The oversizing ratio: one number per room
The table above answers "what does this radiator give?" The more useful question runs the other way: "how hot must the water be for this room?" Turn the formula around and it depends on one ratio per room:
Oversizing ratio = radiator rated output ÷ room heat loss at the design outdoor temperature
If the radiator in a room is rated at exactly the room's design heat loss, the ratio is 1.0 and the water has to reach rating conditions on the coldest day. If the radiator is twice the room's loss, the ratio is 2.0 and the water can be much cooler. The required temperature difference is the rated ΔT × (1 ÷ ratio) raised to the power 1 ÷ 1.3.
| Oversizing ratio |
Required ΔT (European rating, 50 K) |
Mean water in a 20 °C room |
Flow temperature with 5 K drop |
Required ΔT (N. American rating, 112 °F) |
Mean water in a 68 °F room |
| 1.00 |
50.0 K |
70.0 °C |
72.5 °C |
112.0 °F |
180 °F |
| 1.25 |
42.1 K |
62.1 °C |
64.6 °C |
94.3 °F |
162 °F |
| 1.50 |
36.6 K |
56.6 °C |
59.1 °C |
82.0 °F |
150 °F |
| 2.00 |
29.3 K |
49.3 °C |
51.8 °C |
65.7 °F |
134 °F |
| 2.50 |
24.7 K |
44.7 °C |
47.2 °C |
55.3 °F |
123 °F |
| 3.00 |
21.5 K |
41.5 °C |
44.0 °C |
48.1 °F |
116 °F |
| 4.00 |
17.2 K |
37.2 °C |
39.7 °C |
38.6 °F |
107 °F |
Read it as a set of thresholds. For 55 °C flow on the coldest day, every room needs a ratio of about 1.75 or more (European rating). For 45 °C flow, about 2.8. For 35 °C flow, about 6.2: radiators six times the room's heat loss, which is where underfloor heating or fan-assisted emitters take over.
Two inputs make or break the ratio, and both can be wrong on an old quote.
The radiator's rating. For a modern panel, the model code (type 11, 21, 22 or 33 plus height and length) leads to the maker's catalogue figure. Check whether the catalogue gives ΔT 50, ΔT 30 or the North American 180 °F point; Zehnder notes that it prints outputs at ΔT 30 on some data sheets, so a "1,000 W" label is not always comparable with another brand's "1,000 W". Old cast-iron column radiators in North American houses are rated in square feet of equivalent direct radiation (EDR) from the maker's or industry tables; the installer measures height, column count and section count and looks the rating up. Our steam radiator guide covers EDR in more detail.
The room's heat loss. It must be calculated room by room at your local design temperature, the way a Manual J or CSA F280 heat loss calculation does it. A whole-house number divided by floor area hides exactly the rooms that fail: the corner bedroom with two outside walls, the bathroom that people want warmer than the rest of the house, the room over the garage.
A six-room house, checked room by room
Here is a hypothetical house to show the method end to end. The inputs are stated so you can replace them with yours: design heat loss 7,500 W (about 25,600 Btu/h) at a design outdoor temperature of -7 °C, 20 °C indoors, and existing steel panel radiators with ratings taken at ΔT 50.
| Room |
Design heat loss |
Existing radiator rating (ΔT 50) |
Ratio |
Lowest flow temperature that holds 20 °C |
Output at 55 °C flow |
Rating needed at 55 °C |
Rating needed at 45 °C |
| Living room |
2,400 W |
5,200 W |
2.17 |
50 °C |
2,970 W |
4,202 W |
6,777 W |
| Kitchen |
1,100 W |
2,600 W |
2.36 |
48 °C |
1,485 W |
1,926 W |
3,106 W |
| Back bedroom (two outside walls) |
1,500 W |
2,000 W |
1.33 |
63 °C |
1,142 W |
2,626 W |
4,236 W |
| Front bedroom |
1,000 W |
2,400 W |
2.40 |
48 °C |
1,371 W |
1,751 W |
2,824 W |
| Bathroom |
600 W |
900 W |
1.50 |
59 °C |
514 W |
1,050 W |
1,694 W |
| Hall and landing |
900 W |
2,700 W |
3.00 |
44 °C |
1,542 W |
1,576 W |
2,541 W |
| House |
7,500 W |
15,800 W |
2.11 |
63 °C (set by the back bedroom) |
|
13,130 W |
21,178 W |
Three things come out of this table that a whole-house shortcut would miss.
First, the house as a whole looks fine. Its radiators add up to 15,800 W of rating against 13,130 W needed at 55 °C, an overall ratio of 2.11. Someone dividing total radiator rating by total heat loss would sign off on 55 °C. Yet the back bedroom would sit at 1,142 W against a 1,500 W loss on the design day, and the bathroom at 514 W against 600 W.
Second, the worst room sets the water temperature for every room. One undersized radiator in the back bedroom forces a 63 °C flow on the coldest day. The Systemair unit used for the COP figures below lists a maximum heating water setting of 65 °C, so this house sits close to the top of that unit's range before anything is changed.
Third, the fix is two radiators, not six. Replace the back bedroom panel with one rated about 3,600 W at ΔT 50 (ratio 2.4) and the bathroom with one rated about 1,500 W (ratio 2.5), and the worst room becomes the living room at about 50 °C flow. Going further, to 45 °C, would mean enlarging five of the six rooms, about 21,300 W of total rating against 15,800 W today. That is where the cost per degree climbs and where a homeowner should ask whether the efficiency gain repays it.
Low water temperature also changes the pipework question. A boiler system designed around a 20 K drop (80 °C out, 60 °C back) moves 7,500 W with about 5.4 litres per minute; a heat pump running a 5 K drop needs about 21.5 litres per minute for the same heat, four times the flow (flow in kg/s = watts ÷ (4,186 × ΔT)). In U.S. units the rule is gallons per minute = Btu/h ÷ (500 × temperature drop in °F): 25,600 Btu/h at a 20 °F drop is about 2.6 gpm, and at a 10 °F drop about 5.1 gpm. Small-bore or microbore branches that were fine for a boiler may not carry the higher flow without noise or a larger circulator, so the installer should check pipe sizes along with radiator sizes.
What a lower flow temperature is worth in electricity
Radiator output is half of the decision. The other half is how much less electricity the heat pump uses when the water is cooler. The cleanest way to see it is a single manufacturer's data sheet that tests the same unit at several water temperatures.
Systemair's engineering data for its SYSHP MINI EVO-P monobloc air-to-water range (R32 refrigerant) lists heating output, electrical input and COP at three outdoor temperatures and three leaving-water temperatures. The 7 kW and 9 kW models:
| Test point (outdoor air / leaving water) |
07 model: heat output |
07 model: COP |
09 model: heat output |
09 model: COP |
| A7 / W35 (7 °C outside, 35 °C water) |
8,400 W |
5.05 |
10,000 W |
4.70 |
| A7 / W45 |
8,500 W |
3.80 |
10,200 W |
3.65 |
| A7 / W55 |
8,200 W |
3.15 |
9,400 W |
3.10 |
| A2 / W35 |
7,100 W |
3.95 |
8,200 W |
3.80 |
| A2 / W45 |
7,500 W |
3.05 |
8,500 W |
2.95 |
| A2 / W55 |
7,600 W |
2.70 |
8,400 W |
2.65 |
| A-7 / W35 |
7,100 W |
3.15 |
8,000 W |
3.00 |
| A-7 / W45 |
6,800 W |
2.50 |
7,400 W |
2.40 |
| A-7 / W55 |
6,600 W |
2.10 |
7,200 W |
2.05 |
Converted into electricity per 1,000 kWh of heat delivered (1,000 ÷ COP), the 07 model needs:
| Outdoor air |
At 55 °C water |
At 45 °C water |
At 35 °C water |
| 7 °C |
317 kWh |
263 kWh |
198 kWh |
| 2 °C |
370 kWh |
328 kWh |
253 kWh |
| -7 °C |
476 kWh |
400 kWh |
317 kWh |
On the coldest test day, dropping from 55 °C to 45 °C cuts the electricity for the same heat by about 16% (476 to 400 kWh), and dropping to 35 °C cuts it by a third. Over a season the gap is larger in relative terms: the same 07 model's seasonal COP for an average European climate is listed at 5.18 for 35 °C systems and 3.67 for 55 °C systems. That means a 55 °C system uses about 41% more electricity for the same annual heat than a 35 °C one (5.18 ÷ 3.67 = 1.41), all else equal.
Back in the six-room house, the two radiator swaps take the design-day flow from about 63 °C, above the sheet's highest test point, to about 50 °C, between the 45 °C and 55 °C columns. The sheet gives no 50 °C figure, so we do not invent one; the honest reading is that the design-day COP moves from below 2.10 (the 07 model's A-7/W55 value) to somewhere between 2.10 and 2.50 for the 07 model.
Two limits on those numbers matter for Canada and the northern U.S.
The coldest test point is -7 °C. Systemair lists an operating range down to -25 °C outdoors, but its performance table stops at -7 °C. If your design temperature is colder, as it is in much of Canada, the table does not cover your design day. The output column shows the trend: the 07 model's heat output at 55 °C falls from 8,200 W at 7 °C to 6,600 W at -7 °C. Ask for the maker's extended capacity table at your design temperature before sizing; if the unit cannot cover the house on the design day at the flow temperature your radiators need, the gap is made up by a backup element or by the old boiler kept as a second stage. Our cold-climate heat pump performance data page explains why capacity, not COP, is usually the binding figure in deep cold.
Design day is a few hours a year. A heat pump with weather compensation (outdoor reset) runs warmer water only when it is cold outside and cooler water the rest of the time. The design-day flow temperature from the room check is a ceiling, not the operating point for most of the winter. That is why the seasonal COPs above are higher than the -7 °C figures.