How to correct radiator output for flow temperature
How to convert a radiator's EN 442 catalogue rating to its actual output at a lower heat pump flow temperature, using the standard's exponential correction.
What this means
A radiator's published rating is measured at a fixed temperature difference above room temperature, 50 degrees Celsius under the EN 442 reference condition. Actual output at a different, lower temperature difference is found by multiplying the rated output by the ratio of the two temperature differences raised to a characteristic exponent, close to 1.3 for an ordinary panel radiator. Run at a 30-degree difference instead of 50, a radiator delivers roughly half its catalogue figure, not the linear three-fifths a straight ratio would suggest.
Equipment and model context
- Panel and column radiators rated to EN 442 and run from a heat pump or condensing boiler
- Worked figures illustrate the method and are not a rating for any specific radiator
This explains the correction and why it is not linear. It does not size a radiator for a specific room. That needs the room's own heat loss, the exponent published for the actual radiator type and construction, and the flow and return temperatures the system will actually run at.
What this covers
- Why a radiator's box rating cannot be scaled by temperature difference alone.
- What the EN 442 exponent represents and why it differs by radiator type.
- How much a panel radiator underdelivers when run at heat pump flow temperatures without recalculation.
- Why mean water temperature, not flow temperature alone, is the figure the correction actually uses.
What changes the result
- Treating radiator output as directly proportional to temperature difference, which understates the loss at low temperature difference because the true relationship is exponential.
- Reusing a boiler-era radiator schedule without recalculating output at the heat pump's lower flow and return temperatures.
- Using flow temperature alone in the correction rather than the mean of flow and return, which is what the standard's temperature difference actually represents.
- Applying a generic exponent to every emitter type when the published exponent differs between panel radiators, column radiators, and trench convectors.
What the catalogue figure actually measures
EN 442 tests a radiator in a controlled chamber and publishes its output at a stated reference condition: a 75 degree Celsius flow, 65 degree return, and 20 degree room temperature under EN 442, giving a mean water to air temperature difference of 50 degrees. That single number on the box is a lab result at one specific condition, not a universal capacity the radiator delivers regardless of how it is run.
Mean temperature difference, not flow temperature alone, is what the rating and the correction both use. It is the average of flow and return temperature minus the room air temperature, which is why a system with a wide temperature drop across the radiator behaves differently from one with a narrow drop even at the same flow temperature.
Why the correction is exponential rather than linear
Heat transfer from a radiator surface to room air does not scale directly with temperature difference, because convective and radiative transfer both depend on that difference in a nonlinear way. EN 442 captures this with an exponent, denoted n, specific to each radiator's construction: a typical single-panel radiator runs close to 1.3, while radiators with more convective fin area can run somewhat differently.
The correction formula is delivered output equals rated output multiplied by the ratio of the new mean temperature difference to the rated mean temperature difference, that ratio raised to the power of n. At n greater than 1, halving the temperature difference cuts output by more than half, which is the mechanism behind the underperformance a heat pump retrofit exposes when radiators are not resized.
Working the correction
Take a radiator rated 2,000 watts at the standard 50 degree mean temperature difference, with an exponent of 1.3, now run at a heat pump flow and return of 45 and 35 degrees Celsius into a 20 degree room. Mean water temperature is 40, and the mean temperature difference above room air is 20 degrees. The ratio of 20 to 50 is 0.4, and 0.4 raised to the power of 1.3 is approximately 0.303.
Delivered output is 2,000 multiplied by 0.303, or roughly 605 watts, about 30 percent of the catalogue rating. If the room's own heat loss at design conditions is more than 605 watts, this radiator is undersized at the intended flow temperature and needs replacing with a larger unit, or the system needs a higher flow temperature, trading away some of the efficiency gain the lower temperature was chosen for in the first place.
Why this changes emitter selection, not just verification
A heat pump retrofit onto an existing radiator system is, in effect, a resizing exercise for every emitter in the building, because the temperature difference the system was originally designed around no longer applies. Skipping this step and simply connecting a heat pump to an unchanged radiator schedule is a documented reason a correctly sized heat pump still fails to hold room temperature on a cold day.
The correction also interacts with the flow temperature decision covered elsewhere: a lower design flow temperature improves heat pump efficiency but demands larger radiators to compensate, while accepting a higher flow temperature keeps existing radiators adequate at the cost of some of the efficiency the heat pump could otherwise deliver. Neither choice is free, and the exponential correction is what makes the trade quantifiable rather than a guess.
An illustrative correction curve for a radiator rated 2,000 watts at a 50 degree Celsius mean temperature difference above room temperature, showing delivered output against mean water to air temperature difference using an exponent of 1.3.
- Delivered output at exponent 1.3
- Output if the relationship were linear
- At a 30 degree mean temperature difference the radiator delivers about 1,024 watts, close to half its 50 degree catalogue rating.
- A straight linear scaling, the dashed line, would predict 1,200 watts at the same 30 degree difference, overstating delivered heat by roughly 17 percent.
- The gap between the two curves widens as temperature difference falls, which is exactly the range a heat pump operates in.
- The same radiator would need to be roughly double the rated size, or the room's design temperature difference roughly doubled, to deliver its full rated output at a low flow temperature.
| Radiator type | Typical exponent | Sensitivity to flow temperature reduction |
|---|---|---|
| Panel radiator, primarily convective | Around 1.3 | Output falls sharply as flow temperature drops |
| Column or cast iron radiator, more radiative | Can run lower than a panel type | Somewhat less sensitive to the same temperature drop |
| Underfloor heating | Different governing relationship entirely | Designed around low flow temperature from the outset |
Questions people ask about this
Where does the exponent for a specific radiator come from?
The manufacturer publishes it alongside the EN 442 rated output for each radiator model, since it depends on the specific construction, panel count, and fin geometry. A generic figure like 1.3 is a reasonable planning estimate for an ordinary panel radiator, but the exact published value for the actual product should be used in a final calculation.
Does this correction apply to underfloor heating the same way?
No. Underfloor heating is designed from the outset around a low flow temperature and a different heat transfer relationship between the floor surface and the room, so it does not use the same radiator correction exponent. Underfloor systems are sized directly for their intended low-temperature operation rather than corrected down from a high-temperature reference rating.
Can a bigger radiator ever be too big for a heat pump system?
A radiator with more capacity than the room needs at the design flow temperature is not itself a fault, and it gives the system margin at low outdoor temperature. The design consideration is control: an oversized emitter on a system without weather compensation or good modulation can overshoot room temperature and cycle the heat pump, so oversizing pairs with a control strategy that manages it rather than being a free improvement on its own.
Why is mean water temperature used instead of flow temperature?
A radiator's output depends on the average condition of the water passing through it, not only the temperature it enters at, because the water cools as it gives up heat along the radiator's length. Using flow temperature alone would overstate output on a system with a wide temperature drop and understate it on a system with a narrow one, so the standard defines the correction against the mean of flow and return.
Evidence recordSource verification pending
standards body publication · editorial review
This page is awaiting source verification against the documentation in its evidence record: European Committee for Standardization and ASHRAE technical literature. Its documentation class and intended scope are shown here while that check is pending.
- Documentation class
- standards body publication
- Scope of the definition
- Confirm against the exact model manual