COP, SCOP and SPF: the Three Figures and Which One Answers the Question

A data sheet quotes a heat pump at COP 4.6 at A7/W35 and, further down in smaller print, an SCOP of 3.4. The customer wants to know what their winter bills will look like. Only one of those numbers helps.

The short answer

COP = heat output (kW) ÷ electrical power input (kW). A unit delivering 9 kW while drawing 2.5 kW has a COP of 3.6 — 3.6 kW of heat for every 1 kW of electricity it consumes. A working heat pump always returns a COP above 1; if your answer comes out below 1, you have divided the wrong way round.

But the customer asked about winter bills, and the answer is the SCOP, because it is calculated across a whole heating season. A7/W35 is a mild autumn morning, not a British January.

Why it can be "400 per cent efficient"

It can, because a heat pump does not make heat. It moves heat that is already there — in the outside air, in the ground, in a pond — and the electricity only pays for the moving. So the number is not an efficiency in the boiler sense at all: it is a ratio.

A COP without its conditions means nothing. The notation tells you which: A7/W35 means air entering at 7 °C and water leaving at 35 °C; B0/W45 means brine entering at 0 °C and water leaving at 45. The first letter is the source: A is air, B is brine, so B0/W45 is a ground source condition.

Manufacturers naturally publish the flattering condition, which is why two headline COPs are only comparable when the conditions beneath them match.

Lift is what really sets the number

Temperature lift is the difference between the source temperature and the flow temperature the unit delivers. A machine collecting at 2 °C and delivering at 45 has a lift of 43 K. That gap is the work the compressor has to do, and the wider it is, the more electricity is burned for each kilowatt of heat.

Set two identical heat pumps to 35 and 55 °C and the 35 °C system returns the higher COP, because a lower flow temperature narrows the lift. It is not about pump speed, water volume or defrosting.

As a working rule, each 1 °C taken off the flow temperature is worth roughly 2 to 2.5 per cent in efficiency. Twenty degrees is therefore not a detail; it is close to half the running cost.

That rule has a consequence on every conversion from a boiler. A radiator's output depends on the difference between its surface temperature and the room temperature, so drop the flow from 75 to 45 °C and the same radiator emits far less heat — often less than half.

So radiators are oversized on a heat pump conversion to emit the required output at a much lower flow temperature. Nothing to do with water volume, flow rate or a Part L rule on surface area. Underfloor heating, with its very large surface, does the same job by design.

SCOP, SPF and system efficiency

A COP is measured at one fixed set of test conditions. The seasonal coefficient of performance (SCOP) is calculated across a whole defined heating season and a spread of outdoor temperatures. The COP is not wrong; it is a laboratory condition, and quoting it as though it were a year-round result is how complaints begin.

The seasonal performance factor (SPF) goes one step further: the measured seasonal result of the installation as built — the heat it actually delivered over a year against the electricity it actually consumed. It therefore takes in everything a headline COP leaves out: circulating pumps, the brine pump, defrost energy, any immersion heater top-up, distribution losses and standing losses from the store.

SPF is the figure the installer most influences, and that is the point of it. Put two identical heat pumps in two identical houses and one can return a far worse SPF. The most likely explanation is not the refrigerant or the pipe run: it is that one runs at a higher flow temperature with the emitters left unchanged. Same machine, wider lift, worse season.

System efficiency draws the boundary wider still: the heat pump plus its auxiliaries, the distribution, the controls and the storage losses. It explains a fact that surprises people: a good machine badly commissioned, on oversized pipework with a poorly controlled cylinder, can be beaten by a lesser machine installed well.

How anybody knows what the SPF actually was

All three figures so far are claims. The SPF is the one that is measured — and it cannot be measured with an electricity meter alone.

The electricity meter records only what the unit consumed; it says nothing whatever about what the unit delivered. Half the sum is missing.

What supplies the other half is a heat meter. It measures the heat given to the system from the flow and return temperatures and the flow rate: the temperature difference tells it how much heat each litre carried, and the flow rate tells it how many litres went past.

The two meters together give the seasonal performance actually achieved — heat out divided by electricity in, over a real year in a real house. That figure does two jobs: it is what an incentive payment is calculated from, and it is what proves the design assumption in service — the one thing that settles an argument about whether a system is underperforming or was simply over-promised.

The ErP energy label supplied with a heat pump exists to let consumers compare the energy efficiency of products consistently, showing an efficiency class alongside figures such as rated output and sound power level. It is not a record of refrigerant charge, not a certificate that the installer belongs to a scheme, and not proof of Building Regulations compliance. A package label can also be produced for the system as installed, reflecting the heat pump together with its controls — a useful reminder that controls are part of the product, not an accessory to it.