What size radiator?
The real calculation, not a rule of thumb — and the reason your radiators will not do the same job on a heat pump.
What size radiator does this room need?
The real calculation, not a rule of thumb. Sized for the Calder Valley at -4°C outside, which is the coldest it sensibly gets here.
1.40 kW radiator, at its catalogue rating
The room loses 1.37 kW at 91 W per square metre. On boiler, 75° flow that is exactly what the catalogue rating means, because the rating is measured on this system — so the radiator you buy is the number above.
| Through | Area | U | Watts |
|---|---|---|---|
| External walls | 17.6 m² | 1.60 | 702 |
| Windows | 2.6 m² | 1.60 | 104 |
| Ground floor | 15.1 m² | 0.70 | 265 |
| Ventilation 1.0 air changes an hour through 36.3 m³ | — | — | 304 |
| Total, holding 21° against -4° | 1,375 | ||
The same room on every system
| Boiler, 75° flow How radiators are rated and how most are still run. | 50.0 K | 100% | 1.40 kW |
|---|---|---|---|
| Boiler, turned down to 60° A condensing boiler actually condensing, which is where its efficiency is. | 32.5 K | 57% | 2.45 kW |
| Heat pump, 50° flow A heat pump pushed hot to suit existing radiators. | 27.5 K | 46% | 3.00 kW |
| Heat pump, 45° flow A sensible heat pump design. | 22.5 K | 35% | 3.90 kW |
| Heat pump, 35° flow What underfloor heating is happy with, and radiators are not. | 12.5 K | 16% | 8.35 kW |
This assumes 1919 to 1975: Cavity wall, unfilled. Little or no loft insulation as built. Those are the standard assumptions for the age, and they are assumptions — a solid wall somebody insulated internally in 2015 will not behave like 1900. We measure the real thing before we order anything.
Why this is not a rule of thumb
The usual shortcut is to multiply the room’s volume by a number. It is wrong in both directions and badly wrong in one of them: it takes no account of whether the room is a mid-terrace bedroom with one external wall or a corner room with three, and no account of the glass, which is the worst surface in most rooms by a long way.
What it is actually doing
Two sums. The fabric loss is the U-value of each surface facing outdoors, times its area, times the temperature difference you are holding. The ventilation loss is the volume of the room, times how many times an hour the air in it is replaced, times the same difference, times the heat capacity of air.
Note what is not in there. Walls between two rooms you are heating to the same temperature carry no load at all, however big they are. And the window comes out of the wall rather than being added to it — counting both at full area is the commonest arithmetic error in this calculation, and it always inflates the answer.
The bit about heat pumps
Radiators are sold at a rating measured with the water 50 degrees hotter than the room. Run the same radiator on a heat pump flowing at 45 and the difference is 22.5 degrees, and output falls with that difference to the power of 1.3 — so it gives about a third of what the box said.
Nothing is broken and nobody has installed it wrongly. It is the same radiator obeying the same physics. But it means a house that is comfortable on a boiler will be cold on a heat pump with the same radiators, and that is worth knowing before the heat pump is in the garden rather than after.