annual heating kWh = delivered heat in Btu ÷ (HSPF2 × 1,000)

The result is an estimate for the heating season, not a full utility bill. It excludes fixed monthly charges, taxes, demand charges, hot water, cooling, and any backup heat. If you prefer thermal kWh and seasonal COP, the equivalent is:

Annual heating kWh = delivered heat in kWh ÷ seasonal COP

If resistance backup heat supplies a share of the load, split the heat delivery. For a backup share s, heat pump kWh is total Btu × (1 – s) ÷ (HSPF2 × 1,000), and resistance kWh is total Btu × s ÷ 3,412. Add the two before multiplying by the price. A variable bill is not the same as a total bill: fixed customer charges, meter fees, and other non-bypassable charges remain even if you use less electricity.

For any heating option, cost per delivered MMBtu can be estimated as fuel price per unit divided by usable Btu per unit. For a heat pump, usable Btu per kWh is HSPF2 × 1,000. For a furnace, usable Btu per therm is AFUE × 100,000. This common basis makes the trade-off visible without pretending that one efficiency label is universal.

Worked example: an illustrative U.S. home

Assume a house needs 48,000,000 Btu (48 MMBtu) of delivered space heat in a year. The installed heat pump has an HSPF2 of 9 Btu/Wh, equivalent to a seasonal COP of about 2.64. The household pays $0.17 per kWh for the relevant electricity, a hypothetical rate chosen for arithmetic. The example excludes cooling, water heating, fixed charges, taxes, and backup heat, and it assumes the house and thermostat behavior stay the same.

Heat pump electricity = 48,000,000 ÷ (9 × 1,000) = 5,333 kWh, rounded. Annual variable cost = 5,333 kWh × $0.17/kWh = $906.61, or about $907. The cost per million Btu delivered is about $18.89: each kWh delivers 9,000 Btu, or 0.009 MMBtu, so $0.17 ÷ 0.009 ≈ $18.89 per MMBtu.

Now suppose 10% of the annual heat comes from electric resistance backup at a COP of about 1. The heat pump supplies 43,200,000 Btu: 43,200,000 ÷ 9,000 = 4,800 kWh. Resistance supplies 4,800,000 Btu: 4,800,000 ÷ 3,412 = 1,407 kWh. Total electricity is 6,207 kWh, and the variable cost is 6,207 × $0.17 = $1,055.19. In this example, a 10% backup share adds roughly $149, or about 16%, to the heating bill.

A gas comparison can be useful, but only with stated assumptions. If a furnace delivers the same 48 MMBtu at 80% AFUE, it needs 60 MMBtu of gas input, or 600 therms. At a hypothetical $1.30 per therm, the gas cost is $780. At $1.50 per therm it is $900; at $1.20 it is $720. These figures exclude fixed charges and non-fuel costs. They show why a heat pump can be cheaper than electric resistance and still be close to or more expensive than gas, depending on local fuel prices and the actual seasonal efficiency.

Scenario table: how load, efficiency, and price interact

The table is not a ranking of climates or products. It shows that a larger heating load, a lower seasonal efficiency, and a higher electricity price all push the same direction. In the example, the cold-load scenario uses twice the electricity of the mild-load scenario: 7,500 kWh versus 3,600 kWh. Price and efficiency differences explain the rest of the cost gap.

What changes the result in a real house

Climate and the annual heating load

Climate affects both the number of hours the heat pump runs and how hard it works. A house in a mild climate may need far less delivered heat than a similar house in a cold climate. Colder outdoor air also tends to lower capacity and COP for an air-source unit, and defrost cycles add load in some conditions. The annual heating load is best estimated from a room-by-room load calculation, past fuel use, or weather-normalized utility data. A thermostat setting is not a load calculation.

Insulation, air sealing, and the distribution system

Heat pumps work best when they can deliver heat at lower temperatures for longer periods. Leaky ducts, uninsulated ducts in attics or crawlspaces, missing insulation, and drafty envelopes waste some of the heat that the machine produces. Energy Advice England’s guidance notes that heat pumps work best with good insulation and lower flow temperatures, and that some properties need changes before a heat pump is suitable. In a U.S. home, duct leakage and duct design can be as important as the outdoor unit’s rating. A ducted system that loses conditioned air to an unconditioned attic is not delivering all its rated heat to the rooms.

Equipment sizing, installation, and commissioning

An oversized heat pump may cycle more, dehumidify poorly in cooling mode, and cost more upfront. An undersized one may lean on backup heat. Installation quality matters: refrigerant charge, airflow, duct static pressure, line-set length, and controls all affect real performance. Ask for a load calculation and a commissioning report rather than a rule-of-thumb replacement based only on the old equipment’s tonnage. This is a professional assessment, not a do-it-yourself check.

Controls, thermostat settings, and backup heat

Many heat pumps have a balance point: the outdoor temperature at which the heat pump’s capacity equals the house’s heat loss. Below that point, supplemental heat may run. If the supplemental heat is electric resistance, its COP is about 1, so its share can raise electricity use quickly. Some systems use a lockout temperature or staged controls to limit unnecessary resistance heat. Defrost cycles, fan settings, and thermostat setbacks also change consumption. A setback that saves energy in a furnace may cause a heat pump to recover with backup heat, depending on the controls and weather, so ask the installer how the system is designed to operate.

Electricity price and rate design

The price that matters for an extra kWh is the marginal price, not necessarily the average price on the bill. If the household pays a flat rate, the marginal energy price is usually close to the supply and delivery charges per kWh, but fixed fees do not change when the heat pump runs. Under time-of-use rates, the price depends on the hour. Under demand charges, a short period of high use can affect the bill. Tiered rates can raise the price of additional use. A heat pump can sometimes shift some operation to cheaper hours, but that depends on the building, controls, and rate design; do not assume load shifting without checking the tariff.

Cooling, hot water, and other loads

If the heat pump also provides cooling, part of its annual electricity use replaces an air conditioner’s use. Compare the combined service, not just heating. Cooling efficiency is rated with SEER2, and cooling cost depends on climate, humidity, thermostat settings, and duct quality. A heat pump water heater is another separate calculation. So is a heat pump clothes dryer, if one is installed. Keep each appliance’s kWh and price separate so the savings from one are not used to justify another.

Maintenance and real-world performance

Clogged filters, blocked outdoor coils, low refrigerant charge, and dirty blower wheels reduce airflow and efficiency. A heat pump needs routine filter care and periodic professional service. The exact maintenance schedule should come from the manufacturer and installer. Maintenance is not a guarantee of savings, but neglect can turn a good rating into poor performance.

Weather, occupancy, and the comparison year

Annual estimates are weather-normalized averages. A colder-than-average winter raises heating load; a milder winter lowers it. Occupancy, thermostat schedules, and the number of rooms conditioned also change use. When comparing a heat pump with an old furnace, use the same weather year and the same indoor temperature. Do not compare a heat pump’s estimated annual use with one unusually cold month of furnace use.

Fuel prices and regional differences

U.S. electricity and natural gas prices vary by region and by utility. A heat pump that beats electric resistance heating may or may not beat a high-efficiency gas furnace on cost. In cooling-dominated parts of the South, the heat pump’s cooling performance and humidity control may matter as much as its heating rating. In colder parts of the Northeast and Midwest, a cold-climate heat pump, a well-insulated envelope, and a clear backup-heat strategy matter more. In mild western climates, the annual heating load may be small, so the fixed cost of the equipment can dominate the decision. Use local prices and local design temperatures, not national averages.

Common mistakes and claims to check

A headline COP or efficiency percentage is not a bill. A heat pump with a rated COP of 4 at one condition does not use one-quarter of the electricity of a resistance heater every hour of the year. Seasonal efficiency is lower, and defrost, fans, and backup heat add consumption. A claim that a heat pump is 300% or 400% efficient can be mathematically meaningful at a condition, but it does not mean the bill falls by 75% or that the system is free to run.

Another common mistake is comparing a furnace’s AFUE with a heat pump’s COP without converting to cost per delivered Btu. Fuel prices are in different units: electricity in dollars per kWh, natural gas in dollars per therm, propane in dollars per gallon, and heating oil in dollars per gallon. Convert to a common basis before comparing. Also check whether a quoted saving includes fixed charges that the household will pay anyway, or whether it assumes a rate, climate, and usage pattern that do not match the home.

Be cautious with savings claims that ignore the backup-heat share, duct leakage, or the cost of electrical upgrades. A heat pump may need a new circuit, panel capacity, or duct modifications. Those are capital costs, not running costs, but they affect the payback. Finally, do not treat an incentive as a reason to skip the load calculation or the commissioning. Incentives vary by location and program and can change; the equipment still has to fit the house.

How to estimate your own running cost

This is an ordered checklist. Use measured data where you have it and label every assumption.

1. Find the marginal electricity price. Use a recent bill and identify the supply and delivery charges that change with each kWh. If the rate is tiered or time-of-use, use the price in the hours when the heat pump is likely to run. Keep fixed charges separate.
2. Estimate annual delivered heat. If you have a gas furnace, multiply annual therms by AFUE and by 100,000 Btu per therm. If you have electric resistance heat, multiply annual heating kWh by 3,412 Btu per kWh. If you have an existing heat pump, use weather-normalized annual kWh and the system’s seasonal performance. A contractor’s room-by-room load calculation is the most direct approach for a new installation.
3. Get the HSPF2 and capacity from the actual model documentation. Use the matched system’s rating, not just the outdoor unit’s label. Ask how the rating was determined and whether the installed configuration matches it.
4. Ask about the balance point and backup heat. Find out what outdoor temperature triggers supplemental heat, whether the backup is electric resistance, and how much of the annual load the installer expects it to carry. If no estimate is available, test a range such as 0%, 10%, and 20% backup share rather than assuming zero.
5. Apply the formula and keep units visible. For each case, calculate heat pump kWh, backup kWh, total kWh, and annual variable cost. Round only at the end. Check that the result is plausible against your past heating bills.
6. Add cooling and water heating separately. Use SEER2 and cooling load for cooling, and the heat pump water heater’s own rating for water heating. Do not attribute the whole combined bill to space heating.
7. Compare quotes on scope, not just price. Ask for the load calculation, equipment model numbers, electrical work, duct work, thermostat, commissioning, warranty, and maintenance terms. A low quote that omits duct sealing or electrical capacity may not be comparable.
8. Check local programs and permits. Incentives, permits, and installation rules vary. Verify them with the program administrator or local authority; do not rely on a sales sheet alone.

Inexpensive checks versus upgrades that need a quote

Start with the low-cost checks because they are cheap and can change the calculation. Move to professional assessment before committing to a capital upgrade. Direct electrical, gas, and refrigerant work should be done by appropriately qualified professionals.

Decision guide: when a heat pump is a sensible next step

A heat pump is often a strong candidate when the household is replacing electric resistance heating, adding central cooling at the same time, or replacing both an old furnace and an old air conditioner. It can also make sense where electricity is relatively inexpensive, the home has a good envelope and duct system, and local incentives reduce the upfront cost. In those cases, the efficiency advantage over resistance heating is large and measurable.

It is reasonable to pause or compare more carefully when natural gas is inexpensive relative to electricity, when the house has leaky ducts or little insulation, when the electrical panel is near its limit, or when no qualified installer has provided a load calculation and commissioning plan. In a cold climate, ask specifically about cold-climate performance, balance point, backup heat, and defrost behavior. A heat pump can work in cold weather, but the installed system must be designed for the local design temperature. Avoid any promise of guaranteed savings; use the formula, local prices, and a range of backup-heat assumptions.

Renters and people facing higher bills can still use the same method. Ask the landlord or property manager for the heating system type, efficiency rating, and past utility data if available. Focus on no-cost and low-cost actions that are within your control, and get permission before making changes. If a heat pump is being installed as a replacement, ask how the thermostat and backup heat are set up; those settings can affect the bill as much as the equipment label.

A sensible next step

Take your last 12 months of electricity or fuel bills and one heat pump specification sheet. Calculate annual heating kWh with the formula above, then run the same calculation with a 10% and 20% backup-heat share. If the result is close to your current cost, ask for a load calculation and a commissioning plan before signing. If a quote cannot state the assumed heating load, HSPF2, and backup-heat strategy, treat it as incomplete rather than comparable.