A heat pump changes the shape of a bill, not only its size

Efficiency above one hundred per cent, honestly

A heat pump doesn’t create heat. It moves it, from outside air or from the ground into the house, using a refrigeration cycle running in reverse. Because moving heat costs less energy than producing it, the useful heat delivered exceeds the electricity consumed, and the ratio between the two is the coefficient of performance.

That ratio isn’t a constant. It falls as the temperature difference the machine has to work across increases, which means it falls when it is cold outside and it falls when the house demands hotter water in the radiators. Both effects are physical and unavoidable, and together they explain almost every disappointing installation.

The number that matters over a year is the seasonal figure rather than a best-case laboratory one, because a heating system spends most of its hours in mild conditions and a minority of them in severe cold. A quotation that leads with a peak coefficient is telling you about the easiest hour of the year.

Flow temperature is the lever

A boiler is comfortable sending very hot water to small radiators. A heat pump is not: pushing the flow temperature up is exactly what degrades its efficiency, and the difference between a system designed to run warm and one designed to run hot is large enough to change whether the installation makes economic sense.

Running cooler means each radiator delivers less heat, so the emitters have to be bigger, or there have to be more of them, or the heat has to come out of a floor with a very large surface area. That is why heat pump installations so often involve replacing radiators, and why an installer who proposes to leave every emitter untouched should be questioned closely.

Fabric matters for the same reason rather than for a different one. A well-insulated house needs less heat, so it can be kept warm at a lower flow temperature, so the heat pump runs at a higher coefficient. Insulation improves the efficiency of the machine, which is a compounding benefit that a boiler never offered.

The electricity bill grows even when total energy falls

This is the part that unsettles households. Moving heating from a fuel to electricity can easily double or triple the electricity passing through the meter, while the total energy the house consumes falls substantially, because the fuel disappears from the accounts entirely.

Whether the change saves money is therefore a question about the ratio between the price of electricity and the price of the fuel being displaced, divided by the efficiency achieved. If electricity costs several times as much per unit as the fuel did, a coefficient of three roughly breaks even, and everything depends on which side of that line your local prices fall. That ratio differs enormously between countries and moves with every energy market shock, which is precisely why no publication can tell you the answer.

It is worth saying plainly that in some markets, at some moments, the arithmetic doesn’t favour the change. The environmental case may still hold, since it depends on how the local electricity is generated, but that is a different argument and should be made as one.

Why the tariff matters more than it used to

A heat pump runs long and steadily rather than in short bursts, which changes how it interacts with time-varying prices. It also has a modest ability to store: heat put into a building or into a water cylinder stays there for hours, so the machine can run harder when electricity is cheap and coast when it is expensive.

That makes a heat pump a good match for time-of-use pricing in a way a boiler never was, and several markets now offer tariffs designed specifically around it. The size of the benefit depends on the thermal mass of the building, since a heavy, well-insulated house holds a pre-charge for far longer than a lightweight, leaky one.

Where a bill includes a charge related to peak demand, the calculation changes again, because a heat pump adds electrical load precisely on the coldest evenings when everyone else is doing the same. In those markets the control strategy is worth as much attention as the equipment.

The seasonal mismatch with the roof

People often assume solar and a heat pump are natural partners. They are, in an annual accounting sense, and they are poorly matched in time. Heat demand peaks in the darkest months, when the array produces least, and the array produces most in months when the heating is off entirely.

That mismatch grows with latitude. Near the equator it barely exists, since neither heating demand nor solar output swings much. In northern Europe the winter output of a rooftop array is a small fraction of its summer output, while heating demand is at its maximum, and no domestic battery bridges a seasonal gap of that size.

What solar does contribute is hot water and shoulder-season heating, which is genuinely worth having. The realistic claim is that an array reduces the annual electricity a heat pump imports by a useful amount. The unrealistic claim is that it powers the heating, and it is worth recognising which claim a quotation is making.