Retrofitting a boiler system with a heat pump requires more than matching a unit's rated capacity. The water temperature at which the radiators, fan coils and underfloor circuits can deliver the required heat determines comfort, efficiency and operating cost.
The guiding principle is to use the lowest supply-water temperature that meets the building's heating demand. Lowering temperature without checking emitters can sacrifice comfort; retaining the old boiler temperature can sacrifice efficiency.
1. Why does water temperature matter?
A heat pump transfers heat from a lower-temperature source to a higher-temperature heating circuit. A larger temperature lift makes the compressor work harder. As outdoor temperature falls, building demand rises; increasing supply-water temperature at the same time makes the operating point more demanding. One catalogue COP cannot describe annual performance.
COP 4.0 means approximately 4 kWh of heat for each 1 kWh of electricity at that operating point. Seasonal performance also depends on weather, water temperature, defrost, circulation pumps, auxiliary heaters and part-load operation.
2. The correct sequence for a retrofit assessment
- Calculate room-by-room heat loss, including design outdoor temperature, indoor targets, envelope construction, infiltration and ventilation.
- Inventory emitters: radiator type and dimensions, fan-coil coils and underfloor circuits.
- Establish output at alternative water temperatures using manufacturer tables and correction factors.
- Check the hydraulics: flow, internal pipe diameter, pressure loss, pumps, valves and balancing.
- Match the heat pump at the actual outdoor-air and water-temperature operating points, checking both output and COP.
Historical fuel use is a useful cross-check, but occupancy, boiler efficiency and domestic hot-water demand affect it. It is not a substitute for design heat loss.
3. How radiator output changes
Radiator output falls as the difference between mean water temperature and room temperature decreases. The following relation is useful for an initial assessment; manufacturer data governs final selection.
A radiator producing 2,000 W at 75/65 °C in a 20 °C room has a reference temperature difference of 50 K. At 45/40 °C, the difference becomes 22.5 K. With an illustrative panel-radiator exponent n=1.3, output falls to about 710 W. This explains why an adequate boiler-era radiator can become inadequate after a low-temperature retrofit.
4. Flow, pipes and pumps
Transporting 10 kW requires about 1.72 m³/h at 5 K, or 0.86 m³/h at 10 K. Increasing flow through the same pipe can increase velocity, noise and pressure loss. Check pipes, strainers, valves and the pump at the new operating point.
Maintain the manufacturer's minimum flow. Do not add a buffer tank or hydraulic separator automatically: identify a need such as hydraulic separation, minimum water volume or defrost energy, and avoid unnecessary mixing.
5. Weather compensation
The system should not operate all season at the water temperature required for the coldest design hour. Weather compensation lowers supply temperature as outdoor temperature rises. Start with calculated design points and adjust the curve using site measurements.
If a room remains cold, check flow, balancing, filters and valves before raising the curve. Frequent cycling calls for investigation of minimum unit output, zoning, water volume and thermostat behaviour.
6. Assess domestic hot water separately
Low-temperature space heating and domestic hot-water production require different operating scenarios. Storage hygiene and Legionella control may require higher temperatures. Select a suitable cylinder coil and coordinate storage temperature, hygiene cycles, electric backup and protection against scalding.
7. Commissioning and monitoring
Measure design flows, clean filters, purge air and verify sensors. Record outdoor, supply, return and room temperatures alongside electricity consumption and heat delivered. Defrost frequency, auxiliary-heater runtime and compressor starts can reveal problems early.
The handover file should include heat-loss calculations, unit operating points, emitter capacities, design flows, the weather-compensation curve, auxiliary-heater settings, the hot-water strategy and a first-season monitoring plan.
8. Common mistakes
- Selecting the heat pump from the old boiler's nameplate capacity.
- Treating a single COP as annual performance.
- Ignoring radiator output at low water temperatures.
- Overlooking the effect of higher flow on pipes and pumps.
- Disabling compensation and maintaining a fixed high temperature.
- Leaving electrical connection capacity and backup heating until the end.
9. Room-by-room emitter check: a worked comparison
Choosing water temperature from total building demand can conceal the weakest room. List each room's design loss and installed radiator together. If the label is missing, measure radiator type, height, length and panel count and verify output against manufacturer data. Covers, curtains, furniture and connection arrangements also affect performance. One deficient room does not automatically justify raising the temperature of the entire building.
Consider a room losing 1,200 W with a radiator rated at 2,000 W under 75/65/20 °C conditions. For this educational example, take n=1.3. The mean-water-to-room difference is 50 K at the reference condition and 22.5 K at 45/40/20. Output is 2,000 × (22.5/50)^1.3 ≈ 708 W. At 55/45/20 it is approximately 1,030 W, and at 60/50/20 approximately 1,258 W. The first two alternatives do not meet the room's demand.
Confirm final output using the manufacturer's actual low-temperature data. The arithmetic mean-water approximation requires particular care with large supply/return differences.
Meeting 1,200 W at 45/40 would require approximately 3,389 W of nominal radiator output under the reference condition. This is a comparison of emitter-upgrade options, not a direct product selection. Compare envelope improvements, alternative emitters and higher system temperatures through both investment and operating costs. If inadequate flow causes the cold room, investigate air, valves and balancing before replacing the radiator.
10. Capacity, electricity and minimum load are separate checks
A high COP does not prove sufficient output at the coldest design condition. If demand is 18 kW and actual unit output at the required outdoor and supply temperatures is 14 kW, there is a 4 kW shortfall. Consider auxiliary heating, envelope improvements or another unit. The model name is not evidence of capacity: document defrost conditions and manufacturer corrections.
At mild-weather demand of 4 kW, a unit with a 7 kW minimum output may cycle excessively. A larger model can worsen this problem.
For illustration, suppose 10,000 kWh of heat is delivered at COP 4, another 2,000 kWh at COP 2.5, and 1,000 kWh by resistance backup. Electricity use is 2,500+800+1,000=4,300 kWh for 13,000 kWh of heat: a ratio of approximately 3.02. If distribution pumps and auxiliaries are excluded, do not present this as whole-system seasonal performance. Real weather and operating data require an hourly or otherwise suitable time-step assessment.
11. Retrofit decision matrix and measurement plan
| Finding | Priority investigation | Verification |
|---|---|---|
| A few rooms remain cold | Local emitter, valve and heat loss | Room calculation and flow measurement |
| The entire building is cold | Envelope and water temperature | Alternative design scenarios |
| Water temperature difference is excessive | Insufficient flow or control issue | Filter, pump and balancing checks |
| Frequent starts and stops | Minimum capacity and zoning | Runtime and water-volume assessment |
| Cold-weather capacity shortfall | Auxiliary source and heat pump | Actual design operating point |
A low-temperature trial using the existing boiler can reveal emitter behaviour, but comfort on a mild day does not prove performance on the coldest day. Record weather, room temperatures, supply/return temperatures, pump settings and runtimes together.
For 12 kW, about 2.07 m³/h is required at 5 K and 1.03 m³/h at 10 K. A smaller temperature difference may require more flow through existing pipes: check internal diameters and the critical circuit's pressure loss. A separator or bypass can mix water so that emitters receive colder water than the heat pump produces.
During the first weeks, identify problem rooms and investigate local causes before lifting the whole compensation curve. Record the date and reason for every curve change. When comparing bills, account for differences in weather, occupancy and hot-water use. Technical background: https://www.caleffi.com/en-us/blog/4-key-design-concepts
12. Conclusion
A successful retrofit provides the required comfort at the lowest practical water temperature. Heat loss, emitters, flow, pipes and controls form one connected design calculation.