Ö-quadrat - Ökologische und ökonomische Konzepte

Heat pump - a climate heating system?

Heat pumps for climate protection? 

Heat pumps are often presented as a solution for energy efficient heating; however, after new investigations and results which are based on the daily use of the heating systems, these statements become increasingly doubtful.  In the following, Dieter Seifried clarifies some of the background to this debate by exploring relevant aspects of CO2 emission estimation and heat pump energy efficiency issues. 

More than 215,000 heat pumps are currently operating nationwide, and in the year 2008 alone 45,000 new devices were installed.  If these heat pump installations saved around 1.5 tonnes of CO2 per year per unit (as calculated by the Department of Commerce for Northrhine-Westphalia), the role of heat pumps for climate protection would be notable.  Unfortunately, these devices cause exactly the opposite effect! 

How does one come to this false estimation?  In short, this is the result of a dangerous mixture of incorrect/misleading statements as well as false advertising by heat pump manufacturers and energy providers.  In addition, this almost certainly reflects a poor understanding of energy economics on the part of politicians, planers, and investors. 

A bit of background on heat pumps is needed to properly understand the situation: when it comes to the characterization of heat pumps, the annual efficiency factor describes the efficiency of an electric heat pump.  For instance, an annual efficiency factor of 4 means that each 4 kWh of heat delivered by the heat pump requires 1 kWh electrical input into the device. The annual efficiency factor should not be confused with the “Coefficient of Performance” (COP), which is the (nominal) ratio of heat delivered and electricity consumed under optimal operating conditions for the device.  It should be emphasized that only the annual efficiency factor is appropriate for determining the electric energy consumed by a heat pump, since this reflects the average operating conditions throughout the year. 

The value of the annual efficiency factor depends on several things: it is (for example) higher if the temperature of the heat source is high, the necessary output temperature of the heating system is low (eg. floor heating), and the quality of the overall heating system design is good.  A recent field test of 33 heat pumps installed after 2002 produced the following values for annual efficiency factors: 

Heat source

Annual efficiency factor with floor heating

Annual efficiency factor with radiator heating


2,8 (7 Installations)

2,3 (5 Installations)


3,4 (11 Installations)

3,3 (2 Installations)

Ground Water

3,0 (6 Installations)

3,4 (1 Installations)

Average annual efficiency factors for electric heat pumps; source:  Local Agenda-21-Gruppe Lahr (state March 2008)

The result:  soil heat pumps (ground) are frontrunners in terms of energy utilization. For floor heating systems with low output temperatures, heat pumps obtained an annual efficiency factor of 3,4 (arithmetic mean); however, when heat was transferred using conventional radiators, the efficiency factor dropped to an average of 3,3. 

The air based heat pumps performed most poorly:  when combined with floor heating, their efficiency factor amounted to only 2,8- and this dropped to 2,3 with radiator systems. 

Therefore, in order to calculate the electricity demand of a heat pump, and particularly for the related task of calculating resulting CO2 emissions, an annual efficiency factor should clearly be much less than 4 if the climate protection element of heat pumps is to be properly assessed. 

Another pair of important considerations for assessing the climate impact of heat pumps is: how much CO2 is associated with the electricity consumed by heat pumps? And, how might this change with a mass deployment of heat pumps? In many publications, including the most recent position paper of Federal Bureau for the Environment, the CO2 emissions resulting from heat pump usage are calculated based on the average relative share of power plants in the German power plant mix. For instance, in 2006, the energy produced was based on 27% nuclear energy, 4% hydro power, 5% wind power, 23% brown coal, 21% hard coal, 2% heating oil and 12% with natural gas. By multiplying these relative shares by their respective emission factors, one can derive a reasonable average CO2 production per unit of electricity consumed. However, one must consider: will this average CO2/kWh still be appropriate with a mass deployment of heat pumps (and associated increase in electricity demand)? To answer this, we need to consider how an increase in electricity demand will affect Germany’s power plant mix, and who will actually produce the extra electricity to fill the additional demand.

Consider first, that nuclear plants, brown coal plants and hydro plants are base load power stations and will produce electricity irrespective of market conditions. This is because their variable cost of production is so low. As a result, these plants are nearly independent of load variations and independent of the power demand both in the winter and in the summer, in other words, they always operate more or less at full capacity.  This means that other power plants including peak-load plants (gas-fired, oil and hard coal plants) must be used more often and to a greater extent to satisfy the additional power demand; they would also have to be expanded and reinforced if the demand increases (especially) from heat pumps during the winter.  Renewable power sources (water, wind, solar energy, and geothermal energy) have been expanded until now independent of the power demand and their costs because it is a declared goal that these shares increase. It should be noted though, that feed-in electricity of renewables is not encouraged or discouraged by heat pumps.

Additional production of solar and wind power would be sure to reduce the specific CO2-emission factor of the total system.  Despite this, the emissions attributed to additional heat pumps must be seen in the context of the specific plants covering their load, which in the case of Germany, would be fossil power plants such as gas, oil, and hard coal plants. Of course, this would be different if the entire power production was coming from renewable energy sources. 

The energy providers and also some institutes calculate the emissions caused by the heat pumps with the specific CO2-emission factor of the entire national German electricity mix.  This amounts to 591 g/kWh.  However, in order to determine the additional emissions caused by heat pumps, one must consider the average CO2 emission factor of the fossil plants that will be covering this increased load, and not the average of the entire system. From this perspective, the “new” CO2-emissions resulting from heat pump power demand works out to around 783 g/kWh.  If one reckons network losses of around 6%[1], the specific CO2 emissions for heat pump energy rises to around 830 g/kWh. 

Let's compare now the emissions that emerge from the heating system of a detached house with 180 meters squared living space and a heating demand of 20,000 kWh/a.  If we assume that a ground-source heat pump system is installed at this house (with a high efficiency factor of 3.4), it would require approximately 5,882 kWh of electricity to heat the house.  This would result in approximately 4,88 tonnes of CO2 per year. 

If the same house were equipped with a natural gas heating system (assuming an annual utilization ratio of 95 % [2] related to the caloric value), this would result in approximately 21,000 kWh of energy consumed for heating purposes. The result of this consumption would be annual emissions of 4.23 tonnes CO2 (with in an emission factor of 201 g CO2/kWh for natural gases relative to caloric value). From this simple example, it can be clearly demonstrated that the heat pump system causes about 15% more CO2 emissions than a modern gas heating system. 

The above results are striking since the assumptions used in the above model actually overstate the benefits of heat pumps, and the results still indicate that heat pumps perform poorly compared with a modern gas heating system. Specifically: 

  • The mathematical model employed to calculate the average specific CO2 factor or power production for addition heat pumps. However, the heat pump demand will be met not with an average power plant. It will rather be an older fossil power plant with a lower efficiency rate and higher emissions. Therefore the emissions to be expected (also emissions such as sulfur dioxide and fine dust) would be higher than suggested.
  • The assessment does not consider that heat pumps typically employ (and may leak) halogenated fluorine hydrocarbons, which are extremely potent greenhouse gases.
  • Additionally, the annual efficiency factor selected for the example used the highest value from the heat pump field study considered previously (3.4). If a weighted average value was taken, which considers the other less efficient systems in a more balanced way, the heat pump would become around 26% worse off (relative to calorific value heating) than the natural gas system.  

What follows from this:  The heat pump strategy pursued by energy providers and heat pump manufacturers is to be sure good for their revenues – however not for the climate.  Homeowner and investors that decide on a heat pump are also shaping the structure and mix of the future energy supply system in Germany.  The power demand of 210,000 heat pumps leads to an increased power demand especially in the winter months.  The way towards a low-carbon and nuclear-free future will require a different approach that embraces heat production centers located near consumption areas, which can take the form of district heating systems, solar energy collectors as well as decentralized small biofuel/cogeneration plants.  These technologies will contribute to a more wise use of energy resources and achieve a better balance with the climate and the environment by avoiding the unnecessary energy consumption and electricity demand caused by heat pumps.

*Dieter Seifried, Ö-quadrat, Freiburg

[1] It was demonstrated for the year 2006 in Germany that avoided demand leads to a corresponding decline of the power production from mid-level and peak load power plants. Additionally, it was determined that there were losses in the system and particularly the low-voltage system in the range of 6% compared with the energy produced at power plants.  Quelle: VDI 2007: "Erstellung der Grundlagen für einen harmonisierten und fortschreibbaren Datensatz des deutschen Strommixes."  Research completed on behalf of the Federal Ministry for Development and Research. 

[2] Using the lower caloric value.  Quelle: Prof. Dr. Wolff, Fachhochschule Braunschweig Wolfenbüttel, Felduntersuchung – Betriebsverhalten von Heizungsanlagen mit Gas-Brennwertkesseln“, Juli 2004