45°C vs 55°C vs 65°C: What Flow Temperature Should Your Heat Pump Run At?
The short answer
A heat pump should run at the **lowest flow temperature that still keeps the coldest room warm on the coldest design day**. For many homes that ends up somewhere between 45°C and 55°C at design conditions, and lower for most of the season. 65°C is achievable with many modern R290 units; whether it makes sense depends on the trade-off between higher running costs and the cost of upgrading radiators.
The rest of this article explains why, and what each of the three temperatures typically means for efficiency, radiator output and running cost.
Why flow temperature matters so much for a heat pump
A heat pump does not create heat; it moves it from outdoor air into the heating water. The harder it has to "lift" that heat — from a cold outdoor temperature up to the water temperature you ask for — the more electricity it uses per kWh of heat delivered.
That ratio is the **coefficient of performance (COP)**. Raising the flow temperature increases the lift, so COP falls. The effect is directional and consistent, but not a fixed percentage per degree: it depends on the outdoor temperature, the unit's compressor and heat exchangers, and how the unit is controlled.
This is very different from a gas boiler, where the flow temperature is mainly a comfort and condensing-efficiency setting. With a heat pump, flow temperature is one of the biggest levers on running cost you can actually control.
**How to read this on a datasheet.** Performance is measured at standard test conditions defined in EN 14511, written as outdoor air / water outlet temperature — for example A7/W35 (7°C air, 35°C water) or A7/W55. Which points a manufacturer publishes varies, so where a datasheet or performance table offers figures at several water temperatures for the same outdoor temperature, comparing them shows how that unit responds to flow temperature.
Seasonal efficiency (SCOP, calculated per EN 14825) is declared under EU ErP rules for a **low-temperature application (35°C)** and/or a **medium-temperature application (55°C)**, depending on the product. These are **reference conditions for comparing products, not recommended design temperatures** for any particular installation. For radiator systems, the 55°C figure is usually the more relevant one; a 35°C figure alone says little about radiator performance.
45°C, 55°C and 65°C compared
The table below assumes typical heat-pump operation (around 5 K difference between flow and return water) and a 20°C room. Radiator figures use the standard EN 442 approach with a typical panel-radiator exponent of about 1.3; actual figures depend on the radiator manufacturer's data.
| | **45°C flow** | **55°C flow** | **65°C flow** |
|---|---|---|---|
| Mean water temp (approx.) | 42.5°C | 52.5°C | 62.5°C |
| Radiator output vs its ΔT50 rating | ~35% | ~57% | ~81% |
| Required ΔT50-rated output multiplier* | ~2.8× | ~1.75× | ~1.2× |
| Heat pump efficiency | Highest of the three | Moderate | Lowest of the three |
| Typical fit | Upgraded or generously sized radiators, fan coils, good insulation | Many retrofits with some radiator upgrades | Existing radiators largely retained |
| Main trade-off | Larger emitters may be needed in high-loss rooms | Balance between efficiency and emitter changes | Higher running cost; capacity can fall in cold weather |
\* *To meet a room's heat loss, the radiator's catalogue ΔT50 output needs to be roughly this multiple of that heat loss. It describes rated heat output, not physical radiator size — a higher-output radiator (for example a double or triple panel) can deliver more without being proportionally larger.*
*Figures are illustrative and calculated from standard radiator equations, not measured product data.*
Two points stand out:
1. **Radiator output falls steeply as water gets cooler.** A radiator that comfortably heated a room on a 70°C-plus boiler system may deliver only around a third of its rated output at a 45°C flow.
2. **Many existing radiators are already oversized.** Boiler-era systems were often sized with generous margins, and homes have since been insulated. That is why a room-by-room check often finds the system can run lower than expected without replacing every radiator.
Illustrative example: one room, three flow temperatures
Take a single room with a design heat loss of **1,200 W** and an existing radiator rated at **2,500 W at ΔT50**, with the same assumptions as above (20°C room, 5 K flow/return difference, exponent 1.3):
| Flow temperature | Estimated radiator output | Meets 1,200 W? |
|---|---|---|
| 45°C | ~885 W | No — about 26% short |
| 55°C | ~1,430 W | Yes |
| 65°C | ~2,020 W | Yes, with margin |
Under these simplified assumptions, the existing radiator would meet the room's design heat loss at approximately 51°C flow temperature, provided the required water flow is maintained. To run at 45°C, it would need a radiator rated at around 3,400 W at ΔT50.
This is a single-room illustration. A whole building is designed room by room, and the room that needs the highest temperature — not the average — sets the system flow temperature. For the full derating method and further examples, see [Can an R290 Heat Pump Replace a Gas Boiler Without Replacing the Radiators?](/blog/r290-heat-pump-vs-gas-boiler-can-you-keep-your-radiators).
45°C: the efficient target for well-matched systems
At 45°C or below, a heat pump typically operates in its most efficient range for radiator-based systems. Underfloor heating usually runs lower still (see [What Is the Best Flow Temperature for Underfloor Heating with a Heat Pump?](/blog/what-is-the-best-flow-temperature-for-underfloor-heating-with-a-heat-pump)).
The cost is emitter output: radiators may need to be upgraded, supplemented, or replaced with low-temperature models or fan coils in the rooms with the highest heat loss. In a well-insulated home, often only a few rooms need changing.
55°C: the common retrofit middle ground
55°C is a practical design point for many existing homes. It reduces the radiator upgrades needed while keeping efficiency at a reasonable level. It is also the ErP medium-temperature reference, so seasonal performance at this temperature is often available for comparison between products — though that makes it a convenient benchmark, not an automatic design choice.
With weather compensation, a system designed for 55°C on the coldest day may run considerably lower for much of the heating season — the design temperature is a ceiling, not a constant.
65°C: an engineering and economic trade-off
Designing for 65°C allows more existing radiators to stay in place, which can reduce installation cost, disruption and project time. Many R290 heat pumps can now deliver this temperature. In some projects — where radiator replacement is difficult, expensive, or restricted — those savings may justify higher running costs.
The trade-off should be evaluated rather than assumed:
- **Running cost over the system's life.** Lower efficiency at 65°C means more electricity for the same heat, every heating season.
- **Capacity at low outdoor temperatures.** A unit's heating output usually falls as outdoor temperature drops and flow temperature rises. Check the capacity at your design outdoor temperature *and* 65°C flow, not the headline rating.
- **How much of the house actually needs it.** If only one or two rooms force the high temperature, upgrading those emitters may change the calculation significantly.
What it means for running costs
The electricity a heat pump uses is the heat delivered divided by COP:
**Electricity (kWh) = Heat demand (kWh) ÷ COP**
**Cost per kWh of heat = Electricity price ÷ COP**
So a change in COP feeds straight through to the bill. As pure arithmetic: if a lower flow temperature lifts the average COP from 3 to 4, the electricity needed for the same heat falls by a quarter. Going the other way, from 4 to 3, it rises by a third.
Because COP varies with weather, the meaningful number is the **seasonal** figure at the flow temperatures the system will actually run, not a single best-case test point. A lower design flow temperature combined with weather compensation tends to improve seasonal performance more than any single setting change.
Selecting an R290 heat pump for higher flow temperatures
R290 (propane) has a critical temperature of around 97°C, compared with around 78°C for R32. That higher critical temperature makes R290 well suited to high-temperature heat pump design. Actual outlet temperatures and efficiency, however, depend on the compressor's operating limits, the overall system design and the test conditions behind the published figures — the refrigerant alone does not determine them. Under EU Regulation 2024/573, R290 has a listed GWP100 of 0.02; older references may quote a different value because they draw on earlier assessment data.
When a project genuinely needs 55–65°C, these selection points help:
1. **Confirm the maximum outlet temperature and the conditions behind it.** A stated maximum may only apply above a certain outdoor temperature. Look for the operating envelope chart.
2. **Size on capacity at design conditions and target flow temperature**, for example A-7/W55 or A-7/W65 where published, not the A7/W35 headline.
3. **Compare seasonal efficiency at 55°C** where declared, which reflects radiator operation better than the 35°C figure.
4. **Check weather compensation and control.** The ability to run a compensation curve, rather than a fixed flow temperature, often matters more than peak capability.
5. **Consider domestic hot water separately.** Hot water cylinders typically need around 55–60°C storage, plus periodic higher-temperature cycles for hygiene where required. A unit capable of high outlet temperatures can cover this, but space heating should not be run at that temperature just because the unit can reach it.
How to find the right flow temperature for a specific home
1. **Calculate room-by-room heat loss** at the local design outdoor temperature.
2. **Check each room's existing emitter output** at candidate flow temperatures (45°C, 50°C, 55°C).
3. **Identify the limiting rooms** — the worst room sets the system flow temperature.
4. **Compare upgrading those emitters against running hotter**, then settle on the lowest design flow temperature that makes sense for the project.
5. **Set a weather compensation curve** so the system only reaches design flow temperature on the coldest days.
A low-temperature boiler trial can provide useful preliminary evidence, but results on mild days cannot establish the flow temperature required at the winter design condition, when heat loss is much higher. A room-by-room heat-loss and emitter assessment is still necessary.
FAQ
**Is 55°C too high for a heat pump?**
Not necessarily. 55°C is a common design temperature for radiator retrofits and a reference point in EU seasonal efficiency declarations. Efficiency is lower than at 45°C, but with weather compensation the system may only reach 55°C on the coldest days.
**Can a heat pump run at 65°C all winter?**
Many R290 units can reach 65°C, but running there continuously usually means noticeably lower efficiency and higher running costs, and output may fall in very cold weather. Whether that is acceptable depends on how it compares with the cost of upgrading the limiting emitters.
**Do I need bigger radiators for a heat pump?**
It depends on the flow temperature and on how generously the existing radiators were sized. Many homes need only a few rooms upgraded, and "bigger" usually means higher rated output rather than physically much larger. A room-by-room heat-loss and emitter check is the only reliable way to know.
**Does a lower flow temperature mean the house heats up more slowly?**
It can, which is why heat pumps are usually run more steadily rather than in short boiler-style bursts. Steady operation at a lower temperature is typically both more comfortable and more efficient.
**Does flow temperature affect hot water?**
Hot water is usually produced in a separate cycle at its own temperature, so space-heating flow temperature and hot water temperature can be set independently.
Continue Exploring
**📖 Related Guides**
- [Can an R290 Heat Pump Replace a Gas Boiler Without Replacing the Radiators?](https://nordthermglobal.com/blog/r290-heat-pump-replace-gas-boiler-keep-radiators)
- [What Is the Best Flow Temperature for Underfloor Heating with a Heat Pump?](https://nordthermglobal.com/blog/best-flow-temperature-underfloor-heating-heat-pump)
- [Do You Need a Buffer Tank with an Air-to-Water Heat Pump?](https://nordthermglobal.com/blog/buffer-tank-air-to-water-heat-pump)
**📸 Real Projects**
- [Air-to-Water Heat Pump Retrofit for an Existing UK House with Underfloor Heating and Fan Coil Cooling](https://nordthermglobal.com/case/air-to-water-heat-pump-retrofit-for-an-existing-uk-house-with-underfloor-heating-and-fan-coil-cooling)
- [UK Home Extension: R290 Heat Pump with Wet Underfloor Heating](https://nordthermglobal.com/case/uk-home-extension-r290-heat-pump-with-wet-underfloor-heating)
**🏠 Related Products**
- [R290 Air-to-Water Heat Pump Range](https://nordthermglobal.com/products/r290-air-to-water-split-heat-pump-for-heating-cooling-and-hot-water)
**Planning a radiator retrofit or selecting an R290 heat pump?**
Whether you are an installer checking flow-temperature requirements, an HVAC contractor sizing a retrofit, or a distributor comparing high-temperature models, share your design outdoor temperature, heating load and target flow temperature. We can review the operating conditions with you and provide the relevant product specifications for suitable R290 air-to-water heat pump options.
