One idea, three measurements
All three numbers express the same idea: units of heat delivered for each unit of electricity used. A figure of 3 means three kilowatt-hours of heat for one kilowatt-hour of electricity. What differs is the conditions, the period and what counts as part of the system.
COP: one moment, one set of conditions
The coefficient of performance is measured at a fixed outdoor temperature and a fixed water temperature, for example mild air and cool water. It is a snapshot. The same heat pump has a higher COP on a mild day with cool radiators and a lower one on a freezing day with hot radiators. Manufacturers publish COPs at several standard test points, which is useful for comparing machines, but no single COP tells you what the system will do over a year.
SCOP: a standard season, on paper
The seasonal coefficient of performance combines test results at several temperatures into a weighted average for a standard heating season. It is the figure on energy labels and data sheets, and it is always quoted for a particular water temperature, commonly a low-temperature and a medium-temperature case. A product's SCOP at a low flow temperature will be noticeably higher than at a medium one. When comparing products, compare like with like: the same climate and the same flow temperature.
SPF: what really happened
The seasonal performance factor is measured, not calculated: meters record the heat delivered and the electricity used by an installed system over a whole season. It includes everything the lab leaves out, from defrost cycles and hot water to the controls and the household's habits. Field trials also define a boundary, which says which parts of the system are inside the measurement. The "H4" boundary used in UK trials includes the heat pump, its backup heater and the circulating pumps, so it reflects what the household pays for.
The best recent British evidence comes from the Electrification of Heat demonstration, run by Energy Systems Catapult with government funding, which installed 742 heat pumps across a range of homes. Its interim results put the median air source SPF at 2.80 at the SPFH4 boundary. Half of the systems did better than that and half did worse, which is the point: design and use matter as much as the machine.
Why the three numbers differ
- Hot water. Heating water for taps needs higher temperatures than heating rooms, which lowers efficiency while the heat pump does it. Lab heating figures may leave it out.
- Flow temperature. A system run hotter than designed loses efficiency every hour it runs.
- Defrosting. In cold, damp weather an air source unit periodically reverses to melt ice on its coil, using energy without heating the home.
- Cycling. An oversized heat pump switches on and off rather than running steadily, which costs efficiency.
- Boundary. Including pumps and backup heaters lowers the figure, but makes it more honest.
What the number means for running cost
The cost of a unit of heat from a heat pump is the electricity price divided by its seasonal efficiency. On the price cap for 1 October to 31 December 2026, electricity costs 26.32p a kWh, so:
| Seasonal efficiency | Pence per kWh of heat |
|---|---|
| 2.5 | 10.53p |
| 2.8 | 9.40p |
| 3.0 | 8.77p |
| 3.5 | 7.52p |
For comparison, a gas boiler at an assumed 90% efficiency turns gas at 7.97p into heat at 8.86p a kWh. That is why, at today's prices, the comparison with gas depends so heavily on the heat pump's real seasonal efficiency. The cost calculator lets you try your own figures.
Questions to ask about any efficiency figure
- Is it a COP, a SCOP or a measured SPF?
- At what flow temperature, and does that match the design for this home?
- Does it include hot water?
- For a measured figure, what boundary was used?
Reading a data sheet
A typical heat pump data sheet lists several numbers in a table: capacity and COP at standard test points (often written as an air temperature and a water temperature, such as A7/W35, meaning 7°C air and 35°C water), and SCOP for a stated climate and flow temperature. The test points let you see how quickly performance falls as the water gets hotter or the air gets colder. A machine whose COP drops sharply between the low and high water temperatures will be more sensitive to how hot its radiators need to run.
Two cautions. Capacity also changes with conditions: a unit rated at a certain output in mild air delivers less in freezing air, which is why designs size to the coldest days. And data-sheet figures describe the heat pump, not the whole system, so the cylinder, the controls and the pipework all sit outside them.
If a quote gives only one efficiency number, ask for the other two kinds of context: the flow temperature it assumes, and whether the installer can show measured results from similar systems. A design that predicts a seasonal figure, and explains how, is more useful than a headline taken from a brochure.
These measures matter most when you weigh up options: air source against ground source sets the two main types side by side, the running costs estimator turns an efficiency figure into a yearly estimate, and is my home suitable? covers the features of a house that pull the real figure up or down.