Researchers from ETH Zurich have carried out the largest field study to date, to find out whether heat pumps are running efficiently. When looked at in operation, it turns out that these units are often not optimally configured. Monitoring systems and legal standards could provide a remedy.
In brief
- More and more house-owners and property managers are choosing heat pumps to heat their buildings.
- A study by ETH Zurich has determined that many of these units have been wrongly configured or set up, and are therefore inefficient.
- In the future, digital tools could be used to monitor installations over a wide area and ensure that they are operating efficiently.
Heat pumps are currently the most sought-after heat source in Switzerland. Since the year 2000, their number has increased many times over and every fifth building is now heated with one. These units make an important contribution to the net zero target of society as a whole. However, they do increase electricity consumption, which makes their efficient operation an important consideration. Most owners cannot tell whether their units have been properly configured. But trust in, and acceptance of, the deployment of climate-friendly technologies depends on their efficiency, because this is directly reflected in energy bills.
Europe-wide field study puts efficiency under the microscope
In the largest field study to date, researchers at ETH Zurich investigated the efficient operation of heat pumps. They spent two years gathering real-time data from 1,023 units in ten European countries, including Germany, Austria, the Netherlands, Sweden, Denmark, France and the United Kingdom. The aim of the study was to develop a system to compare the units to each other in a standardised manner, even though they were installed in different types of buildings. The study's authors only had access to data from one manufacturer, but they assume that their findings will also apply to others. "Our results can be interpreted universally, as the collected data and algorithms are based exclusively on physical measurements such as flow temperature and energy consumption," says Tobias Brudermüller, one of the main authors.
Incorrect configurations lead to considerable energy wastage
During the observation period, considerable differences in performance came to light. The units with the lowest levels of efficiency were, in some cases, two or three times less efficient that those with the highest efficiency levels. In fact, 17 percent of air-source heat pumps fell short of the European efficiency standards.
Optimising the heating curve setting would solve this problem for many units, they determined. The ETH researchers also showed in their study that one unit in ten was too large. A less powerful unit would have sufficed for the prevailing conditions in the building, and would have consumed less electricity. However, only 1 percent of units were too small, which can lead to units becoming overworked and breaking down more readily. "Even though we knew that errors in planning and configuring heat pumps were not uncommon, it surprised us how strongly this was reflected in the actual efficiency levels achieved, and how clearly that showed up in the data," says Thorsten Staake, Co-Chair of the Bits to Energy Lab at ETH Zurich.
Heat pumps in Switzerland could be more efficient
Back in 2023, Tobias Brudermüller carried out similar studies in Switzerland in collaboration with partners from industry, the Swiss Federal Office of Energy and the electricity utility provider of the canton of Zurich (EKZ). These were based on electricity consumption data from Swiss households equipped with smart meters and evaluated using artificial intelligence. These studies determined that 41 percent of the 410 units under investigation had their heating curve set too high, so that the heat pumps had to generate an unnecessarily high rise in temperature. In 36 percent of cases, the reduced-temperature night mode was activated, which led to the building cooling overnight and then requiring reheating in the morning. In 26 percent of cases, the heat threshold was set too high, meaning that the units were in operation for longer than necessary in light of the outdoor temperature.
Efficiency tips for heat pump owners:
- A flow temperature that is as low and consistent as possible (between 35°C and 55°C, depending on the building) is an indicator of efficient operation.
- You can tell whether you have a good ratio of electricity consumption to heat generated by the annual performance factor, which should be over three. This can generally be checked on an app or on the heat pump's control panel.
- Consistent operation points to a well-configured unit. Make sure that the unit does not switch itself on and off too frequently.
Digital tools needed
In order to guarantee efficiency over the entire life cycle, the study's authors suggest standardising the monitoring of the units across the whole of Europe. "We still do not have the framework conditions for technical standards that would suit the variety of situations affecting real-life buildings," says Brudermüller.
By developing their algorithms, they want to provide information to households specifically on the potential for optimising their heat pumps. Sensor data from web-enabled heating appliances and smart meters, in particular, opens up new possibilities. "This data will help owners to get to know their units better, reduce their energy bills and improve satisfaction with their units. That in turn will lead to these systems catching on more quickly and climate targets being reached sooner," says Brudermüller. Elgar Fleisch adds: "Heat pumps are already an advanced technology for climate-friendly heating. The next step is to make their operation more intelligent and efficient. Our field studies create the basis for bringing this innovation into households quickly and making its mark directly in the form of lower energy bills."
References
Brudermüller T, Potthoff U, Fleisch E, Wortmann F, Staake T: Estimation of Energy Efficiency of Heat Pumps in Residential Buildings Using Real Operation Data. Nature Communications, 22 March 2025, doi: external page https://doi.org/10.1038/s41467-025-58014-y
Brudermüller T, Staake T: KI in der Wärmepumpenberatung [AI in heat pump advisory services]. Bundesamt für Energie [Swiss Federal Office of Energy], 18 December 2024, external page https://www.aramis.admin.ch/Default?DocumentID=72956&Load=true