Fig. 20 Measuring a heat pump as part of the "WP Effizienz" project.

Fig. 21 Comparison of the monitoring projects

Fig. 22 Average seasonal performance factors (SPF) for all systems. Above: Air source heating, below: Ground source heating.

Fig. 23 Schematic showing an air-water heat pump.
© Bundesverband Wärmepumpe e. V., Berlin

Fig. 24 Schematic showing a ground-water heat pump
© Bundesverband Wärmepumpe e. V., Berlin

Fig. 25 Bandwidths for the annual performance factors determined in the field test for ground source and air source heat pumps in new and old buildings.

Fig. 26 Installed measurement technology.

Fig. 27 Test rig.

Fig. 28 Online visualisation of operating data.
4 / 12

Efficiency under real conditions

In three broad-based monitoring projects, the Fraunhofer Institute for Solar Energy Systems has investigated the efficiency of electrically driven heat pump systems for space heating and domestic hot water heating in old and new buildings. These field tests have not only determined the efficiency achieved in actual operation but have also investigated the operating conditions and analysed the system behaviour. This was used for assessing the respective systems and for determining potential for optimisation.

The monitoring projects are based on recording heat and electrical energies as well as volume flows and temperatures at one-minute intervals, and also include the daily retrieval of remote data and its storage and evaluation at the institute. The generated thermal energy is balanced directly after the heat pump – separately for the space heating and domestic hot water heating. To calculate the seasonal performance factor, the thermal energy is divided by the energy input of the electrical components. This takes into account the compressor and control system of the heat pump as well as the motor of the heat source circuit’s fan, brine or well pump and the electrical back-up heater.

As part of the “WP Effizienz” project, Fraunhofer ISE measured around 100 heat pumps in newly built singlefamily homes that had a specific heating consumption between 30 and 150 kWh/m² p.a. In a similar project (“WP im Bestand”), Fraunhofer ISE investigated heat pumps in existing buildings that had not been or only partly refurbished. Here around 70 systems were measured that had been installed as a replacement for oilfired boilers.

Seasonal cycle of performance factors

The efficiency that a heat pump can achieve in operation is largely determined by the temperature level on the heat source and heat sink side. The smaller the difference between the heat source and the heat sink, the higher the efficiency of the heat pump. This correlation becomes clear when examining the monthly performance factors for the “WP Effizienz” project.

Fig. 22 depicts the supply temperatures for the heat sink circuit (separated for space heating and domestic water heating) and the monthly performance factors during the course of a year (July 2009 to June 2010) as a mean value for all evaluated ground-to-water heat pumps. It can be clearly seen how the monthly performance factors fluctuate across the course of the year. These reflect the changing operating conditions. Due to the almost exclusively used radiant heating, the supply temperatures for space heating achieve an average of 36 °C, whereas the hot water operation achieves an average value of 51 °C.

During the course of the year, the ratio of heating energy that is respectively provided for space heating and domestic water heating varies in accordance with the heating demand for the particular months. During the core heating season, an average of around 10 % was used for hot water heating in the systems investigated, which is a rather low value for new buildings. The weighted mean value of the heat sink temperature is therefore close to the temperature for space heating in the winter months and close to the temperature for domestic hot water generation during the summer months. The source temperatures (Fig. 29) also show seasonal variations. However, the higher brine temperatures in the summer months cannot “compensate” for the high sink temperatures. In addition to the higher temperature lift in the summer months, the increased proportion of electrical energy used by the control systems as a result of the short operating times of the heat pump also has a negative effect on the seasonal performance factor. The mean monthly performance factors in July/August amount to just over 3.0 and increase at the beginning of the heating season to values above 4.0. During this period the heat pump benefits from the still regenerated ground and the lower average sink temperatures. During the core heating season, the monthly performance factors drop slightly to below 4.0. Considering the whole year, the average annual performance factor (SPF) was 3.9.

The investigated air-to-water heat pumps are also installed in buildings that on average show a similarly high proportion of space heating with low heating circuit (HC) temperatures. This means that there are comparable conditions on the heat sink side (Fig. 22). However, there is a considerable difference in the monthly performance factors compared with ground-to-water heat pumps. Whereas with ground source heat pumps the heating seasons can be clearly recognised by the greater monthly performance factors, it is particularly during the winter months that the air source heat pumps work most inefficiently, since only low source temperatures are available to the heat pumps. For example, the lowest average monthly performance factor (2.6) is in January – which correlates with the outside temperature. Although the air temperatures are considerably higher in the summer months, the weighted mean sink temperatures are also higher due to the high share of domestic hot water. The highest monthly performance factors can therefore be found in the transitional periods (November 2009 and May 2010 with 3.2). When the outside temperature is slightly below the heating limit, the average heat sink temperatures are lower than in summer due to the low temperatures for the space heating operation, whereas the outside temperatures are still relatively high. As with ground-to-water pumps, the “standby consumption” of the heat pumps in the summer months with shorter operating times has a greater (and therefore negative) effect on the monthly performance factor than in the months with longer operating times. The average annual performance factor is 2.9.

Comparison of old and new buildings

Fig. 25 compares the respectively achieved annual performance factors for the two projects. The diagram shows the average annual performance factors determined across the described evaluation periods for all systems in the four combinations comprising old and new buildings and air and ground as the heat source. As described above, the ground source heat pump systems in the “WP Effizienz” project were operated with an average annual performance factor of 3.9, whereas the ground source heat pump systems in the “WP im Bestand” project achieved an average annual performance factor of 3.3. This reflects the installed heat distribution systems: the new buildings are almost exclusively equipped with underfloor heating whereas the space heating in most of the old buildings is provided using radiators. With an average of 54 °C, the heating supply temperatures in the old buildings were almost 20 K above those in the new buildings. By trend, the same difference between the two projects is also shown with the measured air source heat pumps: whereas an average annual performance factor of 2.9 was achieved in the new buildings, the average annual performance factor with the old buildings was 2.6.

The examination of the mean values only reveals the general tendency of the differences between the two projects. Within each project, however, there are some considerable differences in the annual performance factors of the individual systems. For example, although three quarters of the ground source heat pumps in the “WP Effizienz” project achieved annual performance factors in the range of +/– 10 % around the project mean value, individual values for the annual performance factors ranged between 3.0 and 5.2. The highest value was achieved by a system that on the one hand was operated with below average sink temperatures and on the other hand also maintained high source temperatures during the core heating season as a result of a generously sized borehole heat exchanger. In the case of the old buildings, the ground source system with the lowest efficiency achieved an annual performance factor of 2.2 and the system with the highest efficiency, which is located in a comprehensively refurbished building with underfloor heating and is not used for domestic water heating, has an annual performance factor of 4.7. The annual performance factors for the air source heat pumps cover roughly the same range with both the existing and new buildings and extend from 2.1 to 3.3 and from 2.3 to 3.4 respectively.

Use of electrical back-up heaters

Fortunately, the use of electrical back-up heaters is very low in most of the measured systems. The results from the “WP Effizienz” project for the year 07/2009 – 06/2010 are presented here. In the case of ground source systems, the majority of the 56 plants did not activate the electrical back-up heaters at all. And only around 10 % of the systems provided more than 1 % of the thermal energy for space heating and domestic water heating with an electrical back-up heater. The detailed evaluation of the activities of the electric back-up heaters indicates that these activities were at least partly due to a deliberate use of the electric back-up heaters to provide support while drying out the building, unfavourable parameterisation of the control systems or short-term heat pump failures.

As expected, air source heat pump systems provided a larger share of the heat with electrical back-up heating. Here around 40 % of the 18 systems provided more than 1 % of the thermal energy with electric back-up heaters. However, with the exception of one system, less than 5 % of the heat was provided by electric back-up heaters. Whereas with ground source systems the heating element activity does hardly correlate with the outside temperatures, this correlation is evident in most air source heat pumps. In almost all investigated systems, activity of back-up heaters was detected during the coldest months. Some systems activated the electric back-up heater also during the hot summer months. With extremely high outside temperatures, operational conditions can lie outside the characteristic diagram of the compressor. Then, the heat pump is not activated and the electric back-up heater takes over the hot water heating.

Electrical energy consumption for fans and pumps

The components on the heat source side, i.e. pump or fan, accounted for a very similar share of energy consumption for both heat sources .The ground source heat pumps required an average of 5.9 % of the heat pump’s electrical energy consumption (without electric back-up heaters) to drive the source pump. The bandwidth for the individual systems was quite evenly spread in a range between 2 and 10 %; in one case the share amounted to 12 %. The reasons for this are diverse and range from the different sizing of the source systems and operating outside of the design flow rates to different pump efficiencies at the respective operating point.

With air source heat pumps, the fans required 6.7 % on average. The bandwidth for the individual systems is comparable to the source pump for ground source systems and ranges from 2 to 11 %.


BINE subscription

Subscribe to publication


Project execution
Fraunhofer ISE


BINE-Themeninfo I/2013
(PDF, 20 pages, 3.3 MB)

More from BINE

Cooling with heat
BINE-Projektinfo 07/2012

Geothermal heating for railway points
BINE-Projektinfo 12/2010

Using geothermal energy in office buildings
BINE-Projektinfo 07/2010

Ground-coupled heat pumps for new buildings
BINE-Projektinfo 03/2010